Breeding progress and preparedness for mass‐scale ...
Transcript of Breeding progress and preparedness for mass‐scale ...
University of Dundee
Breeding progress and preparedness for mass-scale deployment of perenniallignocellulosic biomass crops switchgrass miscanthus willow and poplarClifton-Brown John Harfouche Antoine Casler Michael D Dylan Jones Huw MacalpineWilliam J Murphy-Bokern DonalPublished inGlobal Change Biology Bioenergy
DOI101111gcbb12566
Publication date2019
LicenceCC BY
Document VersionPublishers PDF also known as Version of record
Link to publication in Discovery Research Portal
Citation for published version (APA)Clifton-Brown J Harfouche A Casler M D Dylan Jones H Macalpine W J Murphy-Bokern D Smart LB Adler A Ashman C Awty-Carroll D Bastien C Bopper S Botnari V Brancourt-Hulmel M Chen ZClark L V Cosentino S Dalton S Davey C Lewandowski I (2019) Breeding progress andpreparedness for mass-scale deployment of perennial lignocellulosic biomass crops switchgrass miscanthuswillow and poplar Global Change Biology Bioenergy 11(1) 118-151 httpsdoiorg101111gcbb12566
General rightsCopyright and moral rights for the publications made accessible in Discovery Research Portal are retained by the authors andor othercopyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated withthese rights
bull Users may download and print one copy of any publication from Discovery Research Portal for the purpose of private study or research bull You may not further distribute the material or use it for any profit-making activity or commercial gain bull You may freely distribute the URL identifying the publication in the public portal
Take down policyIf you believe that this document breaches copyright please contact us providing details and we will remove access to the work immediatelyand investigate your claim
Download date 14 Nov 2021
P L A T FORM
Breeding progress and preparedness for mass‐scale deploymentof perennial lignocellulosic biomass crops switchgrassmiscanthus willow and poplar
John Clifton‐Brown1 | Antoine Harfouche2 | Michael D Casler3 | HuwDylan
Jones1 | William JMacalpine4 | DonalMurphy‐Bokern5 | Lawrence B Smart6 |
Anneli Adler78 | Chris Ashman1 | Danny Awty‐Carroll1 | Catherine Bastien9 |
Sebastian Bopper10 | Vasile Botnari11 | Maryse Brancourt‐Hulmel12 | Zhiyong Chen13 |
Lindsay V Clark14 | Salvatore Cosentino15 | Sue Dalton1 | Chris Davey1 |
Oene Dolstra16 | Iain Donnison1 | Richard Flavell17 | Joerg Greef18 | Steve Hanley4 |
AstleyHastings19 | MagnusHertzberg7 | Tsai‐WenHsu20 | Lin SHuang1 |
Antonella Iurato1 | Elaine Jensen1 | Xiaoli Jin21 | Uffe Joslashrgensen22 | AndreasKiesel23 |
Do‐SoonKim24 | Jianxiu Liu25 | JonPMcCalmont1 | BernardGMcMahon26 |
MichalMos27 | Paul Robson1 | Erik J Sacks14 | Anatolii Sandu11 | Giovanni Scalici15 |
Kai Schwarz18 | Danilo Scordia15 | Reza Shafiei28 | Ian Shield4 | Gancho Slavov4 | Brian
J Stanton29 | Kankshita Swaminathan30 | Gail Taylor31 | Andres F Torres16 | Luisa
M Trindade16 | TimothyTschaplinski32 | GeraldA Tuskan32 | ToshihikoYamada33 |
ChangYeonYu34 | Ronald S ZalesnyJr35 | JunqinZong25 | Iris Lewandowski23
1Institute of Biological Environmental and Rural Sciences Aberystwyth University Aberystwyth UK2Department for Innovation in Biological Agrofood and Forest systems University of Tuscia Viterbo Italy3USDA‐ARS US Dairy Forage Research Center Madison Wisconsin4Rothamsted Research Harpenden UK5LohneGermany6Horticulture Section School of Integrative Plant Science Cornell University Geneva New York7SweTree Technologies AB Umearing Sweden8Institute of Crop Production Ecology Swedish University of Agricultural Sciences Uppsala Sweden9INRA‐BIOFORA Orleacuteans France10Department of Seed Science and Technology Institute of Plant Breeding Seed Science and Population Genetics University of Hohenheim StuttgartGermany11Institute of Genetics Physiology and Plant Protection (IGFPP) of Academy of Sciences of Moldova Chisinau Moldova12INRA‐AgroImpact Peacuteronne cedex France13Insitute of Miscanthus Hunan Agricultural University Hunan Changsha China
These authors contributed equally to this work
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -This is an open access article under the terms of the Creative Commons Attribution License which permits use distribution and reproduction in any medium provided theoriginal work is properly citedcopy 2018 The Authors GCB Bioenergy Published by John Wiley amp Sons Ltd
Received 29 May 2018 | Accepted 18 July 2018
DOI 101111gcbb12566
GCB Bioenergy 20181ndash34 wileyonlinelibrarycomjournalgcbb | 1
14Department of Crop Sciences amp Center for Advanced Bioenergy and Bioproducts Innovation 279 Edward R Madigan Laboratory University ofIllinois Urbana Illinois15Dipartimento di Agricoltura Alimentazione e AmbienteUniversitagrave degli Studi di CataniaCataniaItaly16Plant Breeding Wageningen University amp Research Wageningen The Netherlands17Battersea London UK18Julius Kuhn‐Institut (JKI) Bundesforschungsinstitut fur Kulturpflanzen Braunschweig Germany19Institute of Biological and Environmental Science University of Aberdeen Aberdeen UK20Taiwan Endemic Species Research Institute (TESRI) Nantou County Taiwan21Department of Agronomy amp The Key Laboratory of Crop Germplasm Resource of Zhejiang Province Zhejiang University Hangzhou China22Department of Agroecology Aarhus University Centre for Circular Bioeconomy Tjele Denmark23Department of Biobased Products and Energy Crops Institute of Crop Science University of Hohenheim Stuttgart Germany24Department of Plant Sciences Research Institute of Agriculture amp Life Sciences CALS Seoul National University Seoul Korea25Institute of Botany Jiangsu Province and Chinese Academy of Sciences Nanjing China26Natural Resources Research Institute University of Minnesota ndash Duluth Duluth Minnesota27Energene sp z oo Wrocław Poland28James Hutton Institute University of Dundee Dundee UK29GreenWood Resources Inc Portland Oregon30Hudson-Alpha Institute for Biotechnology Huntsville Alabama31Biological Sciences University of Southampton Southampton UK32The Center for Bioenergy Innovation Oak Ridge National Laboratory Oak Ridge Tennessee33Field Science Centre for the Northern Biosphere Hokkaido University Sapporo Japan34College of Agriculture and Life Sciences 2 Kangwon National University Chuncheon South Korea35USDA Forest Service Northern Research Station Rhinelander Wisconsin
CorrespondenceJohn Clifton‐Brown Institute ofBiological Environmental and RuralSciences Aberystwyth UniversityAberystwyth UKEmail jhcaberacuk
Funding informationFP7 Food Agriculture and FisheriesBiotechnology GrantAward NumberOPTIMISCFP7-289159 WATBIOFP7-311929 H2020 Environment GrantAward Number GRACE-745012Bio-Based Industries Joint Undertaking ItalianMinistry of Education Brain GainProgram (Rientro dei cervelli) USDepartment of Energy GrantAwardNumber DE-AC05-00OR22725 DE-SC0006634 DE-SC0012379 and DE-SC0018420 Department of Trade andIndustry (UK) GrantAward Number BW6005990000 Biotechnology andBiological Sciences Research CouncilGrantAward Number CSP17301 BBN0161491 N016149 LK0863 K01711X1 10963A01 G0162161 E0068331G00580X1 000I0410 Department forEnvironment Food and Rural AffairsGrantAward Number LK0863 NF0424NF0426 US Department of AgricultureNational Institute of Food and Agriculture
AbstractGenetic improvement through breeding is one of the key approaches to increasing
biomass supply This paper documents the breeding progress to date for four
perennial biomass crops (PBCs) that have high outputndashinput energy ratios namely
Panicum virgatum (switchgrass) species of the genera Miscanthus (miscanthus)
Salix (willow) and Populus (poplar) For each crop we report on the size of
germplasm collections the efforts to date to phenotype and genotype the diversity
available for breeding and on the scale of breeding work as indicated by number
of attempted crosses We also report on the development of faster and more pre-
cise breeding using molecular breeding techniques Poplar is the model tree for
genetic studies and is furthest ahead in terms of biological knowledge and genetic
resources Linkage maps transgenesis and genome editing methods are now being
used in commercially focused poplar breeding These are in development in
switchgrass miscanthus and willow generating large genetic and phenotypic data
sets requiring concomitant efforts in informatics to create summaries that can be
accessed and used by practical breeders Cultivars of switchgrass and miscanthus
can be seed‐based synthetic populations semihybrids or clones Willow and
poplar cultivars are commercially deployed as clones At local and regional level
the most advanced cultivars in each crop are at technology readiness levels which
could be scaled to planting rates of thousands of hectares per year in about
5 years with existing commercial developers Investment in further development
of better cultivars is subject to current market failure and the long breeding cycles
2 | CLIFTON‐BROWN ET AL
We conclude that sustained public investment in breeding plays a key role in
delivering future mass‐scale deployment of PBCs
KEYWORD S
bioenergy feedstocks lignocellulose M sacchariflorus M sinensis Miscanthus Panicum virgatum
perennial biomass crop Populus spp Salix spp
1 | INTRODUCTION
Increasing sustainable biomass production is an importantcomponent of the transition from a fossil fuel‐based econ-omy to renewables Taking the United Kingdom as an exam-ple Lovett Suumlnnenberg and Dockerty (2014) suggested that14 million ha of marginal agricultural land could be used forbiomass production without compromising food productionAssuming a biomass dry matter (DM) yield of 10 Mgha anda calorific value of 18 GJMg DM 14 million ha woulddeliver around 28 TWh of electricity (with 40 biomassconversion efficiency) which would be ~8 of primary UKelectricity generation (336 TWh in 2017 (DUKES 2017))To achieve this by 2050 planting rates of ~35000 hayearwould be needed from 2022 in line with calculations byEvans (2017) The current annual planting rates in the UnitedKingdom are orders of magnitude short of these levels atonly several hundred hectares per year Similar scenarioshave been generated for other countries (BMU 2009 Scar-lat Dallemand Monforti‐Ferrario amp Nita 2015)
If perennial biomass crops (PBCs) are to make a real con-tribution to sustainable development they should be grownon agricultural land which is less suitable for food crops(Lewandowski 2015) This economically ldquomarginalrdquo land istypically characterized by abiotic stresses (drought floodingstoniness steep slope exposure to wind and sub‐optimalaspect) low nutrients andor contaminated soils (Toacuteth et al2016) In these challenging environments PBCs need resili-ence traits They also need high outputinput ratios forenergy (typically 20ndash50) to deliver large carbon savingsLand may also be marginal due to environmental vulnerabil-ity Much of the value for society from the genetic improve-ment of these crops depends on positive effects arising fromhighly productive perennial systems In addition to produc-ing biomass as a carbon source to replace fossil carbon thesecrops reduce nitrate leaching (Pugesgaard Schelde LarsenLaeligrke amp Joslashrgensen 2015) making them good candidates tohelp fulfil Water Framework Directive (200060EC) and canincrease soil carbon storage during their production (McCal-mont et al 2017)
The objective of this paper was to report on the prepared-ness for wide deployment by summarizing the technicalstate of the art in breeding of four important PBCs namelyswitchgrass miscanthus willow and poplar These four
crops are the most promising and advanced PBCs for tem-perate regions and have therefore the focus here Switch-grass and miscanthus are both rhizomatous grasses with C4
photosynthesis while willow and poplar are trees with C3
photosynthesis Specifically this paper (a) reviews availablecrop trait genetic diversity information (b) assesses the pro-gress of conventional breeding technologies for yield resili-ence and biomass quality (c) reports on progress with newmolecular‐based breeding technologies to increase speedand precision of selection and (d) discusses the require-ments and next steps for breeding of PBCs including com-mercial considerations in order to sustainably meet thebiomass requirements of a growing worldwide bioeconomy
We summarize the crop‐specific attributes the location ofbreeding programmes the current availability of commercialcultivars and yield expectations in selected environments(Table 1) and the generalized breeding targets for all PBCs(Table 2) Economic information relating to the current mar-ket value of the biomass and the investment in breeding arepresented for different countriesregions in Table 3 We alsopresent a comparison of the prebreeding and conventionalbreeding efforts step‐by‐step starting with wild germplasmcollection and evaluation before wide crossing of wild rela-tives (Table 4) Hybridization is followed by at least 6 yearsof selection and evaluation before commercial upscaling canbegin (Figure 1) Recurrent selection often over decades isused within parent populations as part of an ongoing long‐term process to produce hybrid vigour (Brummer 1999) Inthe following sections the state of the art and new opportuni-ties of breeding switchgrass miscanthus willow and poplarare described The application of modern breeding technolo-gies is compared for the four crops in Table 5 It is mostadvanced in poplar and is therefore described in most detail
2 | SWITCHGRASS
Switchgrass is indigenous to the North American prairiesIt is grown from seed and harvested annually using tech-nology similar to that used for pastures Based on collec-tions from thousands of wild prairie remnants the geneticresources are roughly divided into lowland and upland eco-types and there are distinct clades within each ecotypewhich occur along both latitudinal and longitudinal gradi-ents (Evans et al 2018 Lu et al 2013 Zhang et al
CLIFTON‐BROWN ET AL | 3
TABLE1
Breeding‐relatedattributes
forfour
leadingperennialbiom
asscrops(PBCs)
Species
Switc
hgrass
Miscanthu
sW
illow
Pop
lar
Typ
eC4mdash
Grass
C4mdash
Grass
C3mdash
SRC
C3mdash
SRCSRF
Sourcesof
indigeno
usgerm
plasm
CAato
MXeastof
Rocky
Mou
ntains
Eastern
AsiaandOceania
Predom
inantly
Northernhemisph
ere
Northernhemisph
ere
Breedingsystem
Mon
oeciou
sou
tcrossing
Mon
oeciou
sou
tcrossing
Dioeciousou
tcrossing
Dioeciousou
tcrossing
Ploidy
4times8
times2times
3times
4times
2timesminus12
times2times
Specieswith
ingenu
s~4
50~1
4~4
00~3
0ndash32
Typ
esused
mainlyfor
breeding
US
Low
land
ecotyp
e(sub
trop
ical
clim
ates)andup
land
ecotyp
e(tem
perate
clim
ates)
EUb JPSK
andUS
MsinandMsac
EUS
viminalistimesschw
erinii
Sda
syclad
ostimesrehd
eriana
S
dasyclad
osS
viminalis
USandUKS
viminalistimesmiyab
eana
S
miyab
eana
US
Spu
rpurea
timesmiyab
eana
S
purpurea
Pop
ulus
tricho
carpa
Pdelto
ides
Pnigra
Psuaveolens
subsp
maximow
icziiPba
lsam
ifera
Palba
andhy
brids
Typ
ical
haploid
geno
mesize
(Mbp
)~1
500
Msin~5
400
andMsac~4
400
(Raybu
rnCrawfordRaybu
rnamp
Juvik
2009
)
~450
~485
plusmn10
Breedingprog
rammes
CA20
00(REAP
Quebec)
US 19
92(N
ebraska
19
92(O
klahom
a)
1992
(Georgia)
19
96(W
isconsin)
19
96(Sou
thDakota)
20
00(Tennessee)
20
02(M
ississippi)
20
07(O
klahom
a)
2008
(RutgersNew
Jersey)
20
12(Cornell
New
York)
20
12(U
rbana‐Champaign
Illin
ois)
DE19
90s(K
lein‐W
anzleben)
NL20
00s(W
ageningen)
UK20
04(A
berystwyth)
CN20
06(Chang
sha)
JP20
06(H
okkaido)
US 20
06(Californiamdash
CERESInc)
20
06(CaliforniaandIndianamdash
MBI)
20
08(U
rbana‐Champaign)
SK20
09(Suw
on‐SNU)and(M
uan
NICSof
RDA)
FR20
11(Estreacutees‐M
ons)
UK19
80s(Lon
gAshton
relocatedto
Rothamsted
Researchin
2002
)SE
19
80s(SvaloumlfWeibu
llSalix
energi
Europ
aAB)
US
1990
s(Cornell
New
York)
PL20
00s(O
lsztyn
University
ofWarmia
andMazury)
SE
19
39(M
ykingeEkebo
Research
Institu
te)
19
90s(U
ppsalaSL
U)
20
10(U
ppsalaST
T)
US 19
27(N
ewYork
OxfordPaper
Com
pany
andNew
YorkBotanical
Garden
Wheeler
etal20
15)
19
79(W
ashing
ton
UW
andOrego
nGWR)
19
80ndash1
995(M
ississippiMSS
tate
and
GWR)
19
96(M
innesotaUMD
NRRIand
GWR)
IT19
83(Piedm
ontAFV
)FR
20
01(O
rleacuteans
ampNancyIN
RA
Charrey‐sur‐Saocircne
ampPierroton
FCBA
Nog
ent‐sur‐V
ernisson
IRST
EA)
DE20
08(G
oumlttin
gen
NW‐FVA)
(Con
tinues)
4 | CLIFTON‐BROWN ET AL
TABLE
1(Contin
ued)
Species
Switc
hgrass
Miscanthu
sW
illow
Pop
lar
Current
commercial
varietieson
the
market
US
Nocommercial
hybrids
CA2
EU1(M
timesgfrom
different
origins)
+selected
Msinforthatching
inDK
US
3(but
2aregenetically
identical)
UK25
US
8EUDElt10
FR44
IT10ndash1
5SE
~1
4US
8ndash12
Southeast8ndash
14Upp
erMidwest10
PacificNorthwest
Precom
mercial
cultivars
expected
tobe
onthemarketin
3years
US
36registered
cultivars
(halfare
rand
omseed
increasesfrom
natural
prairiesandhalfarebred
varieties)
Mostarepu
blic
releasesfew
are
protected
patented
orlicensed
NL8
seeded
hybridsvanDinterSemo
MTA
UK4
seeded
hybridsCERES(Land
OLakes)andTerravestaLtdMTA
US
Non
eFR
Non
e
EU53
registered
with
CPV
OforPB
RUK20
registered
with
CPV
OforPB
RUS
18clon
esfrom
Cornellin
multisite
trials
CAun
quantified
UAlberta
andQuebec
FR8ndash
12SR
Fclon
eslicence‐based
IT5SR
Cand6SR
FAFV
MTA
orlicence
SE14
STTlicence‐based
andMTA
US
~19So
utheast5ndash7Upp
erMidwest
from
UMD
NRRIMTA6ndash12
North
Central
from
GWRMTAPacific
Northwest8clon
esGWRMTA24
inmultilocationyieldtrialsGWR
Com
mercial
yield
(tDM
haminus1year
minus1 )
US
3ndash18
EU8ndash
12CN20
ndash30
US
10ndash25
EU7ndash20
UK8ndash
14US
8ndash14
EU5ndash20
(SEandBaltic
Cou
ntries8ndash
12)
US
10ndash2
2(10ndash
12North
Central12
ndash16
Southeast15ndash2
2PacificNortheast)
Harvestrotatio
nand
commercial
stand
lifespan
Ann
ualfor10
ndash12years
Ann
ualfor10ndash2
5years(M
sac
hasbeen
used
for~3
0yearsin
China)
2‐to
4‐year
cyclefor22
ndash30years
SRC3‐
to7‐year
cyclefor20
years
SRF
10‐to12‐yearcycleforgt50
years
Adaptiverang
eOpen‐po
llinatedandsynthetic
cultivars
arelim
itedin
adaptatio
nby
temperature
andprecipitatio
n(~8breeding
zonesin
USandCA)
Standard
Mtimesgiswidelyadaptedin
EU
butislim
itedin
theUnitedStates
byinsufficient
winterh
ardiness
forN
orthern
Midwestand
heatintolerancein
the
southNovelMsactimesMsinhyb
rids
andMsintimesMsinhybrids
selected
incontinentalD
Ehave
show
nawide
adaptiv
erang
ein
EU(K
alininaetal
2017
)Ongoing
trialson
heavymetal
contam
inated
soils
indicatetoleranceby
exclusion(K
rzyżak
etal20
17)
Different
hybridsareneeded
fordifferent
zones
Besthy
bridsshow
someG
timesE(U
S)
andsomehy
bridslow
GtimesE(Fabio
etal20
17)
Different
hybridsareneeded
ford
ifferent
clim
aticzonesHyb
rids
thatareadapted
forg
rowingseason
sof
~6mon
thsand
relativ
elyshortd
aysin
Southern
Europ
earemaladaptedto
shortg
rowing
season
sof
~4mon
thsandrelativ
ely
long
days
inNorthernEU
The
mostb
roadly
adaptedvarietiescome
from
thePcana
densistaxo
n
Note
AFV
AlasiaFranco
VivaiC
ornell
CornellUniversity
CPV
OC
ommunity
PlantV
ariety
OfficeFC
BAF
orestCelluloseW
ood
ConstructionandFu
rnitu
reTechnologyInstitu
teG
timesEg
enotype‐by
‐environmentinterac-
tion
GWRG
reenWoodResourcesINRAF
renchNationalInstituteforA
griculturalR
esearch
IRST
EAN
ationalR
esearchUnito
fScience
andTechnologyforE
nvironmentand
AgricultureM
sacMsacchariflo
rusandM
timesg
(Mtimes
giganteus)M
sin
MiscanthussinensisM
BIMendelB
iotechnology
IncM
bpm
egabase
pairM
SStateM
ississippi
StateUniversity
MTAm
aterialtransferagreem
entNICS
NationalInstituteof
CropScience
NW‐
FVANorthwestGerman
ForestResearchInstitu
tePB
Rplantbreeders
rightRDARural
DevelopmentAdm
inistration
REAP
ResourceEfficient
Agriculture
Productio
nSL
USw
edishUniversity
ofAgriculturalSciences
SNUS
eoul
NationalU
niSR
Cshort‐ro
tatio
ncoppice
SRF
short‐rotationforestryS
TTS
weT
reeTech
tDM
haminus1year
minus1 tons
ofdrymatterperhectareperyearU
AlbertaUniversity
ofAlbertaU
MD
NRRIUniversity
ofMinnesotaDuluthsNaturalResources
ResearchInstitu
teU
MNU
niversity
ofMinnesotaU
WU
niversity
ofWashington
UWMU
niversity
ofWarmiaandMazury
a ISO
Alpha‐2
lettercountrycodes
b EU
isused
forEurope
CLIFTON‐BROWN ET AL | 5
2011) Genotype‐by‐environment interactions (G times E) arestrong and must be considered in breeding (Casler 2012Casler Mitchell amp Vogel 2012) Adaptation to environ-ment is regulated principally by responses to day‐lengthand temperature There are also strong genotype times environ-ment interactions between the drier western regions and thewetter eastern regions (Casler et al 2017)
The growing regions of North America are divided intofour adaptation zones for switchgrass each roughly corre-sponding to two official hardiness zones The lowland eco-types are generally late flowering high yielding andadapted to warmer climates but have lower drought andcold resistance than upland ecotypes (Casler 2012 Casleret al 2012)
In 2015 the US Department of Agriculture (USDA)National Plant Germplasm System GRIN (httpswwwars-gringovnpgs) had 181 switchgrass accessions of whichonly 96 were available for distribution due to limitationsassociated with seed multiplication (Casler Vogel amp Har-rison 2015) There are well over 2000 additional uncata-logued accessions (Table 1) held by various universities butthe USDA access to these is also constrained by the effortneeded in seed multiplication Switchgrass is a model herba-ceous species for conducting scientific research on biomass(Sanderson Adler Boateng Casler amp Sarath 2006) but lit-tle funding is available for the critical prebreeding work thatis necessary to link this biological research to commercialbreeding More than a million genotypes from ~2000 acces-sions (seed accessions contain many genotypes) have beenphenotypically screened in spaced plant nurseries and tenthousand of the most useful have been genotyped with differ-ent technologies depending on the technology available at
the time when these were performed From these character-ized genotypes parents are selected for exploratory pairwisecrosses to produce synthetic populations within ecotypesSwitchgrass like many grasses is outcrossing due to astrong genetically controlled self‐incompatibly (akin to theS‐Z‐locus system of other grasses (Martinez‐Reyna ampVogel 2002)) Thus the normal breeding approaches usedare F1 wide crosses and recurrent selection cycles within syn-thetic populations
The scale of these programmes varies from small‐scaleconventional breeding based solely on phenotypic selec-tion (eg REAP Canada Montreal Quebec) to large pro-grammes incorporating modern molecular breedingmethods (eg USDA‐ARS Madison Wisconsin) Earlyagronomic research and biomass production efforts werefocused on the seed‐based multiplication of promising wildaccessions from natural prairies Cultivars Alamo Kanlowand Cave‐in‐Rock were popular due to high yield andmoderate‐to‐wide adaptation Conventional breedingapproaches focussed on biomass production traits and haveled to the development of five cultivars particularly suitedto biomass production Cimarron EG1101 EG1102EG2101 and Liberty The first four of these represent thelowland ecotype and were developed either in Oklahomaor Georgia Liberty is a derivative of lowland times uplandhybrids developed in Nebraska following selection for lateflowering the high yield of the lowland ecotype and coldtolerance of the upland ecotype (Vogel et al 2014) Thesefive cultivars were all approximately 25ndash30 years in themaking counting from the initiation of these breeding pro-grammes Many more biomass‐type cultivars are expectedwithin the next few years as these and other breeding
TABLE 2 Generalized improvement targets for perennial biomass crops (PBCs)
Net energy yield per hectare
Increased yield
Reduced moisture content at harvest
Physical and chemical composition for different end‐use applications
Increased lignin content and decreased corrosive elements for thermal conversion
Reduced recalcitrance through decreased lignin content andor modified lignin monomer composition to reduce pretreatment requirementsfor next‐generation biofuels by saccharification and fermentation
Plant morphological differences which influence biomass harvest transport and storage (eg stem thickness)
Propagation costs
Improved cloning systems (trees and grasses)
Seed systems (grasses)
Optimizing agronomy for each new cultivar
Resilience through enhanced
Abiotic stress toleranceresistance (eg drought salinity and high and low temperature )
Biotic stress resistance (eg insects fungal bacterial and viral diseases)
Site adaptability especially to those of marginalcontaminated agricultural land
6 | CLIFTON‐BROWN ET AL
TABLE3
Preparedness
formassupscaling
currentmarketvalueandresearch
investmentin
four
perennialbiom
asscrops(PBCs)
Species
Switc
hgrass
Miscanthu
sW
illow
Pop
lar
Current
commercial
plantin
gcostsperha
USa20
0USD
bin
the
establishm
entyear
DE3
375
Eurob
(XueK
alininaamp
Lew
ando
wski20
15)
UK2
153
GBPb
(Evans201
6)redu
cedto
116
9GBPwith
Mxg
rhizom
ewhere
thefarm
erdo
estheland
preparation(w
wwterravestacom
)US
173
0ndash222
5USD
UK150
0ndash173
9GBPplus
land
preparation(Evans20
16)
US
197
6USD
(=80
0USD
acre)
IT110
0Euro
FRSE100
0ndash200
0Euro
US
863USD
ha(LazarusHeadleeamp
Zalesny
20
15)
Current
marketvalue
ofthebiom
asspert
DM
US
80ndash1
00USD
UK~8
0GBP(bales)(Terravesta
person
alcommun
ication)
US
94USD
(chipp
ed)
UK49
41GBP(chipp
ed)(Evans20
16)
US
55ndash7
0USD
IT10
0Euro
US
80ndash90USD
deliv
ered
basedon
40milesof
haulage
Scienceforg
enetic
improv
ementprojects
inthelast10
years
USgt50
projectsfund
edby
USDOEand
USD
ANIFA
CNgt
30projectsfund
edby
CN‐N
SFCandMBI
EUc 6projects(O
PTIM
ISCO
PTIM
A
WATBIO
SUNLIBBF
IBRAandGRACE)
FRB
FFSK
2projectsfund
edby
IPETandPM
BC
US
EBICABBImultip
leDOEFeedstock
Genom
icsandUSD
AAFR
Iprojects
UKB
SBECB
EGIN
(200
3ndash20
10)
US
USDOEJG
IGenom
eSequ
encing
(200
9ndash20
122
015ndash
2018
)USD
ANortheastSu
nGrant
(200
9ndash20
12)
USD
ANIFANEWBio
Con
sortium
(201
2ndash20
17)USD
ANIFAWillow
SkyC
AP(201
8ndash20
21)
EUF
P7(EnergyPo
plarN
ovelTreeWATBIO
Tree4Fu
ture)
SEC
limate‐adaptedpo
plarsandNanow
ood
SLU
andST
TUS
Bioenergy
Science(O
RNL)USD
Afeedstocks
geno
micsprog
rammeBRDIa
ndBRC‐CBI
Major
projects
supp
ortin
gcrossing
andselectioncycles
inthelast10
years
USandCA12
projects
NLRUEmiscanthu
sPP
P20
15ndash2
019
50ndash
100K
Euroyear
UKGIA
NT‐LIN
KPP
I20
11ndash2
016
13M
GBPyear
US
EBICABBI~0
5M
USD
year
UKBEGIN
2000
ndash201
0US
USD
ANortheastSu
nGrant
(200
9ndash20
12)USD
ANIFA
NEWBio
Con
sortium
(201
2ndash20
17)USD
ANIFA
Willow
SkyC
AP(201
8ndash20
21)
FR1
4region
alandnatio
nalp
rojects10
0ndash15
0k
Euroyear
SES
TTo
nebreeding
effortas
aninternalproject
in20
10with
aresulting
prog
enytrial
US
DOESu
nGrant
feedstockdevelopm
ent
partnershipprog
ramme(including
Willow
)
Current
annu
alinvestmentin
projects
fortranslationinto
commercial
hybrids
USgt20
MUSD
year
UKMUST
201
6ndash20
1905M
GBPyear
US
CABBIUSD
AAFR
I20
17ndash2
02205M
USD
year
US
USD
ANIFA
NEWBio
~04
MUSD
year
FRScienceforim
prov
ement~2
0kEuroyear
US
USD
Aun
derAFR
ICo‐ordinatedAg
Prod
ucer
projectsAnd
severaltranslational
geno
micsprojectswith
outpu
blic
fund
ing
Upscalin
gtim
e(years)
toproducesufficient
propagules
toplantgt10
0ha
US
Using
seed
2ndash3years
UKRhizomefor1ndash200
0hectares
canbe
ready
in6mon
ths
UKNL2
0ha
ofseed
hybridsplantedin
2018
In
2019
sufficientseedfor5
0ndash10
0ha
isexpected
UK3yearsusingconv
entio
nalcutting
sfaster
usingmicroprop
agation
FRIT3yearsby
vegetativ
eprop
agation
SEgt3yearsby
vegetativ
eprop
agation(cuttin
gs)
and3yearsby
microprop
agation
Note
AFR
IAgriculture
andFo
odResearchInitiativein
theUnitedStatesBEGIN
BiomassforEnergyGenetic
Improvem
entNetwork
BFF
BiomassFo
rtheFu
tureBRC‐CBI
Bioenergy
ResearchCentre‐Centrefor
Bioenergy
Innovatio
nBRDIBiomassresearch
developm
entinitiative
BSB
ECBBSR
CSu
stainableBioenergy
Centre
CABBICenterforadvanced
bioenergyandbioproductsinnovatio
nin
theUnitedStatesCN‐N
SFC
Natural
ScienceFo
undatio
nof
ChinaDOEDepartm
entof
Energyin
theUnitedStatesEBIEnergyBiosciences
Institu
teFIBRAFibrecropsas
sustainablesource
ofbiobased
materialforindustrial
productsin
Europeand
ChinaFP
7SeventhFram
eworkProgrammein
theEUGIA
NT‐LIN
KGenetic
improvem
entof
miscanthusas
asustainablefeedstockforbioenergyin
theUnitedKingdom
GRACEGRow
ingAdvancedindustrial
Crops
onmarginallandsforbiorEfineriesIPETKorea
Institu
teof
Planning
andEvaluationforTechnologyin
FoodA
gricultureFo
restry
andFisheriesJG
IJointGenom
eInstitu
teMBIMendelBiotechnology
IncMUST
Miscant-
husUpS
calin
gTechnology
NEWBioNortheast
WoodyW
arm‐seasonBiomassConsortiumNIFANationalInstitu
teof
Food
andAgriculture
intheUnitedStatesNovelTree
Novel
tree
breeding
strategiesOPT
IMAOpti-
mizationof
perennialgrassesforbiom
assproductio
nin
theMediterraneanareaOPT
IMISCOptim
izingbioenergyproductio
nfrom
Miscanthus
ORNLOak
Ridge
NationalLabPM
BCPlantMolecular
BreedingCenterof
theNextGenerationBiogreenResearchCenters
oftheRepublic
ofKoreaPP
Ipublicndashp
rivate
investmentPP
Ppublicndashp
rivate
partnership
RUEradiationuseefficiencySk
yCAP
Coordinated
AgriculturalProjectSL
U
SwedishUniversity
ofAgriculturalSciencesST
TSw
eTreeTech
SUNLIBBSu
stainableLiquidBiofuelsfrom
BiomassBiorefiningTree4Fu
tureDesigning
Trees
forthefutureUSD
AUSDepartm
entof
Agriculture
WATBIO
Developmentof
improved
perennialnonfoodbiom
assandbioproduct
cropsforwater
stressed
environm
ents
a ISO
Alpha‐2
lettercountrycodes
b Local
currencies
areused
asat
2018Euro
GBP(G
reat
Britain
Pound)USD
(USDollar)c EU
isused
forEurope
CLIFTON‐BROWN ET AL | 7
TABLE4
Prebreedingresearch
andthestatus
ofconventio
nalbreeding
infour
leadingperennialbiom
asscrops(PBCs)
Breeding
techno
logy
step
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sWillow
Poplar
1Collected
wild
accessions
or
secondarysources
availablefor
breeding
Provideabroadbase
of
useful
traits
Respect
forCBD
and
Nagoyaprotocol
on
collections
after2014
Not
allindigenous
genetic
resourcesareaccessible
under
CBD
forpolitical
reasons
USa~2
000
(181
inofficial
GRIN
gene
bank
ofwhich
96
wereavailable
others
in
variousunofficial
collections)
CNChangsha
~1000
andNanjin
g
2000from
China
DK~1
20from
JP
FRworking
collectionof
~100
(mainlyM
sin)
JP~1
500
from
JPS
KandCN
NLworking
collectionof
~300
Msin
SK~7
00from
SKC
NJP
andRU
UK~1
500
accessions
(from
~500
sites)
ofwhich
1000arein
field
nurseriesin
2018
US
14in
GRIN
global
US
Illin
ois~1
500
collections
made
inAsiawith
25
currently
available
inUnitedStates
forbreeding
UK1500with
about20
incommon
with
the
UnitedStates
US
350largelyunique
FR3370Populus
delto
ides
(650)Pnigra(2000)
Ptrichocarpa(600)and
Pmaximow
iczii(120)
IT30000P
delto
idsPnigra
Pmaximow
icziiand
Ptrichocarpa
SE13000
accessionsmainly
Ptrichocarpa
byST
TandSL
U
US
GWR150Ptrichocarpaand
300Pmaximow
icziiaccessions
forPacificNorthwestprogram
535Pdelto
ides
and~2
00
Pnigraaccessions
for
Southeastprogrammeand77
Psimoniifrom
Mongolia
UMD
NRRI550+
clonal
accessions
(primarily
Ptimescanadnesisa
long
with
Pdelto
ides
andPnigra
parents)
forUpper
Midwest
program
2Wild
accessions
which
have
undergone
phenotypic
screeningin
field
trials
Selectionof
parental
lines
with
useful
traits
Strong
partnerships
torun
multilocationfieldtrials
tophenotypeconsistently
theaccessionsgenotypes
indifferentenvironm
ents
anddatabases
Costof
runningmultilocation
US
gt500000genotypes
CN~1
250
inChangsha
~1700
in
HunanJiangsuS
handong
Hainan
NL~250genotypes
JP~1
200
SK~4
00
UK‐le
d~1
000
genotypesin
a
replicated
trial
US‐led
~1200
genotypesin
multi‐
locatio
nreplicated
trialsin
Asiaand
North
America
MsinSK
CNJP
CA
(Ontario)US(Illinoisand
Colorado)MsacSK
(Kangw
on)
CN
(Zhejiang)JP
(Sapporo)CA
(Onatario)U
S(Illinois)DK
(Aarhus)
UKgt400
US
180
FR2720
IT10000
SE~1
50
US
GWR~1
500
wild
Ptrichocarpaphenotypes
and
OPseedlots(total
of70)from
thePmaximow
icziispecies
complex
(suaveolens
cathayana
koreana
ussuriensismaximow
iczii)and
2000ndash3000Pdelto
ides
collections
(based
on104s‐
generatio
nfamilies)
UMD
NRRIscreened
seedlin
gsfrom
aPdelto
ides
(OPor
CP)
breeding
programme(1996ndash2016)
gt7200OPseedlin
gsof
PnigraunderMinnesota
clim
atic
conditions(improve
parental
populatio
n)
3Wild
germ
plasm
genotyping
Amanaged
living
collectionof
clonal
types
US
~20000genotypes
CNChangsha
~1000Nanjin
g37
FR44
bycpDNA
(Fenget
al
UK~4
00
US
225
(Con
tinues)
8 | CLIFTON‐BROWN ET AL
TABLE
4(Contin
ued)
Breeding
techno
logy
step
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sWillow
Poplar
Construct
phylogenetic
treesanddissim
ilarity
indices
The
type
ofmolecular
analysis
mdashAFL
PcpDNAR
ADseq
whole
genomesequencing
2014)
JP~1
200
NL250genotypesby
RADseq
UK~1
000
byRADseq
SK~3
00
US
Illin
oisRADseqon
allwild
accessions
FR2310
SE150
US
1500
4Exploratory
crossing
and
progenytests
Discovergood
parental
combinatio
nsmdashgeneral
combining
ability
Geographicseparatio
n
speciesor
phylogenetic
trees
Costly
long‐te
rmmultilocation
trialsforprogeny
US
gt10000
CN~1
0
FR~2
00
JP~3
0
NL3600
SK~2
0
UK~4
000
(~1500Msin)
US
500ndash1000
UK gt700since2003
gt800EWBP(1996ndash
2002)
US
800
FRCloned13
times13
factorial
matingdesign
with
3species[10
Pdelto
ides8Ptrichocarpa
8
Pnigra)
US
GWR1000exploratory
crossingsof
Pfrem
ontii
Psimoniiforim
proved
droughttolerance
UMD
NRRIcrossednativ
e
Pdelto
ides
with
non‐nativ
e
Ptrichocarpaand
Pmaximow
icziiMajority
of
progenypopulatio
nssuffered
from
clim
atic
andor
pathogenic
susceptib
ility
issues
5Wideintraspecies
hybridization
Discovergood
parental
combinatio
nsmdashgeneral
combining
ability
1to
4above
inform
atics
Flow
eringsynchronizationand
low
seed
set
US
~200
CN~1
0
NL1800
SK~1
0
UK~5
00
UK276
US
400
IT500
SE120
US
50ndash100
6With
inspecies
recurrentselection
Concentratio
nof
positiv
e
traits
Identificationof
theright
heterotic
groups
insteps
1ndash5
Difficultto
introducenew
germ
plasm
with
outdilutio
n
ofbesttraits
US
gt40
populatio
ns
undergoing
recurrentselection
NL2populatio
ns
UK6populatio
ns
US
~12populatio
nsto
beestablished
by2019
UK4
US
150
FRPdelto
ides
(gt30
FSfamilies
ndash900clones)Pnigra(gt40
FS
families
ndash1600clones)and
Ptrichocarpa(15FS
Families
minus350clones)
IT80
SEca50
parent
breeding
populatio
nswith
in
Ptrichocarpa
US
Goalsis~1
00parent
breeding
populatio
nswith
inPtrichocarpa
Pdelto
ides
PnigraandPmaximow
iczii
7Interspecies
hybrid
breeding
Com
bining
complem
entary
traitsto
producegood
morphotypes
and
heterosiseffects
Allthesteps1ndash6
Inearlystage
improvem
ents
areunpredictable
Awide
base
iscostly
tomanage
None
CNChangsha
3Nanjin
g
120MsintimesMflo
r
JP~5
with
Saccharum
spontaneum
SK5
UK420
US
250
FRPcanadensis(1800)
Pdelto
ides
timesPtrichocarpa
(800)and
PtrichocarpatimesPmaximow
iczii
(1200)
(Con
tinues)
CLIFTON‐BROWN ET AL | 9
TABLE
4(Contin
ued)
Breeding
techno
logy
step
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sWillow
Poplar
IT~1
20
US
525genotypesin
eastside
hybrid
programme
205westside
hybrid
programmeand~1
200
in
SoutheastPdelto
ides
program
8Chrom
osom
e
doublin
g
Arouteto
triploid
seeded
typesdoublin
gadiploid
parent
ofknow
n
breeding
valuefrom
recurrentselection
Needs
1ndash7to
help
identify
therightparental
lines
Doubled
plantsarenotorious
forrevertingto
diploid
US
~50plantstakenfrom
tetraploid
tooctaploid
which
arenow
infieldtrials
CN~1
0Msactriploids
DKSeveralattemptsin
mid
90s
FR5genotypesof
Msin
UK2Msin1Mflora
nd1Msac
US
~30
UKAttempted
butnot
routinelyused
US
Not
partof
theregular
breeding
becausethere
existnaturalpolyploids
NA
9DoubleHaploids
Arouteto
producing
homogeneous
progeny
andalso
forthe
introductio
nof
transgenes
orforgenome
editing
Needs
1ndash7to
help
identify
therightparental
lines
Fully
homozygousplantsare
weakandeasily
die
andmay
notflow
ersynchronously
US
Noneyetattempteddueto
poor
vigour
andviability
of
haploids
CN2genotypesof
Msin
andMflor
UKDihaploidsof
MsinMflor
and
Msacand2transgenic
Msinvia
anther
cultu
re
US
6createdandused
toidentify
miss‐identifying
paralogueloci
(causedby
recent
genome
duplicationin
miscanthus)Usedin
creatin
gthereferencegenomefor
MsinVeryweakplantsand
difficultto
retain
thelin
es
NA
NA
10Embryo
rescue
Anattractiv
etechniquefor
recovering
plantsfrom
sexual
crosses
The
majority
ofem
bryos
cannot
survivein
vivo
or
becomelong‐timedorm
ant
None
UKone3times
hybrid
UKEmbryosrescue
protocol
proved
robustat
8days
post‐pollin
ation
FRAnem
bryo
rescue
protocol
proved
robustandim
proved
hybridizationsuccessfor
Pdelto
ides
timesPtrichocarpa
crosses
11Po
llenstorage
Flexibility
tocross
interestingparents
with
outneed
of
flow
ering
synchronization
None
Nosuccessto
daten
otoriously
difficultwith
grassesincluding
sugarcane
Freshpollencommonly
used
incrossing
Pollenstoragein
some
interspecificcrosses
FRBothstored
(cryobank)
and
freshindividual
pollen
Note
AFL
Pam
plifiedfragmentlength
polymorphismCBDconventio
non
biological
diversity
CP
controlledpollinatio
nCpD
NAchloroplastDNAEWBP
Europeanwillow
breeding
programme
FSfull‐sib
GtimesE
genotype‐by‐environm
entGRIN
germ
plasm
resourcesinform
ationnetwork
GWRGreenWoodResourcesMflorMiscanthusflo
ridulusMsacMsacchariflo
rus
MsinMsinensisNAnotapplicableNRRINatural
Resources
ResearchInstitu
teOP
open
pollinatio
nRADseq
restrictionsite‐associatedDNA
sequencingSL
USw
edishUniversity
ofAgriculturalSciencesST
TSw
eTreeTech
UMDUniversity
ofMinnesota
Duluth
a ISO
Alpha‐2
lettercountrycodes
10 | CLIFTON‐BROWN ET AL
programmes mature The average rate of gain for biomassyield in long‐term switchgrass breeding programmes hasbeen 1ndash4 per year depending on ecotype populationand location of the breeding programme (Casler amp Vogel2014 Casler et al 2018) The hybrid derivative Libertyhas a biomass yield 43 higher than the better of its twoparents (Casler amp Vogel 2014 Vogel et al 2014) Thedevelopment of cold‐tolerant and late‐flowering lowland‐ecotype populations for the northern United States hasincreased biomass yields by 27 (Casler et al 2018)
Currently more than 20 recurrent selection populationsare being managed in the United States to select parentsfor improved yield yield resilience and compositional qual-ity of the biomass For the agronomic development andupscaling high seed multiplication rates need to be com-bined with lower seed dormancy to reduce both crop estab-lishment costs and risks Expresso is the first cultivar withsignificantly reduced seed dormancy which is the first steptowards development of domesticated populations (Casleret al 2015) Most phenotypic traits of interest to breeders
require a minimum of 2 years to be fully expressed whichresults in a breeding cycle that is at least two years Morecomplicated breeding programmes or traits that requiremore time to evaluate can extend the breeding cycle to 4ndash8 years per generation for example progeny testing forbiomass yield or field‐based selection for cold toleranceBreeding for a range of traits with such long cycles callsfor the development of molecular methods to reduce time-scales and improve breeding efficiency
Two association panels of switchgrass have been pheno-typically and genotypically characterized to identify quanti-tative trait loci (QTLs) that control important biomasstraits The northern panel consists of 60 populationsapproximately 65 from the upland ecotype The southernpanel consists of 48 populations approximately 65 fromthe lowland ecotype Numerous QTLs have been identifiedwithin the northern panel to date (Grabowski et al 2017)Both panels are the subject of additional studies focused onbiomass quality flowering and phenology and cold toler-ance Additionally numerous linkage maps have been
FIGURE 1 Cumulative minimum years needed for the conventional breeding cycle through the steps from wild germplasm to thecommercial hybrids in switchgrass miscanthus willow and poplar Information links between the steps are indicated by dotted arrows andhighlight the importance of long‐term informatics to maximize breeding gain
CLIFTON‐BROWN ET AL | 11
created by the pairwise crossing of individuals with diver-gent characteristics often to generate four‐way crosses thatare analysed as pseudo‐F2 crosses (Liu Wu Wang ampSamuels 2012 Okada et al 2010 Serba et al 2013Tornqvist et al 2018) Individual markers and QTLs iden-tified can be used to design marker‐assisted selection(MAS) programmes to accelerate breeding and increase itsefficiency Genomic prediction and selection (GS) holdseven more promise with the potential to double or triplethe rate of gain for biomass yield and other highly complexquantitative traits of switchgrass (Casler amp Ramstein 2018Ramstein et al 2016) The genome of switchgrass hasrecently been made public through the Joint Genome Insti-tute (httpsphytozomejgidoegov)
Transgenic approaches have been heavily relied upon togenerate unique genetic variants principally for traitsrelated to biomass quality (Merrick amp Fei 2015) Switch-grass is highly transformable using either Agrobacterium‐mediated transformation or biolistics bombardment butregeneration of plants is the bottleneck to these systemsTraditionally plants from the cultivar Alamo were the onlyregenerable genotypes but recent efforts have begun toidentify more genotypes from different populations that arecapable of both transformation and subsequent regeneration(King Bray Lafayette amp Parrott 2014 Li amp Qu 2011Ogawa et al 2014 Ogawa Honda Kondo amp Hara‐Nishi-mura 2016) Cell wall recalcitrance and improved sugarrelease are the most common targets for modification (Bis-wal et al 2018 Fu et al 2011) Transgenic approacheshave the potential to provide traits that cannot be bredusing natural genetic variability However they will stillrequire about 10ndash15 years and will cost $70ndash100 millionfor cultivar development and deployment (Harfouche Mei-lan amp Altman 2011) In addition there is commercialuncertainty due to the significant costs and unpredictabletimescales and outcomes of the regulatory approval processin the countries targeted for seed sales As seen in maizeone advantage of transgenic approaches is that they caneasily be incorporated into F1 hybrid cultivars (Casler2012 Casler et al 2012) but this does not decrease thetime required for cultivar development due to field evalua-tion and seed multiplication requirements
The potential impacts of unintentional gene flow andestablishment of non‐native transgene sequences in nativeprairie species via cross‐pollination are also major issues forthe environmental risk assessment These limit further thecommercialization of varieties made using these technolo-gies Although there is active research into switchgrass steril-ity mechanisms to curb unintended pollen‐mediated genetransfer it is likely that the first transgenic cultivars proposedfor release in the United States will be met with considerableopposition due to the potential for pollen flow to remainingwild prairie sites which account for lt1 of the original
prairie land area and are highly protected by various govern-mental and nongovernmental organizations (Casler et al2015) Evidence for landscape‐level pollen‐mediated geneflow from genetically modified Agrostis seed multiplicationfields (over a mountain range) to pollinate wild relatives(Watrud et al 2004) confirms the challenge of using trans-genic approaches Looking ahead genome editing technolo-gies hold considerable promise for creating targeted changesin phenotype (Burris Dlugosz Collins Stewart amp Lena-ghan 2016 Liu et al 2018) and at least in some jurisdic-tions it is likely that cultivars resulting from gene editingwill not need the same regulatory approval as GMOs (Jones2015a) However in July 2018 the European Court of Justice(ECJ) ruled that cultivars carrying mutations resulting fromgene editing should be regulated in the same way as GMOsThe ECJ ruled that such cultivars be distinguished from thosearising from untargeted mutation breeding which isexempted from regulation under Directive 200118EC
3 | MISCANTHUS
Miscanthus is indigenous to eastern Asia and Oceania whereit is traditionally used for forage thatching and papermaking(Xi 2000 Xi amp Jezowkski 2004) In the 1960s the highbiomass potential of a Japanese genotype introduced to Eur-ope by Danish nurseryman Aksel Olsen in 1935 was firstrecognized in Denmark (Linde‐Laursen 1993) Later thisaccession was characterized described and named asldquoM times giganteusrdquo (Greef amp Deuter 1993 Hodkinson ampRenvoize 2001) commonly abbreviated as Mxg It is a natu-rally occurring interspecies triploid hybrid between tetraploidM sacchariflorus (2n = 4x) and diploid M sinensis(2n = 2x) Despite its favourable agronomic characteristicsand ability to produce high yields in a wide range of environ-ments in Europe (Kalinina et al 2017) the risks of relianceon it as a single clone have been recognized Miscanthuslike switchgrass is outcrossing due to self‐incompatibility(Jiang et al 2017) Thus seeded hybrids are an option forcommercial breeding Miscanthus can also be vegetativelypropagated by rhizome or in vitro culture which allows thedevelopment of clones The breeding approaches are usuallybased on F1 crosses and recurrent selection cycles within thesynthetic populations There are several breeding pro-grammes that target improvement of miscanthus traitsincluding stress resilience targeted regional adaptation agro-nomic ldquoscalabilityrdquo through cheaper propagation fasterestablishment lower moisture and ash contents and greaterusable yield (Clifton‐Brown et al 2017)
Germplasm collections specifically to support breedingfor biomass started in the late 1980s and early 1990s inDenmark Germany and the United Kingdom (Clifton‐Brown Schwarz amp Hastings 2015) These collectionshave continued with successive expeditions from European
12 | CLIFTON‐BROWN ET AL
and US teams assembling diverse collections from a widegeographic range in eastern Asia including from ChinaJapan South Korea Russia and Taiwan (Hodkinson KlaasJones Prickett amp Barth 2015 Stewart et al 2009) Threekey miscanthus species for biomass production are M si-nensis M floridulus and M sacchariflorus M sinensis iswidely distributed throughout eastern Asia with an adap-tive range from the subtropics to southern Russia (Zhaoet al 2013) This species has small rhizomes and producesmany tightly packed shoots forming a ldquotuftrdquo M floridulushas a more southerly adaptive range with a rather similarmorphology to M sinensis but grows taller with thickerstems and is evergreen and less cold‐tolerant than the othermiscanthus species M sacchariflorus is the most northern‐adapted species ranging to 50 degN in eastern Russia (Clarket al 2016) Populations of diploid and tetraploid M sac-chariflorus are found in China (Xi 2000) and South Korea(Yook 2016) and eastern Russia but only tetraploids havebeen found in Japan (Clark Jin amp Petersen 2018)
Germplasm has been assembled from multiple collec-tions over the last century though some early collectionsare poorly documented This historical germplasm has beenused to initiate breeding programmes largely based on phe-notypic and genotypic characterization As many of theaccessions from these collections are ldquoorigin unknownrdquocrucial environmental envelope data are not available UK‐led expeditions started in 2006 and continued until 2011with European and Asian partners and have built up a com-prehensive collection of 1500 accessions from 500 sitesacross Eastern Asia including China Japan South Koreaand Taiwan These collections were guided using spatialclimatic data to identify variation in abiotic stress toleranceAccessions from these recent collections were planted fol-lowing quarantine in multilocation nursery trials at severallocations in Europe to examine trait expression in differentenvironments Based on the resulting phenotypic andmolecular marker data several studies (a) characterized pat-terns of population genetic structure (Slavov et al 20132014) (b) evaluated the statistical power of genomewideassociation studies (GWASs) and identified preliminarymarkerndashtrait associations (Slavov et al 2013 2014) and(c) assessed the potential of genomic prediction (Daveyet al 2017 Slavov et al 2014 2018b) Genomic indexselection in particular offers the possibility of exploringscenarios for different locations or industrial markets (Sla-vov et al 2018a 2018b)
Separately US‐led expeditions also collected about1500 accessions between 2010 and 2014 (Clark et al2014 2016 2018 2015) A comprehensive genetic analy-sis of the population structure has been produced by RAD-seq for M sinensis (Clark et al 2015 Van der WeijdeKamei et al 2017) and M sacchariflorus (Clark et al2018) Multilocation replicated field trials have also been
conducted on these materials in North America and inAsia GWAS has been conducted for both M sinensis anda subset of M sacchariflorus accessions (Clark et al2016) To date about 75 of these recent US‐led collec-tions are in nursery trials outside the United States Due tolengthy US quarantine procedures these are not yet avail-able for breeding in the United States However molecularanalyses have allowed us to identify and prioritize sets ofgenotypes that best encompass the genetic variation in eachspecies
While mostM sinensis accessions flower in northern Eur-ope very few M sacchariflorus accessions flower even inheated glasshouses For this reason the European pro-grammes in the United Kingdom the Netherlands and Francehave performed mainly M sinensis (intraspecies) hybridiza-tions (Table 4) Selected progeny become the parents of latergenerations (recurrent selection as in switchgrass) Seed setsof up to 400 seed per panicle occur inM sinensis In Aberyst-wyth and Illinois significant efforts to induce synchronousflowering in M sacchariflorus and M sinensis have beenmade because interspecies hybrids have proven higher yieldperformance and wide adaptability (Kalinina et al 2017) Ininterspecies pairwise crosses in glasshouses breathable bagsandor large crossing tubes or chambers in which two or morewhole plants fit are used for pollination control Seed sets arelower in bags than in the open air because bags restrict pollenmovement while increasing temperatures and reducinghumidity (Clifton‐Brown Senior amp Purdy 2018) About30 of attempted crosses produced 10 to 60 seeds per baggedpanicle The seed (thousand seed mass ranges from 05 to09 g) is threshed from the inflorescences and sown into mod-ular trays to produce plug plants which are then planted infield nurseries to identify key parental combinations
A breeding programme of this scale must serve theneeds of different environments accepting the commonpurpose is to optimize the interception of solar radiationAn ideal hybrid for a given environment combines adapta-tion to date of emergence with optimization of traits suchas height number of stems per plant flowering and senes-cence time to optimize solar interception to produce a highbiomass yield with low moisture content at harvest (Rob-son Farrar et al 2013 Robson Jensen et al 2013) By20132014 conventional breeding in Europe had producedintra‐ and interspecific fertile seeded hybrids When acohort (typically about 5) of outstanding crosses have beenidentified it is important to work on related upscaling mat-ters in parallel These are as follows
Assessment of the yield and critical traits in selectedhybrids using a network of field trials
Efficient cloning of the seed parents While in vitro andmacro‐cloning techniques are used some genotypes areamenable to neither technique
CLIFTON‐BROWN ET AL | 13
TABLE5
Status
ofmod
ernplantbreeding
techniqu
esin
four
leadingperennialbiom
asscrop
s(PBCs)
Breeding
techno
logy
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sW
illow
Pop
lar
MAS
Use
ofmarker
sequ
encesthat
correlatewith
atrait
allowingearly
prog
enyselectionor
rejectionlocatin
gkn
owngenesfor
useful
traits(suchas
height)from
other
speciesin
your
crop
Breeding
prog
ramme
relevant
biparental
crosses(Table
4)to
create
ldquomapping
popu
latio
nsrdquoand
QTLidentification
andestim
ationto
identifyrobu
stmarkers
alternatively
acomprehensive
panelof
unrelated
geno
typesfor
GWAS
Ineffectivewhere
traits
areaffected
bymany
geneswith
small
effects
US
Literally
dozens
ofstud
iesto
identify
SNPs
andmarkers
ofinterestNoattempts
atMASas
yetlargely
dueto
popu
latio
nspecificity
CNS
SRISS
Rmarkers
forM
sinMsacand
Mflor
FR2
Msin
mapping
popu
latio
ns(BFF
project)
NL2
Msin
mapping
popu
latio
ns(A
tienza
Ram
irezetal
2003
AtienzaSatovic
PetersenD
olstraamp
Martin
200
3Van
Der
Weijdeetal20
17a
Van
derW
eijde
Hux
ley
etal20
17
Van
derW
eijdeKam
ei
etal20
17V
ander
WeijdeKieseletal
2017
)SK
2GWASpanels(1
collectionof
Msin1
diploidbiparentalF 1
popu
latio
n)in
the
pipelin
eUK1
2families
tofind
QTLin
common
indifferentfam
ilies
ofthe
sameanddifferent
speciesandhy
brids
US
2largeGWAS
panels4
diploidbi‐
parentalF 1
popu
latio
ns
1F 2
popu
latio
n(add
ition
alin
pipelin
e)
UK16
families
for
QTLdiscov
ery
2crossesMASscreened
1po
tentialvariety
selected
US
9families
geno
typedby
GBS(8
areF 1o
neisF 2)a
numberof
prom
ising
QTLdevelopm
entof
markers
inprog
ress
FR(INRA)tested
inlargeFS
families
and
infactorialmating
design
low
efficiency
dueto
family
specificity
US
49families
GS
Metho
dto
accelerate
breeding
throug
hredu
cing
the
resourcesforcross
attemptsby
Aldquotraining
popu
latio
nrdquoof
individu
alsfrom
stages
abov
ethat
have
been
both
Risks
ofpo
orpredictio
nProg
eny
testingneedsto
becontinueddu
ring
the
timethetraining
setis
US
One
prog
rammeso
farmdash
USD
Ain
WisconsinThree
cycles
ofgeno
mic
selectioncompleted
CN~1
000
geno
types
NLprelim
inary
mod
elsforMsin
developed
UK~1
000
geno
types
Verysuitable
application
not
attempted
yet
Maybe
challeng
ingfor
interspecies
hybrids
FR(INRA)120
0geno
type
ofPop
ulus
nigraarebeingused
todevelop
intraspecificGS
(Con
tinues)
14 | CLIFTON‐BROWN ET AL
TABLE
5(Contin
ued)
Breeding
techno
logy
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sW
illow
Pop
lar
predictin
gthe
performance
ofprog
enyof
crosses
(and
inresearch
topredictbestparents
tousein
biparental
crosses)
geno
typedand
phenotyp
edto
developamod
elthattakesgeno
typic
datafrom
aldquocandidate
popu
latio
nrdquoof
untested
individu
als
andprod
uces
GEBVs(Jannink
Lorenzamp
Iwata
2010
)
beingph
enotyp
edand
geno
typed
Inthis
time
next‐generation
germ
plasm
andreadyto
beginthe
second
roun
dof
training
and
recalib
ratio
nof
geno
mic
predictio
nmod
els
US
Prelim
inary
mod
elsforMsac
and
Msin
developed
Work
isun
derw
ayto
deal
moreeffectivelywith
polyploidgenetics
calib
ratio
nsUS
125
0geno
types
arebeingused
todevelopGS
calib
ratio
ns
Traditio
nal
transgenesis
Efficient
introd
uctio
nof
ldquoforeign
rdquotraits
(possiblyfrom
other
genu
sspecies)
into
anelite
plant(ega
prov
enparent
ora
hybrid)that
needsa
simpletraitto
beim
prov
edValidate
cand
idategenesfrom
QTLstud
ies
MASand
know
ledg
eof
the
biolog
yof
thetrait
andsource
ofgenesto
confer
the
relevant
changesto
phenotyp
eWorking
transformation
protocol
IPissuescostof
regu
latory
approv
al
GMO
labelling
marketin
gissuestricky
tousetransgenes
inou
t‐breedersbecause
ofcomplexity
oftransformingandgene
flow
risks
US
Manyprog
ramsin
UnitedStates
are
creatin
gtransgenic
plantsusingAlamoas
asource
oftransformable
geno
typesManytraits
ofinterestNothing
commercial
yet
CNC
hang
shaBtg
ene
transformed
into
Msac
in20
04M
sin
transformed
with
MINAC2gene
and
Msactransform
edwith
Cry2A
ageneb
othwith
amarkerg
eneby
Agrob
acterium
JP1
Msintransgenic
forlow
temperature
tolerancewith
increased
expression
offructans
(unp
ublished)
NLImprov
ingprotocol
forM
sin
transformation
SK2
Msin
with
Arabido
psisand
Brachypod
ium
PhytochromeBgenes
(Hwang
Cho
etal
2014
Hwang
Lim
etal20
14)
UK2
Msin
and
1Mflorg
enotyp
es
UKRou
tine
transformationno
tyet
possibleResearchto
overcomerecalcitrance
ongo
ing
Currently
trying
differentspecies
andcond
ition
sPo
plar
transformationused
atpresent
US
Frequently
attempted
with
very
limitedsuccess
US
600transgenic
lines
have
been
characterized
mostly
performed
inthe
aspenhy
brids
reprod
uctiv
esterility
drou
ghttolerance
with
Kno
ckdo
wns
(Con
tinues)
CLIFTON‐BROWN ET AL | 15
TABLE
5(Contin
ued)
Breeding
techno
logy
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sW
illow
Pop
lar
variou
slytransformed
with
four
cellwall
genesiptand
uidA
genesusingtwo
selectionsystem
sby
biolisticsand
Agrob
acterium
Transform
edplantsare
beinganalysed
US
Prelim
inarywork
will
betakenforw
ardin
2018
ndash202
2in
CABBI
Genom
eediting
CRISPR
Refinem
entofexisting
traits
inusefulparents
orprom
isinghybrids
bygeneratingtargeted
mutations
ingenes
know
ntocontrolthe
traitof
interestF
irst
adouble‐stranded
breakismadeinthe
DNAwhich
isrepairedby
natural
DNArepair
machineryL
eads
tofram
eshiftSN
Por
can
usealdquorepair
templaterdquo
orcanbe
used
toinserta
transgene
intoaldquosafe
harbourlocusrdquoIt
couldbe
used
todeleterepressors(or
transcriptionfactors)
Identification
mapping
and
sequ
encing
oftarget
genes(from
DNA
sequ
ence)
avoiding
screening
outof
unintend
ededits
Manyregu
latory
authorities
have
not
decidedwhether
CRISPR
andother
geno
meediting
techno
logies
are
GMOsor
notIf
GMOsseecomment
onstage10
abov
eIf
noteditedcrop
swill
beregu
latedas
conv
entio
nalvarieties
CATechn
olog
yisstill
toonewand
switchg
rass
geno
meis
very
complexOther
labo
ratories
are
interestedbu
tno
tyet
mov
ingon
this
US
One
prog
rammeso
farmdash
USD
Ain
Albany
FRInitiated
in20
16(M
ISEDIT
project)
NLInitiatingin
2018
US
Initiatingin
2018
in
CABBI
UKNot
started
Not
possible
until
transformation
achieved
US
CRISPR
‐Cas9and
Cpf1have
been
successfully
developedin
Pop
ulus
andarehigh
lyefficientExamples
forlig
nin
biosyn
thesis
Note
BFF
BiomassFo
rtheFu
tureBtBacillus
thuringiensisCABBICenterforadvanced
bioenergyandbioproductsinnovatio
nCpf1
CRISPR
from
Prevotella
andFrancisella
1CRISPR
clusteredregularlyinterspaced
palin
drom
icrepeatsCRISPR
‐CasC
RISPR
‐associated
Cry2A
acrystaltoxins2A
asubfam
ilyproduced
byBtFS
full‐sibG
BS
genotyping
bysequencingG
EBVsgenomic‐estim
ated
breeding
valuesG
MOg
enetically
modi-
fied
organism
GS
genomicselection
GWAS
genomew
ideassociationstudy
INRAF
renchNationalInstituteforAgriculturalR
esearch
IPintellectualp
roperty
IPTisopentenyltransferase
ISSR
inter‐SSR
MflorMiscant-
husflo
ridulusMsacMsacchariflo
rusMsinMsinensisM
AS
marker‐a
ssistedselection
MISEDITm
iscanthusgene
editing
forseed‐propagatedtriploidsNACn
oapicalmeristemA
TAF1
2and
cup‐shaped
cotyledon2
‐lik
eQTLq
uantitativ
etraitlocusS
NP
singlenucleotid
epolymorphismS
SRsim
plesequence
repeatsUSD
AU
SDepartm
ento
fAgriculture
a ISO
Alpha‐2
lettercountrycodes
16 | CLIFTON‐BROWN ET AL
High seed production from field crossing trials con-ducted in locations where flowering in both seed andpollen parents is likely to happen synchronously
Scalable and adapted harvesting threshing and seed pro-cessing methods for producing high seed quality
The results of these parallel activities need to becombined to identify the upscaling pathway for eachhybrid if this cannot be achieved the hybrid will likelynot be commercially viable The UK‐led programme withpartners in Italy and Germany shows that seedbased mul-tiplication rates of 12000 are achievable several inter-specific hybrids (Clifton‐Brown et al 2017) Themultiplication rate of M sinensis is higher probably15000ndash10000 Conventional cloning from rhizome islimited to around 120 that is one ha could provide rhi-zomes for around 20 ha of new plantation
Multilocation field testing of wild and novel miscanthushybrids selected by breeding programmes in the Nether-lands and the United Kingdom was performed as part ofthe project Optimizing Miscanthus Biomass Production(OPTIMISC 2012ndash2016) These trials showed that com-mercial yields and biomass qualities (Kiesel et al 2017Van der Weijde et al 2017a Van der Weijde Kieselet al 2017) could be produced in a wide range of climatesand soil conditions from the temperate maritime climate ofwestern Wales to the continental climate of eastern Russiaand the Ukraine (Kalinina et al 2017) Extensive environ-mental measurements of soil and climate combined withgrowth monitoring are being used to understand abioticstresses (Nunn et al 2017 Van der Weijde Huxley et al2017) and develop genotype‐specific scenarios similar tothose reported earlier in Hastings et al (2009) Phenomicsexperiments on drought tolerance have been conducted onwild and improved germplasm (Malinowska Donnison ampRobson 2017 Van der Weijde Huxley et al 2017)Recently produced interspecific hybrids displaying excep-tional yield under drought (~30 greater than control Mxg)in field trials in Poland and Moldova are being furtherstudied in detail in the phenomics and genomics facility atAberystwyth to better understand genendashtrait associationswhich can be fed back into breeding
Intraspecific seeded hybrids of M sinensis produced inthe Netherlands and interspecific M sacchariflorus times M si-nensis hybrids produced by the UK‐led breeding programmehave entered yield testing in 2018 with the recently EU‐funded project ldquoGRowing Advanced industrial Crops on mar-ginal lands for biorEfineries (GRACE)rdquo (httpswwwgrace-bbieu) Substantial variation in biomass quality for sacchari-fication efficiency (glucose release as of dry matter) ashcontent and melting point has already been generated inintraspecific M sinensis hybrids (Van der Weijde Kiesel
et al 2017) across environments (Weijde Dolstra et al2017a) GRACE aims to establish more than 20 hectares ofnew inter‐ and intraspecific seeded hybrids across six Euro-pean countries This project is building the know‐how andagronomy needed to transition from small research plots tocommercial‐scale field sites and linking biomass productiondirectly to industrial applications The biomass produced byhybrids in different locations will be supplied to innovativeindustrial end‐users making a wide range of biobased prod-ucts both for chemicals and for energy In the United Statesmulti‐location yield were initiated in 2018 to evaluate new tri-ploid M times giganteus genotypes developed at Illinois Cur-rently infertile hybrids are favoured in the United Statesbecause this eliminates the risk of invasiveness from naturallydispersed viable seed The precautionary principle is appliedas fertile miscanthus has naturalized in several states (QuinnAllen amp Stewart 2010) North European multilocation fieldtrials in the EMI and OPTIMISC projects have shown thereis minimal risk of invasiveness even in years when fertileflowering hybrids produce viable seed Naturalized standshave not established here due perhaps to low dormancy pooroverwintering and low seedling competitive strength In addi-tion to breeding for nonshattering or sterile seeded hybridsQuinn et al (2010) suggest management strategies which canfurther minimize environmental opportunities to manage therisk of invasiveness
31 | Molecular breeding and biotechnology
In miscanthus new plant breeding techniques (Table 5)have focussed on developing molecular markers forbreeding in Europe the United States South Korea andJapan There are several publications on QTL mappingpopulations for key traits such as flowering (AtienzaRamirez amp Martin 2003) and compositional traits(Atienza Satovic Petersen Dolstra amp Martin 2003) Inthe United States and United Kingdom independent andinterconnected bi‐parental ldquomappingrdquo families have beenstudied (Dong et al 2018 Gifford Chae SwaminathanMoose amp Juvik 2015) alongside panels of diverse germ-plasm accessions for GWAS (Slavov et al 2013) Fur-ther developments calibrating GS with very large panelsof parents and cross progeny are underway (Daveyet al 2017) The recently completed first miscanthus ref-erence genome sequence is expected to improve the effi-ciency of MAS strategies and especially GWAS(httpsphytozomejgidoegovpzportalhtmlinfoalias=Org_Msinensis_er) For example without a referencegenome sequence Clark et al (2014) obtained 21207RADseq SNPs (single nucleotide polymorphisms) on apanel of 767 miscanthus genotypes (mostly M sinensis)but subsequent reanalysis of the RADseq data using the
CLIFTON‐BROWN ET AL | 17
new reference genome resulted in hundreds of thousandsof SNPs being called
Robust and effective in vitro regeneration systems havebeen developed for Miscanthus sinensis M times giganteusand M sacchariflorus (Dalton 2013 Guo et al 2013Hwang Cho et al 2014 Rambaud et al 2013 Ślusarkie-wicz‐Jarzina et al 2017 Wang et al 2011 Zhang et al2012) However there is still significant genotype speci-ficity and these methods need ldquoin‐houserdquo optimization anddevelopment to be used routinely They provide potentialroutes for rapid clonal propagation and also as a basis forgenetic transformation Stable transformation using bothbiolistics (Wang et al 2011) and Agrobacterium tumefa-ciens DNA delivery methods (Hwang Cho et al 2014Hwang Lim et al 2014) has been achieved in M sinen-sis The development of miscanthus transformation andgene editing to generate diplogametes for producing seed‐propagated triploid hybrids are performed as part of theFrench project MISEDIT (miscanthus gene editing forseed‐propagated triploids) There are no reports of genomeediting in any miscanthus species but new breeding inno-vations including genome editing are particularly relevantin this slow‐to‐breed nonfood bioenergy crop (Table 4)
4 | WILLOW
Willow (Salix spp) is a very diverse group of catkin‐bear-ing trees and shrubs Willow belongs to the family Sali-caceae which also includes the Populus genus There areapproximately 350 willow species (Argus 2007) foundmostly in temperate and arctic zones in the northern hemi-sphere A few are adapted to subtropical and tropicalzones The centre of diversity is believed to be in Asiawith over 200 species in China Around 120 species arefound in the former Soviet Union over 100 in NorthAmerica and around 65 species in Europe and one speciesis native to South America (Karp et al 2011) Willows aredioecious thus obligate outcrossers and highly heterozy-gous The haploid chromosome number is 19 (Hanley ampKarp 2014) Around 40 of species are polyploid (Sudaamp Argus 1968) ranging from triploids to the atypicaldodecaploid S maxxaliana with 2n=190 (Zsuffa et al1984)
Although almost exclusively native to the NorthernHemisphere willow has been grown around the globe formany thousands of years to support a wide range of appli-cations (Kuzovkina amp Quigley 2005 Stott 1992) How-ever it has been the focus of domestication for bioenergypurposes for only a relatively short period since the 1970sin North America and Europe For bioenergy breedershave focused their efforts on the shrub willows (subgenusVetix) because of their rapid juvenile growth rates as aresponse to coppicing on a 2‐ to 4‐year cycle that can be
accomplished using farm machinery rather than forestryequipment (Shield Macalpine Hanley amp Karp 2015Smart amp Cameron 2012)
Since shrub willow was not generally recognized as anagricultural crop until very recently there has been littlecommitment to building and maintaining germplasm reposi-tories of willow to support long‐term breeding One excep-tion is the United Kingdom where a large and well‐characterized Salix germplasm collection comprising over1500 accessions is held at Rothamsted Research (Stott1992 Trybush et al 2008) Originally initiated for use inbasketry in 1923 accessions have been added ever sinceIn the United States a germplasm collection of gt350accessions is located at Cornell University to support thebreeding programme there The UK and Cornell collectionshave a relatively small number of accessions in common(around 20) Taken together they represent much of thespecies diversity but only a small fraction of the overallgenetic diversity within the genus There are three activewillow breeding programmes in Europe RothamstedResearch (UK) Salixenergi Europa AB (SEE) and a pro-gramme at the University of Warmia and Mazury in Olsz-tyn (Poland) (abbreviations used in Table 1) There is oneactive US programme based at Cornell University Culti-vars are still being marketed by the European WillowBreeding Programme (EWBP) (UK) which was activelybreeding biomass varieties from 1996 to 2002 Cultivarsare protected by plant breedersrsquo rights (PBRs) in Europeand by plant patents in the United States The sharing ofgenetic resources in the willow community is generallyregulated by material transfer agreements (MTA) and tai-lored licensing agreements although the import of cuttingsinto North America is prohibited except under special quar-antine permit conditions
Efforts to augment breeding germplasm collection fromnature are continuing with phenotypic screening of wildgermplasm performed in field experiments with 177 S pur-purea genotypes in the United States (at sites in Genevaand Portland NY and Morgantown WV) that have beengenotyped using genotyping by sequencing (GBS) (Elshireet al 2011) In addition there are approximately 400accessions of S viminalis in Europe (near Pustnaumls Upp-sala Sweden and Woburn UK (Berlin et al 2014 Hal-lingbaumlck et al 2016) The S viminalis accessions wereinitially genotyped using 38 simple sequence repeats (SSR)markers to assess genetic diversity and screened with~1600 SNPs in genes of potential interest for phenologyand biomass traits Genetics and genomics combined withextensive phenotyping have substantially improved thegenetic basis of biomass‐related traits in willow and arenow being developed in targeted breeding via MAS Thisunderpinning work has been conducted on large specifi-cally developed biparental Salix mapping populations
18 | CLIFTON‐BROWN ET AL
(Hanley amp Karp 2014 Zhou et al 2018) as well asGWAS panels (Hallingbaumlck et al 2016)
Once promising parental combinations are identifiedcrosses are usually performed using fresh pollen frommaterial that has been subject to a phased removal fromcold storage (minus4degC) (Lindegaard amp Barker 1997 Macal-pine Shield Trybush Hayes amp Karp 2008 Mosseler1990) Pollen storage is useful in certain interspecific com-binations where flowering is not naturally synchronizedThis can be overcome by using pollen collection and stor-age protocol which involves extracting pollen using toluene(Kopp Maynard Niella Smart amp Abrahamson 2002)
The main breeding approach to improve willow yieldsrelies on species hybridization to capture hybrid vigour(Fabio et al 2017 Serapiglia Gouker amp Smart 2014) Inthe absence of genotypic models for heterosis breedershave extensively tested general and specific combiningability of parents to produce superior progeny The UKbreeding programmes (EWBP 1996ndash2002 and RothamstedResearch from 2003 on) have performed more than 1500exploratory cross‐pollinations The Cornell programme hassuccessfully completed about 550 crosses since 1998Investment into the characterization of genetic diversitycombined with progeny tests from exploratory crosses hasbeen used to produce hundreds of targeted intraspeciescrosses in the United Kingdom and United States respec-tively (see Table 1) To achieve long‐term gains beyond F1hybrids four intraspecific recurrent selection populationshave been created in the United Kingdom (for S dasycla-dos S viminalis and S miyabeana) and Cornell is pursu-ing recurrent selection of S purpurea Interspecifichybridizations with genotypes selected from the recurrentselection cycles are well advanced in willow with suchcrosses to date totalling 420 in the United Kingdom andover 100 in the United States
While species hybridization is common in Salix it isnot universal Of the crosses attempted about 50 hybri-dize and produce seed (Macalpine Shield amp Karp 2010)As the viability of seed from successful crosses is short (amatter of days at ambient temperatures) proper seed rear-ing and storage protocols are essential (Maroder PregoFacciuto amp Maldonado 2000)
Progeny from crosses are treated in different waysamong the breeding programmes at the seedling stage Inthe United States seedlings are planted into an irrigatedfield where plants are screened for two seasons beforebeing progressed to further field trials In the United King-dom seedlings are planted into trays of compost wherethey remain containerized in an irrigated nursery for theremainder of year one In the United Kingdom seedlingsare subject to two rounds of selection in the nursery yearThe first round takes place in September to select againstsusceptibility to rust infection (Melampsora spp) A second
round of selection in winter assesses tip damage from frostand giant willow aphid infestation In the United Stateswhere the rust pressure is lower screening for Melampsoraspp cannot be performed at the nursery stage Both pro-grammes monitor Melampsora spp pest susceptibilityyield and architecture over multiple years in field trialsSelected material is subject to two rounds of field trials fol-lowed by a final multilocation yield trial to identify vari-eties for commercialization
Promising selections (ie potential cultivars) need to beclonally propagated A rapid in vitro tissue culture propa-gation method has been developed (Palomo‐Riacuteos et al2015) This method can generate about 5000 viable trans-plantable clones from a single plant in just 24 weeks Anin vitro system can also accommodate early selection viamolecular or biochemical markers to increase selectionspeed Conventional breeding systems take 13 years viafour rounds of selection from crossing to selecting a variety(Figure 1) but this has the potential to be reduced to7 years if micropropagation and MAS selection are adopted(Hanley amp Karp 2014 Palomo‐Riacuteos et al 2015)
Willows are currently propagated commercially byplanting winter‐dormant stem cuttings in spring Commer-cial planting systems for willow use mechanical plantersthat cut and insert stem sections from whips into a well‐prepared soil One hectare of stock plants grown in specificmultiplication beds planted at 40000 plants per ha pro-duces planting material for 80 hectares of commercialshort‐rotation coppice willow annually (planted at 15000cuttings per hectare) (Whittaker et al 2016) When com-mercial plantations are established the industry standard isto plant intimate mixtures of ~5 diverse rust (Melampsoraspp)‐resistant varieties (McCracken amp Dawson 1997 VanDen Broek et al 2001)
The foundations for using new plant breeding tech-niques have been established with funding from both thepublic and the private sectors To establish QTL maps 16mapping populations from biparental crosses are understudy in the United Kingdom Nine are under study in theUnited States The average number of individuals in thesefamilies ranges from 150 to 947 (Hanley amp Karp 2014)GS is also being evaluated in S viminalis and preliminaryresults indicate that multiomic approaches combininggenomic and metabolomic data have great potential (Sla-vov amp Davey 2017) For both QTL and GS approachesthe field phenotyping demands are large as several thou-sand individuals need to be phenotyped for a wide rangeof traits These include the following dates of bud burstand growth succession stem height stem density wooddensity and disease resistance The greater the number ofindividuals the more precise the QTL marker maps andGS models are However the logistical and financial chal-lenges of phenotyping large numbers of individuals are
CLIFTON‐BROWN ET AL | 19
considerable because the willow crop is gt5 m tall in thesecond year There is tremendous potential to improve thethroughput of phenotyping using unmanned aerial systemswhich is being tested in the USDA National Institute ofFood and Agriculture (NIFA) Willow SkyCAP project atCornell Further investment in these approaches needs tobe sustained over many years fully realizes the potentialof a marker‐assisted selection programme for willow
To date despite considerable efforts in Europe and theUnited States to establish a routine transformation systemthere has not been a breakthrough in willow but attemptsare ongoing As some form of transformation is typically aprerequisite for genome editing techniques these have notyet been applied to willow
In Europe there are 53 short‐rotation coppice (SRC) bio-mass willow cultivars registered with the Community PlantVariety Office (CPVO) for PBRs of which ~25 are availablecommercially in the United Kingdom There are eightpatented cultivars commercially available in the UnitedStates In Sweden there are nine commercial cultivars regis-tered in Europe and two others which are unregistered(httpssalixenergiseplanting-material) Furthermore thereare about 20 precommercial hybrids in final yield trials inboth the United States and the United Kingdom It has beenestimated that it would take two years to produce the stockrequired to plant 50 ha commercially from the plant stock inthe final yield trials Breeding programmes have alreadydelivered rust‐resistant varieties and increases in yield to themarket The adoption of advanced breeding technologies willlikely lead to a step change in improving traits of interest
5 | POPLAR
Poplar a fast‐growing tree from the northern hemispherewith a small genome size has been adopted for commercialforestry and scientific purposes The genus Populus con-sists of about 29 species classified in six different sectionsPopulus (formerly Leuce) Tacamahaca Aigeiros AbasoTuranga and Leucoides (Eckenwalder 1996) The Populusspecies of most interest for breeding and testing in the Uni-ted States and Europe are P nigra P deltoides P maxi-mowiczii and P trichocarpa (Stanton 2014) Populusclones for biomass production are being developed byintra‐ and interspecies hybridization (DeWoody Trewin ampTaylor 2015 Richardson Isebrands amp Ball 2014 van derSchoot et al 2000) Recurrent selection approaches areused for gradual population improvement and to create eliteclonal lines for commercialization (Berguson McMahon ampRiemenschneider 2017 Neale amp Kremer 2011) Currentlypoplar breeding in the United States occurs in industrialand academic programmes located in the Southeast theMidwest and the Pacific Northwest These use six speciesand five interspecific taxa (Stanton 2014)
The southeastern programme historically focused onrecurrent selection of P deltoides from accessions made inthe lower Mississippi River alluvial plain (Robison Rous-seau amp Zhang 2006) More recently the genetic base hasbeen broadened to produce interspecific hybrids with resis-tance to the fungal infection Septoria musiva which causescankers
In the midwest of the United States populationimprovement efforts are focused on P deltoides selectionsfrom native provenances and hybrid crosses with acces-sions introduced from Europe Interspecific intercontinental(Europe and America) hybrid crosses between P nigra andP deltoides (P times canadensis) are behind many of theleading commercial hybrids which are the most advancedbreeding materials for many applications and regions InMinnesota previous breeding experience and efforts utiliz-ing P maximowiczii and P trichocarpa have been discon-tinued due to Septoria susceptibility and a lack of coldhardiness (Berguson et al 2017) Traits targeted forimprovement include yieldgrowth rate cold hardinessadventitious rooting resistance to Septoria and Melamp-sora leaf rust and stem form The Upper Midwest pro-gramme also carries out wide hybridizations within thesection Populus The P times wettsteinii (P trem-ula times P tremuloides) taxon is bred for gains in growthrate wood quality and resistance to the fungus Entoleucamammata which causes hypoxylon canker (David ampAnderson 2002)
In the Pacific Northwest GreenWood Resources Incleads poplar breeding that emphasizes interspecifichybrid improvement of P times generosa (P deltoides timesP trichocarpa and reciprocal) and P deltoides times P maxi-mowiczii taxa for coastal regions and the P times canadensistaxon for the drier continental regions Intraspecificimprovement of second‐generation breeding populations ofP deltoides P nigra P maximowiczii and P trichocarpaare also involved (Stanton et al 2010) The present focusof GreenWood Resourcesrsquo hybridization is bioenergy feed-stock improvement concentrating on coppice yield wood‐specific gravity and rate of sugar release
Industrial interest in poplar in the United States has his-torically come from the pulp and paper sector althoughveneer and dimensional lumber markets have been pursuedat times Currently the biomass market for liquid trans-portation fuels is being emphasized along with the use oftraditional and improved poplar genotypes for ecosystemservices such as phytoremediation (Tuskan amp Walsh 2001Zalesny et al 2016)
In Europe there are breeding programmes in FranceGermany Italy and Sweden These include the following(a) Alasia Franco Vivai (AFV) programme in northernItaly (b) the French programme led by the poplar Scien-tific Interest Group (GIS Peuplier) and carried out
20 | CLIFTON‐BROWN ET AL
collaboratively between the National Institute for Agricul-tural Research (INRA) the National Research Unit ofScience and Technology for Environment and Agriculture(IRSTEA) and the Forest Cellulose Wood Constructionand Furniture Technology Institute (FCBA) (c) the Ger-man programme at Northwest German Forest Research Sta-tion (NW‐FVA) at Hannoversch Muumlnden and (d) theSwedish programme at the Swedish University of Agricul-tural Sciences and SweTree Technologies AB (Table 1)
AFV leads an Italian poplar breeding programme usingextensive field‐grown germplasm collections of P albaP deltoides P nigra and P trichocarpa While interspeci-fic hybridization uses several taxa the focus is onP times canadensis The breeding programme addresses dis-ease resistance (Marssonina brunnea Melampsora larici‐populina Discosporium populeum and poplar mosaicvirus) growth rate and photoperiod adaptation AFV andGreenWood Resources collaborate in poplar improvementin Europe through the exchange of frozen pollen and seedfor reciprocal breeding projects Plantations in Poland andRomania are currently the focus of the collaboration
The ongoing French GIS Peuplier is developing a long‐term breeding programme based on intraspecific recurrentselection for the four parental species (P deltoides P tri-chocarpa P nigra and P maximowiczii) designed to betterbenefit from hybrid vigour demonstrated by the interspeci-fic crosses P canadensis P deltoides times P trichocarpaand P trichocarpa times P maximowiczii Current selectionpriorities are targeting adaptation to soil and climate condi-tions resistance and tolerance to the most economicallyimportant diseases and pests high volume production underSRC and traditional poplar cultivation regimes as well aswood quality of interest by different markets Currentlygenomic selection is under exploration to increase selectionaccuracy and selection intensity while maintaining geneticdiversity over generations
The German NW‐FVA programme is breeding intersec-tional AigeirosndashTacamahaca hybrids with a focus on resis-tance to Pollaccia elegans Xanthomonas populiDothichiza spp Marssonina brunnea and Melampsoraspp (Stanton 2014) Various cross combinations ofP maximowiczii P trichocarpa P nigra and P deltoideshave led to new cultivars suitable for deployment in vari-etal mixtures of five to ten genotypes of complementarystature high productivity and phenotypic stability (Weis-gerber 1993) The current priority is the selection of culti-vars for high‐yield short‐rotation biomass production Sixhundred P nigra genotypes are maintained in an ex situconservation programme An in situ P nigra conservationeffort involves an inventory of native stands which havebeen molecular fingerprinted for identity and diversity
The Swedish programme is concentrating on locallyadapted genotypes used for short‐rotation forestry (SRF)
because these meet the needs of the current pulping mar-kets Several field trials have shown that commercial poplarclones tested and deployed in Southern and Central Europeare not well adapted to photoperiods and low temperaturesin Sweden and in the Baltics Consequently SwedishUniversity of Agricultural Sciences and SweTree Technolo-gies AB started breeding in Sweden in 1990s to producepoplar clones better adapted to local climates and markets
51 | Molecular breeding technologies
Poplar genetic improvement cannot be rapidly achievedthrough traditional methods alone because of the longbreeding cycles outcrossing breeding systems and highheterozygosity Integrating modern genetic genomic andphenomics techniques with conventional breeding has thepotential to expedite poplar improvement
The genome of poplar has been sequenced (Tuskanet al 2006) It has an estimated genome size of485 plusmn 10 Mbp divided into 19 chromosomes This is smal-ler than other PBCs and makes poplar more amenable togenetic engineering (transgenesis) GS and genome editingPoplar has seen major investment in both the United Statesand Europe being the model system for woody perennialplant genetics and genomics research
52 | Targets for genetic modification
Traits targeted include wood properties (lignin content andcomposition) earlylate flowering male sterility to addressbiosafety regulation issues enhanced yield traits and herbi-cide tolerance These extensive transgenic experiments haveshown differences in recalcitrance to in vitro regenerationand genetic transformation in some of the most importantcommercial hybrid poplars (Alburquerque et al 2016)Further transgene expression stability is being studied Sofar China is the only country known to have commerciallyused transgenic insect‐resistant poplar A precommercialherbicide‐tolerant poplar was trialled for 8 years in the Uni-ted States (Li Meilan Ma Barish amp Strauss 2008) butcould not be released due to stringent environmental riskassessments required for regulatory approval This increasestranslation costs and delays reducing investor confidencefor commercial deployment (Harfouche et al 2011)
The first field trials of transgenic poplar were performedin France in 1987 (Fillatti Sellmer Mccown Haissig ampComai 1987) and in Belgium in 1988 (Deblock 1990)Although there have been a total of 28 research‐scale GMpoplar field trials approved in the European Union underCouncil Directive 90220EEC since October 1991 (inPoland Belgium Finland France Germany Spain Swe-den and in the United Kingdom (Pilate et al 2016) onlyauthorizations in Poland and Belgium are in place today In
CLIFTON‐BROWN ET AL | 21
the United States regulatory notifications and permits fornearly 20000 transgenic poplar trees derived from approxi-mately 600 different constructs have been issued since1995 by the USDAs Animal and Plant Health InspectionService (APHIS) (Strauss et al 2016)
53 | Genome editing CRISPR technologies
Clustered regularly interspaced palindromic repeats(CRISPR) and the CRISPR‐associated (CRISPR‐Cas)nucleases are a groundbreaking genome‐engineering toolthat complements classical plant breeding and transgenicmethods (Moreno‐Mateos et al 2017) Only two publishedstudies in poplar have applied the CRISPRCas9 technol-ogy One is in P tomentosa in which an endogenous phy-toene desaturase gene (PtoPDS) was successfully disruptedsite specifically in the first generation of transgenic plantsresulting in an albino and dwarf phenotype (Fan et al2015) The second was in P tremula times alba in whichhigh CRISPR‐Cas9 mutational efficiency was achieved forthree 4‐coumarateCoA ligase (4Cl) genes 4CL1 4CL2and 4CL5 associated with lignin and flavonoid biosynthe-sis (Zhou et al 2015) Due to its low cost precision andrapidness it is very probable that cultivars or clones pro-duced using CRISPR technology will be ready for market-ing in the near future (Yin et al 2017) Recently aCRISPR with a smaller associated endonuclease has beendiscovered from Prevotella and Francisella 1 (Cpf1) whichmay have advantages over Cas9 In addition there arereports of DNA‐free editing in plants using both CRISPRCpf1 and CRISPR Cas9 for example Ref (Kim et al2017 Mahfouz 2017 Zaidi et al 2017)
It remains unresolved whether plants modified by genomeediting will be regulated as genetically modified organisms(GMOs) by the relevant authorities in different countries(Lozano‐Juste amp Cutler 2014) Regulations to cover thesenew breeding techniques are still evolving but those coun-tries who have published specific guidance (including UnitedStates Argentina and Chile) are indicating that plants pos-sessing simple genome edits will not be regulated as conven-tional transgenesis (Jones 2015b) The first generation ofgenome‐edited crops will likely be phenocopy gene knock-outs that already exist to produce ldquonature identicalrdquo traits thatis traits that could also be derived by conventional breedingDespite this confidence in applying these new powerfulbreeding tools remains limited owing to the uncertain regula-tory environment in many parts of the world (Gao 2018)including the recent ECJ 2018 rulings mentioned earlier
54 | Genomics‐based breeding technologies
Poplar breeding programs are becoming well equipped withuseful genomics tools and resources that are critical to
explore genomewide variability and make use of the varia-tion for enhancing genetic gains Deep transcriptomesequencing resequencing of alternate genomes and GBStechnology for genomewide marker detection using next‐generation sequencing (NGS) are yielding valuable geno-mics tools GWAS with NGS‐based markers facilitatesmarker identification for MAS breeding by design and GS
GWAS approaches have provided a deeper understand-ing of genome function as well as allelic architectures ofcomplex traits (Huang et al 2010) and have been widelyimplemented in poplar for wood characteristics (Porthet al 2013) stomatal patterning carbon gain versus dis-ease resistance (McKown et al 2014) height and phenol-ogy (Evans et al 2014) cell wall chemistry (Mucheroet al 2015) growth and cell walls traits (Fahrenkroget al 2017) bark roughness (Bdeir et al 2017) andheight and diameter growth (Liu et al 2018) Using high‐throughput sequencing and genotyping platforms an enor-mous amount of SNP markers have been used to character-ize the linkage disequilibrium (LD) in poplar (eg Slavovet al 2012 discussed below)
The genetic architecture of photoperiodic traits in peren-nial trees is complex involving many loci However itshows high levels of conservation during evolution (Mau-rya amp Bhalerao 2017) These genomics tools can thereforebe used to address adaptation issues and fine‐tune themovement of elite lines into new environments For exam-ple poor timing of spring bud burst and autumn bud setcan result in frost damage resulting in yield losses (Ilstedt1996) These have been studied in P tremula genotypesalong a latitudinal cline in Sweden (~56ndash66oN) and haverevealed high nucleotide polymorphism in two nonsynony-mous SNPs within and around the phytochrome B2 locus(Ingvarsson Garcia Hall Luquez amp Jansson 2006 Ing-varsson Garcia Luquez Hall amp Jansson 2008) Rese-quencing 94 of these P tremula genotypes for GWASshowed that noncoding variation of a single genomicregion containing the PtFT2 gene described 65 ofobserved genetic variation in bud set along the latitudinalcline (Tan 2018)
Resequencing genomes is currently the most rapid andeffective method detecting genetic differences betweenvariants and for linking loci to complex and importantagronomical and biomass traits thus addressing breedingchallenges associated with long‐lived plants like poplars
To date whole genome resequencing initiatives havebeen launched for several poplar species and genotypes InPopulus LD studies based on genome resequencing sug-gested the feasibility of GWAS in undomesticated popula-tions (Slavov et al 2012) This plant population is beingused to inform breeding for bioenergy development Forexample the detection of reliable phenotypegenotype asso-ciations and molecular signatures of selection requires a
22 | CLIFTON‐BROWN ET AL
detailed knowledge about genomewide patterns of allelefrequency variation LD and recombination suggesting thatGWAS and GS in undomesticated populations may bemore feasible in Populus than previously assumed Slavovet al (2012) have resequenced 16 genomes of P tri-chocarpa and genotyped 120 trees from 10 subpopulationsusing 29213 SNPs (Geraldes et al 2013) The largest everSNP data set of genetic variations in poplar has recentlybeen released providing useful information for breedinghttpswwwbioenergycenterorgbescgwasindexcfmAlso deep sequencing of transcriptomes using RNA‐Seqhas been used for identification of functional genes andmolecular markers that is polymorphism markers andSSRs A multitissue and multiple experimental data set forP trichocarpa RNA‐Seq is publicly available (httpsjgidoegovdoe-jgi-plant-flagship-gene-atlas)
The availability of genomic information of DNA‐con-taining cell organelles (nucleus chloroplast and mitochon-dria) will also allow a holistic approach in poplarmolecular breeding in the near future (Kersten et al2016) Complete Populus genome sequences are availablefor nucleus (P trichocarpa section Tacamahaca) andchloroplasts (seven species and two clones from P trem-ula W52 and P tremula times P alba 717ndash1B4) A compara-tive approach revealed structural and functionalinformation broadening the knowledge base of PopuluscpDNA and stimulating future diagnostic marker develop-ment The availability of whole genome sequences of thesecellular compartments of P tremula holds promise forboosting marker‐assisted poplar breeding Other nucleargenome sequences from additional Populus species arenow available (eg P deltoides (httpsphytozomejgidoegovpz) and will become available in the forthcomingyears (eg P tremula and P tremuloidesmdashPopGenIE(Sjodin Street Sandberg Gustafsson amp Jansson 2009))Recently the characterization of the poplar pan‐genome bygenomewide identification of structural variation in threecrossable poplar species P nigra P deltoides and P tri-chocarpa revealed a deeper understanding of the role ofinter‐ and intraspecific structural variants in poplar pheno-type and may have important implications for breedingparticularly interspecific hybrids (Pinosio et al 2016)
GS has been proposed as an alternative to MAS in cropimprovement (Bernardo amp Yu 2007 Heffner Sorrells ampJannink 2009) GS is particularly well suited for specieswith long generation times for characteristics that displaymoderate‐to‐low heritability for traits that are expensive tomeasure and for selection of traits expressed late in the lifecycle as is the case for most traits of commercial value inforestry (Harfouche et al 2012) Current joint genomesequencing efforts to implement GS in poplar using geno-mic‐estimated breeding values for bioenergy conversiontraits from 49 P trichocarpa families and 20 full‐sib
progeny are taking place at the Oak Ridge National Labo-ratory and GreenWood Resources (Brian Stanton personalcommunication httpscbiornlgov) These data togetherwith the resequenced GWAS population data will be thebasis for developing GS algorithms Genomic breedingtools have been developed for the intraspecific programmetargeting yield resistance to Venturia shoot blightMelampsora leaf rust resistance to Cryptorhynchus lapathistem form wood‐specific gravity and wind firmness (Evanset al 2014 Guerra et al 2016) A newly developedldquobreeding with rare defective allelesrdquo (BRDA) technologyhas been developed to exploit natural variation of P nigraand identify defective variants of genes predicted by priortransgenic research to impact lignin properties Individualtrees carrying naturally defective alleles can then be incor-porated directly into breeding programs thereby bypassingthe need for transgenics (Vanholme et al 2013) Thisnovel breeding technology offers a reverse genetics com-plement to emerging GS for targeted improvement of quan-titative traits (Tsai 2013)
55 | Phenomics‐assisted breeding technology
Phenomics involves the characterization of phenomesmdashthefull set of phenotypes of given individual plants (HouleGovindaraju amp Omholt 2010) Traditional phenotypingtools which inefficiently measure a limited set of pheno-types have become a bottleneck in plant breeding studiesHigh‐throughput plant phenotyping facilities provide accu-rate screening of thousands of plant breeding lines clones orpopulations over time (Fu 2015) are critical for acceleratinggenomics‐based breeding Automated image collection andanalysis phenomics technologies allow accurate and nonde-structive measurements of a diversity of phenotypic traits inlarge breeding populations (Gegas Gay Camargo amp Doo-nan 2014 Goggin Lorence amp Topp 2015 Ludovisi et al2017 Shakoor Lee amp Mockler 2017) One important con-sideration is the identification of relevant and quantifiabletarget traits that are early diagnostic indicators of biomassyield Good progress has been made in elucidating theseunderpinning morpho‐physiological traits that are amenableto remote sensing in Populus (Harfouche Meilan amp Altman2014 Rae Robinson Street amp Taylor 2004) Morerecently Ludovisi et al (2017) developed a novel methodol-ogy for field phenomics of drought stress in a P nigra F2partially inbred population using thermal infrared imagesrecorded from an unmanned aerial vehicle‐based platform
Energy is the current main market for poplar biomass butthe market return provided is not sufficient to support pro-duction expansion even with added demand for environmen-tal and land management ldquoecosystem servicesrdquo such as thetreatment of effluent phytoremediation riparian buffer zonesand agro‐forestry plantings Aviation fuel is a significant
CLIFTON‐BROWN ET AL | 23
target market (Crawford et al 2016) To serve this marketand to reduce current carbon costs of production (Budsberget al 2016) key improvement traits in addition to yield(eg coppice regeneration pestdisease resistance water‐and nutrient‐use efficiencies) will be trace greenhouse gas(GHG) emissions (eg isoprene volatiles) site adaptabilityand biomass conversion efficiency Efforts are underway tohave national environmental protection agenciesrsquo approvalfor poplar hybrids qualifying for renewable energy credits
6 | REFLECTIONS ON THECOMMERCIALIZATIONCHALLENGE
The research and innovation activities reviewed in thispaper aim to advance the genetic improvement of speciesthat can provide feedstocks for bioenergy applicationsshould those markets eventually develop These marketsneed to generate sufficient revenue and adequately dis-tribute it to the actors along the value chain The work onall four crops shares one thing in common long‐termefforts to integrate fundamental knowledge into breedingand crop development along a research and development(RampD) pipeline The development of miscanthus led byAberystwyth University exemplifies the concerted researcheffort that has integrated the RampD activities from eight pro-jects over 14 years with background core research fundingalong an emerging innovation chain (Figure 2) This pro-gramme has produced a first range of conventionally bredseeded interspecies hybrids which are now in upscaling tri-als (Table 3) The application of molecular approaches(Table 5) with further conventional breeding (Table 4)offers the prospect of a second range of improved seededhybrids This example shows that research‐based support ofthe development of new crops or crop types requires along‐term commitment that goes beyond that normallyavailable from project‐based funding (Figure 1) Innovationin this sector requires continuous resourcing of conven-tional breeding operations and capability to minimize timeand investment losses caused by funding discontinuities
This challenge is increased further by the well‐knownmarket failure in the breeding of many agricultural cropspecies The UK Department for Environment Food andRural Affairs (Defra) examined the role of geneticimprovement in relation to nonmarket outcomes such asenvironmental protection and concluded that public invest-ment in breeding was required if profound market failure isto be addressed (Defra 2002) With the exception ofwidely grown hybrid crops such as maize and some high‐value horticultural crops royalties arising from plant breed-ersrsquo rights or other returns to breeders fail to adequatelycompensate for the full cost for research‐based plant breed-ing The result even for major crops such as wheat is sub‐
optimal investment and suboptimal returns for society Thismarket failure is especially acute for perennial crops devel-oped for improved sustainability rather than consumerappeal (Tracy et al 2016) Figure 3 illustrates the underly-ing challenge of capturing value for the breeding effortThe ldquovalley of deathrdquo that results from the low and delayedreturns to investment applies generally to the research‐to‐product innovation pipeline (Beard Ford Koutsky amp Spi-wak 2009) and certainly to most agricultural crop speciesHowever this schematic is particularly relevant to PBCsMost of the value for society from the improved breedingof these crops comes from changes in how agricultural landis used that is it depends on the increased production ofthese crops The value for society includes many ecosys-tems benefits the effects of a return to seminatural peren-nial crop cover that protects soils the increase in soilcarbon storage the protection of vulnerable land or the cul-tivation of polluted soils and the reductions in GHG emis-sions (Lewandowski 2016) By its very nature theproduction of biomass on agricultural land marginal forfood production challenges farm‐level profitability Thecosts of planting material and one‐time nature of cropestablishment are major early‐stage costs and therefore theopportunities for conventional royalty capture by breedersthat are manifold for annual crops are limited for PBCs(Hastings et al 2017) Public investment in developingPBCs for the nonfood biobased sector needs to providemore long‐term support for this critical foundation to a sus-tainable bioeconomy
7 | CONCLUSIONS
This paper provides an overview of research‐based plantbreeding in four leading PBCs For all four PBC generasignificant progress has been made in genetic improvementthrough collaboration between research scientists and thoseoperating ongoing breeding programmes Compared withthe main food crops most PBC breeding programmes dateback only a few decades (Table 1) This breeding efforthas thus co‐evolved with molecular biology and the result-ing ‐omics technologies that can support breeding Thedevelopment of all four PBCs has depended strongly onpublic investment in research and innovation The natureand driver of the investment varied In close associationwith public research organizations or universities all theseprogrammes started with germplasm collection and charac-terization which underpin the selection of parents forexploratory wide crosses for progeny testing (Figure 1Table 4)
Public support for switchgrass in North America wasexplicitly linked to plant breeding with 12 breeding pro-grammes supported in the United States and CanadaSwitchgrass breeding efforts to date using conventional
24 | CLIFTON‐BROWN ET AL
breeding have resulted in over 36 registered cultivars inthe United States (Table 1) with the development of dedi-cated biomass‐type cultivars coming within the past fewyears While ‐omics technologies have been incorporatedinto several of these breeding programmes they have notyet led to commercial deployment in either conventional orhybrid cultivars
Willow genetic improvement was led by the researchcommunity closely linked to plant breeding programmesWillow and poplar have the longest record of public invest-ment in genetic improvement that can be traced back to1920s in the United Kingdom and United States respec-tively Like switchgrass breeding programmes for willoware connected to public research efforts The United King-dom in partnership with the programme based at CornellUniversity remains the European leader in willowimprovement with a long‐term breeding effort closelylinked to supporting biological research at Rothamsted Inwillow F1 hybrids have produced impressive yield gainsover parental germplasm by capturing hybrid vigour Over30 willow clones are commercially available in the UnitedStates and Europe and a further ~90 are under precommer-cial testing (Table 1)
Compared with willow and poplar miscanthus is a rela-tive newcomer with all the current breeding programmesstarting in the 2000s Clonal M times giganteus propagated byrhizomes is expected to be replaced by more readily scal-able seeded hybrids from intra‐ (M sinensis) and inter‐(M sacchariflorus times M sinensis) species crosses with highseed multiplication rates (of gt2000) The first group ofhybrid cultivars is expected to be market‐ready around2022
Of the four genera used as PBCs Populus is the mostadvanced in terms of achievements in biological researchas a result of its use as a model for basic research of treesMuch of this biological research is not directly connectedto plant breeding Nevertheless reflecting the fact thatpoplar is widely grown as a single‐stem tree in SRF thereare about 60 commercially available clones and an addi-tional 80 clones in commercial pipelines (Table 1) Trans-genic poplar hybrids have moved beyond proof of conceptto commercial reality in China
Many PBC programmes have initiated long‐term con-ventional recurrent selection breeding cycles for populationimprovement which is a key process in increasing yieldthrough hybrid vigour As this approach requires manyyears most programmes are experimenting with molecularbreeding methods as these have the potential to accelerateprecision breeding For all four PBCs investments in basicgenetic and genomic resources including the developmentof mapping populations for QTLs and whole genomesequences are available to support long‐term advancesMore recently association genetics with panels of diversegermplasm are being used as training populations for GSmodels (Table 5) These efforts are benefitting from pub-licly available DNA sequences and whole genome assem-blies in crop databases Key to these accelerated breedingtechnologies are developments in novel phenomics tech-nologies to bridge the genotypephenotype gap In poplarnovel remote sensing field phenotyping is now beingdeployed to assist breeders These advances are being com-bined with in vitro and in planta modern molecular breed-ing techniques such as CRISPR (Table 5) CRISPRtechnology for genome editing has been proven in poplar
FIGURE 2 A schematic developmentpathway for miscanthus in the UnitedKingdom related to the investment in RampDprojects at Aberystwyth (top coloured areasfor projects in the three categories basicresearch breeding and commercialupscaling) leading to a projected croppingarea of 350000 ha by 2030 with clonal andsuccessive ranges of improved seed‐basedhybrids Purple represents theBiotechnology and Biological SciencesResearch Council (BBSRC) and brown theDepartment for Environment Food andRural Affairs (Defra) (UK Nationalfunding) blue bars represent EU fundingand green private sector funding (Terravestaand CERES) and GIANT‐LINK andMiscanthus Upscaling Technology (MUST)are publicndashprivate‐initiatives (PPI)
CLIFTON‐BROWN ET AL | 25
This technology is also being applied in switchgrass andmiscanthus (Table 5) but the future of CRISPR in com-mercial breeding for the European market is uncertain inthe light of recent ECJ 2018 rulings
There is integration of research and plant breeding itselfin all four PBCs Therefore estimating the ongoing costs ofmaintaining these breeding programmes is difficult Invest-ment in research also seeks wider benefits associated withtechnological advances in plant science rather than cultivardevelopment per se However in all cases the conventionalbreeding cycle shown in Figure 1 is the basic ldquoenginerdquo withmolecular technologies (‐omics) serving to accelerate thisengine The history of the development shows that the exis-tence of these breeding programmes is essential to gain ben-efits from the biological research Despite this it is thisessential step that is at most risk from reductions in invest-ment A conventional breeding programme typicallyrequires a breeder and several technicians who are supported
over the long term (20ndash30 years Figure 1) at costs of about05 to 10 million Euro per year (as of 2018) The analysisreported here shows that the time needed to perform onecycle of conventional breeding bringing germplasm fromthe wild to a commercial hybrid ranged from 11 years inswitchgrass to 26 years in poplar (Figure 1) In a maturecrop grown on over 100000 ha with effective cultivar pro-tection and a suitable business model this level of revenuecould come from royalties Until such levels are reachedPBCs lie in the innovation valley of death (Figure 3) andneed public support
Applying industrial ldquotechnology readiness levelsrdquo (TRL)originally developed for aerospace (Heacuteder 2017) to ourplant breeding efforts we estimate many promising hybridscultivars are at TRL levels of 3ndash4 In Table 3 experts ineach crop estimate that it would take 3 years from now toupscale planting material from leading cultivars in plot tri-als to 100 ha
FIGURE 3 A schematic relating some of the steps in the innovation chain from relatively basic crop science research through to thedeployment in commercial cropping systems and value chains The shape of the funnel above the expanding development and deploymentrepresents the availability of investment along the development chain from relatively basic research at the top to the upscaled deployment at thebottom Plant breeding links the research effort with the development of cropping systems The constriction represents the constrained fundingfor breeding that links conventional public research investment and the potential returns from commercial development The handover pointsbetween publicly funded work to develop the germplasm resources (often known as prebreeding) the breeding and the subsequent cropdevelopment are shown on the left The constriction point is aggravated by the lack academic rewards for this essential breeding activity Theoutcome is such that this innovation system is constrained by the precarious resourcing of plant breeding The authorsrsquo assessment ofdevelopment status of the four species is shown (poplar having two one for short‐rotation coppice (SRC) poplar and one for the more traditionalshort‐rotation forestry (SRF)) The four new perennial biomass crops (PBCs) are now in the critical phase of depending of plant breedingprogress without the income stream from a large crop production base
26 | CLIFTON‐BROWN ET AL
Taking the UK example mentioned in the introductionplanting rates of ~35000 ha per year from 2022 onwardsare needed to reach over 1 m ha by 2050 Ongoing workin the UK‐funded Miscanthus Upscaling Technology(MUST) project shows that ramping annual hybrid seedproduction from the current level of sufficient seed for10 ha in 2018 to 35000 ha would take about 5 yearsassuming no setbacks If current hybrids of any of the fourPBCs in the upscaling pipeline fail on any step for exam-ple lower than expected multiplication rates or unforeseenagronomic barriers then further selections from ongoingbreeding are needed to replace earlier candidates
In conclusion the breeding foundations have been laidwell for switchgrass miscanthus willow and poplar owingto public funding over the long time periods necessaryImproved cultivars or genotypes are available that could bescaled up over a few years if real sustained market oppor-tunities emerged in response to sustained favourable poli-cies and industrial market pull The potential contributionsof growing and using these PBCs for socioeconomic andenvironmental benefits are clear but how farmers andothers in commercial value chains are rewarded for mass‐scale deployment as is necessary is not obvious at presentTherefore mass‐scale deployment of these lignocellulosecrops needs developments outside the breeding arenas todrive breeding activities more rapidly and extensively
8 | UNCITED REFERENCE
Huang et al (2019)
ACKNOWLEDGEMENTS
UK The UK‐led miscanthus research and breeding wasmainly supported by the Biotechnology and BiologicalSciences Research Council (BBSRC) Department for Envi-ronment Food and Rural Affairs (Defra) the BBSRC CSPstrategic funding grant BBCSP17301 Innovate UKBBSRC ldquoMUSTrdquo BBN0161491 CERES Inc and Ter-ravesta Ltd through the GIANT‐LINK project (LK0863)Genomic selection and genomewide association study activi-ties were supported by BBSRC grant BBK01711X1 theBBSRC strategic programme grant on Energy Grasses ampBio‐refining BBSEW10963A01 The UK‐led willow RampDwork reported here was supported by BBSRC (BBSEC00005199 BBSEC00005201 BBG0162161 BBE0068331 BBG00580X1 and BBSEC000I0410) Defra(NF0424 NF0426) and the Department of Trade and Indus-try (DTI) (BW6005990000) IT The Brain Gain Program(Rientro dei cervelli) of the Italian Ministry of EducationUniversity and Research supports Antoine Harfouche USContributions by Gerald Tuskan to this manuscript were sup-ported by the Center for Bioenergy Innovation a US
Department of Energy Bioenergy Research Center supportedby the Office of Biological and Environmental Research inthe DOE Office of Science under contract number DE‐AC05‐00OR22725 Willow breeding efforts at CornellUniversity have been supported by grants from the USDepartment of Agriculture National Institute of Food andAgriculture Contributions by the University of Illinois weresupported primarily by the DOE Office of Science Office ofBiological and Environmental Research (BER) grant nosDE‐SC0006634 DE‐SC0012379 and DE‐SC0018420 (Cen-ter for Advanced Bioenergy and Bioproducts Innovation)and the Energy Biosciences Institute EU We would like tofurther acknowledge contributions from the EU projectsldquoOPTIMISCrdquo FP7‐289159 on miscanthus and ldquoWATBIOrdquoFP7‐311929 on poplar and miscanthus as well as ldquoGRACErdquoH2020‐EU326 Biobased Industries Joint Technology Ini-tiative (BBI‐JTI) Project ID 745012 on miscanthus
CONFLICT OF INTEREST
The authors declare that progress reported in this paperwhich includes input from industrial partners is not biasedby their business interests
ORCID
John Clifton‐Brown httporcidorg0000-0001-6477-5452Antoine Harfouche httporcidorg0000-0003-3998-0694Lawrence B Smart httporcidorg0000-0002-7812-7736Danny Awty‐Carroll httporcidorg0000-0001-5855-0775VasileBotnari httpsorcidorg0000-0002-0470-0384Lindsay V Clark httporcidorg0000-0002-3881-9252SalvatoreCosentino httporcidorg0000-0001-7076-8777Lin SHuang httporcidorg0000-0003-3079-326XJon P McCalmont httporcidorg0000-0002-5978-9574Paul Robson httporcidorg0000-0003-1841-3594AnatoliiSandu httporcidorg0000-0002-7900-448XDaniloScordia httporcidorg0000-0002-3822-788XRezaShafiei httporcidorg0000-0003-0891-0730Gail Taylor httporcidorg0000-0001-8470-6390IrisLewandowski httporcidorg0000-0002-0388-4521
REFERENCES
Alburquerque N Baldacci‐Cresp F Baucher M Casacuberta JM Collonnier C ElJaziri M hellip Burgos L (2016) New trans-formation technologies for trees In C Vettori F Gallardo H
CLIFTON‐BROWN ET AL | 27
Haumlggman V Kazana F Migliacci G Pilateamp M Fladung(Eds) Biosafety of forest transgenic trees (pp 31ndash66) DordrechtNetherlands Springer
Argus G W (2007) Salix (Salicaceae) distribution maps and a syn-opsis of their classification in North America north of MexicoHarvard Papers in Botany 12 335ndash368 httpsdoiorg1031001043-4534(2007)12[335SSDMAA]20CO2
Atienza S G Ramirez M C amp Martin A (2003) Mapping‐QTLscontrolling flowering date in Miscanthus sinensis Anderss CerealResearch Communications 31 265ndash271
Atienza S G Satovic Z Petersen K K Dolstra O amp Martin A(2003) Identification of QTLs influencing agronomic traits inMiscanthus sinensis Anderss I Total height flag‐leaf height andstem diameter Theoretical and Applied Genetics 107 123ndash129
Atienza S G Satovic Z Petersen K K Dolstra O amp Martin A(2003) Influencing combustion quality in Miscanthus sinensisAnderss Identification of QTLs for calcium phosphorus and sul-phur content Plant Breeding 122 141ndash145
Bdeir R Muchero W Yordanov Y Tuskan G A Busov V ampGailing O (2017) Quantitative trait locus mapping of Populusbark features and stem diameter BMC Plant Biology 17 224httpsdoiorg101186s12870-017-1166-4
Beard T R Ford G S Koutsky T M amp Spiwak L J (2009) AValley of Death in the innovation sequence An economic investi-gation Research Evaluation 18 343ndash356 httpsdoiorg103152095820209X481057
Berguson W McMahon B amp Riemenschneider D (2017) Addi-tive and non‐additive genetic variances for tree growth in severalhybrid poplar populations and implications regarding breedingstrategy Silvae Genetica 66(1) 33ndash39 httpsdoiorg101515sg-2017-0005
Berlin S Trybush S O Fogelqvist J Gyllenstrand N Halling-baumlck H R Aringhman I hellip Hanley S J (2014) Genetic diversitypopulation structure and phenotypic variation in European Salixviminalis L (Salicaceae) Tree Genetics amp Genomes 10 1595ndash1610 httpsdoiorg101007s11295-014-0782-5
Bernardo R amp Yu J (2007) Prospects for genomewide selectionfor quantitative traits in maize Crop Science 47 1082ndash1090httpsdoiorg102135cropsci2006110690
Biswal A K Atmodjo M A Li M Baxter H L Yoo C GPu Y hellip Mohnen D (2018) Sugar release and growth of bio-fuel crops are improved by downregulation of pectin biosynthesisNature Biotechnology 36 249ndash257
Bmu (2009) National biomass action plan for Germany (p 17) Bun-desministerium fuumlr Umwelt Naturschutz und ReaktorsicherheitBundesministerium fuumlr Ernaumlhrung (BMU) amp Landwirtschaft undVerbraucherschutz (BMELV) 11055 Berlin | Germany Retrievedfrom httpwwwbmeldeSharedDocsDownloadsENPublicationsBiomassActionPlanpdf__blob=publicationFile
Brummer E C (1999) Capturing heterosis in forage crop cultivardevelopment Crop Science 39 943ndash954 httpsdoiorg102135cropsci19990011183X003900040001x
Budsberg E Crawford J T Morgan H Chin W S Bura R ampGustafson R (2016) Hydrocarbon bio‐jet fuel from bioconver-sion of poplar biomass Life cycle assessment Biotechnology forBiofuels 9 170 httpsdoiorg101186s13068-016-0582-2
Burris K P Dlugosz E M Collins A G Stewart C N amp Lena-ghan S C (2016) Development of a rapid low‐cost protoplasttransfection system for switchgrass (Panicum virgatum L) Plant
Cell Reports 35 693ndash704 httpsdoiorg101007s00299-015-1913-7
Casler M (2012) Switchgrass breeding genetics and genomics InA Monti (Ed) Switchgrass (pp 29ndash54) London UK Springer
Casler M Mitchell R amp Vogel K (2012) Switchgrass In CKole C P Joshi amp D R Shonnard (Eds) Handbook of bioen-ergy crop plants Vol 2 New York NY Taylor amp Francis
Casler M D amp Ramstein G P (2018) Breeding for biomass yieldin switchgrass using surrogate measures of yield BioEnergyResearch 11 6ndash12
Casler M D Sosa S Hoffman L Mayton H Ernst C Adler PR hellip Bonos S A (2017) Biomass Yield of Switchgrass Culti-vars under High‐ versus Low‐Input Conditions Crop Science 57821ndash832 httpsdoiorg102135cropsci2016080698
Casler M D amp Vogel K P (2014) Selection for biomass yield inupland lowland and hybrid switchgrass Crop Science 54 626ndash636 httpsdoiorg102135cropsci2013040239
Casler M D Vogel K P amp Harrison M (2015) Switchgrassgermplasm resources Crop Science 55 2463ndash2478 httpsdoiorg102135cropsci2015020076
Casler M D Vogel K P Lee D K Mitchell R B Adler P RSulc R M hellip Moore K J (2018) 30 Years of progress towardincreasing biomass yield of switchgrass and big bluestem CropScience 58 1242
Clark L V Brummer J E Głowacka K Hall M C Heo K PengJ hellip Sacks E J (2014) A footprint of past climate change on thediversity and population structure of Miscanthus sinensis Annals ofBotany 114 97ndash107 httpsdoiorg101093aobmcu084
Clark L V Dzyubenko E Dzyubenko N Bagmet L SabitovA Chebukin P hellip Sacks E J (2016) Ecological characteristicsand in situ genetic associations for yield‐component traits of wildMiscanthus from eastern Russia Annals of Botany 118 941ndash955
Clark L V Jin X Petersen K K Anzoua K G Bagmet LChebukin P hellip Sacks E J(2018) Population structure of Mis-canthus sacchariflorus reveals two major polyploidization eventstetraploid‐mediated unidirectional introgression from diploid Msinensis and diversity centred around the Yellow Sea Annals ofBotany 20 1ndash18 httpsdoiorg101093aobmcy1161
Clark L V Stewart J R Nishiwaki A Toma Y Kjeldsen J BJoslashrgensen U hellip Sacks E J (2015) Genetic structure ofMiscanthus sinensis and Miscanthus sacchariflorus in Japan indi-cates a gradient of bidirectional but asymmetric introgressionJournal of Experimental Botany 66 4213ndash4225
Clifton‐Brown J Hastings A Mos M McCalmont J P Ashman CAwty‐Carroll DhellipCracroft‐EleyW (2017) Progress in upscalingMis-canthus biomass production for the European bio‐economy with seed‐based hybridsGlobal Change Biology Bioenergy 9 6ndash17
Clifton‐Brown J Schwarz K U amp Hastings A (2015) History ofthe development of Miscanthus as a bioenergy crop From smallbeginnings to potential realisation Biology and Environment‐Pro-ceedings of the Royal Irish Academy 115B 45ndash57
Clifton‐Brown J Senior H Purdy S Horsnell R Lankamp BMuumlennekhoff A-K hellip Bentley A R (2018) Investigating thepotential of novel non‐woven fabrics to increase pollination effi-ciency in plant breeding PloS One (Accepted)
Crawford J T Shan CW Budsberg E Morgan H Bura R ampGus-tafson R (2016) Hydrocarbon bio‐jet fuel from bioconversion ofpoplar biomass Techno‐economic assessment Biotechnology forBiofuels 9 141 httpsdoiorg101186s13068-016-0545-7
28 | CLIFTON‐BROWN ET AL
Dalton S (2013) Biotechnology of Miscanthus In S M Jain amp SD Gupta (Eds) Biotechnology of neglected and underutilizedcrops (pp 243ndash294) Dordrecht Netherlands Springer
Davey C L Robson P Hawkins S Farrar K Clifton‐Brown J CDonnison I S amp Slavov G T (2017) Genetic relationshipsbetween spring emergence canopy phenology and biomass yieldincrease the accuracy of genomic prediction in Miscanthus Journalof Experimental Botany 68 5093ndash5102 httpsdoiorg101093jxberx339
David X amp Anderson X (2002) Aspen and larch genetics cooper-ative annual report 13 St Paul MN Department of ForestResources University of Minnesota
Deblock M (1990) Factors influencing the tissue‐culture and theAgrobacterium‐Tumefaciens‐mediated transformation of hybridaspen and poplar clones Plant Physiology 93 1110ndash1116httpsdoiorg101104pp9331110
Defra (2002) The role of future public research investment in thegenetic improvement of UK‐grown crops (p 220) LondonRetrieved from sciencesearchdefragovukDefaultaspxMenu=MenuampModule=MoreampLocation=NoneampCompleted=220ampProjectID=10412
Dewoody J Trewin H amp Taylor G (2015) Genetic and morpho-logical differentiation in Populus nigra L Isolation by coloniza-tion or isolation by adaptation Molecular Ecology 24 2641ndash2655
Dong H Liu S Clark L V Sharma S Gifford J M Juvik JA hellip Sacks E J (2018) Genetic mapping of biomass yield inthree interconnected Miscanthus populations Global Change Biol-ogy Bioenergy 10 165ndash185
Dukes (2017) Digest of UK Energy Statistics (DUKES) (p 264) Lon-don UK Department for Business Energy amp Industrial StrategyRetrieved from wwwgovukgovernmentcollectionsdigest-of-uk-energy-statistics-dukes
Eckenwalder J (1996) Systematics and evolution of Populus In RStettler H Bradshaw J Heilman amp T Hinckley (Eds) Biologyof Populus and its implications for management and conservation(pp 7‐32) Ottawa ON National Research Council of Canada
Elshire R J Glaubitz J C Sun Q Poland J A Kawamoto KBuckler E S amp Mitchell S E (2011) A robust simple geno-typing‐by‐sequencing (GBS) approach for high diversity speciesPloS One 6 e19379
Evans H (2016) Bioenergy crops in the UK Case studies of suc-cessful whole farm integration (p 23) Loughborough UKEnergy Technologies Institute Retrieved from httpswwweticouklibrarybioenergy-crops-in-the-uk-case-studies-on-successful-whole-farm-integration-evidence-pack
Evans H (2017) Increasing UK biomass production through moreproductive use of land (p 7) Loughborough UK Energy Tech-nologies Institute Retrieved from httpswwweticouklibraryan-eti-perspective-increasing-uk-biomass-production-through-more-productive-use-of-land
Evans J Sanciangco M D Lau K H Crisovan E Barry KDaum C hellip Buell C R (2018) Extensive genetic diversity ispresent within North American switchgrass germplasm The PlantGenome 11 1ndash16 httpsdoiorg103835plantgenome2017060055
Evans L M Slavov G T Rodgers‐Melnick E Martin J RanjanP Muchero W hellip DiFazio S P (2014) Population genomicsof Populus trichocarpa identifies signatures of selection and
adaptive trait associations Nature Genetics 46 1089ndash1096httpsdoiorg101038ng3075
Fabio E S Volk T A Miller R O Serapiglia M J Gauch HG Van Rees K C J hellip Smart L B (2017) Genotypetimes envi-ronment interaction analysis of North American shrub willowyield trials confirms superior performance of triploid hybrids Glo-bal Change Biology Bioenergy 9 445ndash459 httpsdoiorg101111gcbb12344
Fahrenkrog A M Neves L G Resende M F R Vazquez A Ide los Campos G Dervinis C hellip Kirst M (2017) Genome‐wide association study reveals putative regulators of bioenergytraits in Populus deltoides New Phytologist 213 799ndash811
Fan D Liu T T Li C F Jiao B Li S Hou Y S amp Luo KM (2015) Efficient CRISPRCas9‐mediated targeted mutagenesisin Populus in the first generation Scientific Reports 5 12217
Feng X P Lourgant K Castric V Saumitou-Laprade P ZhengB S Jiang D M amp Brancourt-Hulmel M (2014) The discov-ery of natural accessions related to Miscanthus x giganteus usingchloroplast DNA Crop Science 54 1645ndash1655
Fillatti J J Sellmer J Mccown B Haissig B amp Comai L(1987) Agrobacterium-mediated transformation and regenerationof Populus Molecular amp General Genetics 206 192ndash199
Fu Y B (2015) Understanding crop genetic diversity under modernplant breeding Theoretical and Applied Genetics 128 2131ndash2142 httpsdoiorg101007s00122-015-2585-y
Fu C Mielenz J R Xiao X Ge Y Hamilton C Y RodriguezM hellip Wang Z‐Y (2011) Genetic manipulation of ligninreduces recalcitrance and improves ethanol production fromswitchgrass Proceedings of the National Academy of Sciences ofthe United States of America 108 3803ndash3808 httpsdoiorg101073pnas1100310108
Gao C (2018) The future of CRISPR technologies in agricultureNature Reviews Molecular Cell Biology 19(5) 275ndash276
Gegas V Gay A Camargo A amp Doonan J (2014) Challengesof Crop Phenomics in the Post-genomic Era In J Hancock (Ed)Phenomics (pp 142ndash171) Boca Raton FL CRC Press
Geraldes A DiFazio S P Slavov G T Ranjan P Muchero WHannemann J hellip Tuskan G A (2013) A 34K SNP genotypingarray for Populus trichocarpa Design application to the study ofnatural populations and transferability to other Populus speciesMolecular Ecology Resources 13 306ndash323
Gifford J M Chae W B Swaminathan K Moose S P amp JuvikJ A (2015) Mapping the genome of Miscanthus sinensis forQTL associated with biomass productivity Global Change Biol-ogy Bioenergy 7 797ndash810
Goggin F L Lorence A amp Topp C N (2015) Applying high‐throughput phenotyping to plant‐insect interactions Picturingmore resistant crops Current Opinion in Insect Science 9 69ndash76httpsdoiorg101016jcois201503002
Grabowski P P Evans J Daum C Deshpande S Barry K WKennedy M hellip Casler M D (2017) Genome‐wide associationswith flowering time in switchgrass using exome‐capture sequenc-ing data New Phytologist 213 154ndash169 httpsdoiorg101111nph14101
Greef J M amp Deuter M (1993) Syntaxonomy of Miscanthus xgiganteus GREEF et DEU Angewandte Botanik 67 87ndash90
Guerra F P Richards J H Fiehn O Famula R Stanton B JShuren R hellip Neale D B (2016) Analysis of the genetic varia-tion in growth ecophysiology and chemical and metabolomic
CLIFTON‐BROWN ET AL | 29
composition of wood of Populus trichocarpa provenances TreeGenetics amp Genomes 12 6 httpsdoiorg101007s11295-015-0965-8
Guo H P Shao R X Hong C T Hu H K Zheng B S ampZhang Q X (2013) Rapid in vitro propagation of bioenergy cropMiscanthus sacchariflorus Applied Mechanics and Materials 260181ndash186
Hallingbaumlck H R Fogelqvist J Powers S J Turrion‐Gomez JRossiter R Amey J hellip Roumlnnberg‐Waumlstljung A‐C (2016)Association mapping in Salix viminalis L (Salicaceae)ndashIdentifica-tion of candidate genes associated with growth and phenologyGlobal Change Biology Bioenergy 8 670ndash685
Hanley S J amp Karp A (2014) Genetic strategies for dissectingcomplex traits in biomass willows (Salix spp) Tree Physiology34 1167ndash1180 httpsdoiorg101093treephystpt089
Harfouche A Meilan R amp Altman A (2011) Tree genetic engi-neering and applications to sustainable forestry and biomass pro-duction Trends in Biotechnology 29 9ndash17 httpsdoiorg101016jtibtech201009003
Harfouche A Meilan R amp Altman A (2014) Molecular and phys-iological responses to abiotic stress in forest trees and their rele-vance to tree improvement Tree Physiology 34 1181ndash1198httpsdoiorg101093treephystpu012
Harfouche A Meilan R Kirst M Morgante M Boerjan WSabatti M amp Mugnozza G S (2012) Accelerating the domesti-cation of forest trees in a changing world Trends in PlantScience 17 64ndash72 httpsdoiorg101016jtplants201111005
Hastings A Clifton‐Brown J Wattenbach M Mitchell C PStampfl P amp Smith P (2009) Future energy potential of Mis-canthus in Europe Global Change Biology Bioenergy 1 180ndash196
Hastings A Mos M Yesufu J AMccalmont J Schwarz KShafei RhellipClifton-Brown J (2017) Economic and environmen-tal assessment of seed and rhizome propagated Miscanthus in theUK Frontiers in Plant Science 8 16 httpsdoiorg103389fpls201701058
Heacuteder M (2017) From NASA to EU The evolution of the TRLscale in Public Sector Innovation The Innovation Journal 22 1ndash23 httpsdoiorg103389fpls201701058
Heffner E L Sorrells M E amp Jannink J L (2009) Genomicselection for crop improvement Crop Science 49 1ndash12 httpsdoiorg102135cropsci2008080512
Hodkinson T R Klaas M Jones M B Prickett R amp Barth S(2015) Miscanthus A case study for the utilization of naturalgenetic variation Plant Genetic Resources‐Characterization andUtilization 13 219ndash237 httpsdoiorg101017S147926211400094X
Hodkinson T R amp Renvoize S (2001) Nomenclature of Miscant-hus xgiganteus (Poaceae) Kew Bulletin 56 759ndash760 httpsdoiorg1023074117709
Houle D Govindaraju D R amp Omholt S (2010) Phenomics Thenext challenge Nature Reviews Genetics 11 855ndash866 httpsdoiorg101038nrg2897
Huang X H Wei X H Sang T Zhao Q Feng Q Zhao Yhellip Han B (2010) Genome‐wide association studies of 14 agro-nomic traits in rice landraces Nature Genetics 42 961ndashU976
Hwang O‐J Cho M‐A Han Y‐J Kim Y‐M Lim S‐H KimD‐S hellip Kim J‐I (2014) Agrobacterium‐mediated genetic trans-formation of Miscanthus sinensis Plant Cell Tissue and OrganCulture (PCTOC) 117 51ndash63
Hwang O‐J Lim S‐H Han Y‐J Shin A‐Y Kim D‐S ampKim J‐I (2014) Phenotypic characterization of transgenic Mis-canthus sinensis plants overexpressing Arabidopsis phytochromeB International Journal of Photoenergy 2014501016
Ilstedt B (1996) Genetics and performance of Belgian poplar clonestested in Sweden International Journal of Forest Genetics 3183ndash195
Ingvarsson P K Garcia M V Hall D Luquez V amp Jansson S(2006) Clinal variation in phyB2 a candidate gene for day‐length‐induced growth cessation and bud set across a latitudinalgradient in European aspen (Populus tremula) Genetics 1721845ndash1853
Ingvarsson P K Garcia M V Luquez V Hall D amp Jansson S(2008) Nucleotide polymorphism and phenotypic associationswithin and around the phytochrome B2 locus in European aspen(Populus tremula Salicaceae) Genetics 178 2217ndash2226 httpsdoiorg101534genetics107082354
Jannink J‐L Lorenz A J amp Iwata H (2010) Genomic selectionin plant breeding From theory to practice Briefings in FunctionalGenomics 9 166ndash177 httpsdoiorg101093bfgpelq001
Jiang J Guan Y Mccormick S Juvik J Lubberstedt T amp FeiS‐Z (2017) Gametophytic self‐incompatibility is operative inMiscanthus sinensis (Poaceae) and is affected by pistil age CropScience 57 1948ndash1956
Jones H D (2015a) Future of breeding by genome editing is in thehands of regulators GM Crops amp Food 6 223ndash232
Jones H D (2015b) Regulatory uncertainty over genome editingNature Plants 1 14011
Kalinina O Nunn C Sanderson R Hastings A F S van der Weijde TOumlzguumlven MhellipClifton‐Brown J C (2017) ExtendingMiscanthus cul-tivation with novel germplasm at six contrasting sites Frontiers in PlantScience 8563 httpsdoiorg103389fpls201700563
Karp A Hanley S J Trybush S O Macalpine W Pei M ampShield I (2011) Genetic improvement of willow for bioenergyand biofuels free access Journal of Integrative Plant Biology 53151ndash165 httpsdoiorg101111j1744-7909201001015x
Kersten B Faivre Rampant P Mader M Le Paslier M‐C Bou-non R Berard A hellip Fladung M (2016) Genome sequences ofPopulus tremula chloroplast and mitochondrion Implications forholistic poplar breeding PloS One 11 e0147209 httpsdoiorg101371journalpone0147209
Kiesel A Nunn C Iqbal Y Van der Weijde T Wagner MOumlzguumlven M hellip Lewandowski I (2017) Site‐specific manage-ment of Miscanthus genotypes for combustion and anaerobicdigestion A comparison of energy yields Frontiers in PlantScience 8 347 httpsdoiorg103389fpls201700347
Kim H Kim S T Ryu J Kang B C Kim J S amp Kim S G(2017) CRISPRCpf1‐mediated DNA‐free plant genome editingNature Communications 8 14406 httpsdoiorg101038ncomms14406
King Z R Bray A L Lafayette P R amp Parrott W A (2014)Biolistic transformation of elite genotypes of switchgrass (Pan-icum virgatum L) Plant Cell Reports 33 313ndash322 httpsdoiorg101007s00299-013-1531-1
Kopp R F Maynard C A De Niella P R Smart L B amp Abra-hamson L P (2002) Collection and storage of pollen from Salix(Salicaceae) American Journal of Botany 89 248ndash252
Krzyżak J Pogrzeba M Rusinowski S Clifton‐Brown J McCal-mont J P Kiesel A hellip Mos M (2017) Heavy metal uptake
30 | CLIFTON‐BROWN ET AL
by novel miscanthus seed‐based hybrids cultivated in heavy metalcontaminated soil Civil and Environmental Engineering Reports26 121ndash132 httpsdoiorg101515ceer-2017-0040
Kuzovkina Y A amp Quigley M F (2005) Willows beyond wet-lands Uses of Salix L species for environmental projects WaterAir and Soil Pollution 162 183ndash204 httpsdoiorg101007s11270-005-6272-5
Lazarus W Headlee W L amp Zalesny R S (2015) Impacts of sup-plyshed‐level differences in productivity and land costs on the eco-nomics of hybrid poplar production in Minnesota USA BioEnergyResearch 8 231ndash248 httpsdoiorg101007s12155-014-9520-y
Lewandowski I (2015) Securing a sustainable biomass supply in agrowing bioeconomy Global Food Security 6 34ndash42 httpsdoiorg101016jgfs201510001
Lewandowski I (2016) The role of perennial biomass crops in agrowing bioeconomy In S Barth D Murphy‐Bokern O Kalin-ina G Taylor amp M Jones (Eds) Perennial biomass crops for aresource‐constrained world (pp 3ndash13) Cham SwitzerlandSpringer Switzerland AG
Li J L Meilan R Ma C Barish M amp Strauss S H (2008)Stability of herbicide resistance over 8 years of coppice in field‐grown genetically engineered poplars Western Journal of AppliedForestry 23 89ndash93
Li R Y amp Qu R D (2011) High throughput Agrobacterium‐medi-ated switchgrass transformation Biomass amp Bioenergy 35 1046ndash1054 httpsdoiorg101016jbiombioe201011025
Lindegaard K N amp Barker J H A (1997) Breeding willows forbiomass Aspects of Applied Biology 49 155ndash162
Linde‐Laursen I B (1993) Cytogenetic analysis of MiscanthusGiganteus an interspecific hybrid Hereditas 119 297ndash300httpsdoiorg101111j1601-5223199300297x
Liu Y Merrick P Zhang Z Z Ji C H Yang B amp Fei S Z(2018) Targeted mutagenesis in tetraploid switchgrass (Panicumvirgatum L) using CRISPRCas9 Plant Biotechnology Journal16 381ndash393
Liu L L Wu Y Q Wang Y W amp Samuels T (2012) A high‐density simple sequence repeat‐based genetic linkage map ofswitchgrass G3 (Bethesda Md) 2 357ndash370
Lovett A Suumlnnenberg G amp Dockerty T (2014) The availabilityof land for perennial energy crops in Great Britain GlobalChange Biology Bioenergy 6 99ndash107 httpsdoiorg101111gcbb12147
Lozano‐Juste J amp Cutler S R (2014) Plant genome engineering infull bloom Trends in Plant Science 19 284ndash287 httpsdoiorg101016jtplants201402014
Lu F Lipka A E Glaubitz J Elshire R Cherney J H CaslerM D hellip Costich D E (2013) Switchgrass Genomic DiversityPloidy and Evolution Novel Insights from a Network‐Based SNPDiscovery Protocol Plos Genetics 9 e1003215
Ludovisi R Tauro F Salvati R Khoury S Mugnozza G S ampHarfouche A (2017) UAV‐based thermal imaging for high‐throughput field phenotyping of black poplar response to droughtFrontiers in Plant Science 81681
Macalpine W Shield I amp Karp A (2010) Seed to near market vari-ety the BEGIN willow breeding pipeline 2003ndash2010 and beyond InBioten conference proceedings Birmingham pp 21ndash23
Macalpine W J Shield I F Trybush S O Hayes C M ampKarp A (2008) Overcoming barriers to crossing in willow (Salixspp) breeding Aspects of Applied Biology 90 173ndash180
Mahfouz M M (2017) The efficient tool CRISPR‐Cpf1 NaturePlants 3 17028
Malinowska M Donnison I S amp Robson P R (2017) Phenomicsanalysis of drought responses in Miscanthus collected from differ-ent geographical locations Global Change Biology Bioenergy 978ndash91
Maroder H L Prego I A Facciuto G R amp Maldonado S B(2000) Storage behaviour of Salix alba and Salix matsudanaseeds Annals of Botany 86 1017ndash1021 httpsdoiorg101006anbo20001265
Martinez‐Reyna J amp Vogel K (2002) Incompatibility systems inswitchgrass Crop Science 42 1800ndash1805 httpsdoiorg102135cropsci20021800
Maurya J P amp Bhalerao R P (2017) Photoperiod‐and tempera-ture‐mediated control of growth cessation and dormancy in treesA molecular perspective Annals of Botany 120 351ndash360httpsdoiorg101093aobmcx061
McCalmont J P Hastings A Mcnamara N P Richter G MRobson P Donnison I S amp Clifton‐Brown J (2017) Environ-mental costs and benefits of growing Miscanthus for bioenergy inthe UK Global Change Biology Bioenergy 9 489ndash507
McCracken A R amp Dawson W M (1997) Using mixtures of wil-low clones as a means of controlling rust disease Aspects ofApplied Biology 49 97ndash103
McKown A D Klaacutepště J Guy R D Geraldes A Porth I Hanne-mann JhellipDouglas C J (2014) Genome‐wide association implicatesnumerous genes underlying ecological trait variation in natural popula-tions ofPopulus trichocarpaNew Phytologist 203 535ndash553
Merrick P amp Fei S Z (2015) Plant regeneration and genetic transforma-tion in switchgrass ndash A review Journal of Integrative Agriculture 14483ndash493 httpsdoiorg101016S2095-3119(14)60921-7
Moreno‐Mateos M A Fernandez J P Rouet R Lane M AVejnar C E Mis E hellip Giraldez A J (2017) CRISPR‐Cpf1mediates efficient homology‐directed repair and temperature‐con-trolled genome editing Nature Communications 8 2024 httpsdoiorg101038s41467-017-01836-2
Mosseler A (1990) Hybrid performance and species crossabilityrelationships in willows (Salix) Canadian Journal of Botany 682329ndash2338
Muchero W Guo J DiFazio S P Chen J‐G Ranjan P Slavov GT hellip Tuskan G A (2015) High‐resolution genetic mapping ofallelic variants associated with cell wall chemistry in Populus BMCGenomics 16 24 httpsdoiorg101186s12864-015-1215-z
Neale D B amp Kremer A (2011) Forest tree genomics Growingresources and applications Nature Reviews Genetics 12 111ndash122 httpsdoiorg101038nrg2931
Nunn C Hastings A F S J Kalinina O Oumlzguumlven M SchuumlleH Tarakanov I G hellip Clifton‐Brown J C (2017) Environmen-tal influences on the growing season duration and ripening ofdiverse Miscanthus germplasm grown in six countries Frontiersin Plant Science 8 907 httpsdoiorg103389fpls201700907
Ogawa Y Honda M Kondo Y amp Hara‐Nishimura I (2016) Anefficient Agrobacterium‐mediated transformation method forswitchgrass genotypes using Type I callus Plant Biotechnology33 19ndash26
Ogawa Y Shirakawa M Koumoto Y Honda M Asami Y KondoY ampHara‐Nishimura I (2014) A simple and reliable multi‐gene trans-formation method for switchgrass Plant Cell Reports 33 1161ndash1172httpsdoiorg101007s00299-014-1605-8
CLIFTON‐BROWN ET AL | 31
Okada M Lanzatella C Saha M C Bouton J Wu R L amp Tobias CM (2010) Complete switchgrass genetic maps reveal subgenomecollinearity preferential pairing and multilocus interactions Genetics185 745ndash760 httpsdoiorg101534genetics110113910
Palomo‐Riacuteos E Macalpine W Shield I Amey J Karaoğlu C WestJ hellip Jones H D (2015) Efficient method for rapid multiplication ofclean and healthy willow clones via in vitro propagation with broadgenotype applicability Canadian Journal of Forest Research 451662ndash1667 httpsdoiorg101139cjfr-2015-0055
Pilate G Allona I Boerjan W Deacutejardin A Fladung M Gal-lardo F hellip Halpin C (2016) Lessons from 25 years of GM treefield trials in Europe and prospects for the future In C Vettori FGallardo H Haumlggman V Kazana F Migliacci G Pilateamp MFladung (Eds) Biosafety of forest transgenic trees (pp 67ndash100)Dordrecht Netherlands Springer Switzerland AG
Pinosio S Giacomello S Faivre-Rampant P Taylor G Jorge VLe Paslier M C amp Marroni F (2016) Characterization of thepoplar pan-genome by genome-wide identification of structuralvariation Molecular Biology and Evolution 33 2706ndash2719
Porth I Klaacutepště J Skyba O Friedmann M C Hannemann JEhlting J hellip Douglas C J (2013) Network analysis reveals therelationship among wood properties gene expression levels andgenotypes of natural Populus trichocarpa accessions New Phytol-ogist 200 727ndash742
Pugesgaard S Schelde K Larsen S U Laeligrke P E amp JoslashrgensenU (2015) Comparing annual and perennial crops for bioenergyproductionndashinfluence on nitrate leaching and energy balance Glo-bal Change Biology Bioenergy 7 1136ndash1149 httpsdoiorg101111gcbb12215
Quinn L D Allen D J amp Stewart J R (2010) Invasivenesspotential of Miscanthus sinensis Implications for bioenergy pro-duction in the United States Global Change Biology Bioenergy2 310ndash320 httpsdoiorg101111j1757-1707201001062x
Rae A M Robinson K M Street N R amp Taylor G (2004)Morphological and physiological traits influencing biomass pro-ductivity in short‐rotation coppice poplar Canadian Journal ofForest Research‐Revue Canadienne De Recherche Forestiere 341488ndash1498 httpsdoiorg101139x04-033
Rambaud C Arnoult S Bluteau A Mansard M C Blassiau Camp Brancourt‐Hulmel M (2013) Shoot organogenesis in threeMiscanthus species and evaluation for genetic uniformity usingAFLP analysis Plant Cell Tissue and Organ Culture (PCTOC)113 437ndash448 httpsdoiorg101007s11240-012-0284-9
Ramstein G P Evans J Kaeppler S M Mitchell R B Vogel K PBuell C R amp Casler M D (2016) Accuracy of genomic prediction inswitchgrass (Panicum virgatum L) improved by accounting for linkagedisequilibriumG3 (Bethesda Md) 6 1049ndash1062
Rayburn A L Crawford J Rayburn C M amp Juvik J A (2009)Genome size of three Miscanthus species Plant Molecular Biol-ogy Reporter 27 184 httpsdoiorg101007s11105-008-0070-3
Richardson J Isebrands J amp Ball J (2014) Ecology and physiol-ogy of poplars and willows In J Isebrands amp J Richardson(Eds) Poplars and willows Trees for society and the environ-ment (pp 92‐123) Boston MA and Rome CABI and FAO
Robison T L Rousseau R J amp Zhang J (2006) Biomass produc-tivity improvement for eastern cottonwood Biomass amp Bioenergy30 735ndash739 httpsdoiorg101016jbiombioe200601012
Robson P R Farrar K Gay A P Jensen E F Clifton‐Brown JC amp Donnison I S (2013) Variation in canopy duration in the
perennial biofuel crop Miscanthus reveals complex associationswith yield Journal of Experimental Botany 64 2373ndash2383
Robson P Jensen E Hawkins S White S R Kenobi K Clif-ton‐Brown J hellip Farrar K (2013) Accelerating the domestica-tion of a bioenergy crop Identifying and modelling morphologicaltargets for sustainable yield increase in Miscanthus Journal ofExperimental Botany 64 4143ndash4155
Sanderson M A Adler P R Boateng A A Casler M D ampSarath G (2006) Switchgrass as a biofuels feedstock in theUSA Canadian Journal of Plant Science 86 1315ndash1325httpsdoiorg104141P06-136
Scarlat N Dallemand J F Monforti‐Ferrario F amp Nita V(2015) The role of biomass and bioenergy in a future bioecon-omy Policies and facts Environmental Development 15 3ndash34httpsdoiorg101016jenvdev201503006
Serapiglia M J Gouker F E amp Smart L B (2014) Early selec-tion of novel triploid hybrids of shrub willow with improved bio-mass yield relative to diploids BMC Plant Biology 14 74httpsdoiorg1011861471-2229-14-74
Serba D Wu L Daverdin G Bahri B A Wang X Kilian Ahellip Devos K M (2013) Linkage maps of lowland and upland tet-raploid switchgrass ecotypes BioEnergy Research 6 953ndash965httpsdoiorg101007s12155-013-9315-6
Shakoor N Lee S amp Mockler T C (2017) High throughput phe-notyping to accelerate crop breeding and monitoring of diseases inthe field Current Opinion in Plant Biology 38 184ndash192 httpsdoiorg101016jpbi201705006
Shield I Macalpine W Hanley S amp Karp A (2015) Breedingwillow for short rotation coppice energy cropping In Industrialcrops Breeding for BioEnergy and Bioproducts (pp 67ndash80) NewYork NY Springer
Sjodin A Street N R Sandberg G Gustafsson P amp Jansson S (2009)The Populus Genome Integrative Explorer (PopGenIE) A new resourcefor exploring the Populus genomeNewPhytologist 182 1013ndash1025
Slavov G amp Davey C (2017) Integrated genomic prediction inbioenergy crops In Abstracts from the International Conferenceon Developing biomass crops for future climates pp 34 24ndash27September 2017 Oriel College Oxford wwwwatbioeu
Slavov G Davey C Bosch M Robson P Donnison I ampMackay I (2018a) Genomic index selection provides a pragmaticframework for setting and refining multi‐objective breeding targetsin Miscanthus Annals of Botany (in press)
Slavov G T DiFazio S P Martin J Schackwitz W MucheroW Rodgers‐Melnick E hellip Tuskan G A (2012) Genome rese-quencing reveals multiscale geographic structure and extensivelinkage disequilibrium in the forest tree Populus trichocarpa NewPhytologist 196 713ndash725
Slavov G Davey C Robson P Donnison I amp Mackay I(2018b) Domestication of Miscanthus through genomic indexselection In Plant and Animal Genome Conference 13 January2018 San Diego CA Retrieved from httpspagconfexcompagxxvimeetingappcgiPaper30622
Slavov G T Nipper R Robson P Farrar K Allison G GBosch M hellip Jensen E (2013) Genome‐wide association studiesand prediction of 17 traits related to phenology biomass and cellwall composition in the energy grass Miscanthus sinensis NewPhytologist 201 1227ndash1239
Slavov G Robson P Jensen E Hodgson E Farrar K AllisonG hellip Donnison I (2014) Contrasting geographic patterns of
32 | CLIFTON‐BROWN ET AL
genetic variation for molecular markers vs phenotypic traits in theenergy grass Miscanthus sinensis Global Change Biology Bioen-ergy 5 562ndash571
Ślusarkiewicz‐Jarzina A Ponitka A Cerazy‐Waliszewska J Woj-ciechowicz M K Sobańska K Jeżowski S amp Pniewski T(2017) Effective and simple in vitro regeneration system of Mis-canthus sinensis Mtimes giganteus and M sacchariflorus for plant-ing and biotechnology purposes Biomass and Bioenergy 107219ndash226 httpsdoiorg101016jbiombioe201710012
Smart L amp Cameron K (2012) Shrub willow In C Kole S Joshiamp D Shonnard (Eds) Handbook of bioenergy crop plants (pp687ndash708) Boca Raton FL Taylor and Francis Group
Stanton B J (2014) The domestication and conservation of Populusand Salix genetic resources (Chapter 4) In Isebrands J ampRichardson J (Eds) Poplars and willows Trees for society andthe environment (pp 124ndash200) Rome Italy CAB InternationalFood and Agricultural Organization of the United Nations
Stanton B J Neale D B amp Li S (2010) Populus breeding Fromthe classical to the genomic approach In S Jansson R Bha-lerao amp A T Groover (Eds) Genetics and genomics of popu-lus (pp 309ndash348) New York NY Springer
Stewart J R Toma Y Fernandez F G Nishiwaki A Yamada T ampBollero G (2009) The ecology and agronomy ofMiscanthus sinensisa species important to bioenergy crop development in its native range inJapan A reviewGlobal Change Biology Bioenergy 1 126ndash153
Stott K G (1992) Willows in the service of man Proceedings of the RoyalSociety of Edinburgh Section B Biological Sciences 98 169ndash182
Strauss S H Ma C Ault K amp Klocko A L (2016) Lessonsfrom two decades of field trials with genetically modified trees inthe USA Biology and regulatory compliance In C Vettori FGallardo H Haumlggman V Kazana F Migliacci G Pilate amp MFladung (Eds) Biosafety of forest transgenic trees (pp 101ndash124)Dordrecht Netherlands Springer
Suda Y amp Argus G W (1968) Chromosome numbers of some NorthAmerican Salix Brittonia 20 191ndash197 httpsdoiorg1023072805440
Tan B (2018) Genomic selection and genome‐wide association studies todissect quantitative traits in forest trees Umearing Sweden University
Tornqvist C Taylor M Jiang Y Evans J Buell C KaepplerS amp Casler M (2018) Quantitative trait locus mapping for flow-ering time in a lowland x upland switchgrass pseudo‐F2 popula-tion The Plant Genome 11 170093
Toacuteth G Hermann T Da Silva M amp Montanarella L (2016)Heavy metals in agricultural soils of the European Union withimplications for food safety Environment International 88 299ndash309 httpsdoiorg101016jenvint201512017
Tracy W Dawson J Moore V amp Fisch J (2016) Intellectual propertyrights and public plant breeding Recommendations and proceedings ofa conference on best practices for intellectual property protection of pub-lically developed plant germplasm (p 70) University of Wisconsin‐Madison Retrieved from httpshostcalswisceduagronomywp-contentuploadssites1620162005Proceedings-IPR-Finalpdf
Trybush S Jahodovaacute Š Macalpine W amp Karp A (2008) A geneticstudy of a Salix germplasm resource reveals new insights into rela-tionships among subgenera sections and species BioEnergyResearch 1 67ndash79 httpsdoiorg101007s12155-008-9007-9
Tsai C J (2013) Next‐generation sequencing for next‐generationbreeding and more New Phytologist 198 635ndash637 httpsdoiorg101111nph12245
Tuskan G A Difazio S Jansson S Bohlmann J Grigoriev I HellstenU hellip Rokhsar D (2006) The genome of black cottonwood Populustrichocarpa (Torr ampGray) Science 313 1596ndash1604
Tuskan G A amp Walsh M E (2001) Short‐rotation woody cropsystems atmospheric carbon dioxide and carbon management AUS case study The Forestry Chronicle 77 259ndash264 httpsdoiorg105558tfc77259-2
Van Den Broek R Treffers D‐J Meeusen M Van Wijk ANieuwlaar E amp Turkenburg W (2001) Green energy or organicfood A life‐cycle assessment comparing two uses of set‐asideland Journal of Industrial Ecology 5 65ndash87 httpsdoiorg101162108819801760049477
Van Der Schoot J Pospiskova M Vosman B amp Smulders M J M(2000) Development and characterization of microsatellite markers inblack poplar (Populus nigra L) Theoretical and Applied Genetics 101317ndash322 httpsdoiorg101007s001220051485
Van Der Weijde T Dolstra O Visser R G amp Trindade L M(2017a) Stability of cell wall composition and saccharificationefficiency in Miscanthus across diverse environments Frontiers inPlant Science 7 2004
Van der Weijde T Huxley L M Hawkins S Sembiring E HFarrar K Dolstra O hellip Trindade L M (2017) Impact ofdrought stress on growth and quality of miscanthus for biofuelproduction Global Change Biology Bioenergy 9 770ndash782
Van der Weijde T Kamei C L A Severing E I Torres A FGomez L D Dolstra O hellip Trindade L M (2017) Geneticcomplexity of miscanthus cell wall composition and biomass qual-ity for biofuels BMC Genomics 18 406
Van der Weijde T Kiesel A Iqbal Y Muylle H Dolstra O Visser RG FhellipTrindade LM (2017) Evaluation ofMiscanthus sinensis bio-mass quality as feedstock for conversion into different bioenergy prod-uctsGlobal Change Biology Bioenergy 9 176ndash190
Vanholme B Cesarino I Goeminne G Kim H Marroni F VanAcker R hellip Boerjan W (2013) Breeding with rare defectivealleles (BRDA) A natural Populus nigra HCT mutant with modi-fied lignin as a case study New Phytologist 198 765ndash776
Vogel K P Mitchell R B Casler M D amp Sarath G (2014)Registration of Liberty Switchgrass Journal of Plant Registra-tions 8 242ndash247 httpsdoiorg103198jpr2013120076crc
Wang X Yamada T Kong F‐J Abe Y Hoshino Y Sato Hhellip Yamada T (2011) Establishment of an efficient in vitro cul-ture and particle bombardment‐mediated transformation systems inMiscanthus sinensis Anderss a potential bioenergy crop GlobalChange Biology Bioenergy 3 322ndash332 httpsdoiorg101111j1757-1707201101090x
Watrud L S Lee E H Fairbrother A Burdick C Reichman JR Bollman M hellip Van de Water P K (2004) Evidence forlandscape‐level pollen‐mediated gene flow from genetically modi-fied creeping bentgrass with CP4 EPSPS as a marker Proceedingsof the National Academy of Sciences of the United States of Amer-ica 101 14533ndash14538
Weisgerber H (1993) Poplar breeding for the purpose of biomassproduction in short‐rotation periods in Germany ndash Problems andfirst findings Forestry Chronicle 69 727ndash729 httpsdoiorg105558tfc69727-6
Wheeler N C Steiner K C Schlarbaum S E amp Neale D B(2015) The evolution of forest genetics and tree improvementresearch in the United States Journal of Forestry 113 500ndash510httpsdoiorg105849jof14-120
CLIFTON‐BROWN ET AL | 33
Whittaker C Macalpine W J Yates N E amp Shield I (2016)Dry matter losses and methane emissions during wood chip stor-age The impact on full life cycle greenhouse gas savings of shortrotation coppice willow for heat BioEnergy Research 9 820ndash835 httpsdoiorg101007s12155-016-9728-0
Xi Q (2000) Investigation on the distribution and potential of giant grassesin China PhD thesis Goettingen Germany Cuvillier Verlag
Xi Q amp Jezowkski S (2004) Plant resources of Triarrhena andMiscanthus species in China and its meaning for Europe PlantBreeding and Seed Science 49 63ndash77
Xue S Kalinina O amp Lewandowski I (2015) Present and future optionsfor the improvement of Miscanthus propagation techniques Renewableand Sustainable Energy Reviews 49 1233ndash1246
Yin K Q Gao C X amp Qiu J L (2017) Progress and prospectsin plant genome editing Nature Plants 3 httpsdoiorg101038nplants2017107
YookM (2016)Genetic diversity and transcriptome analysis for salt toler-ance inMiscanthus PhD thesis Seoul National University Korea
Zaidi S S E A Mahfouz M M amp Mansoor S (2017) CRISPR‐Cpf1 A new tool for plant genome editing Trends in PlantScience 22 550ndash553 httpsdoiorg101016jtplants201705001
Zalesny R S Stanturf J A Gardiner E S Perdue J H YoungT M Coyle D R hellip Hass A (2016) Ecosystem services ofwoody crop production systems BioEnergy Research 9 465ndash491 httpsdoiorg101007s12155-016-9737-z
Zhang Q X Sun Y Hu H K Chen B Hong C T Guo H Phellip Zheng B S (2012) Micropropagation and plant regenerationfrom embryogenic callus of Miscanthus sinensis Vitro Cellular ampDevelopmental Biology‐Plant 48 50ndash57 httpsdoiorg101007s11627-011-9387-y
Zhang Y Zalapa J E Jakubowski A R Price D L AcharyaA Wei Y hellip Casler M D (2011) Post‐glacial evolution of
Panicum virgatum Centers of diversity and gene pools revealedby SSR markers and cpDNA sequences Genetica 139 933ndash948httpsdoiorg101007s10709-011-9597-6
Zhao H Wang B He J Yang J Pan L Sun D amp Peng J(2013) Genetic diversity and population structure of Miscanthussinensis germplasm in China PloS One 8 e75672
Zhou X H Jacobs T B Xue L J Harding S A amp Tsai C J(2015) Exploiting SNPs for biallelic CRISPR mutations in theoutcrossing woody perennial Populus reveals 4‐coumarate CoAligase specificity and redundancy New Phytologist 208 298ndash301
Zhou R Macaya‐Sanz D Rodgers‐Melnick E Carlson C HGouker F E Evans L M hellip DiFazio S P (2018) Characteri-zation of a large sex determination region in Salix purpurea L(Salicaceae) Molecular Genetics and Genomics 1ndash16 httpsdoiorg101007s00438-018-1473-y
Zsuffa L Mosseler A amp Raj Y (1984) Prospects for interspecifichybridization in willow for biomass production In K L Perttu(Ed) Ecology and management of forest biomass production sys-tems Report 15 (pp 261ndash281) Uppsala Sweden SwedishUniversity of Agricultural Sciences
How to cite this article Clifton‐Brown J HarfoucheA Casler MD et al Breeding progress andpreparedness for mass‐scale deployment of perenniallignocellulosic biomass crops switchgrassmiscanthus willow and poplar GCB Bioenergy201800000ndash000 httpsdoiorg101111gcbb12566
34 | CLIFTON‐BROWN ET AL
Graphical AbstractThe contents of this page will be used as part of the graphical abstract of html only
It will not be published as part of main article
Plant breeding links the research effort with commercial mass upscaling The authorsrsquo assessment of development status ofthe four species is shown (poplar having two one for short rotation coppice (SRC) poplar and one for the more traditionalshort rotation forestry (SRF)) Mass scale deployment needs developments outside the breeding arenas to drive breedingactivities more rapidly and extensively
P L A T FORM
Breeding progress and preparedness for mass‐scale deploymentof perennial lignocellulosic biomass crops switchgrassmiscanthus willow and poplar
John Clifton‐Brown1 | Antoine Harfouche2 | Michael D Casler3 | HuwDylan
Jones1 | William JMacalpine4 | DonalMurphy‐Bokern5 | Lawrence B Smart6 |
Anneli Adler78 | Chris Ashman1 | Danny Awty‐Carroll1 | Catherine Bastien9 |
Sebastian Bopper10 | Vasile Botnari11 | Maryse Brancourt‐Hulmel12 | Zhiyong Chen13 |
Lindsay V Clark14 | Salvatore Cosentino15 | Sue Dalton1 | Chris Davey1 |
Oene Dolstra16 | Iain Donnison1 | Richard Flavell17 | Joerg Greef18 | Steve Hanley4 |
AstleyHastings19 | MagnusHertzberg7 | Tsai‐WenHsu20 | Lin SHuang1 |
Antonella Iurato1 | Elaine Jensen1 | Xiaoli Jin21 | Uffe Joslashrgensen22 | AndreasKiesel23 |
Do‐SoonKim24 | Jianxiu Liu25 | JonPMcCalmont1 | BernardGMcMahon26 |
MichalMos27 | Paul Robson1 | Erik J Sacks14 | Anatolii Sandu11 | Giovanni Scalici15 |
Kai Schwarz18 | Danilo Scordia15 | Reza Shafiei28 | Ian Shield4 | Gancho Slavov4 | Brian
J Stanton29 | Kankshita Swaminathan30 | Gail Taylor31 | Andres F Torres16 | Luisa
M Trindade16 | TimothyTschaplinski32 | GeraldA Tuskan32 | ToshihikoYamada33 |
ChangYeonYu34 | Ronald S ZalesnyJr35 | JunqinZong25 | Iris Lewandowski23
1Institute of Biological Environmental and Rural Sciences Aberystwyth University Aberystwyth UK2Department for Innovation in Biological Agrofood and Forest systems University of Tuscia Viterbo Italy3USDA‐ARS US Dairy Forage Research Center Madison Wisconsin4Rothamsted Research Harpenden UK5LohneGermany6Horticulture Section School of Integrative Plant Science Cornell University Geneva New York7SweTree Technologies AB Umearing Sweden8Institute of Crop Production Ecology Swedish University of Agricultural Sciences Uppsala Sweden9INRA‐BIOFORA Orleacuteans France10Department of Seed Science and Technology Institute of Plant Breeding Seed Science and Population Genetics University of Hohenheim StuttgartGermany11Institute of Genetics Physiology and Plant Protection (IGFPP) of Academy of Sciences of Moldova Chisinau Moldova12INRA‐AgroImpact Peacuteronne cedex France13Insitute of Miscanthus Hunan Agricultural University Hunan Changsha China
These authors contributed equally to this work
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -This is an open access article under the terms of the Creative Commons Attribution License which permits use distribution and reproduction in any medium provided theoriginal work is properly citedcopy 2018 The Authors GCB Bioenergy Published by John Wiley amp Sons Ltd
Received 29 May 2018 | Accepted 18 July 2018
DOI 101111gcbb12566
GCB Bioenergy 20181ndash34 wileyonlinelibrarycomjournalgcbb | 1
14Department of Crop Sciences amp Center for Advanced Bioenergy and Bioproducts Innovation 279 Edward R Madigan Laboratory University ofIllinois Urbana Illinois15Dipartimento di Agricoltura Alimentazione e AmbienteUniversitagrave degli Studi di CataniaCataniaItaly16Plant Breeding Wageningen University amp Research Wageningen The Netherlands17Battersea London UK18Julius Kuhn‐Institut (JKI) Bundesforschungsinstitut fur Kulturpflanzen Braunschweig Germany19Institute of Biological and Environmental Science University of Aberdeen Aberdeen UK20Taiwan Endemic Species Research Institute (TESRI) Nantou County Taiwan21Department of Agronomy amp The Key Laboratory of Crop Germplasm Resource of Zhejiang Province Zhejiang University Hangzhou China22Department of Agroecology Aarhus University Centre for Circular Bioeconomy Tjele Denmark23Department of Biobased Products and Energy Crops Institute of Crop Science University of Hohenheim Stuttgart Germany24Department of Plant Sciences Research Institute of Agriculture amp Life Sciences CALS Seoul National University Seoul Korea25Institute of Botany Jiangsu Province and Chinese Academy of Sciences Nanjing China26Natural Resources Research Institute University of Minnesota ndash Duluth Duluth Minnesota27Energene sp z oo Wrocław Poland28James Hutton Institute University of Dundee Dundee UK29GreenWood Resources Inc Portland Oregon30Hudson-Alpha Institute for Biotechnology Huntsville Alabama31Biological Sciences University of Southampton Southampton UK32The Center for Bioenergy Innovation Oak Ridge National Laboratory Oak Ridge Tennessee33Field Science Centre for the Northern Biosphere Hokkaido University Sapporo Japan34College of Agriculture and Life Sciences 2 Kangwon National University Chuncheon South Korea35USDA Forest Service Northern Research Station Rhinelander Wisconsin
CorrespondenceJohn Clifton‐Brown Institute ofBiological Environmental and RuralSciences Aberystwyth UniversityAberystwyth UKEmail jhcaberacuk
Funding informationFP7 Food Agriculture and FisheriesBiotechnology GrantAward NumberOPTIMISCFP7-289159 WATBIOFP7-311929 H2020 Environment GrantAward Number GRACE-745012Bio-Based Industries Joint Undertaking ItalianMinistry of Education Brain GainProgram (Rientro dei cervelli) USDepartment of Energy GrantAwardNumber DE-AC05-00OR22725 DE-SC0006634 DE-SC0012379 and DE-SC0018420 Department of Trade andIndustry (UK) GrantAward Number BW6005990000 Biotechnology andBiological Sciences Research CouncilGrantAward Number CSP17301 BBN0161491 N016149 LK0863 K01711X1 10963A01 G0162161 E0068331G00580X1 000I0410 Department forEnvironment Food and Rural AffairsGrantAward Number LK0863 NF0424NF0426 US Department of AgricultureNational Institute of Food and Agriculture
AbstractGenetic improvement through breeding is one of the key approaches to increasing
biomass supply This paper documents the breeding progress to date for four
perennial biomass crops (PBCs) that have high outputndashinput energy ratios namely
Panicum virgatum (switchgrass) species of the genera Miscanthus (miscanthus)
Salix (willow) and Populus (poplar) For each crop we report on the size of
germplasm collections the efforts to date to phenotype and genotype the diversity
available for breeding and on the scale of breeding work as indicated by number
of attempted crosses We also report on the development of faster and more pre-
cise breeding using molecular breeding techniques Poplar is the model tree for
genetic studies and is furthest ahead in terms of biological knowledge and genetic
resources Linkage maps transgenesis and genome editing methods are now being
used in commercially focused poplar breeding These are in development in
switchgrass miscanthus and willow generating large genetic and phenotypic data
sets requiring concomitant efforts in informatics to create summaries that can be
accessed and used by practical breeders Cultivars of switchgrass and miscanthus
can be seed‐based synthetic populations semihybrids or clones Willow and
poplar cultivars are commercially deployed as clones At local and regional level
the most advanced cultivars in each crop are at technology readiness levels which
could be scaled to planting rates of thousands of hectares per year in about
5 years with existing commercial developers Investment in further development
of better cultivars is subject to current market failure and the long breeding cycles
2 | CLIFTON‐BROWN ET AL
We conclude that sustained public investment in breeding plays a key role in
delivering future mass‐scale deployment of PBCs
KEYWORD S
bioenergy feedstocks lignocellulose M sacchariflorus M sinensis Miscanthus Panicum virgatum
perennial biomass crop Populus spp Salix spp
1 | INTRODUCTION
Increasing sustainable biomass production is an importantcomponent of the transition from a fossil fuel‐based econ-omy to renewables Taking the United Kingdom as an exam-ple Lovett Suumlnnenberg and Dockerty (2014) suggested that14 million ha of marginal agricultural land could be used forbiomass production without compromising food productionAssuming a biomass dry matter (DM) yield of 10 Mgha anda calorific value of 18 GJMg DM 14 million ha woulddeliver around 28 TWh of electricity (with 40 biomassconversion efficiency) which would be ~8 of primary UKelectricity generation (336 TWh in 2017 (DUKES 2017))To achieve this by 2050 planting rates of ~35000 hayearwould be needed from 2022 in line with calculations byEvans (2017) The current annual planting rates in the UnitedKingdom are orders of magnitude short of these levels atonly several hundred hectares per year Similar scenarioshave been generated for other countries (BMU 2009 Scar-lat Dallemand Monforti‐Ferrario amp Nita 2015)
If perennial biomass crops (PBCs) are to make a real con-tribution to sustainable development they should be grownon agricultural land which is less suitable for food crops(Lewandowski 2015) This economically ldquomarginalrdquo land istypically characterized by abiotic stresses (drought floodingstoniness steep slope exposure to wind and sub‐optimalaspect) low nutrients andor contaminated soils (Toacuteth et al2016) In these challenging environments PBCs need resili-ence traits They also need high outputinput ratios forenergy (typically 20ndash50) to deliver large carbon savingsLand may also be marginal due to environmental vulnerabil-ity Much of the value for society from the genetic improve-ment of these crops depends on positive effects arising fromhighly productive perennial systems In addition to produc-ing biomass as a carbon source to replace fossil carbon thesecrops reduce nitrate leaching (Pugesgaard Schelde LarsenLaeligrke amp Joslashrgensen 2015) making them good candidates tohelp fulfil Water Framework Directive (200060EC) and canincrease soil carbon storage during their production (McCal-mont et al 2017)
The objective of this paper was to report on the prepared-ness for wide deployment by summarizing the technicalstate of the art in breeding of four important PBCs namelyswitchgrass miscanthus willow and poplar These four
crops are the most promising and advanced PBCs for tem-perate regions and have therefore the focus here Switch-grass and miscanthus are both rhizomatous grasses with C4
photosynthesis while willow and poplar are trees with C3
photosynthesis Specifically this paper (a) reviews availablecrop trait genetic diversity information (b) assesses the pro-gress of conventional breeding technologies for yield resili-ence and biomass quality (c) reports on progress with newmolecular‐based breeding technologies to increase speedand precision of selection and (d) discusses the require-ments and next steps for breeding of PBCs including com-mercial considerations in order to sustainably meet thebiomass requirements of a growing worldwide bioeconomy
We summarize the crop‐specific attributes the location ofbreeding programmes the current availability of commercialcultivars and yield expectations in selected environments(Table 1) and the generalized breeding targets for all PBCs(Table 2) Economic information relating to the current mar-ket value of the biomass and the investment in breeding arepresented for different countriesregions in Table 3 We alsopresent a comparison of the prebreeding and conventionalbreeding efforts step‐by‐step starting with wild germplasmcollection and evaluation before wide crossing of wild rela-tives (Table 4) Hybridization is followed by at least 6 yearsof selection and evaluation before commercial upscaling canbegin (Figure 1) Recurrent selection often over decades isused within parent populations as part of an ongoing long‐term process to produce hybrid vigour (Brummer 1999) Inthe following sections the state of the art and new opportuni-ties of breeding switchgrass miscanthus willow and poplarare described The application of modern breeding technolo-gies is compared for the four crops in Table 5 It is mostadvanced in poplar and is therefore described in most detail
2 | SWITCHGRASS
Switchgrass is indigenous to the North American prairiesIt is grown from seed and harvested annually using tech-nology similar to that used for pastures Based on collec-tions from thousands of wild prairie remnants the geneticresources are roughly divided into lowland and upland eco-types and there are distinct clades within each ecotypewhich occur along both latitudinal and longitudinal gradi-ents (Evans et al 2018 Lu et al 2013 Zhang et al
CLIFTON‐BROWN ET AL | 3
TABLE1
Breeding‐relatedattributes
forfour
leadingperennialbiom
asscrops(PBCs)
Species
Switc
hgrass
Miscanthu
sW
illow
Pop
lar
Typ
eC4mdash
Grass
C4mdash
Grass
C3mdash
SRC
C3mdash
SRCSRF
Sourcesof
indigeno
usgerm
plasm
CAato
MXeastof
Rocky
Mou
ntains
Eastern
AsiaandOceania
Predom
inantly
Northernhemisph
ere
Northernhemisph
ere
Breedingsystem
Mon
oeciou
sou
tcrossing
Mon
oeciou
sou
tcrossing
Dioeciousou
tcrossing
Dioeciousou
tcrossing
Ploidy
4times8
times2times
3times
4times
2timesminus12
times2times
Specieswith
ingenu
s~4
50~1
4~4
00~3
0ndash32
Typ
esused
mainlyfor
breeding
US
Low
land
ecotyp
e(sub
trop
ical
clim
ates)andup
land
ecotyp
e(tem
perate
clim
ates)
EUb JPSK
andUS
MsinandMsac
EUS
viminalistimesschw
erinii
Sda
syclad
ostimesrehd
eriana
S
dasyclad
osS
viminalis
USandUKS
viminalistimesmiyab
eana
S
miyab
eana
US
Spu
rpurea
timesmiyab
eana
S
purpurea
Pop
ulus
tricho
carpa
Pdelto
ides
Pnigra
Psuaveolens
subsp
maximow
icziiPba
lsam
ifera
Palba
andhy
brids
Typ
ical
haploid
geno
mesize
(Mbp
)~1
500
Msin~5
400
andMsac~4
400
(Raybu
rnCrawfordRaybu
rnamp
Juvik
2009
)
~450
~485
plusmn10
Breedingprog
rammes
CA20
00(REAP
Quebec)
US 19
92(N
ebraska
19
92(O
klahom
a)
1992
(Georgia)
19
96(W
isconsin)
19
96(Sou
thDakota)
20
00(Tennessee)
20
02(M
ississippi)
20
07(O
klahom
a)
2008
(RutgersNew
Jersey)
20
12(Cornell
New
York)
20
12(U
rbana‐Champaign
Illin
ois)
DE19
90s(K
lein‐W
anzleben)
NL20
00s(W
ageningen)
UK20
04(A
berystwyth)
CN20
06(Chang
sha)
JP20
06(H
okkaido)
US 20
06(Californiamdash
CERESInc)
20
06(CaliforniaandIndianamdash
MBI)
20
08(U
rbana‐Champaign)
SK20
09(Suw
on‐SNU)and(M
uan
NICSof
RDA)
FR20
11(Estreacutees‐M
ons)
UK19
80s(Lon
gAshton
relocatedto
Rothamsted
Researchin
2002
)SE
19
80s(SvaloumlfWeibu
llSalix
energi
Europ
aAB)
US
1990
s(Cornell
New
York)
PL20
00s(O
lsztyn
University
ofWarmia
andMazury)
SE
19
39(M
ykingeEkebo
Research
Institu
te)
19
90s(U
ppsalaSL
U)
20
10(U
ppsalaST
T)
US 19
27(N
ewYork
OxfordPaper
Com
pany
andNew
YorkBotanical
Garden
Wheeler
etal20
15)
19
79(W
ashing
ton
UW
andOrego
nGWR)
19
80ndash1
995(M
ississippiMSS
tate
and
GWR)
19
96(M
innesotaUMD
NRRIand
GWR)
IT19
83(Piedm
ontAFV
)FR
20
01(O
rleacuteans
ampNancyIN
RA
Charrey‐sur‐Saocircne
ampPierroton
FCBA
Nog
ent‐sur‐V
ernisson
IRST
EA)
DE20
08(G
oumlttin
gen
NW‐FVA)
(Con
tinues)
4 | CLIFTON‐BROWN ET AL
TABLE
1(Contin
ued)
Species
Switc
hgrass
Miscanthu
sW
illow
Pop
lar
Current
commercial
varietieson
the
market
US
Nocommercial
hybrids
CA2
EU1(M
timesgfrom
different
origins)
+selected
Msinforthatching
inDK
US
3(but
2aregenetically
identical)
UK25
US
8EUDElt10
FR44
IT10ndash1
5SE
~1
4US
8ndash12
Southeast8ndash
14Upp
erMidwest10
PacificNorthwest
Precom
mercial
cultivars
expected
tobe
onthemarketin
3years
US
36registered
cultivars
(halfare
rand
omseed
increasesfrom
natural
prairiesandhalfarebred
varieties)
Mostarepu
blic
releasesfew
are
protected
patented
orlicensed
NL8
seeded
hybridsvanDinterSemo
MTA
UK4
seeded
hybridsCERES(Land
OLakes)andTerravestaLtdMTA
US
Non
eFR
Non
e
EU53
registered
with
CPV
OforPB
RUK20
registered
with
CPV
OforPB
RUS
18clon
esfrom
Cornellin
multisite
trials
CAun
quantified
UAlberta
andQuebec
FR8ndash
12SR
Fclon
eslicence‐based
IT5SR
Cand6SR
FAFV
MTA
orlicence
SE14
STTlicence‐based
andMTA
US
~19So
utheast5ndash7Upp
erMidwest
from
UMD
NRRIMTA6ndash12
North
Central
from
GWRMTAPacific
Northwest8clon
esGWRMTA24
inmultilocationyieldtrialsGWR
Com
mercial
yield
(tDM
haminus1year
minus1 )
US
3ndash18
EU8ndash
12CN20
ndash30
US
10ndash25
EU7ndash20
UK8ndash
14US
8ndash14
EU5ndash20
(SEandBaltic
Cou
ntries8ndash
12)
US
10ndash2
2(10ndash
12North
Central12
ndash16
Southeast15ndash2
2PacificNortheast)
Harvestrotatio
nand
commercial
stand
lifespan
Ann
ualfor10
ndash12years
Ann
ualfor10ndash2
5years(M
sac
hasbeen
used
for~3
0yearsin
China)
2‐to
4‐year
cyclefor22
ndash30years
SRC3‐
to7‐year
cyclefor20
years
SRF
10‐to12‐yearcycleforgt50
years
Adaptiverang
eOpen‐po
llinatedandsynthetic
cultivars
arelim
itedin
adaptatio
nby
temperature
andprecipitatio
n(~8breeding
zonesin
USandCA)
Standard
Mtimesgiswidelyadaptedin
EU
butislim
itedin
theUnitedStates
byinsufficient
winterh
ardiness
forN
orthern
Midwestand
heatintolerancein
the
southNovelMsactimesMsinhyb
rids
andMsintimesMsinhybrids
selected
incontinentalD
Ehave
show
nawide
adaptiv
erang
ein
EU(K
alininaetal
2017
)Ongoing
trialson
heavymetal
contam
inated
soils
indicatetoleranceby
exclusion(K
rzyżak
etal20
17)
Different
hybridsareneeded
fordifferent
zones
Besthy
bridsshow
someG
timesE(U
S)
andsomehy
bridslow
GtimesE(Fabio
etal20
17)
Different
hybridsareneeded
ford
ifferent
clim
aticzonesHyb
rids
thatareadapted
forg
rowingseason
sof
~6mon
thsand
relativ
elyshortd
aysin
Southern
Europ
earemaladaptedto
shortg
rowing
season
sof
~4mon
thsandrelativ
ely
long
days
inNorthernEU
The
mostb
roadly
adaptedvarietiescome
from
thePcana
densistaxo
n
Note
AFV
AlasiaFranco
VivaiC
ornell
CornellUniversity
CPV
OC
ommunity
PlantV
ariety
OfficeFC
BAF
orestCelluloseW
ood
ConstructionandFu
rnitu
reTechnologyInstitu
teG
timesEg
enotype‐by
‐environmentinterac-
tion
GWRG
reenWoodResourcesINRAF
renchNationalInstituteforA
griculturalR
esearch
IRST
EAN
ationalR
esearchUnito
fScience
andTechnologyforE
nvironmentand
AgricultureM
sacMsacchariflo
rusandM
timesg
(Mtimes
giganteus)M
sin
MiscanthussinensisM
BIMendelB
iotechnology
IncM
bpm
egabase
pairM
SStateM
ississippi
StateUniversity
MTAm
aterialtransferagreem
entNICS
NationalInstituteof
CropScience
NW‐
FVANorthwestGerman
ForestResearchInstitu
tePB
Rplantbreeders
rightRDARural
DevelopmentAdm
inistration
REAP
ResourceEfficient
Agriculture
Productio
nSL
USw
edishUniversity
ofAgriculturalSciences
SNUS
eoul
NationalU
niSR
Cshort‐ro
tatio
ncoppice
SRF
short‐rotationforestryS
TTS
weT
reeTech
tDM
haminus1year
minus1 tons
ofdrymatterperhectareperyearU
AlbertaUniversity
ofAlbertaU
MD
NRRIUniversity
ofMinnesotaDuluthsNaturalResources
ResearchInstitu
teU
MNU
niversity
ofMinnesotaU
WU
niversity
ofWashington
UWMU
niversity
ofWarmiaandMazury
a ISO
Alpha‐2
lettercountrycodes
b EU
isused
forEurope
CLIFTON‐BROWN ET AL | 5
2011) Genotype‐by‐environment interactions (G times E) arestrong and must be considered in breeding (Casler 2012Casler Mitchell amp Vogel 2012) Adaptation to environ-ment is regulated principally by responses to day‐lengthand temperature There are also strong genotype times environ-ment interactions between the drier western regions and thewetter eastern regions (Casler et al 2017)
The growing regions of North America are divided intofour adaptation zones for switchgrass each roughly corre-sponding to two official hardiness zones The lowland eco-types are generally late flowering high yielding andadapted to warmer climates but have lower drought andcold resistance than upland ecotypes (Casler 2012 Casleret al 2012)
In 2015 the US Department of Agriculture (USDA)National Plant Germplasm System GRIN (httpswwwars-gringovnpgs) had 181 switchgrass accessions of whichonly 96 were available for distribution due to limitationsassociated with seed multiplication (Casler Vogel amp Har-rison 2015) There are well over 2000 additional uncata-logued accessions (Table 1) held by various universities butthe USDA access to these is also constrained by the effortneeded in seed multiplication Switchgrass is a model herba-ceous species for conducting scientific research on biomass(Sanderson Adler Boateng Casler amp Sarath 2006) but lit-tle funding is available for the critical prebreeding work thatis necessary to link this biological research to commercialbreeding More than a million genotypes from ~2000 acces-sions (seed accessions contain many genotypes) have beenphenotypically screened in spaced plant nurseries and tenthousand of the most useful have been genotyped with differ-ent technologies depending on the technology available at
the time when these were performed From these character-ized genotypes parents are selected for exploratory pairwisecrosses to produce synthetic populations within ecotypesSwitchgrass like many grasses is outcrossing due to astrong genetically controlled self‐incompatibly (akin to theS‐Z‐locus system of other grasses (Martinez‐Reyna ampVogel 2002)) Thus the normal breeding approaches usedare F1 wide crosses and recurrent selection cycles within syn-thetic populations
The scale of these programmes varies from small‐scaleconventional breeding based solely on phenotypic selec-tion (eg REAP Canada Montreal Quebec) to large pro-grammes incorporating modern molecular breedingmethods (eg USDA‐ARS Madison Wisconsin) Earlyagronomic research and biomass production efforts werefocused on the seed‐based multiplication of promising wildaccessions from natural prairies Cultivars Alamo Kanlowand Cave‐in‐Rock were popular due to high yield andmoderate‐to‐wide adaptation Conventional breedingapproaches focussed on biomass production traits and haveled to the development of five cultivars particularly suitedto biomass production Cimarron EG1101 EG1102EG2101 and Liberty The first four of these represent thelowland ecotype and were developed either in Oklahomaor Georgia Liberty is a derivative of lowland times uplandhybrids developed in Nebraska following selection for lateflowering the high yield of the lowland ecotype and coldtolerance of the upland ecotype (Vogel et al 2014) Thesefive cultivars were all approximately 25ndash30 years in themaking counting from the initiation of these breeding pro-grammes Many more biomass‐type cultivars are expectedwithin the next few years as these and other breeding
TABLE 2 Generalized improvement targets for perennial biomass crops (PBCs)
Net energy yield per hectare
Increased yield
Reduced moisture content at harvest
Physical and chemical composition for different end‐use applications
Increased lignin content and decreased corrosive elements for thermal conversion
Reduced recalcitrance through decreased lignin content andor modified lignin monomer composition to reduce pretreatment requirementsfor next‐generation biofuels by saccharification and fermentation
Plant morphological differences which influence biomass harvest transport and storage (eg stem thickness)
Propagation costs
Improved cloning systems (trees and grasses)
Seed systems (grasses)
Optimizing agronomy for each new cultivar
Resilience through enhanced
Abiotic stress toleranceresistance (eg drought salinity and high and low temperature )
Biotic stress resistance (eg insects fungal bacterial and viral diseases)
Site adaptability especially to those of marginalcontaminated agricultural land
6 | CLIFTON‐BROWN ET AL
TABLE3
Preparedness
formassupscaling
currentmarketvalueandresearch
investmentin
four
perennialbiom
asscrops(PBCs)
Species
Switc
hgrass
Miscanthu
sW
illow
Pop
lar
Current
commercial
plantin
gcostsperha
USa20
0USD
bin
the
establishm
entyear
DE3
375
Eurob
(XueK
alininaamp
Lew
ando
wski20
15)
UK2
153
GBPb
(Evans201
6)redu
cedto
116
9GBPwith
Mxg
rhizom
ewhere
thefarm
erdo
estheland
preparation(w
wwterravestacom
)US
173
0ndash222
5USD
UK150
0ndash173
9GBPplus
land
preparation(Evans20
16)
US
197
6USD
(=80
0USD
acre)
IT110
0Euro
FRSE100
0ndash200
0Euro
US
863USD
ha(LazarusHeadleeamp
Zalesny
20
15)
Current
marketvalue
ofthebiom
asspert
DM
US
80ndash1
00USD
UK~8
0GBP(bales)(Terravesta
person
alcommun
ication)
US
94USD
(chipp
ed)
UK49
41GBP(chipp
ed)(Evans20
16)
US
55ndash7
0USD
IT10
0Euro
US
80ndash90USD
deliv
ered
basedon
40milesof
haulage
Scienceforg
enetic
improv
ementprojects
inthelast10
years
USgt50
projectsfund
edby
USDOEand
USD
ANIFA
CNgt
30projectsfund
edby
CN‐N
SFCandMBI
EUc 6projects(O
PTIM
ISCO
PTIM
A
WATBIO
SUNLIBBF
IBRAandGRACE)
FRB
FFSK
2projectsfund
edby
IPETandPM
BC
US
EBICABBImultip
leDOEFeedstock
Genom
icsandUSD
AAFR
Iprojects
UKB
SBECB
EGIN
(200
3ndash20
10)
US
USDOEJG
IGenom
eSequ
encing
(200
9ndash20
122
015ndash
2018
)USD
ANortheastSu
nGrant
(200
9ndash20
12)
USD
ANIFANEWBio
Con
sortium
(201
2ndash20
17)USD
ANIFAWillow
SkyC
AP(201
8ndash20
21)
EUF
P7(EnergyPo
plarN
ovelTreeWATBIO
Tree4Fu
ture)
SEC
limate‐adaptedpo
plarsandNanow
ood
SLU
andST
TUS
Bioenergy
Science(O
RNL)USD
Afeedstocks
geno
micsprog
rammeBRDIa
ndBRC‐CBI
Major
projects
supp
ortin
gcrossing
andselectioncycles
inthelast10
years
USandCA12
projects
NLRUEmiscanthu
sPP
P20
15ndash2
019
50ndash
100K
Euroyear
UKGIA
NT‐LIN
KPP
I20
11ndash2
016
13M
GBPyear
US
EBICABBI~0
5M
USD
year
UKBEGIN
2000
ndash201
0US
USD
ANortheastSu
nGrant
(200
9ndash20
12)USD
ANIFA
NEWBio
Con
sortium
(201
2ndash20
17)USD
ANIFA
Willow
SkyC
AP(201
8ndash20
21)
FR1
4region
alandnatio
nalp
rojects10
0ndash15
0k
Euroyear
SES
TTo
nebreeding
effortas
aninternalproject
in20
10with
aresulting
prog
enytrial
US
DOESu
nGrant
feedstockdevelopm
ent
partnershipprog
ramme(including
Willow
)
Current
annu
alinvestmentin
projects
fortranslationinto
commercial
hybrids
USgt20
MUSD
year
UKMUST
201
6ndash20
1905M
GBPyear
US
CABBIUSD
AAFR
I20
17ndash2
02205M
USD
year
US
USD
ANIFA
NEWBio
~04
MUSD
year
FRScienceforim
prov
ement~2
0kEuroyear
US
USD
Aun
derAFR
ICo‐ordinatedAg
Prod
ucer
projectsAnd
severaltranslational
geno
micsprojectswith
outpu
blic
fund
ing
Upscalin
gtim
e(years)
toproducesufficient
propagules
toplantgt10
0ha
US
Using
seed
2ndash3years
UKRhizomefor1ndash200
0hectares
canbe
ready
in6mon
ths
UKNL2
0ha
ofseed
hybridsplantedin
2018
In
2019
sufficientseedfor5
0ndash10
0ha
isexpected
UK3yearsusingconv
entio
nalcutting
sfaster
usingmicroprop
agation
FRIT3yearsby
vegetativ
eprop
agation
SEgt3yearsby
vegetativ
eprop
agation(cuttin
gs)
and3yearsby
microprop
agation
Note
AFR
IAgriculture
andFo
odResearchInitiativein
theUnitedStatesBEGIN
BiomassforEnergyGenetic
Improvem
entNetwork
BFF
BiomassFo
rtheFu
tureBRC‐CBI
Bioenergy
ResearchCentre‐Centrefor
Bioenergy
Innovatio
nBRDIBiomassresearch
developm
entinitiative
BSB
ECBBSR
CSu
stainableBioenergy
Centre
CABBICenterforadvanced
bioenergyandbioproductsinnovatio
nin
theUnitedStatesCN‐N
SFC
Natural
ScienceFo
undatio
nof
ChinaDOEDepartm
entof
Energyin
theUnitedStatesEBIEnergyBiosciences
Institu
teFIBRAFibrecropsas
sustainablesource
ofbiobased
materialforindustrial
productsin
Europeand
ChinaFP
7SeventhFram
eworkProgrammein
theEUGIA
NT‐LIN
KGenetic
improvem
entof
miscanthusas
asustainablefeedstockforbioenergyin
theUnitedKingdom
GRACEGRow
ingAdvancedindustrial
Crops
onmarginallandsforbiorEfineriesIPETKorea
Institu
teof
Planning
andEvaluationforTechnologyin
FoodA
gricultureFo
restry
andFisheriesJG
IJointGenom
eInstitu
teMBIMendelBiotechnology
IncMUST
Miscant-
husUpS
calin
gTechnology
NEWBioNortheast
WoodyW
arm‐seasonBiomassConsortiumNIFANationalInstitu
teof
Food
andAgriculture
intheUnitedStatesNovelTree
Novel
tree
breeding
strategiesOPT
IMAOpti-
mizationof
perennialgrassesforbiom
assproductio
nin
theMediterraneanareaOPT
IMISCOptim
izingbioenergyproductio
nfrom
Miscanthus
ORNLOak
Ridge
NationalLabPM
BCPlantMolecular
BreedingCenterof
theNextGenerationBiogreenResearchCenters
oftheRepublic
ofKoreaPP
Ipublicndashp
rivate
investmentPP
Ppublicndashp
rivate
partnership
RUEradiationuseefficiencySk
yCAP
Coordinated
AgriculturalProjectSL
U
SwedishUniversity
ofAgriculturalSciencesST
TSw
eTreeTech
SUNLIBBSu
stainableLiquidBiofuelsfrom
BiomassBiorefiningTree4Fu
tureDesigning
Trees
forthefutureUSD
AUSDepartm
entof
Agriculture
WATBIO
Developmentof
improved
perennialnonfoodbiom
assandbioproduct
cropsforwater
stressed
environm
ents
a ISO
Alpha‐2
lettercountrycodes
b Local
currencies
areused
asat
2018Euro
GBP(G
reat
Britain
Pound)USD
(USDollar)c EU
isused
forEurope
CLIFTON‐BROWN ET AL | 7
TABLE4
Prebreedingresearch
andthestatus
ofconventio
nalbreeding
infour
leadingperennialbiom
asscrops(PBCs)
Breeding
techno
logy
step
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sWillow
Poplar
1Collected
wild
accessions
or
secondarysources
availablefor
breeding
Provideabroadbase
of
useful
traits
Respect
forCBD
and
Nagoyaprotocol
on
collections
after2014
Not
allindigenous
genetic
resourcesareaccessible
under
CBD
forpolitical
reasons
USa~2
000
(181
inofficial
GRIN
gene
bank
ofwhich
96
wereavailable
others
in
variousunofficial
collections)
CNChangsha
~1000
andNanjin
g
2000from
China
DK~1
20from
JP
FRworking
collectionof
~100
(mainlyM
sin)
JP~1
500
from
JPS
KandCN
NLworking
collectionof
~300
Msin
SK~7
00from
SKC
NJP
andRU
UK~1
500
accessions
(from
~500
sites)
ofwhich
1000arein
field
nurseriesin
2018
US
14in
GRIN
global
US
Illin
ois~1
500
collections
made
inAsiawith
25
currently
available
inUnitedStates
forbreeding
UK1500with
about20
incommon
with
the
UnitedStates
US
350largelyunique
FR3370Populus
delto
ides
(650)Pnigra(2000)
Ptrichocarpa(600)and
Pmaximow
iczii(120)
IT30000P
delto
idsPnigra
Pmaximow
icziiand
Ptrichocarpa
SE13000
accessionsmainly
Ptrichocarpa
byST
TandSL
U
US
GWR150Ptrichocarpaand
300Pmaximow
icziiaccessions
forPacificNorthwestprogram
535Pdelto
ides
and~2
00
Pnigraaccessions
for
Southeastprogrammeand77
Psimoniifrom
Mongolia
UMD
NRRI550+
clonal
accessions
(primarily
Ptimescanadnesisa
long
with
Pdelto
ides
andPnigra
parents)
forUpper
Midwest
program
2Wild
accessions
which
have
undergone
phenotypic
screeningin
field
trials
Selectionof
parental
lines
with
useful
traits
Strong
partnerships
torun
multilocationfieldtrials
tophenotypeconsistently
theaccessionsgenotypes
indifferentenvironm
ents
anddatabases
Costof
runningmultilocation
US
gt500000genotypes
CN~1
250
inChangsha
~1700
in
HunanJiangsuS
handong
Hainan
NL~250genotypes
JP~1
200
SK~4
00
UK‐le
d~1
000
genotypesin
a
replicated
trial
US‐led
~1200
genotypesin
multi‐
locatio
nreplicated
trialsin
Asiaand
North
America
MsinSK
CNJP
CA
(Ontario)US(Illinoisand
Colorado)MsacSK
(Kangw
on)
CN
(Zhejiang)JP
(Sapporo)CA
(Onatario)U
S(Illinois)DK
(Aarhus)
UKgt400
US
180
FR2720
IT10000
SE~1
50
US
GWR~1
500
wild
Ptrichocarpaphenotypes
and
OPseedlots(total
of70)from
thePmaximow
icziispecies
complex
(suaveolens
cathayana
koreana
ussuriensismaximow
iczii)and
2000ndash3000Pdelto
ides
collections
(based
on104s‐
generatio
nfamilies)
UMD
NRRIscreened
seedlin
gsfrom
aPdelto
ides
(OPor
CP)
breeding
programme(1996ndash2016)
gt7200OPseedlin
gsof
PnigraunderMinnesota
clim
atic
conditions(improve
parental
populatio
n)
3Wild
germ
plasm
genotyping
Amanaged
living
collectionof
clonal
types
US
~20000genotypes
CNChangsha
~1000Nanjin
g37
FR44
bycpDNA
(Fenget
al
UK~4
00
US
225
(Con
tinues)
8 | CLIFTON‐BROWN ET AL
TABLE
4(Contin
ued)
Breeding
techno
logy
step
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sWillow
Poplar
Construct
phylogenetic
treesanddissim
ilarity
indices
The
type
ofmolecular
analysis
mdashAFL
PcpDNAR
ADseq
whole
genomesequencing
2014)
JP~1
200
NL250genotypesby
RADseq
UK~1
000
byRADseq
SK~3
00
US
Illin
oisRADseqon
allwild
accessions
FR2310
SE150
US
1500
4Exploratory
crossing
and
progenytests
Discovergood
parental
combinatio
nsmdashgeneral
combining
ability
Geographicseparatio
n
speciesor
phylogenetic
trees
Costly
long‐te
rmmultilocation
trialsforprogeny
US
gt10000
CN~1
0
FR~2
00
JP~3
0
NL3600
SK~2
0
UK~4
000
(~1500Msin)
US
500ndash1000
UK gt700since2003
gt800EWBP(1996ndash
2002)
US
800
FRCloned13
times13
factorial
matingdesign
with
3species[10
Pdelto
ides8Ptrichocarpa
8
Pnigra)
US
GWR1000exploratory
crossingsof
Pfrem
ontii
Psimoniiforim
proved
droughttolerance
UMD
NRRIcrossednativ
e
Pdelto
ides
with
non‐nativ
e
Ptrichocarpaand
Pmaximow
icziiMajority
of
progenypopulatio
nssuffered
from
clim
atic
andor
pathogenic
susceptib
ility
issues
5Wideintraspecies
hybridization
Discovergood
parental
combinatio
nsmdashgeneral
combining
ability
1to
4above
inform
atics
Flow
eringsynchronizationand
low
seed
set
US
~200
CN~1
0
NL1800
SK~1
0
UK~5
00
UK276
US
400
IT500
SE120
US
50ndash100
6With
inspecies
recurrentselection
Concentratio
nof
positiv
e
traits
Identificationof
theright
heterotic
groups
insteps
1ndash5
Difficultto
introducenew
germ
plasm
with
outdilutio
n
ofbesttraits
US
gt40
populatio
ns
undergoing
recurrentselection
NL2populatio
ns
UK6populatio
ns
US
~12populatio
nsto
beestablished
by2019
UK4
US
150
FRPdelto
ides
(gt30
FSfamilies
ndash900clones)Pnigra(gt40
FS
families
ndash1600clones)and
Ptrichocarpa(15FS
Families
minus350clones)
IT80
SEca50
parent
breeding
populatio
nswith
in
Ptrichocarpa
US
Goalsis~1
00parent
breeding
populatio
nswith
inPtrichocarpa
Pdelto
ides
PnigraandPmaximow
iczii
7Interspecies
hybrid
breeding
Com
bining
complem
entary
traitsto
producegood
morphotypes
and
heterosiseffects
Allthesteps1ndash6
Inearlystage
improvem
ents
areunpredictable
Awide
base
iscostly
tomanage
None
CNChangsha
3Nanjin
g
120MsintimesMflo
r
JP~5
with
Saccharum
spontaneum
SK5
UK420
US
250
FRPcanadensis(1800)
Pdelto
ides
timesPtrichocarpa
(800)and
PtrichocarpatimesPmaximow
iczii
(1200)
(Con
tinues)
CLIFTON‐BROWN ET AL | 9
TABLE
4(Contin
ued)
Breeding
techno
logy
step
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sWillow
Poplar
IT~1
20
US
525genotypesin
eastside
hybrid
programme
205westside
hybrid
programmeand~1
200
in
SoutheastPdelto
ides
program
8Chrom
osom
e
doublin
g
Arouteto
triploid
seeded
typesdoublin
gadiploid
parent
ofknow
n
breeding
valuefrom
recurrentselection
Needs
1ndash7to
help
identify
therightparental
lines
Doubled
plantsarenotorious
forrevertingto
diploid
US
~50plantstakenfrom
tetraploid
tooctaploid
which
arenow
infieldtrials
CN~1
0Msactriploids
DKSeveralattemptsin
mid
90s
FR5genotypesof
Msin
UK2Msin1Mflora
nd1Msac
US
~30
UKAttempted
butnot
routinelyused
US
Not
partof
theregular
breeding
becausethere
existnaturalpolyploids
NA
9DoubleHaploids
Arouteto
producing
homogeneous
progeny
andalso
forthe
introductio
nof
transgenes
orforgenome
editing
Needs
1ndash7to
help
identify
therightparental
lines
Fully
homozygousplantsare
weakandeasily
die
andmay
notflow
ersynchronously
US
Noneyetattempteddueto
poor
vigour
andviability
of
haploids
CN2genotypesof
Msin
andMflor
UKDihaploidsof
MsinMflor
and
Msacand2transgenic
Msinvia
anther
cultu
re
US
6createdandused
toidentify
miss‐identifying
paralogueloci
(causedby
recent
genome
duplicationin
miscanthus)Usedin
creatin
gthereferencegenomefor
MsinVeryweakplantsand
difficultto
retain
thelin
es
NA
NA
10Embryo
rescue
Anattractiv
etechniquefor
recovering
plantsfrom
sexual
crosses
The
majority
ofem
bryos
cannot
survivein
vivo
or
becomelong‐timedorm
ant
None
UKone3times
hybrid
UKEmbryosrescue
protocol
proved
robustat
8days
post‐pollin
ation
FRAnem
bryo
rescue
protocol
proved
robustandim
proved
hybridizationsuccessfor
Pdelto
ides
timesPtrichocarpa
crosses
11Po
llenstorage
Flexibility
tocross
interestingparents
with
outneed
of
flow
ering
synchronization
None
Nosuccessto
daten
otoriously
difficultwith
grassesincluding
sugarcane
Freshpollencommonly
used
incrossing
Pollenstoragein
some
interspecificcrosses
FRBothstored
(cryobank)
and
freshindividual
pollen
Note
AFL
Pam
plifiedfragmentlength
polymorphismCBDconventio
non
biological
diversity
CP
controlledpollinatio
nCpD
NAchloroplastDNAEWBP
Europeanwillow
breeding
programme
FSfull‐sib
GtimesE
genotype‐by‐environm
entGRIN
germ
plasm
resourcesinform
ationnetwork
GWRGreenWoodResourcesMflorMiscanthusflo
ridulusMsacMsacchariflo
rus
MsinMsinensisNAnotapplicableNRRINatural
Resources
ResearchInstitu
teOP
open
pollinatio
nRADseq
restrictionsite‐associatedDNA
sequencingSL
USw
edishUniversity
ofAgriculturalSciencesST
TSw
eTreeTech
UMDUniversity
ofMinnesota
Duluth
a ISO
Alpha‐2
lettercountrycodes
10 | CLIFTON‐BROWN ET AL
programmes mature The average rate of gain for biomassyield in long‐term switchgrass breeding programmes hasbeen 1ndash4 per year depending on ecotype populationand location of the breeding programme (Casler amp Vogel2014 Casler et al 2018) The hybrid derivative Libertyhas a biomass yield 43 higher than the better of its twoparents (Casler amp Vogel 2014 Vogel et al 2014) Thedevelopment of cold‐tolerant and late‐flowering lowland‐ecotype populations for the northern United States hasincreased biomass yields by 27 (Casler et al 2018)
Currently more than 20 recurrent selection populationsare being managed in the United States to select parentsfor improved yield yield resilience and compositional qual-ity of the biomass For the agronomic development andupscaling high seed multiplication rates need to be com-bined with lower seed dormancy to reduce both crop estab-lishment costs and risks Expresso is the first cultivar withsignificantly reduced seed dormancy which is the first steptowards development of domesticated populations (Casleret al 2015) Most phenotypic traits of interest to breeders
require a minimum of 2 years to be fully expressed whichresults in a breeding cycle that is at least two years Morecomplicated breeding programmes or traits that requiremore time to evaluate can extend the breeding cycle to 4ndash8 years per generation for example progeny testing forbiomass yield or field‐based selection for cold toleranceBreeding for a range of traits with such long cycles callsfor the development of molecular methods to reduce time-scales and improve breeding efficiency
Two association panels of switchgrass have been pheno-typically and genotypically characterized to identify quanti-tative trait loci (QTLs) that control important biomasstraits The northern panel consists of 60 populationsapproximately 65 from the upland ecotype The southernpanel consists of 48 populations approximately 65 fromthe lowland ecotype Numerous QTLs have been identifiedwithin the northern panel to date (Grabowski et al 2017)Both panels are the subject of additional studies focused onbiomass quality flowering and phenology and cold toler-ance Additionally numerous linkage maps have been
FIGURE 1 Cumulative minimum years needed for the conventional breeding cycle through the steps from wild germplasm to thecommercial hybrids in switchgrass miscanthus willow and poplar Information links between the steps are indicated by dotted arrows andhighlight the importance of long‐term informatics to maximize breeding gain
CLIFTON‐BROWN ET AL | 11
created by the pairwise crossing of individuals with diver-gent characteristics often to generate four‐way crosses thatare analysed as pseudo‐F2 crosses (Liu Wu Wang ampSamuels 2012 Okada et al 2010 Serba et al 2013Tornqvist et al 2018) Individual markers and QTLs iden-tified can be used to design marker‐assisted selection(MAS) programmes to accelerate breeding and increase itsefficiency Genomic prediction and selection (GS) holdseven more promise with the potential to double or triplethe rate of gain for biomass yield and other highly complexquantitative traits of switchgrass (Casler amp Ramstein 2018Ramstein et al 2016) The genome of switchgrass hasrecently been made public through the Joint Genome Insti-tute (httpsphytozomejgidoegov)
Transgenic approaches have been heavily relied upon togenerate unique genetic variants principally for traitsrelated to biomass quality (Merrick amp Fei 2015) Switch-grass is highly transformable using either Agrobacterium‐mediated transformation or biolistics bombardment butregeneration of plants is the bottleneck to these systemsTraditionally plants from the cultivar Alamo were the onlyregenerable genotypes but recent efforts have begun toidentify more genotypes from different populations that arecapable of both transformation and subsequent regeneration(King Bray Lafayette amp Parrott 2014 Li amp Qu 2011Ogawa et al 2014 Ogawa Honda Kondo amp Hara‐Nishi-mura 2016) Cell wall recalcitrance and improved sugarrelease are the most common targets for modification (Bis-wal et al 2018 Fu et al 2011) Transgenic approacheshave the potential to provide traits that cannot be bredusing natural genetic variability However they will stillrequire about 10ndash15 years and will cost $70ndash100 millionfor cultivar development and deployment (Harfouche Mei-lan amp Altman 2011) In addition there is commercialuncertainty due to the significant costs and unpredictabletimescales and outcomes of the regulatory approval processin the countries targeted for seed sales As seen in maizeone advantage of transgenic approaches is that they caneasily be incorporated into F1 hybrid cultivars (Casler2012 Casler et al 2012) but this does not decrease thetime required for cultivar development due to field evalua-tion and seed multiplication requirements
The potential impacts of unintentional gene flow andestablishment of non‐native transgene sequences in nativeprairie species via cross‐pollination are also major issues forthe environmental risk assessment These limit further thecommercialization of varieties made using these technolo-gies Although there is active research into switchgrass steril-ity mechanisms to curb unintended pollen‐mediated genetransfer it is likely that the first transgenic cultivars proposedfor release in the United States will be met with considerableopposition due to the potential for pollen flow to remainingwild prairie sites which account for lt1 of the original
prairie land area and are highly protected by various govern-mental and nongovernmental organizations (Casler et al2015) Evidence for landscape‐level pollen‐mediated geneflow from genetically modified Agrostis seed multiplicationfields (over a mountain range) to pollinate wild relatives(Watrud et al 2004) confirms the challenge of using trans-genic approaches Looking ahead genome editing technolo-gies hold considerable promise for creating targeted changesin phenotype (Burris Dlugosz Collins Stewart amp Lena-ghan 2016 Liu et al 2018) and at least in some jurisdic-tions it is likely that cultivars resulting from gene editingwill not need the same regulatory approval as GMOs (Jones2015a) However in July 2018 the European Court of Justice(ECJ) ruled that cultivars carrying mutations resulting fromgene editing should be regulated in the same way as GMOsThe ECJ ruled that such cultivars be distinguished from thosearising from untargeted mutation breeding which isexempted from regulation under Directive 200118EC
3 | MISCANTHUS
Miscanthus is indigenous to eastern Asia and Oceania whereit is traditionally used for forage thatching and papermaking(Xi 2000 Xi amp Jezowkski 2004) In the 1960s the highbiomass potential of a Japanese genotype introduced to Eur-ope by Danish nurseryman Aksel Olsen in 1935 was firstrecognized in Denmark (Linde‐Laursen 1993) Later thisaccession was characterized described and named asldquoM times giganteusrdquo (Greef amp Deuter 1993 Hodkinson ampRenvoize 2001) commonly abbreviated as Mxg It is a natu-rally occurring interspecies triploid hybrid between tetraploidM sacchariflorus (2n = 4x) and diploid M sinensis(2n = 2x) Despite its favourable agronomic characteristicsand ability to produce high yields in a wide range of environ-ments in Europe (Kalinina et al 2017) the risks of relianceon it as a single clone have been recognized Miscanthuslike switchgrass is outcrossing due to self‐incompatibility(Jiang et al 2017) Thus seeded hybrids are an option forcommercial breeding Miscanthus can also be vegetativelypropagated by rhizome or in vitro culture which allows thedevelopment of clones The breeding approaches are usuallybased on F1 crosses and recurrent selection cycles within thesynthetic populations There are several breeding pro-grammes that target improvement of miscanthus traitsincluding stress resilience targeted regional adaptation agro-nomic ldquoscalabilityrdquo through cheaper propagation fasterestablishment lower moisture and ash contents and greaterusable yield (Clifton‐Brown et al 2017)
Germplasm collections specifically to support breedingfor biomass started in the late 1980s and early 1990s inDenmark Germany and the United Kingdom (Clifton‐Brown Schwarz amp Hastings 2015) These collectionshave continued with successive expeditions from European
12 | CLIFTON‐BROWN ET AL
and US teams assembling diverse collections from a widegeographic range in eastern Asia including from ChinaJapan South Korea Russia and Taiwan (Hodkinson KlaasJones Prickett amp Barth 2015 Stewart et al 2009) Threekey miscanthus species for biomass production are M si-nensis M floridulus and M sacchariflorus M sinensis iswidely distributed throughout eastern Asia with an adap-tive range from the subtropics to southern Russia (Zhaoet al 2013) This species has small rhizomes and producesmany tightly packed shoots forming a ldquotuftrdquo M floridulushas a more southerly adaptive range with a rather similarmorphology to M sinensis but grows taller with thickerstems and is evergreen and less cold‐tolerant than the othermiscanthus species M sacchariflorus is the most northern‐adapted species ranging to 50 degN in eastern Russia (Clarket al 2016) Populations of diploid and tetraploid M sac-chariflorus are found in China (Xi 2000) and South Korea(Yook 2016) and eastern Russia but only tetraploids havebeen found in Japan (Clark Jin amp Petersen 2018)
Germplasm has been assembled from multiple collec-tions over the last century though some early collectionsare poorly documented This historical germplasm has beenused to initiate breeding programmes largely based on phe-notypic and genotypic characterization As many of theaccessions from these collections are ldquoorigin unknownrdquocrucial environmental envelope data are not available UK‐led expeditions started in 2006 and continued until 2011with European and Asian partners and have built up a com-prehensive collection of 1500 accessions from 500 sitesacross Eastern Asia including China Japan South Koreaand Taiwan These collections were guided using spatialclimatic data to identify variation in abiotic stress toleranceAccessions from these recent collections were planted fol-lowing quarantine in multilocation nursery trials at severallocations in Europe to examine trait expression in differentenvironments Based on the resulting phenotypic andmolecular marker data several studies (a) characterized pat-terns of population genetic structure (Slavov et al 20132014) (b) evaluated the statistical power of genomewideassociation studies (GWASs) and identified preliminarymarkerndashtrait associations (Slavov et al 2013 2014) and(c) assessed the potential of genomic prediction (Daveyet al 2017 Slavov et al 2014 2018b) Genomic indexselection in particular offers the possibility of exploringscenarios for different locations or industrial markets (Sla-vov et al 2018a 2018b)
Separately US‐led expeditions also collected about1500 accessions between 2010 and 2014 (Clark et al2014 2016 2018 2015) A comprehensive genetic analy-sis of the population structure has been produced by RAD-seq for M sinensis (Clark et al 2015 Van der WeijdeKamei et al 2017) and M sacchariflorus (Clark et al2018) Multilocation replicated field trials have also been
conducted on these materials in North America and inAsia GWAS has been conducted for both M sinensis anda subset of M sacchariflorus accessions (Clark et al2016) To date about 75 of these recent US‐led collec-tions are in nursery trials outside the United States Due tolengthy US quarantine procedures these are not yet avail-able for breeding in the United States However molecularanalyses have allowed us to identify and prioritize sets ofgenotypes that best encompass the genetic variation in eachspecies
While mostM sinensis accessions flower in northern Eur-ope very few M sacchariflorus accessions flower even inheated glasshouses For this reason the European pro-grammes in the United Kingdom the Netherlands and Francehave performed mainly M sinensis (intraspecies) hybridiza-tions (Table 4) Selected progeny become the parents of latergenerations (recurrent selection as in switchgrass) Seed setsof up to 400 seed per panicle occur inM sinensis In Aberyst-wyth and Illinois significant efforts to induce synchronousflowering in M sacchariflorus and M sinensis have beenmade because interspecies hybrids have proven higher yieldperformance and wide adaptability (Kalinina et al 2017) Ininterspecies pairwise crosses in glasshouses breathable bagsandor large crossing tubes or chambers in which two or morewhole plants fit are used for pollination control Seed sets arelower in bags than in the open air because bags restrict pollenmovement while increasing temperatures and reducinghumidity (Clifton‐Brown Senior amp Purdy 2018) About30 of attempted crosses produced 10 to 60 seeds per baggedpanicle The seed (thousand seed mass ranges from 05 to09 g) is threshed from the inflorescences and sown into mod-ular trays to produce plug plants which are then planted infield nurseries to identify key parental combinations
A breeding programme of this scale must serve theneeds of different environments accepting the commonpurpose is to optimize the interception of solar radiationAn ideal hybrid for a given environment combines adapta-tion to date of emergence with optimization of traits suchas height number of stems per plant flowering and senes-cence time to optimize solar interception to produce a highbiomass yield with low moisture content at harvest (Rob-son Farrar et al 2013 Robson Jensen et al 2013) By20132014 conventional breeding in Europe had producedintra‐ and interspecific fertile seeded hybrids When acohort (typically about 5) of outstanding crosses have beenidentified it is important to work on related upscaling mat-ters in parallel These are as follows
Assessment of the yield and critical traits in selectedhybrids using a network of field trials
Efficient cloning of the seed parents While in vitro andmacro‐cloning techniques are used some genotypes areamenable to neither technique
CLIFTON‐BROWN ET AL | 13
TABLE5
Status
ofmod
ernplantbreeding
techniqu
esin
four
leadingperennialbiom
asscrop
s(PBCs)
Breeding
techno
logy
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sW
illow
Pop
lar
MAS
Use
ofmarker
sequ
encesthat
correlatewith
atrait
allowingearly
prog
enyselectionor
rejectionlocatin
gkn
owngenesfor
useful
traits(suchas
height)from
other
speciesin
your
crop
Breeding
prog
ramme
relevant
biparental
crosses(Table
4)to
create
ldquomapping
popu
latio
nsrdquoand
QTLidentification
andestim
ationto
identifyrobu
stmarkers
alternatively
acomprehensive
panelof
unrelated
geno
typesfor
GWAS
Ineffectivewhere
traits
areaffected
bymany
geneswith
small
effects
US
Literally
dozens
ofstud
iesto
identify
SNPs
andmarkers
ofinterestNoattempts
atMASas
yetlargely
dueto
popu
latio
nspecificity
CNS
SRISS
Rmarkers
forM
sinMsacand
Mflor
FR2
Msin
mapping
popu
latio
ns(BFF
project)
NL2
Msin
mapping
popu
latio
ns(A
tienza
Ram
irezetal
2003
AtienzaSatovic
PetersenD
olstraamp
Martin
200
3Van
Der
Weijdeetal20
17a
Van
derW
eijde
Hux
ley
etal20
17
Van
derW
eijdeKam
ei
etal20
17V
ander
WeijdeKieseletal
2017
)SK
2GWASpanels(1
collectionof
Msin1
diploidbiparentalF 1
popu
latio
n)in
the
pipelin
eUK1
2families
tofind
QTLin
common
indifferentfam
ilies
ofthe
sameanddifferent
speciesandhy
brids
US
2largeGWAS
panels4
diploidbi‐
parentalF 1
popu
latio
ns
1F 2
popu
latio
n(add
ition
alin
pipelin
e)
UK16
families
for
QTLdiscov
ery
2crossesMASscreened
1po
tentialvariety
selected
US
9families
geno
typedby
GBS(8
areF 1o
neisF 2)a
numberof
prom
ising
QTLdevelopm
entof
markers
inprog
ress
FR(INRA)tested
inlargeFS
families
and
infactorialmating
design
low
efficiency
dueto
family
specificity
US
49families
GS
Metho
dto
accelerate
breeding
throug
hredu
cing
the
resourcesforcross
attemptsby
Aldquotraining
popu
latio
nrdquoof
individu
alsfrom
stages
abov
ethat
have
been
both
Risks
ofpo
orpredictio
nProg
eny
testingneedsto
becontinueddu
ring
the
timethetraining
setis
US
One
prog
rammeso
farmdash
USD
Ain
WisconsinThree
cycles
ofgeno
mic
selectioncompleted
CN~1
000
geno
types
NLprelim
inary
mod
elsforMsin
developed
UK~1
000
geno
types
Verysuitable
application
not
attempted
yet
Maybe
challeng
ingfor
interspecies
hybrids
FR(INRA)120
0geno
type
ofPop
ulus
nigraarebeingused
todevelop
intraspecificGS
(Con
tinues)
14 | CLIFTON‐BROWN ET AL
TABLE
5(Contin
ued)
Breeding
techno
logy
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sW
illow
Pop
lar
predictin
gthe
performance
ofprog
enyof
crosses
(and
inresearch
topredictbestparents
tousein
biparental
crosses)
geno
typedand
phenotyp
edto
developamod
elthattakesgeno
typic
datafrom
aldquocandidate
popu
latio
nrdquoof
untested
individu
als
andprod
uces
GEBVs(Jannink
Lorenzamp
Iwata
2010
)
beingph
enotyp
edand
geno
typed
Inthis
time
next‐generation
germ
plasm
andreadyto
beginthe
second
roun
dof
training
and
recalib
ratio
nof
geno
mic
predictio
nmod
els
US
Prelim
inary
mod
elsforMsac
and
Msin
developed
Work
isun
derw
ayto
deal
moreeffectivelywith
polyploidgenetics
calib
ratio
nsUS
125
0geno
types
arebeingused
todevelopGS
calib
ratio
ns
Traditio
nal
transgenesis
Efficient
introd
uctio
nof
ldquoforeign
rdquotraits
(possiblyfrom
other
genu
sspecies)
into
anelite
plant(ega
prov
enparent
ora
hybrid)that
needsa
simpletraitto
beim
prov
edValidate
cand
idategenesfrom
QTLstud
ies
MASand
know
ledg
eof
the
biolog
yof
thetrait
andsource
ofgenesto
confer
the
relevant
changesto
phenotyp
eWorking
transformation
protocol
IPissuescostof
regu
latory
approv
al
GMO
labelling
marketin
gissuestricky
tousetransgenes
inou
t‐breedersbecause
ofcomplexity
oftransformingandgene
flow
risks
US
Manyprog
ramsin
UnitedStates
are
creatin
gtransgenic
plantsusingAlamoas
asource
oftransformable
geno
typesManytraits
ofinterestNothing
commercial
yet
CNC
hang
shaBtg
ene
transformed
into
Msac
in20
04M
sin
transformed
with
MINAC2gene
and
Msactransform
edwith
Cry2A
ageneb
othwith
amarkerg
eneby
Agrob
acterium
JP1
Msintransgenic
forlow
temperature
tolerancewith
increased
expression
offructans
(unp
ublished)
NLImprov
ingprotocol
forM
sin
transformation
SK2
Msin
with
Arabido
psisand
Brachypod
ium
PhytochromeBgenes
(Hwang
Cho
etal
2014
Hwang
Lim
etal20
14)
UK2
Msin
and
1Mflorg
enotyp
es
UKRou
tine
transformationno
tyet
possibleResearchto
overcomerecalcitrance
ongo
ing
Currently
trying
differentspecies
andcond
ition
sPo
plar
transformationused
atpresent
US
Frequently
attempted
with
very
limitedsuccess
US
600transgenic
lines
have
been
characterized
mostly
performed
inthe
aspenhy
brids
reprod
uctiv
esterility
drou
ghttolerance
with
Kno
ckdo
wns
(Con
tinues)
CLIFTON‐BROWN ET AL | 15
TABLE
5(Contin
ued)
Breeding
techno
logy
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sW
illow
Pop
lar
variou
slytransformed
with
four
cellwall
genesiptand
uidA
genesusingtwo
selectionsystem
sby
biolisticsand
Agrob
acterium
Transform
edplantsare
beinganalysed
US
Prelim
inarywork
will
betakenforw
ardin
2018
ndash202
2in
CABBI
Genom
eediting
CRISPR
Refinem
entofexisting
traits
inusefulparents
orprom
isinghybrids
bygeneratingtargeted
mutations
ingenes
know
ntocontrolthe
traitof
interestF
irst
adouble‐stranded
breakismadeinthe
DNAwhich
isrepairedby
natural
DNArepair
machineryL
eads
tofram
eshiftSN
Por
can
usealdquorepair
templaterdquo
orcanbe
used
toinserta
transgene
intoaldquosafe
harbourlocusrdquoIt
couldbe
used
todeleterepressors(or
transcriptionfactors)
Identification
mapping
and
sequ
encing
oftarget
genes(from
DNA
sequ
ence)
avoiding
screening
outof
unintend
ededits
Manyregu
latory
authorities
have
not
decidedwhether
CRISPR
andother
geno
meediting
techno
logies
are
GMOsor
notIf
GMOsseecomment
onstage10
abov
eIf
noteditedcrop
swill
beregu
latedas
conv
entio
nalvarieties
CATechn
olog
yisstill
toonewand
switchg
rass
geno
meis
very
complexOther
labo
ratories
are
interestedbu
tno
tyet
mov
ingon
this
US
One
prog
rammeso
farmdash
USD
Ain
Albany
FRInitiated
in20
16(M
ISEDIT
project)
NLInitiatingin
2018
US
Initiatingin
2018
in
CABBI
UKNot
started
Not
possible
until
transformation
achieved
US
CRISPR
‐Cas9and
Cpf1have
been
successfully
developedin
Pop
ulus
andarehigh
lyefficientExamples
forlig
nin
biosyn
thesis
Note
BFF
BiomassFo
rtheFu
tureBtBacillus
thuringiensisCABBICenterforadvanced
bioenergyandbioproductsinnovatio
nCpf1
CRISPR
from
Prevotella
andFrancisella
1CRISPR
clusteredregularlyinterspaced
palin
drom
icrepeatsCRISPR
‐CasC
RISPR
‐associated
Cry2A
acrystaltoxins2A
asubfam
ilyproduced
byBtFS
full‐sibG
BS
genotyping
bysequencingG
EBVsgenomic‐estim
ated
breeding
valuesG
MOg
enetically
modi-
fied
organism
GS
genomicselection
GWAS
genomew
ideassociationstudy
INRAF
renchNationalInstituteforAgriculturalR
esearch
IPintellectualp
roperty
IPTisopentenyltransferase
ISSR
inter‐SSR
MflorMiscant-
husflo
ridulusMsacMsacchariflo
rusMsinMsinensisM
AS
marker‐a
ssistedselection
MISEDITm
iscanthusgene
editing
forseed‐propagatedtriploidsNACn
oapicalmeristemA
TAF1
2and
cup‐shaped
cotyledon2
‐lik
eQTLq
uantitativ
etraitlocusS
NP
singlenucleotid
epolymorphismS
SRsim
plesequence
repeatsUSD
AU
SDepartm
ento
fAgriculture
a ISO
Alpha‐2
lettercountrycodes
16 | CLIFTON‐BROWN ET AL
High seed production from field crossing trials con-ducted in locations where flowering in both seed andpollen parents is likely to happen synchronously
Scalable and adapted harvesting threshing and seed pro-cessing methods for producing high seed quality
The results of these parallel activities need to becombined to identify the upscaling pathway for eachhybrid if this cannot be achieved the hybrid will likelynot be commercially viable The UK‐led programme withpartners in Italy and Germany shows that seedbased mul-tiplication rates of 12000 are achievable several inter-specific hybrids (Clifton‐Brown et al 2017) Themultiplication rate of M sinensis is higher probably15000ndash10000 Conventional cloning from rhizome islimited to around 120 that is one ha could provide rhi-zomes for around 20 ha of new plantation
Multilocation field testing of wild and novel miscanthushybrids selected by breeding programmes in the Nether-lands and the United Kingdom was performed as part ofthe project Optimizing Miscanthus Biomass Production(OPTIMISC 2012ndash2016) These trials showed that com-mercial yields and biomass qualities (Kiesel et al 2017Van der Weijde et al 2017a Van der Weijde Kieselet al 2017) could be produced in a wide range of climatesand soil conditions from the temperate maritime climate ofwestern Wales to the continental climate of eastern Russiaand the Ukraine (Kalinina et al 2017) Extensive environ-mental measurements of soil and climate combined withgrowth monitoring are being used to understand abioticstresses (Nunn et al 2017 Van der Weijde Huxley et al2017) and develop genotype‐specific scenarios similar tothose reported earlier in Hastings et al (2009) Phenomicsexperiments on drought tolerance have been conducted onwild and improved germplasm (Malinowska Donnison ampRobson 2017 Van der Weijde Huxley et al 2017)Recently produced interspecific hybrids displaying excep-tional yield under drought (~30 greater than control Mxg)in field trials in Poland and Moldova are being furtherstudied in detail in the phenomics and genomics facility atAberystwyth to better understand genendashtrait associationswhich can be fed back into breeding
Intraspecific seeded hybrids of M sinensis produced inthe Netherlands and interspecific M sacchariflorus times M si-nensis hybrids produced by the UK‐led breeding programmehave entered yield testing in 2018 with the recently EU‐funded project ldquoGRowing Advanced industrial Crops on mar-ginal lands for biorEfineries (GRACE)rdquo (httpswwwgrace-bbieu) Substantial variation in biomass quality for sacchari-fication efficiency (glucose release as of dry matter) ashcontent and melting point has already been generated inintraspecific M sinensis hybrids (Van der Weijde Kiesel
et al 2017) across environments (Weijde Dolstra et al2017a) GRACE aims to establish more than 20 hectares ofnew inter‐ and intraspecific seeded hybrids across six Euro-pean countries This project is building the know‐how andagronomy needed to transition from small research plots tocommercial‐scale field sites and linking biomass productiondirectly to industrial applications The biomass produced byhybrids in different locations will be supplied to innovativeindustrial end‐users making a wide range of biobased prod-ucts both for chemicals and for energy In the United Statesmulti‐location yield were initiated in 2018 to evaluate new tri-ploid M times giganteus genotypes developed at Illinois Cur-rently infertile hybrids are favoured in the United Statesbecause this eliminates the risk of invasiveness from naturallydispersed viable seed The precautionary principle is appliedas fertile miscanthus has naturalized in several states (QuinnAllen amp Stewart 2010) North European multilocation fieldtrials in the EMI and OPTIMISC projects have shown thereis minimal risk of invasiveness even in years when fertileflowering hybrids produce viable seed Naturalized standshave not established here due perhaps to low dormancy pooroverwintering and low seedling competitive strength In addi-tion to breeding for nonshattering or sterile seeded hybridsQuinn et al (2010) suggest management strategies which canfurther minimize environmental opportunities to manage therisk of invasiveness
31 | Molecular breeding and biotechnology
In miscanthus new plant breeding techniques (Table 5)have focussed on developing molecular markers forbreeding in Europe the United States South Korea andJapan There are several publications on QTL mappingpopulations for key traits such as flowering (AtienzaRamirez amp Martin 2003) and compositional traits(Atienza Satovic Petersen Dolstra amp Martin 2003) Inthe United States and United Kingdom independent andinterconnected bi‐parental ldquomappingrdquo families have beenstudied (Dong et al 2018 Gifford Chae SwaminathanMoose amp Juvik 2015) alongside panels of diverse germ-plasm accessions for GWAS (Slavov et al 2013) Fur-ther developments calibrating GS with very large panelsof parents and cross progeny are underway (Daveyet al 2017) The recently completed first miscanthus ref-erence genome sequence is expected to improve the effi-ciency of MAS strategies and especially GWAS(httpsphytozomejgidoegovpzportalhtmlinfoalias=Org_Msinensis_er) For example without a referencegenome sequence Clark et al (2014) obtained 21207RADseq SNPs (single nucleotide polymorphisms) on apanel of 767 miscanthus genotypes (mostly M sinensis)but subsequent reanalysis of the RADseq data using the
CLIFTON‐BROWN ET AL | 17
new reference genome resulted in hundreds of thousandsof SNPs being called
Robust and effective in vitro regeneration systems havebeen developed for Miscanthus sinensis M times giganteusand M sacchariflorus (Dalton 2013 Guo et al 2013Hwang Cho et al 2014 Rambaud et al 2013 Ślusarkie-wicz‐Jarzina et al 2017 Wang et al 2011 Zhang et al2012) However there is still significant genotype speci-ficity and these methods need ldquoin‐houserdquo optimization anddevelopment to be used routinely They provide potentialroutes for rapid clonal propagation and also as a basis forgenetic transformation Stable transformation using bothbiolistics (Wang et al 2011) and Agrobacterium tumefa-ciens DNA delivery methods (Hwang Cho et al 2014Hwang Lim et al 2014) has been achieved in M sinen-sis The development of miscanthus transformation andgene editing to generate diplogametes for producing seed‐propagated triploid hybrids are performed as part of theFrench project MISEDIT (miscanthus gene editing forseed‐propagated triploids) There are no reports of genomeediting in any miscanthus species but new breeding inno-vations including genome editing are particularly relevantin this slow‐to‐breed nonfood bioenergy crop (Table 4)
4 | WILLOW
Willow (Salix spp) is a very diverse group of catkin‐bear-ing trees and shrubs Willow belongs to the family Sali-caceae which also includes the Populus genus There areapproximately 350 willow species (Argus 2007) foundmostly in temperate and arctic zones in the northern hemi-sphere A few are adapted to subtropical and tropicalzones The centre of diversity is believed to be in Asiawith over 200 species in China Around 120 species arefound in the former Soviet Union over 100 in NorthAmerica and around 65 species in Europe and one speciesis native to South America (Karp et al 2011) Willows aredioecious thus obligate outcrossers and highly heterozy-gous The haploid chromosome number is 19 (Hanley ampKarp 2014) Around 40 of species are polyploid (Sudaamp Argus 1968) ranging from triploids to the atypicaldodecaploid S maxxaliana with 2n=190 (Zsuffa et al1984)
Although almost exclusively native to the NorthernHemisphere willow has been grown around the globe formany thousands of years to support a wide range of appli-cations (Kuzovkina amp Quigley 2005 Stott 1992) How-ever it has been the focus of domestication for bioenergypurposes for only a relatively short period since the 1970sin North America and Europe For bioenergy breedershave focused their efforts on the shrub willows (subgenusVetix) because of their rapid juvenile growth rates as aresponse to coppicing on a 2‐ to 4‐year cycle that can be
accomplished using farm machinery rather than forestryequipment (Shield Macalpine Hanley amp Karp 2015Smart amp Cameron 2012)
Since shrub willow was not generally recognized as anagricultural crop until very recently there has been littlecommitment to building and maintaining germplasm reposi-tories of willow to support long‐term breeding One excep-tion is the United Kingdom where a large and well‐characterized Salix germplasm collection comprising over1500 accessions is held at Rothamsted Research (Stott1992 Trybush et al 2008) Originally initiated for use inbasketry in 1923 accessions have been added ever sinceIn the United States a germplasm collection of gt350accessions is located at Cornell University to support thebreeding programme there The UK and Cornell collectionshave a relatively small number of accessions in common(around 20) Taken together they represent much of thespecies diversity but only a small fraction of the overallgenetic diversity within the genus There are three activewillow breeding programmes in Europe RothamstedResearch (UK) Salixenergi Europa AB (SEE) and a pro-gramme at the University of Warmia and Mazury in Olsz-tyn (Poland) (abbreviations used in Table 1) There is oneactive US programme based at Cornell University Culti-vars are still being marketed by the European WillowBreeding Programme (EWBP) (UK) which was activelybreeding biomass varieties from 1996 to 2002 Cultivarsare protected by plant breedersrsquo rights (PBRs) in Europeand by plant patents in the United States The sharing ofgenetic resources in the willow community is generallyregulated by material transfer agreements (MTA) and tai-lored licensing agreements although the import of cuttingsinto North America is prohibited except under special quar-antine permit conditions
Efforts to augment breeding germplasm collection fromnature are continuing with phenotypic screening of wildgermplasm performed in field experiments with 177 S pur-purea genotypes in the United States (at sites in Genevaand Portland NY and Morgantown WV) that have beengenotyped using genotyping by sequencing (GBS) (Elshireet al 2011) In addition there are approximately 400accessions of S viminalis in Europe (near Pustnaumls Upp-sala Sweden and Woburn UK (Berlin et al 2014 Hal-lingbaumlck et al 2016) The S viminalis accessions wereinitially genotyped using 38 simple sequence repeats (SSR)markers to assess genetic diversity and screened with~1600 SNPs in genes of potential interest for phenologyand biomass traits Genetics and genomics combined withextensive phenotyping have substantially improved thegenetic basis of biomass‐related traits in willow and arenow being developed in targeted breeding via MAS Thisunderpinning work has been conducted on large specifi-cally developed biparental Salix mapping populations
18 | CLIFTON‐BROWN ET AL
(Hanley amp Karp 2014 Zhou et al 2018) as well asGWAS panels (Hallingbaumlck et al 2016)
Once promising parental combinations are identifiedcrosses are usually performed using fresh pollen frommaterial that has been subject to a phased removal fromcold storage (minus4degC) (Lindegaard amp Barker 1997 Macal-pine Shield Trybush Hayes amp Karp 2008 Mosseler1990) Pollen storage is useful in certain interspecific com-binations where flowering is not naturally synchronizedThis can be overcome by using pollen collection and stor-age protocol which involves extracting pollen using toluene(Kopp Maynard Niella Smart amp Abrahamson 2002)
The main breeding approach to improve willow yieldsrelies on species hybridization to capture hybrid vigour(Fabio et al 2017 Serapiglia Gouker amp Smart 2014) Inthe absence of genotypic models for heterosis breedershave extensively tested general and specific combiningability of parents to produce superior progeny The UKbreeding programmes (EWBP 1996ndash2002 and RothamstedResearch from 2003 on) have performed more than 1500exploratory cross‐pollinations The Cornell programme hassuccessfully completed about 550 crosses since 1998Investment into the characterization of genetic diversitycombined with progeny tests from exploratory crosses hasbeen used to produce hundreds of targeted intraspeciescrosses in the United Kingdom and United States respec-tively (see Table 1) To achieve long‐term gains beyond F1hybrids four intraspecific recurrent selection populationshave been created in the United Kingdom (for S dasycla-dos S viminalis and S miyabeana) and Cornell is pursu-ing recurrent selection of S purpurea Interspecifichybridizations with genotypes selected from the recurrentselection cycles are well advanced in willow with suchcrosses to date totalling 420 in the United Kingdom andover 100 in the United States
While species hybridization is common in Salix it isnot universal Of the crosses attempted about 50 hybri-dize and produce seed (Macalpine Shield amp Karp 2010)As the viability of seed from successful crosses is short (amatter of days at ambient temperatures) proper seed rear-ing and storage protocols are essential (Maroder PregoFacciuto amp Maldonado 2000)
Progeny from crosses are treated in different waysamong the breeding programmes at the seedling stage Inthe United States seedlings are planted into an irrigatedfield where plants are screened for two seasons beforebeing progressed to further field trials In the United King-dom seedlings are planted into trays of compost wherethey remain containerized in an irrigated nursery for theremainder of year one In the United Kingdom seedlingsare subject to two rounds of selection in the nursery yearThe first round takes place in September to select againstsusceptibility to rust infection (Melampsora spp) A second
round of selection in winter assesses tip damage from frostand giant willow aphid infestation In the United Stateswhere the rust pressure is lower screening for Melampsoraspp cannot be performed at the nursery stage Both pro-grammes monitor Melampsora spp pest susceptibilityyield and architecture over multiple years in field trialsSelected material is subject to two rounds of field trials fol-lowed by a final multilocation yield trial to identify vari-eties for commercialization
Promising selections (ie potential cultivars) need to beclonally propagated A rapid in vitro tissue culture propa-gation method has been developed (Palomo‐Riacuteos et al2015) This method can generate about 5000 viable trans-plantable clones from a single plant in just 24 weeks Anin vitro system can also accommodate early selection viamolecular or biochemical markers to increase selectionspeed Conventional breeding systems take 13 years viafour rounds of selection from crossing to selecting a variety(Figure 1) but this has the potential to be reduced to7 years if micropropagation and MAS selection are adopted(Hanley amp Karp 2014 Palomo‐Riacuteos et al 2015)
Willows are currently propagated commercially byplanting winter‐dormant stem cuttings in spring Commer-cial planting systems for willow use mechanical plantersthat cut and insert stem sections from whips into a well‐prepared soil One hectare of stock plants grown in specificmultiplication beds planted at 40000 plants per ha pro-duces planting material for 80 hectares of commercialshort‐rotation coppice willow annually (planted at 15000cuttings per hectare) (Whittaker et al 2016) When com-mercial plantations are established the industry standard isto plant intimate mixtures of ~5 diverse rust (Melampsoraspp)‐resistant varieties (McCracken amp Dawson 1997 VanDen Broek et al 2001)
The foundations for using new plant breeding tech-niques have been established with funding from both thepublic and the private sectors To establish QTL maps 16mapping populations from biparental crosses are understudy in the United Kingdom Nine are under study in theUnited States The average number of individuals in thesefamilies ranges from 150 to 947 (Hanley amp Karp 2014)GS is also being evaluated in S viminalis and preliminaryresults indicate that multiomic approaches combininggenomic and metabolomic data have great potential (Sla-vov amp Davey 2017) For both QTL and GS approachesthe field phenotyping demands are large as several thou-sand individuals need to be phenotyped for a wide rangeof traits These include the following dates of bud burstand growth succession stem height stem density wooddensity and disease resistance The greater the number ofindividuals the more precise the QTL marker maps andGS models are However the logistical and financial chal-lenges of phenotyping large numbers of individuals are
CLIFTON‐BROWN ET AL | 19
considerable because the willow crop is gt5 m tall in thesecond year There is tremendous potential to improve thethroughput of phenotyping using unmanned aerial systemswhich is being tested in the USDA National Institute ofFood and Agriculture (NIFA) Willow SkyCAP project atCornell Further investment in these approaches needs tobe sustained over many years fully realizes the potentialof a marker‐assisted selection programme for willow
To date despite considerable efforts in Europe and theUnited States to establish a routine transformation systemthere has not been a breakthrough in willow but attemptsare ongoing As some form of transformation is typically aprerequisite for genome editing techniques these have notyet been applied to willow
In Europe there are 53 short‐rotation coppice (SRC) bio-mass willow cultivars registered with the Community PlantVariety Office (CPVO) for PBRs of which ~25 are availablecommercially in the United Kingdom There are eightpatented cultivars commercially available in the UnitedStates In Sweden there are nine commercial cultivars regis-tered in Europe and two others which are unregistered(httpssalixenergiseplanting-material) Furthermore thereare about 20 precommercial hybrids in final yield trials inboth the United States and the United Kingdom It has beenestimated that it would take two years to produce the stockrequired to plant 50 ha commercially from the plant stock inthe final yield trials Breeding programmes have alreadydelivered rust‐resistant varieties and increases in yield to themarket The adoption of advanced breeding technologies willlikely lead to a step change in improving traits of interest
5 | POPLAR
Poplar a fast‐growing tree from the northern hemispherewith a small genome size has been adopted for commercialforestry and scientific purposes The genus Populus con-sists of about 29 species classified in six different sectionsPopulus (formerly Leuce) Tacamahaca Aigeiros AbasoTuranga and Leucoides (Eckenwalder 1996) The Populusspecies of most interest for breeding and testing in the Uni-ted States and Europe are P nigra P deltoides P maxi-mowiczii and P trichocarpa (Stanton 2014) Populusclones for biomass production are being developed byintra‐ and interspecies hybridization (DeWoody Trewin ampTaylor 2015 Richardson Isebrands amp Ball 2014 van derSchoot et al 2000) Recurrent selection approaches areused for gradual population improvement and to create eliteclonal lines for commercialization (Berguson McMahon ampRiemenschneider 2017 Neale amp Kremer 2011) Currentlypoplar breeding in the United States occurs in industrialand academic programmes located in the Southeast theMidwest and the Pacific Northwest These use six speciesand five interspecific taxa (Stanton 2014)
The southeastern programme historically focused onrecurrent selection of P deltoides from accessions made inthe lower Mississippi River alluvial plain (Robison Rous-seau amp Zhang 2006) More recently the genetic base hasbeen broadened to produce interspecific hybrids with resis-tance to the fungal infection Septoria musiva which causescankers
In the midwest of the United States populationimprovement efforts are focused on P deltoides selectionsfrom native provenances and hybrid crosses with acces-sions introduced from Europe Interspecific intercontinental(Europe and America) hybrid crosses between P nigra andP deltoides (P times canadensis) are behind many of theleading commercial hybrids which are the most advancedbreeding materials for many applications and regions InMinnesota previous breeding experience and efforts utiliz-ing P maximowiczii and P trichocarpa have been discon-tinued due to Septoria susceptibility and a lack of coldhardiness (Berguson et al 2017) Traits targeted forimprovement include yieldgrowth rate cold hardinessadventitious rooting resistance to Septoria and Melamp-sora leaf rust and stem form The Upper Midwest pro-gramme also carries out wide hybridizations within thesection Populus The P times wettsteinii (P trem-ula times P tremuloides) taxon is bred for gains in growthrate wood quality and resistance to the fungus Entoleucamammata which causes hypoxylon canker (David ampAnderson 2002)
In the Pacific Northwest GreenWood Resources Incleads poplar breeding that emphasizes interspecifichybrid improvement of P times generosa (P deltoides timesP trichocarpa and reciprocal) and P deltoides times P maxi-mowiczii taxa for coastal regions and the P times canadensistaxon for the drier continental regions Intraspecificimprovement of second‐generation breeding populations ofP deltoides P nigra P maximowiczii and P trichocarpaare also involved (Stanton et al 2010) The present focusof GreenWood Resourcesrsquo hybridization is bioenergy feed-stock improvement concentrating on coppice yield wood‐specific gravity and rate of sugar release
Industrial interest in poplar in the United States has his-torically come from the pulp and paper sector althoughveneer and dimensional lumber markets have been pursuedat times Currently the biomass market for liquid trans-portation fuels is being emphasized along with the use oftraditional and improved poplar genotypes for ecosystemservices such as phytoremediation (Tuskan amp Walsh 2001Zalesny et al 2016)
In Europe there are breeding programmes in FranceGermany Italy and Sweden These include the following(a) Alasia Franco Vivai (AFV) programme in northernItaly (b) the French programme led by the poplar Scien-tific Interest Group (GIS Peuplier) and carried out
20 | CLIFTON‐BROWN ET AL
collaboratively between the National Institute for Agricul-tural Research (INRA) the National Research Unit ofScience and Technology for Environment and Agriculture(IRSTEA) and the Forest Cellulose Wood Constructionand Furniture Technology Institute (FCBA) (c) the Ger-man programme at Northwest German Forest Research Sta-tion (NW‐FVA) at Hannoversch Muumlnden and (d) theSwedish programme at the Swedish University of Agricul-tural Sciences and SweTree Technologies AB (Table 1)
AFV leads an Italian poplar breeding programme usingextensive field‐grown germplasm collections of P albaP deltoides P nigra and P trichocarpa While interspeci-fic hybridization uses several taxa the focus is onP times canadensis The breeding programme addresses dis-ease resistance (Marssonina brunnea Melampsora larici‐populina Discosporium populeum and poplar mosaicvirus) growth rate and photoperiod adaptation AFV andGreenWood Resources collaborate in poplar improvementin Europe through the exchange of frozen pollen and seedfor reciprocal breeding projects Plantations in Poland andRomania are currently the focus of the collaboration
The ongoing French GIS Peuplier is developing a long‐term breeding programme based on intraspecific recurrentselection for the four parental species (P deltoides P tri-chocarpa P nigra and P maximowiczii) designed to betterbenefit from hybrid vigour demonstrated by the interspeci-fic crosses P canadensis P deltoides times P trichocarpaand P trichocarpa times P maximowiczii Current selectionpriorities are targeting adaptation to soil and climate condi-tions resistance and tolerance to the most economicallyimportant diseases and pests high volume production underSRC and traditional poplar cultivation regimes as well aswood quality of interest by different markets Currentlygenomic selection is under exploration to increase selectionaccuracy and selection intensity while maintaining geneticdiversity over generations
The German NW‐FVA programme is breeding intersec-tional AigeirosndashTacamahaca hybrids with a focus on resis-tance to Pollaccia elegans Xanthomonas populiDothichiza spp Marssonina brunnea and Melampsoraspp (Stanton 2014) Various cross combinations ofP maximowiczii P trichocarpa P nigra and P deltoideshave led to new cultivars suitable for deployment in vari-etal mixtures of five to ten genotypes of complementarystature high productivity and phenotypic stability (Weis-gerber 1993) The current priority is the selection of culti-vars for high‐yield short‐rotation biomass production Sixhundred P nigra genotypes are maintained in an ex situconservation programme An in situ P nigra conservationeffort involves an inventory of native stands which havebeen molecular fingerprinted for identity and diversity
The Swedish programme is concentrating on locallyadapted genotypes used for short‐rotation forestry (SRF)
because these meet the needs of the current pulping mar-kets Several field trials have shown that commercial poplarclones tested and deployed in Southern and Central Europeare not well adapted to photoperiods and low temperaturesin Sweden and in the Baltics Consequently SwedishUniversity of Agricultural Sciences and SweTree Technolo-gies AB started breeding in Sweden in 1990s to producepoplar clones better adapted to local climates and markets
51 | Molecular breeding technologies
Poplar genetic improvement cannot be rapidly achievedthrough traditional methods alone because of the longbreeding cycles outcrossing breeding systems and highheterozygosity Integrating modern genetic genomic andphenomics techniques with conventional breeding has thepotential to expedite poplar improvement
The genome of poplar has been sequenced (Tuskanet al 2006) It has an estimated genome size of485 plusmn 10 Mbp divided into 19 chromosomes This is smal-ler than other PBCs and makes poplar more amenable togenetic engineering (transgenesis) GS and genome editingPoplar has seen major investment in both the United Statesand Europe being the model system for woody perennialplant genetics and genomics research
52 | Targets for genetic modification
Traits targeted include wood properties (lignin content andcomposition) earlylate flowering male sterility to addressbiosafety regulation issues enhanced yield traits and herbi-cide tolerance These extensive transgenic experiments haveshown differences in recalcitrance to in vitro regenerationand genetic transformation in some of the most importantcommercial hybrid poplars (Alburquerque et al 2016)Further transgene expression stability is being studied Sofar China is the only country known to have commerciallyused transgenic insect‐resistant poplar A precommercialherbicide‐tolerant poplar was trialled for 8 years in the Uni-ted States (Li Meilan Ma Barish amp Strauss 2008) butcould not be released due to stringent environmental riskassessments required for regulatory approval This increasestranslation costs and delays reducing investor confidencefor commercial deployment (Harfouche et al 2011)
The first field trials of transgenic poplar were performedin France in 1987 (Fillatti Sellmer Mccown Haissig ampComai 1987) and in Belgium in 1988 (Deblock 1990)Although there have been a total of 28 research‐scale GMpoplar field trials approved in the European Union underCouncil Directive 90220EEC since October 1991 (inPoland Belgium Finland France Germany Spain Swe-den and in the United Kingdom (Pilate et al 2016) onlyauthorizations in Poland and Belgium are in place today In
CLIFTON‐BROWN ET AL | 21
the United States regulatory notifications and permits fornearly 20000 transgenic poplar trees derived from approxi-mately 600 different constructs have been issued since1995 by the USDAs Animal and Plant Health InspectionService (APHIS) (Strauss et al 2016)
53 | Genome editing CRISPR technologies
Clustered regularly interspaced palindromic repeats(CRISPR) and the CRISPR‐associated (CRISPR‐Cas)nucleases are a groundbreaking genome‐engineering toolthat complements classical plant breeding and transgenicmethods (Moreno‐Mateos et al 2017) Only two publishedstudies in poplar have applied the CRISPRCas9 technol-ogy One is in P tomentosa in which an endogenous phy-toene desaturase gene (PtoPDS) was successfully disruptedsite specifically in the first generation of transgenic plantsresulting in an albino and dwarf phenotype (Fan et al2015) The second was in P tremula times alba in whichhigh CRISPR‐Cas9 mutational efficiency was achieved forthree 4‐coumarateCoA ligase (4Cl) genes 4CL1 4CL2and 4CL5 associated with lignin and flavonoid biosynthe-sis (Zhou et al 2015) Due to its low cost precision andrapidness it is very probable that cultivars or clones pro-duced using CRISPR technology will be ready for market-ing in the near future (Yin et al 2017) Recently aCRISPR with a smaller associated endonuclease has beendiscovered from Prevotella and Francisella 1 (Cpf1) whichmay have advantages over Cas9 In addition there arereports of DNA‐free editing in plants using both CRISPRCpf1 and CRISPR Cas9 for example Ref (Kim et al2017 Mahfouz 2017 Zaidi et al 2017)
It remains unresolved whether plants modified by genomeediting will be regulated as genetically modified organisms(GMOs) by the relevant authorities in different countries(Lozano‐Juste amp Cutler 2014) Regulations to cover thesenew breeding techniques are still evolving but those coun-tries who have published specific guidance (including UnitedStates Argentina and Chile) are indicating that plants pos-sessing simple genome edits will not be regulated as conven-tional transgenesis (Jones 2015b) The first generation ofgenome‐edited crops will likely be phenocopy gene knock-outs that already exist to produce ldquonature identicalrdquo traits thatis traits that could also be derived by conventional breedingDespite this confidence in applying these new powerfulbreeding tools remains limited owing to the uncertain regula-tory environment in many parts of the world (Gao 2018)including the recent ECJ 2018 rulings mentioned earlier
54 | Genomics‐based breeding technologies
Poplar breeding programs are becoming well equipped withuseful genomics tools and resources that are critical to
explore genomewide variability and make use of the varia-tion for enhancing genetic gains Deep transcriptomesequencing resequencing of alternate genomes and GBStechnology for genomewide marker detection using next‐generation sequencing (NGS) are yielding valuable geno-mics tools GWAS with NGS‐based markers facilitatesmarker identification for MAS breeding by design and GS
GWAS approaches have provided a deeper understand-ing of genome function as well as allelic architectures ofcomplex traits (Huang et al 2010) and have been widelyimplemented in poplar for wood characteristics (Porthet al 2013) stomatal patterning carbon gain versus dis-ease resistance (McKown et al 2014) height and phenol-ogy (Evans et al 2014) cell wall chemistry (Mucheroet al 2015) growth and cell walls traits (Fahrenkroget al 2017) bark roughness (Bdeir et al 2017) andheight and diameter growth (Liu et al 2018) Using high‐throughput sequencing and genotyping platforms an enor-mous amount of SNP markers have been used to character-ize the linkage disequilibrium (LD) in poplar (eg Slavovet al 2012 discussed below)
The genetic architecture of photoperiodic traits in peren-nial trees is complex involving many loci However itshows high levels of conservation during evolution (Mau-rya amp Bhalerao 2017) These genomics tools can thereforebe used to address adaptation issues and fine‐tune themovement of elite lines into new environments For exam-ple poor timing of spring bud burst and autumn bud setcan result in frost damage resulting in yield losses (Ilstedt1996) These have been studied in P tremula genotypesalong a latitudinal cline in Sweden (~56ndash66oN) and haverevealed high nucleotide polymorphism in two nonsynony-mous SNPs within and around the phytochrome B2 locus(Ingvarsson Garcia Hall Luquez amp Jansson 2006 Ing-varsson Garcia Luquez Hall amp Jansson 2008) Rese-quencing 94 of these P tremula genotypes for GWASshowed that noncoding variation of a single genomicregion containing the PtFT2 gene described 65 ofobserved genetic variation in bud set along the latitudinalcline (Tan 2018)
Resequencing genomes is currently the most rapid andeffective method detecting genetic differences betweenvariants and for linking loci to complex and importantagronomical and biomass traits thus addressing breedingchallenges associated with long‐lived plants like poplars
To date whole genome resequencing initiatives havebeen launched for several poplar species and genotypes InPopulus LD studies based on genome resequencing sug-gested the feasibility of GWAS in undomesticated popula-tions (Slavov et al 2012) This plant population is beingused to inform breeding for bioenergy development Forexample the detection of reliable phenotypegenotype asso-ciations and molecular signatures of selection requires a
22 | CLIFTON‐BROWN ET AL
detailed knowledge about genomewide patterns of allelefrequency variation LD and recombination suggesting thatGWAS and GS in undomesticated populations may bemore feasible in Populus than previously assumed Slavovet al (2012) have resequenced 16 genomes of P tri-chocarpa and genotyped 120 trees from 10 subpopulationsusing 29213 SNPs (Geraldes et al 2013) The largest everSNP data set of genetic variations in poplar has recentlybeen released providing useful information for breedinghttpswwwbioenergycenterorgbescgwasindexcfmAlso deep sequencing of transcriptomes using RNA‐Seqhas been used for identification of functional genes andmolecular markers that is polymorphism markers andSSRs A multitissue and multiple experimental data set forP trichocarpa RNA‐Seq is publicly available (httpsjgidoegovdoe-jgi-plant-flagship-gene-atlas)
The availability of genomic information of DNA‐con-taining cell organelles (nucleus chloroplast and mitochon-dria) will also allow a holistic approach in poplarmolecular breeding in the near future (Kersten et al2016) Complete Populus genome sequences are availablefor nucleus (P trichocarpa section Tacamahaca) andchloroplasts (seven species and two clones from P trem-ula W52 and P tremula times P alba 717ndash1B4) A compara-tive approach revealed structural and functionalinformation broadening the knowledge base of PopuluscpDNA and stimulating future diagnostic marker develop-ment The availability of whole genome sequences of thesecellular compartments of P tremula holds promise forboosting marker‐assisted poplar breeding Other nucleargenome sequences from additional Populus species arenow available (eg P deltoides (httpsphytozomejgidoegovpz) and will become available in the forthcomingyears (eg P tremula and P tremuloidesmdashPopGenIE(Sjodin Street Sandberg Gustafsson amp Jansson 2009))Recently the characterization of the poplar pan‐genome bygenomewide identification of structural variation in threecrossable poplar species P nigra P deltoides and P tri-chocarpa revealed a deeper understanding of the role ofinter‐ and intraspecific structural variants in poplar pheno-type and may have important implications for breedingparticularly interspecific hybrids (Pinosio et al 2016)
GS has been proposed as an alternative to MAS in cropimprovement (Bernardo amp Yu 2007 Heffner Sorrells ampJannink 2009) GS is particularly well suited for specieswith long generation times for characteristics that displaymoderate‐to‐low heritability for traits that are expensive tomeasure and for selection of traits expressed late in the lifecycle as is the case for most traits of commercial value inforestry (Harfouche et al 2012) Current joint genomesequencing efforts to implement GS in poplar using geno-mic‐estimated breeding values for bioenergy conversiontraits from 49 P trichocarpa families and 20 full‐sib
progeny are taking place at the Oak Ridge National Labo-ratory and GreenWood Resources (Brian Stanton personalcommunication httpscbiornlgov) These data togetherwith the resequenced GWAS population data will be thebasis for developing GS algorithms Genomic breedingtools have been developed for the intraspecific programmetargeting yield resistance to Venturia shoot blightMelampsora leaf rust resistance to Cryptorhynchus lapathistem form wood‐specific gravity and wind firmness (Evanset al 2014 Guerra et al 2016) A newly developedldquobreeding with rare defective allelesrdquo (BRDA) technologyhas been developed to exploit natural variation of P nigraand identify defective variants of genes predicted by priortransgenic research to impact lignin properties Individualtrees carrying naturally defective alleles can then be incor-porated directly into breeding programs thereby bypassingthe need for transgenics (Vanholme et al 2013) Thisnovel breeding technology offers a reverse genetics com-plement to emerging GS for targeted improvement of quan-titative traits (Tsai 2013)
55 | Phenomics‐assisted breeding technology
Phenomics involves the characterization of phenomesmdashthefull set of phenotypes of given individual plants (HouleGovindaraju amp Omholt 2010) Traditional phenotypingtools which inefficiently measure a limited set of pheno-types have become a bottleneck in plant breeding studiesHigh‐throughput plant phenotyping facilities provide accu-rate screening of thousands of plant breeding lines clones orpopulations over time (Fu 2015) are critical for acceleratinggenomics‐based breeding Automated image collection andanalysis phenomics technologies allow accurate and nonde-structive measurements of a diversity of phenotypic traits inlarge breeding populations (Gegas Gay Camargo amp Doo-nan 2014 Goggin Lorence amp Topp 2015 Ludovisi et al2017 Shakoor Lee amp Mockler 2017) One important con-sideration is the identification of relevant and quantifiabletarget traits that are early diagnostic indicators of biomassyield Good progress has been made in elucidating theseunderpinning morpho‐physiological traits that are amenableto remote sensing in Populus (Harfouche Meilan amp Altman2014 Rae Robinson Street amp Taylor 2004) Morerecently Ludovisi et al (2017) developed a novel methodol-ogy for field phenomics of drought stress in a P nigra F2partially inbred population using thermal infrared imagesrecorded from an unmanned aerial vehicle‐based platform
Energy is the current main market for poplar biomass butthe market return provided is not sufficient to support pro-duction expansion even with added demand for environmen-tal and land management ldquoecosystem servicesrdquo such as thetreatment of effluent phytoremediation riparian buffer zonesand agro‐forestry plantings Aviation fuel is a significant
CLIFTON‐BROWN ET AL | 23
target market (Crawford et al 2016) To serve this marketand to reduce current carbon costs of production (Budsberget al 2016) key improvement traits in addition to yield(eg coppice regeneration pestdisease resistance water‐and nutrient‐use efficiencies) will be trace greenhouse gas(GHG) emissions (eg isoprene volatiles) site adaptabilityand biomass conversion efficiency Efforts are underway tohave national environmental protection agenciesrsquo approvalfor poplar hybrids qualifying for renewable energy credits
6 | REFLECTIONS ON THECOMMERCIALIZATIONCHALLENGE
The research and innovation activities reviewed in thispaper aim to advance the genetic improvement of speciesthat can provide feedstocks for bioenergy applicationsshould those markets eventually develop These marketsneed to generate sufficient revenue and adequately dis-tribute it to the actors along the value chain The work onall four crops shares one thing in common long‐termefforts to integrate fundamental knowledge into breedingand crop development along a research and development(RampD) pipeline The development of miscanthus led byAberystwyth University exemplifies the concerted researcheffort that has integrated the RampD activities from eight pro-jects over 14 years with background core research fundingalong an emerging innovation chain (Figure 2) This pro-gramme has produced a first range of conventionally bredseeded interspecies hybrids which are now in upscaling tri-als (Table 3) The application of molecular approaches(Table 5) with further conventional breeding (Table 4)offers the prospect of a second range of improved seededhybrids This example shows that research‐based support ofthe development of new crops or crop types requires along‐term commitment that goes beyond that normallyavailable from project‐based funding (Figure 1) Innovationin this sector requires continuous resourcing of conven-tional breeding operations and capability to minimize timeand investment losses caused by funding discontinuities
This challenge is increased further by the well‐knownmarket failure in the breeding of many agricultural cropspecies The UK Department for Environment Food andRural Affairs (Defra) examined the role of geneticimprovement in relation to nonmarket outcomes such asenvironmental protection and concluded that public invest-ment in breeding was required if profound market failure isto be addressed (Defra 2002) With the exception ofwidely grown hybrid crops such as maize and some high‐value horticultural crops royalties arising from plant breed-ersrsquo rights or other returns to breeders fail to adequatelycompensate for the full cost for research‐based plant breed-ing The result even for major crops such as wheat is sub‐
optimal investment and suboptimal returns for society Thismarket failure is especially acute for perennial crops devel-oped for improved sustainability rather than consumerappeal (Tracy et al 2016) Figure 3 illustrates the underly-ing challenge of capturing value for the breeding effortThe ldquovalley of deathrdquo that results from the low and delayedreturns to investment applies generally to the research‐to‐product innovation pipeline (Beard Ford Koutsky amp Spi-wak 2009) and certainly to most agricultural crop speciesHowever this schematic is particularly relevant to PBCsMost of the value for society from the improved breedingof these crops comes from changes in how agricultural landis used that is it depends on the increased production ofthese crops The value for society includes many ecosys-tems benefits the effects of a return to seminatural peren-nial crop cover that protects soils the increase in soilcarbon storage the protection of vulnerable land or the cul-tivation of polluted soils and the reductions in GHG emis-sions (Lewandowski 2016) By its very nature theproduction of biomass on agricultural land marginal forfood production challenges farm‐level profitability Thecosts of planting material and one‐time nature of cropestablishment are major early‐stage costs and therefore theopportunities for conventional royalty capture by breedersthat are manifold for annual crops are limited for PBCs(Hastings et al 2017) Public investment in developingPBCs for the nonfood biobased sector needs to providemore long‐term support for this critical foundation to a sus-tainable bioeconomy
7 | CONCLUSIONS
This paper provides an overview of research‐based plantbreeding in four leading PBCs For all four PBC generasignificant progress has been made in genetic improvementthrough collaboration between research scientists and thoseoperating ongoing breeding programmes Compared withthe main food crops most PBC breeding programmes dateback only a few decades (Table 1) This breeding efforthas thus co‐evolved with molecular biology and the result-ing ‐omics technologies that can support breeding Thedevelopment of all four PBCs has depended strongly onpublic investment in research and innovation The natureand driver of the investment varied In close associationwith public research organizations or universities all theseprogrammes started with germplasm collection and charac-terization which underpin the selection of parents forexploratory wide crosses for progeny testing (Figure 1Table 4)
Public support for switchgrass in North America wasexplicitly linked to plant breeding with 12 breeding pro-grammes supported in the United States and CanadaSwitchgrass breeding efforts to date using conventional
24 | CLIFTON‐BROWN ET AL
breeding have resulted in over 36 registered cultivars inthe United States (Table 1) with the development of dedi-cated biomass‐type cultivars coming within the past fewyears While ‐omics technologies have been incorporatedinto several of these breeding programmes they have notyet led to commercial deployment in either conventional orhybrid cultivars
Willow genetic improvement was led by the researchcommunity closely linked to plant breeding programmesWillow and poplar have the longest record of public invest-ment in genetic improvement that can be traced back to1920s in the United Kingdom and United States respec-tively Like switchgrass breeding programmes for willoware connected to public research efforts The United King-dom in partnership with the programme based at CornellUniversity remains the European leader in willowimprovement with a long‐term breeding effort closelylinked to supporting biological research at Rothamsted Inwillow F1 hybrids have produced impressive yield gainsover parental germplasm by capturing hybrid vigour Over30 willow clones are commercially available in the UnitedStates and Europe and a further ~90 are under precommer-cial testing (Table 1)
Compared with willow and poplar miscanthus is a rela-tive newcomer with all the current breeding programmesstarting in the 2000s Clonal M times giganteus propagated byrhizomes is expected to be replaced by more readily scal-able seeded hybrids from intra‐ (M sinensis) and inter‐(M sacchariflorus times M sinensis) species crosses with highseed multiplication rates (of gt2000) The first group ofhybrid cultivars is expected to be market‐ready around2022
Of the four genera used as PBCs Populus is the mostadvanced in terms of achievements in biological researchas a result of its use as a model for basic research of treesMuch of this biological research is not directly connectedto plant breeding Nevertheless reflecting the fact thatpoplar is widely grown as a single‐stem tree in SRF thereare about 60 commercially available clones and an addi-tional 80 clones in commercial pipelines (Table 1) Trans-genic poplar hybrids have moved beyond proof of conceptto commercial reality in China
Many PBC programmes have initiated long‐term con-ventional recurrent selection breeding cycles for populationimprovement which is a key process in increasing yieldthrough hybrid vigour As this approach requires manyyears most programmes are experimenting with molecularbreeding methods as these have the potential to accelerateprecision breeding For all four PBCs investments in basicgenetic and genomic resources including the developmentof mapping populations for QTLs and whole genomesequences are available to support long‐term advancesMore recently association genetics with panels of diversegermplasm are being used as training populations for GSmodels (Table 5) These efforts are benefitting from pub-licly available DNA sequences and whole genome assem-blies in crop databases Key to these accelerated breedingtechnologies are developments in novel phenomics tech-nologies to bridge the genotypephenotype gap In poplarnovel remote sensing field phenotyping is now beingdeployed to assist breeders These advances are being com-bined with in vitro and in planta modern molecular breed-ing techniques such as CRISPR (Table 5) CRISPRtechnology for genome editing has been proven in poplar
FIGURE 2 A schematic developmentpathway for miscanthus in the UnitedKingdom related to the investment in RampDprojects at Aberystwyth (top coloured areasfor projects in the three categories basicresearch breeding and commercialupscaling) leading to a projected croppingarea of 350000 ha by 2030 with clonal andsuccessive ranges of improved seed‐basedhybrids Purple represents theBiotechnology and Biological SciencesResearch Council (BBSRC) and brown theDepartment for Environment Food andRural Affairs (Defra) (UK Nationalfunding) blue bars represent EU fundingand green private sector funding (Terravestaand CERES) and GIANT‐LINK andMiscanthus Upscaling Technology (MUST)are publicndashprivate‐initiatives (PPI)
CLIFTON‐BROWN ET AL | 25
This technology is also being applied in switchgrass andmiscanthus (Table 5) but the future of CRISPR in com-mercial breeding for the European market is uncertain inthe light of recent ECJ 2018 rulings
There is integration of research and plant breeding itselfin all four PBCs Therefore estimating the ongoing costs ofmaintaining these breeding programmes is difficult Invest-ment in research also seeks wider benefits associated withtechnological advances in plant science rather than cultivardevelopment per se However in all cases the conventionalbreeding cycle shown in Figure 1 is the basic ldquoenginerdquo withmolecular technologies (‐omics) serving to accelerate thisengine The history of the development shows that the exis-tence of these breeding programmes is essential to gain ben-efits from the biological research Despite this it is thisessential step that is at most risk from reductions in invest-ment A conventional breeding programme typicallyrequires a breeder and several technicians who are supported
over the long term (20ndash30 years Figure 1) at costs of about05 to 10 million Euro per year (as of 2018) The analysisreported here shows that the time needed to perform onecycle of conventional breeding bringing germplasm fromthe wild to a commercial hybrid ranged from 11 years inswitchgrass to 26 years in poplar (Figure 1) In a maturecrop grown on over 100000 ha with effective cultivar pro-tection and a suitable business model this level of revenuecould come from royalties Until such levels are reachedPBCs lie in the innovation valley of death (Figure 3) andneed public support
Applying industrial ldquotechnology readiness levelsrdquo (TRL)originally developed for aerospace (Heacuteder 2017) to ourplant breeding efforts we estimate many promising hybridscultivars are at TRL levels of 3ndash4 In Table 3 experts ineach crop estimate that it would take 3 years from now toupscale planting material from leading cultivars in plot tri-als to 100 ha
FIGURE 3 A schematic relating some of the steps in the innovation chain from relatively basic crop science research through to thedeployment in commercial cropping systems and value chains The shape of the funnel above the expanding development and deploymentrepresents the availability of investment along the development chain from relatively basic research at the top to the upscaled deployment at thebottom Plant breeding links the research effort with the development of cropping systems The constriction represents the constrained fundingfor breeding that links conventional public research investment and the potential returns from commercial development The handover pointsbetween publicly funded work to develop the germplasm resources (often known as prebreeding) the breeding and the subsequent cropdevelopment are shown on the left The constriction point is aggravated by the lack academic rewards for this essential breeding activity Theoutcome is such that this innovation system is constrained by the precarious resourcing of plant breeding The authorsrsquo assessment ofdevelopment status of the four species is shown (poplar having two one for short‐rotation coppice (SRC) poplar and one for the more traditionalshort‐rotation forestry (SRF)) The four new perennial biomass crops (PBCs) are now in the critical phase of depending of plant breedingprogress without the income stream from a large crop production base
26 | CLIFTON‐BROWN ET AL
Taking the UK example mentioned in the introductionplanting rates of ~35000 ha per year from 2022 onwardsare needed to reach over 1 m ha by 2050 Ongoing workin the UK‐funded Miscanthus Upscaling Technology(MUST) project shows that ramping annual hybrid seedproduction from the current level of sufficient seed for10 ha in 2018 to 35000 ha would take about 5 yearsassuming no setbacks If current hybrids of any of the fourPBCs in the upscaling pipeline fail on any step for exam-ple lower than expected multiplication rates or unforeseenagronomic barriers then further selections from ongoingbreeding are needed to replace earlier candidates
In conclusion the breeding foundations have been laidwell for switchgrass miscanthus willow and poplar owingto public funding over the long time periods necessaryImproved cultivars or genotypes are available that could bescaled up over a few years if real sustained market oppor-tunities emerged in response to sustained favourable poli-cies and industrial market pull The potential contributionsof growing and using these PBCs for socioeconomic andenvironmental benefits are clear but how farmers andothers in commercial value chains are rewarded for mass‐scale deployment as is necessary is not obvious at presentTherefore mass‐scale deployment of these lignocellulosecrops needs developments outside the breeding arenas todrive breeding activities more rapidly and extensively
8 | UNCITED REFERENCE
Huang et al (2019)
ACKNOWLEDGEMENTS
UK The UK‐led miscanthus research and breeding wasmainly supported by the Biotechnology and BiologicalSciences Research Council (BBSRC) Department for Envi-ronment Food and Rural Affairs (Defra) the BBSRC CSPstrategic funding grant BBCSP17301 Innovate UKBBSRC ldquoMUSTrdquo BBN0161491 CERES Inc and Ter-ravesta Ltd through the GIANT‐LINK project (LK0863)Genomic selection and genomewide association study activi-ties were supported by BBSRC grant BBK01711X1 theBBSRC strategic programme grant on Energy Grasses ampBio‐refining BBSEW10963A01 The UK‐led willow RampDwork reported here was supported by BBSRC (BBSEC00005199 BBSEC00005201 BBG0162161 BBE0068331 BBG00580X1 and BBSEC000I0410) Defra(NF0424 NF0426) and the Department of Trade and Indus-try (DTI) (BW6005990000) IT The Brain Gain Program(Rientro dei cervelli) of the Italian Ministry of EducationUniversity and Research supports Antoine Harfouche USContributions by Gerald Tuskan to this manuscript were sup-ported by the Center for Bioenergy Innovation a US
Department of Energy Bioenergy Research Center supportedby the Office of Biological and Environmental Research inthe DOE Office of Science under contract number DE‐AC05‐00OR22725 Willow breeding efforts at CornellUniversity have been supported by grants from the USDepartment of Agriculture National Institute of Food andAgriculture Contributions by the University of Illinois weresupported primarily by the DOE Office of Science Office ofBiological and Environmental Research (BER) grant nosDE‐SC0006634 DE‐SC0012379 and DE‐SC0018420 (Cen-ter for Advanced Bioenergy and Bioproducts Innovation)and the Energy Biosciences Institute EU We would like tofurther acknowledge contributions from the EU projectsldquoOPTIMISCrdquo FP7‐289159 on miscanthus and ldquoWATBIOrdquoFP7‐311929 on poplar and miscanthus as well as ldquoGRACErdquoH2020‐EU326 Biobased Industries Joint Technology Ini-tiative (BBI‐JTI) Project ID 745012 on miscanthus
CONFLICT OF INTEREST
The authors declare that progress reported in this paperwhich includes input from industrial partners is not biasedby their business interests
ORCID
John Clifton‐Brown httporcidorg0000-0001-6477-5452Antoine Harfouche httporcidorg0000-0003-3998-0694Lawrence B Smart httporcidorg0000-0002-7812-7736Danny Awty‐Carroll httporcidorg0000-0001-5855-0775VasileBotnari httpsorcidorg0000-0002-0470-0384Lindsay V Clark httporcidorg0000-0002-3881-9252SalvatoreCosentino httporcidorg0000-0001-7076-8777Lin SHuang httporcidorg0000-0003-3079-326XJon P McCalmont httporcidorg0000-0002-5978-9574Paul Robson httporcidorg0000-0003-1841-3594AnatoliiSandu httporcidorg0000-0002-7900-448XDaniloScordia httporcidorg0000-0002-3822-788XRezaShafiei httporcidorg0000-0003-0891-0730Gail Taylor httporcidorg0000-0001-8470-6390IrisLewandowski httporcidorg0000-0002-0388-4521
REFERENCES
Alburquerque N Baldacci‐Cresp F Baucher M Casacuberta JM Collonnier C ElJaziri M hellip Burgos L (2016) New trans-formation technologies for trees In C Vettori F Gallardo H
CLIFTON‐BROWN ET AL | 27
Haumlggman V Kazana F Migliacci G Pilateamp M Fladung(Eds) Biosafety of forest transgenic trees (pp 31ndash66) DordrechtNetherlands Springer
Argus G W (2007) Salix (Salicaceae) distribution maps and a syn-opsis of their classification in North America north of MexicoHarvard Papers in Botany 12 335ndash368 httpsdoiorg1031001043-4534(2007)12[335SSDMAA]20CO2
Atienza S G Ramirez M C amp Martin A (2003) Mapping‐QTLscontrolling flowering date in Miscanthus sinensis Anderss CerealResearch Communications 31 265ndash271
Atienza S G Satovic Z Petersen K K Dolstra O amp Martin A(2003) Identification of QTLs influencing agronomic traits inMiscanthus sinensis Anderss I Total height flag‐leaf height andstem diameter Theoretical and Applied Genetics 107 123ndash129
Atienza S G Satovic Z Petersen K K Dolstra O amp Martin A(2003) Influencing combustion quality in Miscanthus sinensisAnderss Identification of QTLs for calcium phosphorus and sul-phur content Plant Breeding 122 141ndash145
Bdeir R Muchero W Yordanov Y Tuskan G A Busov V ampGailing O (2017) Quantitative trait locus mapping of Populusbark features and stem diameter BMC Plant Biology 17 224httpsdoiorg101186s12870-017-1166-4
Beard T R Ford G S Koutsky T M amp Spiwak L J (2009) AValley of Death in the innovation sequence An economic investi-gation Research Evaluation 18 343ndash356 httpsdoiorg103152095820209X481057
Berguson W McMahon B amp Riemenschneider D (2017) Addi-tive and non‐additive genetic variances for tree growth in severalhybrid poplar populations and implications regarding breedingstrategy Silvae Genetica 66(1) 33ndash39 httpsdoiorg101515sg-2017-0005
Berlin S Trybush S O Fogelqvist J Gyllenstrand N Halling-baumlck H R Aringhman I hellip Hanley S J (2014) Genetic diversitypopulation structure and phenotypic variation in European Salixviminalis L (Salicaceae) Tree Genetics amp Genomes 10 1595ndash1610 httpsdoiorg101007s11295-014-0782-5
Bernardo R amp Yu J (2007) Prospects for genomewide selectionfor quantitative traits in maize Crop Science 47 1082ndash1090httpsdoiorg102135cropsci2006110690
Biswal A K Atmodjo M A Li M Baxter H L Yoo C GPu Y hellip Mohnen D (2018) Sugar release and growth of bio-fuel crops are improved by downregulation of pectin biosynthesisNature Biotechnology 36 249ndash257
Bmu (2009) National biomass action plan for Germany (p 17) Bun-desministerium fuumlr Umwelt Naturschutz und ReaktorsicherheitBundesministerium fuumlr Ernaumlhrung (BMU) amp Landwirtschaft undVerbraucherschutz (BMELV) 11055 Berlin | Germany Retrievedfrom httpwwwbmeldeSharedDocsDownloadsENPublicationsBiomassActionPlanpdf__blob=publicationFile
Brummer E C (1999) Capturing heterosis in forage crop cultivardevelopment Crop Science 39 943ndash954 httpsdoiorg102135cropsci19990011183X003900040001x
Budsberg E Crawford J T Morgan H Chin W S Bura R ampGustafson R (2016) Hydrocarbon bio‐jet fuel from bioconver-sion of poplar biomass Life cycle assessment Biotechnology forBiofuels 9 170 httpsdoiorg101186s13068-016-0582-2
Burris K P Dlugosz E M Collins A G Stewart C N amp Lena-ghan S C (2016) Development of a rapid low‐cost protoplasttransfection system for switchgrass (Panicum virgatum L) Plant
Cell Reports 35 693ndash704 httpsdoiorg101007s00299-015-1913-7
Casler M (2012) Switchgrass breeding genetics and genomics InA Monti (Ed) Switchgrass (pp 29ndash54) London UK Springer
Casler M Mitchell R amp Vogel K (2012) Switchgrass In CKole C P Joshi amp D R Shonnard (Eds) Handbook of bioen-ergy crop plants Vol 2 New York NY Taylor amp Francis
Casler M D amp Ramstein G P (2018) Breeding for biomass yieldin switchgrass using surrogate measures of yield BioEnergyResearch 11 6ndash12
Casler M D Sosa S Hoffman L Mayton H Ernst C Adler PR hellip Bonos S A (2017) Biomass Yield of Switchgrass Culti-vars under High‐ versus Low‐Input Conditions Crop Science 57821ndash832 httpsdoiorg102135cropsci2016080698
Casler M D amp Vogel K P (2014) Selection for biomass yield inupland lowland and hybrid switchgrass Crop Science 54 626ndash636 httpsdoiorg102135cropsci2013040239
Casler M D Vogel K P amp Harrison M (2015) Switchgrassgermplasm resources Crop Science 55 2463ndash2478 httpsdoiorg102135cropsci2015020076
Casler M D Vogel K P Lee D K Mitchell R B Adler P RSulc R M hellip Moore K J (2018) 30 Years of progress towardincreasing biomass yield of switchgrass and big bluestem CropScience 58 1242
Clark L V Brummer J E Głowacka K Hall M C Heo K PengJ hellip Sacks E J (2014) A footprint of past climate change on thediversity and population structure of Miscanthus sinensis Annals ofBotany 114 97ndash107 httpsdoiorg101093aobmcu084
Clark L V Dzyubenko E Dzyubenko N Bagmet L SabitovA Chebukin P hellip Sacks E J (2016) Ecological characteristicsand in situ genetic associations for yield‐component traits of wildMiscanthus from eastern Russia Annals of Botany 118 941ndash955
Clark L V Jin X Petersen K K Anzoua K G Bagmet LChebukin P hellip Sacks E J(2018) Population structure of Mis-canthus sacchariflorus reveals two major polyploidization eventstetraploid‐mediated unidirectional introgression from diploid Msinensis and diversity centred around the Yellow Sea Annals ofBotany 20 1ndash18 httpsdoiorg101093aobmcy1161
Clark L V Stewart J R Nishiwaki A Toma Y Kjeldsen J BJoslashrgensen U hellip Sacks E J (2015) Genetic structure ofMiscanthus sinensis and Miscanthus sacchariflorus in Japan indi-cates a gradient of bidirectional but asymmetric introgressionJournal of Experimental Botany 66 4213ndash4225
Clifton‐Brown J Hastings A Mos M McCalmont J P Ashman CAwty‐Carroll DhellipCracroft‐EleyW (2017) Progress in upscalingMis-canthus biomass production for the European bio‐economy with seed‐based hybridsGlobal Change Biology Bioenergy 9 6ndash17
Clifton‐Brown J Schwarz K U amp Hastings A (2015) History ofthe development of Miscanthus as a bioenergy crop From smallbeginnings to potential realisation Biology and Environment‐Pro-ceedings of the Royal Irish Academy 115B 45ndash57
Clifton‐Brown J Senior H Purdy S Horsnell R Lankamp BMuumlennekhoff A-K hellip Bentley A R (2018) Investigating thepotential of novel non‐woven fabrics to increase pollination effi-ciency in plant breeding PloS One (Accepted)
Crawford J T Shan CW Budsberg E Morgan H Bura R ampGus-tafson R (2016) Hydrocarbon bio‐jet fuel from bioconversion ofpoplar biomass Techno‐economic assessment Biotechnology forBiofuels 9 141 httpsdoiorg101186s13068-016-0545-7
28 | CLIFTON‐BROWN ET AL
Dalton S (2013) Biotechnology of Miscanthus In S M Jain amp SD Gupta (Eds) Biotechnology of neglected and underutilizedcrops (pp 243ndash294) Dordrecht Netherlands Springer
Davey C L Robson P Hawkins S Farrar K Clifton‐Brown J CDonnison I S amp Slavov G T (2017) Genetic relationshipsbetween spring emergence canopy phenology and biomass yieldincrease the accuracy of genomic prediction in Miscanthus Journalof Experimental Botany 68 5093ndash5102 httpsdoiorg101093jxberx339
David X amp Anderson X (2002) Aspen and larch genetics cooper-ative annual report 13 St Paul MN Department of ForestResources University of Minnesota
Deblock M (1990) Factors influencing the tissue‐culture and theAgrobacterium‐Tumefaciens‐mediated transformation of hybridaspen and poplar clones Plant Physiology 93 1110ndash1116httpsdoiorg101104pp9331110
Defra (2002) The role of future public research investment in thegenetic improvement of UK‐grown crops (p 220) LondonRetrieved from sciencesearchdefragovukDefaultaspxMenu=MenuampModule=MoreampLocation=NoneampCompleted=220ampProjectID=10412
Dewoody J Trewin H amp Taylor G (2015) Genetic and morpho-logical differentiation in Populus nigra L Isolation by coloniza-tion or isolation by adaptation Molecular Ecology 24 2641ndash2655
Dong H Liu S Clark L V Sharma S Gifford J M Juvik JA hellip Sacks E J (2018) Genetic mapping of biomass yield inthree interconnected Miscanthus populations Global Change Biol-ogy Bioenergy 10 165ndash185
Dukes (2017) Digest of UK Energy Statistics (DUKES) (p 264) Lon-don UK Department for Business Energy amp Industrial StrategyRetrieved from wwwgovukgovernmentcollectionsdigest-of-uk-energy-statistics-dukes
Eckenwalder J (1996) Systematics and evolution of Populus In RStettler H Bradshaw J Heilman amp T Hinckley (Eds) Biologyof Populus and its implications for management and conservation(pp 7‐32) Ottawa ON National Research Council of Canada
Elshire R J Glaubitz J C Sun Q Poland J A Kawamoto KBuckler E S amp Mitchell S E (2011) A robust simple geno-typing‐by‐sequencing (GBS) approach for high diversity speciesPloS One 6 e19379
Evans H (2016) Bioenergy crops in the UK Case studies of suc-cessful whole farm integration (p 23) Loughborough UKEnergy Technologies Institute Retrieved from httpswwweticouklibrarybioenergy-crops-in-the-uk-case-studies-on-successful-whole-farm-integration-evidence-pack
Evans H (2017) Increasing UK biomass production through moreproductive use of land (p 7) Loughborough UK Energy Tech-nologies Institute Retrieved from httpswwweticouklibraryan-eti-perspective-increasing-uk-biomass-production-through-more-productive-use-of-land
Evans J Sanciangco M D Lau K H Crisovan E Barry KDaum C hellip Buell C R (2018) Extensive genetic diversity ispresent within North American switchgrass germplasm The PlantGenome 11 1ndash16 httpsdoiorg103835plantgenome2017060055
Evans L M Slavov G T Rodgers‐Melnick E Martin J RanjanP Muchero W hellip DiFazio S P (2014) Population genomicsof Populus trichocarpa identifies signatures of selection and
adaptive trait associations Nature Genetics 46 1089ndash1096httpsdoiorg101038ng3075
Fabio E S Volk T A Miller R O Serapiglia M J Gauch HG Van Rees K C J hellip Smart L B (2017) Genotypetimes envi-ronment interaction analysis of North American shrub willowyield trials confirms superior performance of triploid hybrids Glo-bal Change Biology Bioenergy 9 445ndash459 httpsdoiorg101111gcbb12344
Fahrenkrog A M Neves L G Resende M F R Vazquez A Ide los Campos G Dervinis C hellip Kirst M (2017) Genome‐wide association study reveals putative regulators of bioenergytraits in Populus deltoides New Phytologist 213 799ndash811
Fan D Liu T T Li C F Jiao B Li S Hou Y S amp Luo KM (2015) Efficient CRISPRCas9‐mediated targeted mutagenesisin Populus in the first generation Scientific Reports 5 12217
Feng X P Lourgant K Castric V Saumitou-Laprade P ZhengB S Jiang D M amp Brancourt-Hulmel M (2014) The discov-ery of natural accessions related to Miscanthus x giganteus usingchloroplast DNA Crop Science 54 1645ndash1655
Fillatti J J Sellmer J Mccown B Haissig B amp Comai L(1987) Agrobacterium-mediated transformation and regenerationof Populus Molecular amp General Genetics 206 192ndash199
Fu Y B (2015) Understanding crop genetic diversity under modernplant breeding Theoretical and Applied Genetics 128 2131ndash2142 httpsdoiorg101007s00122-015-2585-y
Fu C Mielenz J R Xiao X Ge Y Hamilton C Y RodriguezM hellip Wang Z‐Y (2011) Genetic manipulation of ligninreduces recalcitrance and improves ethanol production fromswitchgrass Proceedings of the National Academy of Sciences ofthe United States of America 108 3803ndash3808 httpsdoiorg101073pnas1100310108
Gao C (2018) The future of CRISPR technologies in agricultureNature Reviews Molecular Cell Biology 19(5) 275ndash276
Gegas V Gay A Camargo A amp Doonan J (2014) Challengesof Crop Phenomics in the Post-genomic Era In J Hancock (Ed)Phenomics (pp 142ndash171) Boca Raton FL CRC Press
Geraldes A DiFazio S P Slavov G T Ranjan P Muchero WHannemann J hellip Tuskan G A (2013) A 34K SNP genotypingarray for Populus trichocarpa Design application to the study ofnatural populations and transferability to other Populus speciesMolecular Ecology Resources 13 306ndash323
Gifford J M Chae W B Swaminathan K Moose S P amp JuvikJ A (2015) Mapping the genome of Miscanthus sinensis forQTL associated with biomass productivity Global Change Biol-ogy Bioenergy 7 797ndash810
Goggin F L Lorence A amp Topp C N (2015) Applying high‐throughput phenotyping to plant‐insect interactions Picturingmore resistant crops Current Opinion in Insect Science 9 69ndash76httpsdoiorg101016jcois201503002
Grabowski P P Evans J Daum C Deshpande S Barry K WKennedy M hellip Casler M D (2017) Genome‐wide associationswith flowering time in switchgrass using exome‐capture sequenc-ing data New Phytologist 213 154ndash169 httpsdoiorg101111nph14101
Greef J M amp Deuter M (1993) Syntaxonomy of Miscanthus xgiganteus GREEF et DEU Angewandte Botanik 67 87ndash90
Guerra F P Richards J H Fiehn O Famula R Stanton B JShuren R hellip Neale D B (2016) Analysis of the genetic varia-tion in growth ecophysiology and chemical and metabolomic
CLIFTON‐BROWN ET AL | 29
composition of wood of Populus trichocarpa provenances TreeGenetics amp Genomes 12 6 httpsdoiorg101007s11295-015-0965-8
Guo H P Shao R X Hong C T Hu H K Zheng B S ampZhang Q X (2013) Rapid in vitro propagation of bioenergy cropMiscanthus sacchariflorus Applied Mechanics and Materials 260181ndash186
Hallingbaumlck H R Fogelqvist J Powers S J Turrion‐Gomez JRossiter R Amey J hellip Roumlnnberg‐Waumlstljung A‐C (2016)Association mapping in Salix viminalis L (Salicaceae)ndashIdentifica-tion of candidate genes associated with growth and phenologyGlobal Change Biology Bioenergy 8 670ndash685
Hanley S J amp Karp A (2014) Genetic strategies for dissectingcomplex traits in biomass willows (Salix spp) Tree Physiology34 1167ndash1180 httpsdoiorg101093treephystpt089
Harfouche A Meilan R amp Altman A (2011) Tree genetic engi-neering and applications to sustainable forestry and biomass pro-duction Trends in Biotechnology 29 9ndash17 httpsdoiorg101016jtibtech201009003
Harfouche A Meilan R amp Altman A (2014) Molecular and phys-iological responses to abiotic stress in forest trees and their rele-vance to tree improvement Tree Physiology 34 1181ndash1198httpsdoiorg101093treephystpu012
Harfouche A Meilan R Kirst M Morgante M Boerjan WSabatti M amp Mugnozza G S (2012) Accelerating the domesti-cation of forest trees in a changing world Trends in PlantScience 17 64ndash72 httpsdoiorg101016jtplants201111005
Hastings A Clifton‐Brown J Wattenbach M Mitchell C PStampfl P amp Smith P (2009) Future energy potential of Mis-canthus in Europe Global Change Biology Bioenergy 1 180ndash196
Hastings A Mos M Yesufu J AMccalmont J Schwarz KShafei RhellipClifton-Brown J (2017) Economic and environmen-tal assessment of seed and rhizome propagated Miscanthus in theUK Frontiers in Plant Science 8 16 httpsdoiorg103389fpls201701058
Heacuteder M (2017) From NASA to EU The evolution of the TRLscale in Public Sector Innovation The Innovation Journal 22 1ndash23 httpsdoiorg103389fpls201701058
Heffner E L Sorrells M E amp Jannink J L (2009) Genomicselection for crop improvement Crop Science 49 1ndash12 httpsdoiorg102135cropsci2008080512
Hodkinson T R Klaas M Jones M B Prickett R amp Barth S(2015) Miscanthus A case study for the utilization of naturalgenetic variation Plant Genetic Resources‐Characterization andUtilization 13 219ndash237 httpsdoiorg101017S147926211400094X
Hodkinson T R amp Renvoize S (2001) Nomenclature of Miscant-hus xgiganteus (Poaceae) Kew Bulletin 56 759ndash760 httpsdoiorg1023074117709
Houle D Govindaraju D R amp Omholt S (2010) Phenomics Thenext challenge Nature Reviews Genetics 11 855ndash866 httpsdoiorg101038nrg2897
Huang X H Wei X H Sang T Zhao Q Feng Q Zhao Yhellip Han B (2010) Genome‐wide association studies of 14 agro-nomic traits in rice landraces Nature Genetics 42 961ndashU976
Hwang O‐J Cho M‐A Han Y‐J Kim Y‐M Lim S‐H KimD‐S hellip Kim J‐I (2014) Agrobacterium‐mediated genetic trans-formation of Miscanthus sinensis Plant Cell Tissue and OrganCulture (PCTOC) 117 51ndash63
Hwang O‐J Lim S‐H Han Y‐J Shin A‐Y Kim D‐S ampKim J‐I (2014) Phenotypic characterization of transgenic Mis-canthus sinensis plants overexpressing Arabidopsis phytochromeB International Journal of Photoenergy 2014501016
Ilstedt B (1996) Genetics and performance of Belgian poplar clonestested in Sweden International Journal of Forest Genetics 3183ndash195
Ingvarsson P K Garcia M V Hall D Luquez V amp Jansson S(2006) Clinal variation in phyB2 a candidate gene for day‐length‐induced growth cessation and bud set across a latitudinalgradient in European aspen (Populus tremula) Genetics 1721845ndash1853
Ingvarsson P K Garcia M V Luquez V Hall D amp Jansson S(2008) Nucleotide polymorphism and phenotypic associationswithin and around the phytochrome B2 locus in European aspen(Populus tremula Salicaceae) Genetics 178 2217ndash2226 httpsdoiorg101534genetics107082354
Jannink J‐L Lorenz A J amp Iwata H (2010) Genomic selectionin plant breeding From theory to practice Briefings in FunctionalGenomics 9 166ndash177 httpsdoiorg101093bfgpelq001
Jiang J Guan Y Mccormick S Juvik J Lubberstedt T amp FeiS‐Z (2017) Gametophytic self‐incompatibility is operative inMiscanthus sinensis (Poaceae) and is affected by pistil age CropScience 57 1948ndash1956
Jones H D (2015a) Future of breeding by genome editing is in thehands of regulators GM Crops amp Food 6 223ndash232
Jones H D (2015b) Regulatory uncertainty over genome editingNature Plants 1 14011
Kalinina O Nunn C Sanderson R Hastings A F S van der Weijde TOumlzguumlven MhellipClifton‐Brown J C (2017) ExtendingMiscanthus cul-tivation with novel germplasm at six contrasting sites Frontiers in PlantScience 8563 httpsdoiorg103389fpls201700563
Karp A Hanley S J Trybush S O Macalpine W Pei M ampShield I (2011) Genetic improvement of willow for bioenergyand biofuels free access Journal of Integrative Plant Biology 53151ndash165 httpsdoiorg101111j1744-7909201001015x
Kersten B Faivre Rampant P Mader M Le Paslier M‐C Bou-non R Berard A hellip Fladung M (2016) Genome sequences ofPopulus tremula chloroplast and mitochondrion Implications forholistic poplar breeding PloS One 11 e0147209 httpsdoiorg101371journalpone0147209
Kiesel A Nunn C Iqbal Y Van der Weijde T Wagner MOumlzguumlven M hellip Lewandowski I (2017) Site‐specific manage-ment of Miscanthus genotypes for combustion and anaerobicdigestion A comparison of energy yields Frontiers in PlantScience 8 347 httpsdoiorg103389fpls201700347
Kim H Kim S T Ryu J Kang B C Kim J S amp Kim S G(2017) CRISPRCpf1‐mediated DNA‐free plant genome editingNature Communications 8 14406 httpsdoiorg101038ncomms14406
King Z R Bray A L Lafayette P R amp Parrott W A (2014)Biolistic transformation of elite genotypes of switchgrass (Pan-icum virgatum L) Plant Cell Reports 33 313ndash322 httpsdoiorg101007s00299-013-1531-1
Kopp R F Maynard C A De Niella P R Smart L B amp Abra-hamson L P (2002) Collection and storage of pollen from Salix(Salicaceae) American Journal of Botany 89 248ndash252
Krzyżak J Pogrzeba M Rusinowski S Clifton‐Brown J McCal-mont J P Kiesel A hellip Mos M (2017) Heavy metal uptake
30 | CLIFTON‐BROWN ET AL
by novel miscanthus seed‐based hybrids cultivated in heavy metalcontaminated soil Civil and Environmental Engineering Reports26 121ndash132 httpsdoiorg101515ceer-2017-0040
Kuzovkina Y A amp Quigley M F (2005) Willows beyond wet-lands Uses of Salix L species for environmental projects WaterAir and Soil Pollution 162 183ndash204 httpsdoiorg101007s11270-005-6272-5
Lazarus W Headlee W L amp Zalesny R S (2015) Impacts of sup-plyshed‐level differences in productivity and land costs on the eco-nomics of hybrid poplar production in Minnesota USA BioEnergyResearch 8 231ndash248 httpsdoiorg101007s12155-014-9520-y
Lewandowski I (2015) Securing a sustainable biomass supply in agrowing bioeconomy Global Food Security 6 34ndash42 httpsdoiorg101016jgfs201510001
Lewandowski I (2016) The role of perennial biomass crops in agrowing bioeconomy In S Barth D Murphy‐Bokern O Kalin-ina G Taylor amp M Jones (Eds) Perennial biomass crops for aresource‐constrained world (pp 3ndash13) Cham SwitzerlandSpringer Switzerland AG
Li J L Meilan R Ma C Barish M amp Strauss S H (2008)Stability of herbicide resistance over 8 years of coppice in field‐grown genetically engineered poplars Western Journal of AppliedForestry 23 89ndash93
Li R Y amp Qu R D (2011) High throughput Agrobacterium‐medi-ated switchgrass transformation Biomass amp Bioenergy 35 1046ndash1054 httpsdoiorg101016jbiombioe201011025
Lindegaard K N amp Barker J H A (1997) Breeding willows forbiomass Aspects of Applied Biology 49 155ndash162
Linde‐Laursen I B (1993) Cytogenetic analysis of MiscanthusGiganteus an interspecific hybrid Hereditas 119 297ndash300httpsdoiorg101111j1601-5223199300297x
Liu Y Merrick P Zhang Z Z Ji C H Yang B amp Fei S Z(2018) Targeted mutagenesis in tetraploid switchgrass (Panicumvirgatum L) using CRISPRCas9 Plant Biotechnology Journal16 381ndash393
Liu L L Wu Y Q Wang Y W amp Samuels T (2012) A high‐density simple sequence repeat‐based genetic linkage map ofswitchgrass G3 (Bethesda Md) 2 357ndash370
Lovett A Suumlnnenberg G amp Dockerty T (2014) The availabilityof land for perennial energy crops in Great Britain GlobalChange Biology Bioenergy 6 99ndash107 httpsdoiorg101111gcbb12147
Lozano‐Juste J amp Cutler S R (2014) Plant genome engineering infull bloom Trends in Plant Science 19 284ndash287 httpsdoiorg101016jtplants201402014
Lu F Lipka A E Glaubitz J Elshire R Cherney J H CaslerM D hellip Costich D E (2013) Switchgrass Genomic DiversityPloidy and Evolution Novel Insights from a Network‐Based SNPDiscovery Protocol Plos Genetics 9 e1003215
Ludovisi R Tauro F Salvati R Khoury S Mugnozza G S ampHarfouche A (2017) UAV‐based thermal imaging for high‐throughput field phenotyping of black poplar response to droughtFrontiers in Plant Science 81681
Macalpine W Shield I amp Karp A (2010) Seed to near market vari-ety the BEGIN willow breeding pipeline 2003ndash2010 and beyond InBioten conference proceedings Birmingham pp 21ndash23
Macalpine W J Shield I F Trybush S O Hayes C M ampKarp A (2008) Overcoming barriers to crossing in willow (Salixspp) breeding Aspects of Applied Biology 90 173ndash180
Mahfouz M M (2017) The efficient tool CRISPR‐Cpf1 NaturePlants 3 17028
Malinowska M Donnison I S amp Robson P R (2017) Phenomicsanalysis of drought responses in Miscanthus collected from differ-ent geographical locations Global Change Biology Bioenergy 978ndash91
Maroder H L Prego I A Facciuto G R amp Maldonado S B(2000) Storage behaviour of Salix alba and Salix matsudanaseeds Annals of Botany 86 1017ndash1021 httpsdoiorg101006anbo20001265
Martinez‐Reyna J amp Vogel K (2002) Incompatibility systems inswitchgrass Crop Science 42 1800ndash1805 httpsdoiorg102135cropsci20021800
Maurya J P amp Bhalerao R P (2017) Photoperiod‐and tempera-ture‐mediated control of growth cessation and dormancy in treesA molecular perspective Annals of Botany 120 351ndash360httpsdoiorg101093aobmcx061
McCalmont J P Hastings A Mcnamara N P Richter G MRobson P Donnison I S amp Clifton‐Brown J (2017) Environ-mental costs and benefits of growing Miscanthus for bioenergy inthe UK Global Change Biology Bioenergy 9 489ndash507
McCracken A R amp Dawson W M (1997) Using mixtures of wil-low clones as a means of controlling rust disease Aspects ofApplied Biology 49 97ndash103
McKown A D Klaacutepště J Guy R D Geraldes A Porth I Hanne-mann JhellipDouglas C J (2014) Genome‐wide association implicatesnumerous genes underlying ecological trait variation in natural popula-tions ofPopulus trichocarpaNew Phytologist 203 535ndash553
Merrick P amp Fei S Z (2015) Plant regeneration and genetic transforma-tion in switchgrass ndash A review Journal of Integrative Agriculture 14483ndash493 httpsdoiorg101016S2095-3119(14)60921-7
Moreno‐Mateos M A Fernandez J P Rouet R Lane M AVejnar C E Mis E hellip Giraldez A J (2017) CRISPR‐Cpf1mediates efficient homology‐directed repair and temperature‐con-trolled genome editing Nature Communications 8 2024 httpsdoiorg101038s41467-017-01836-2
Mosseler A (1990) Hybrid performance and species crossabilityrelationships in willows (Salix) Canadian Journal of Botany 682329ndash2338
Muchero W Guo J DiFazio S P Chen J‐G Ranjan P Slavov GT hellip Tuskan G A (2015) High‐resolution genetic mapping ofallelic variants associated with cell wall chemistry in Populus BMCGenomics 16 24 httpsdoiorg101186s12864-015-1215-z
Neale D B amp Kremer A (2011) Forest tree genomics Growingresources and applications Nature Reviews Genetics 12 111ndash122 httpsdoiorg101038nrg2931
Nunn C Hastings A F S J Kalinina O Oumlzguumlven M SchuumlleH Tarakanov I G hellip Clifton‐Brown J C (2017) Environmen-tal influences on the growing season duration and ripening ofdiverse Miscanthus germplasm grown in six countries Frontiersin Plant Science 8 907 httpsdoiorg103389fpls201700907
Ogawa Y Honda M Kondo Y amp Hara‐Nishimura I (2016) Anefficient Agrobacterium‐mediated transformation method forswitchgrass genotypes using Type I callus Plant Biotechnology33 19ndash26
Ogawa Y Shirakawa M Koumoto Y Honda M Asami Y KondoY ampHara‐Nishimura I (2014) A simple and reliable multi‐gene trans-formation method for switchgrass Plant Cell Reports 33 1161ndash1172httpsdoiorg101007s00299-014-1605-8
CLIFTON‐BROWN ET AL | 31
Okada M Lanzatella C Saha M C Bouton J Wu R L amp Tobias CM (2010) Complete switchgrass genetic maps reveal subgenomecollinearity preferential pairing and multilocus interactions Genetics185 745ndash760 httpsdoiorg101534genetics110113910
Palomo‐Riacuteos E Macalpine W Shield I Amey J Karaoğlu C WestJ hellip Jones H D (2015) Efficient method for rapid multiplication ofclean and healthy willow clones via in vitro propagation with broadgenotype applicability Canadian Journal of Forest Research 451662ndash1667 httpsdoiorg101139cjfr-2015-0055
Pilate G Allona I Boerjan W Deacutejardin A Fladung M Gal-lardo F hellip Halpin C (2016) Lessons from 25 years of GM treefield trials in Europe and prospects for the future In C Vettori FGallardo H Haumlggman V Kazana F Migliacci G Pilateamp MFladung (Eds) Biosafety of forest transgenic trees (pp 67ndash100)Dordrecht Netherlands Springer Switzerland AG
Pinosio S Giacomello S Faivre-Rampant P Taylor G Jorge VLe Paslier M C amp Marroni F (2016) Characterization of thepoplar pan-genome by genome-wide identification of structuralvariation Molecular Biology and Evolution 33 2706ndash2719
Porth I Klaacutepště J Skyba O Friedmann M C Hannemann JEhlting J hellip Douglas C J (2013) Network analysis reveals therelationship among wood properties gene expression levels andgenotypes of natural Populus trichocarpa accessions New Phytol-ogist 200 727ndash742
Pugesgaard S Schelde K Larsen S U Laeligrke P E amp JoslashrgensenU (2015) Comparing annual and perennial crops for bioenergyproductionndashinfluence on nitrate leaching and energy balance Glo-bal Change Biology Bioenergy 7 1136ndash1149 httpsdoiorg101111gcbb12215
Quinn L D Allen D J amp Stewart J R (2010) Invasivenesspotential of Miscanthus sinensis Implications for bioenergy pro-duction in the United States Global Change Biology Bioenergy2 310ndash320 httpsdoiorg101111j1757-1707201001062x
Rae A M Robinson K M Street N R amp Taylor G (2004)Morphological and physiological traits influencing biomass pro-ductivity in short‐rotation coppice poplar Canadian Journal ofForest Research‐Revue Canadienne De Recherche Forestiere 341488ndash1498 httpsdoiorg101139x04-033
Rambaud C Arnoult S Bluteau A Mansard M C Blassiau Camp Brancourt‐Hulmel M (2013) Shoot organogenesis in threeMiscanthus species and evaluation for genetic uniformity usingAFLP analysis Plant Cell Tissue and Organ Culture (PCTOC)113 437ndash448 httpsdoiorg101007s11240-012-0284-9
Ramstein G P Evans J Kaeppler S M Mitchell R B Vogel K PBuell C R amp Casler M D (2016) Accuracy of genomic prediction inswitchgrass (Panicum virgatum L) improved by accounting for linkagedisequilibriumG3 (Bethesda Md) 6 1049ndash1062
Rayburn A L Crawford J Rayburn C M amp Juvik J A (2009)Genome size of three Miscanthus species Plant Molecular Biol-ogy Reporter 27 184 httpsdoiorg101007s11105-008-0070-3
Richardson J Isebrands J amp Ball J (2014) Ecology and physiol-ogy of poplars and willows In J Isebrands amp J Richardson(Eds) Poplars and willows Trees for society and the environ-ment (pp 92‐123) Boston MA and Rome CABI and FAO
Robison T L Rousseau R J amp Zhang J (2006) Biomass produc-tivity improvement for eastern cottonwood Biomass amp Bioenergy30 735ndash739 httpsdoiorg101016jbiombioe200601012
Robson P R Farrar K Gay A P Jensen E F Clifton‐Brown JC amp Donnison I S (2013) Variation in canopy duration in the
perennial biofuel crop Miscanthus reveals complex associationswith yield Journal of Experimental Botany 64 2373ndash2383
Robson P Jensen E Hawkins S White S R Kenobi K Clif-ton‐Brown J hellip Farrar K (2013) Accelerating the domestica-tion of a bioenergy crop Identifying and modelling morphologicaltargets for sustainable yield increase in Miscanthus Journal ofExperimental Botany 64 4143ndash4155
Sanderson M A Adler P R Boateng A A Casler M D ampSarath G (2006) Switchgrass as a biofuels feedstock in theUSA Canadian Journal of Plant Science 86 1315ndash1325httpsdoiorg104141P06-136
Scarlat N Dallemand J F Monforti‐Ferrario F amp Nita V(2015) The role of biomass and bioenergy in a future bioecon-omy Policies and facts Environmental Development 15 3ndash34httpsdoiorg101016jenvdev201503006
Serapiglia M J Gouker F E amp Smart L B (2014) Early selec-tion of novel triploid hybrids of shrub willow with improved bio-mass yield relative to diploids BMC Plant Biology 14 74httpsdoiorg1011861471-2229-14-74
Serba D Wu L Daverdin G Bahri B A Wang X Kilian Ahellip Devos K M (2013) Linkage maps of lowland and upland tet-raploid switchgrass ecotypes BioEnergy Research 6 953ndash965httpsdoiorg101007s12155-013-9315-6
Shakoor N Lee S amp Mockler T C (2017) High throughput phe-notyping to accelerate crop breeding and monitoring of diseases inthe field Current Opinion in Plant Biology 38 184ndash192 httpsdoiorg101016jpbi201705006
Shield I Macalpine W Hanley S amp Karp A (2015) Breedingwillow for short rotation coppice energy cropping In Industrialcrops Breeding for BioEnergy and Bioproducts (pp 67ndash80) NewYork NY Springer
Sjodin A Street N R Sandberg G Gustafsson P amp Jansson S (2009)The Populus Genome Integrative Explorer (PopGenIE) A new resourcefor exploring the Populus genomeNewPhytologist 182 1013ndash1025
Slavov G amp Davey C (2017) Integrated genomic prediction inbioenergy crops In Abstracts from the International Conferenceon Developing biomass crops for future climates pp 34 24ndash27September 2017 Oriel College Oxford wwwwatbioeu
Slavov G Davey C Bosch M Robson P Donnison I ampMackay I (2018a) Genomic index selection provides a pragmaticframework for setting and refining multi‐objective breeding targetsin Miscanthus Annals of Botany (in press)
Slavov G T DiFazio S P Martin J Schackwitz W MucheroW Rodgers‐Melnick E hellip Tuskan G A (2012) Genome rese-quencing reveals multiscale geographic structure and extensivelinkage disequilibrium in the forest tree Populus trichocarpa NewPhytologist 196 713ndash725
Slavov G Davey C Robson P Donnison I amp Mackay I(2018b) Domestication of Miscanthus through genomic indexselection In Plant and Animal Genome Conference 13 January2018 San Diego CA Retrieved from httpspagconfexcompagxxvimeetingappcgiPaper30622
Slavov G T Nipper R Robson P Farrar K Allison G GBosch M hellip Jensen E (2013) Genome‐wide association studiesand prediction of 17 traits related to phenology biomass and cellwall composition in the energy grass Miscanthus sinensis NewPhytologist 201 1227ndash1239
Slavov G Robson P Jensen E Hodgson E Farrar K AllisonG hellip Donnison I (2014) Contrasting geographic patterns of
32 | CLIFTON‐BROWN ET AL
genetic variation for molecular markers vs phenotypic traits in theenergy grass Miscanthus sinensis Global Change Biology Bioen-ergy 5 562ndash571
Ślusarkiewicz‐Jarzina A Ponitka A Cerazy‐Waliszewska J Woj-ciechowicz M K Sobańska K Jeżowski S amp Pniewski T(2017) Effective and simple in vitro regeneration system of Mis-canthus sinensis Mtimes giganteus and M sacchariflorus for plant-ing and biotechnology purposes Biomass and Bioenergy 107219ndash226 httpsdoiorg101016jbiombioe201710012
Smart L amp Cameron K (2012) Shrub willow In C Kole S Joshiamp D Shonnard (Eds) Handbook of bioenergy crop plants (pp687ndash708) Boca Raton FL Taylor and Francis Group
Stanton B J (2014) The domestication and conservation of Populusand Salix genetic resources (Chapter 4) In Isebrands J ampRichardson J (Eds) Poplars and willows Trees for society andthe environment (pp 124ndash200) Rome Italy CAB InternationalFood and Agricultural Organization of the United Nations
Stanton B J Neale D B amp Li S (2010) Populus breeding Fromthe classical to the genomic approach In S Jansson R Bha-lerao amp A T Groover (Eds) Genetics and genomics of popu-lus (pp 309ndash348) New York NY Springer
Stewart J R Toma Y Fernandez F G Nishiwaki A Yamada T ampBollero G (2009) The ecology and agronomy ofMiscanthus sinensisa species important to bioenergy crop development in its native range inJapan A reviewGlobal Change Biology Bioenergy 1 126ndash153
Stott K G (1992) Willows in the service of man Proceedings of the RoyalSociety of Edinburgh Section B Biological Sciences 98 169ndash182
Strauss S H Ma C Ault K amp Klocko A L (2016) Lessonsfrom two decades of field trials with genetically modified trees inthe USA Biology and regulatory compliance In C Vettori FGallardo H Haumlggman V Kazana F Migliacci G Pilate amp MFladung (Eds) Biosafety of forest transgenic trees (pp 101ndash124)Dordrecht Netherlands Springer
Suda Y amp Argus G W (1968) Chromosome numbers of some NorthAmerican Salix Brittonia 20 191ndash197 httpsdoiorg1023072805440
Tan B (2018) Genomic selection and genome‐wide association studies todissect quantitative traits in forest trees Umearing Sweden University
Tornqvist C Taylor M Jiang Y Evans J Buell C KaepplerS amp Casler M (2018) Quantitative trait locus mapping for flow-ering time in a lowland x upland switchgrass pseudo‐F2 popula-tion The Plant Genome 11 170093
Toacuteth G Hermann T Da Silva M amp Montanarella L (2016)Heavy metals in agricultural soils of the European Union withimplications for food safety Environment International 88 299ndash309 httpsdoiorg101016jenvint201512017
Tracy W Dawson J Moore V amp Fisch J (2016) Intellectual propertyrights and public plant breeding Recommendations and proceedings ofa conference on best practices for intellectual property protection of pub-lically developed plant germplasm (p 70) University of Wisconsin‐Madison Retrieved from httpshostcalswisceduagronomywp-contentuploadssites1620162005Proceedings-IPR-Finalpdf
Trybush S Jahodovaacute Š Macalpine W amp Karp A (2008) A geneticstudy of a Salix germplasm resource reveals new insights into rela-tionships among subgenera sections and species BioEnergyResearch 1 67ndash79 httpsdoiorg101007s12155-008-9007-9
Tsai C J (2013) Next‐generation sequencing for next‐generationbreeding and more New Phytologist 198 635ndash637 httpsdoiorg101111nph12245
Tuskan G A Difazio S Jansson S Bohlmann J Grigoriev I HellstenU hellip Rokhsar D (2006) The genome of black cottonwood Populustrichocarpa (Torr ampGray) Science 313 1596ndash1604
Tuskan G A amp Walsh M E (2001) Short‐rotation woody cropsystems atmospheric carbon dioxide and carbon management AUS case study The Forestry Chronicle 77 259ndash264 httpsdoiorg105558tfc77259-2
Van Den Broek R Treffers D‐J Meeusen M Van Wijk ANieuwlaar E amp Turkenburg W (2001) Green energy or organicfood A life‐cycle assessment comparing two uses of set‐asideland Journal of Industrial Ecology 5 65ndash87 httpsdoiorg101162108819801760049477
Van Der Schoot J Pospiskova M Vosman B amp Smulders M J M(2000) Development and characterization of microsatellite markers inblack poplar (Populus nigra L) Theoretical and Applied Genetics 101317ndash322 httpsdoiorg101007s001220051485
Van Der Weijde T Dolstra O Visser R G amp Trindade L M(2017a) Stability of cell wall composition and saccharificationefficiency in Miscanthus across diverse environments Frontiers inPlant Science 7 2004
Van der Weijde T Huxley L M Hawkins S Sembiring E HFarrar K Dolstra O hellip Trindade L M (2017) Impact ofdrought stress on growth and quality of miscanthus for biofuelproduction Global Change Biology Bioenergy 9 770ndash782
Van der Weijde T Kamei C L A Severing E I Torres A FGomez L D Dolstra O hellip Trindade L M (2017) Geneticcomplexity of miscanthus cell wall composition and biomass qual-ity for biofuels BMC Genomics 18 406
Van der Weijde T Kiesel A Iqbal Y Muylle H Dolstra O Visser RG FhellipTrindade LM (2017) Evaluation ofMiscanthus sinensis bio-mass quality as feedstock for conversion into different bioenergy prod-uctsGlobal Change Biology Bioenergy 9 176ndash190
Vanholme B Cesarino I Goeminne G Kim H Marroni F VanAcker R hellip Boerjan W (2013) Breeding with rare defectivealleles (BRDA) A natural Populus nigra HCT mutant with modi-fied lignin as a case study New Phytologist 198 765ndash776
Vogel K P Mitchell R B Casler M D amp Sarath G (2014)Registration of Liberty Switchgrass Journal of Plant Registra-tions 8 242ndash247 httpsdoiorg103198jpr2013120076crc
Wang X Yamada T Kong F‐J Abe Y Hoshino Y Sato Hhellip Yamada T (2011) Establishment of an efficient in vitro cul-ture and particle bombardment‐mediated transformation systems inMiscanthus sinensis Anderss a potential bioenergy crop GlobalChange Biology Bioenergy 3 322ndash332 httpsdoiorg101111j1757-1707201101090x
Watrud L S Lee E H Fairbrother A Burdick C Reichman JR Bollman M hellip Van de Water P K (2004) Evidence forlandscape‐level pollen‐mediated gene flow from genetically modi-fied creeping bentgrass with CP4 EPSPS as a marker Proceedingsof the National Academy of Sciences of the United States of Amer-ica 101 14533ndash14538
Weisgerber H (1993) Poplar breeding for the purpose of biomassproduction in short‐rotation periods in Germany ndash Problems andfirst findings Forestry Chronicle 69 727ndash729 httpsdoiorg105558tfc69727-6
Wheeler N C Steiner K C Schlarbaum S E amp Neale D B(2015) The evolution of forest genetics and tree improvementresearch in the United States Journal of Forestry 113 500ndash510httpsdoiorg105849jof14-120
CLIFTON‐BROWN ET AL | 33
Whittaker C Macalpine W J Yates N E amp Shield I (2016)Dry matter losses and methane emissions during wood chip stor-age The impact on full life cycle greenhouse gas savings of shortrotation coppice willow for heat BioEnergy Research 9 820ndash835 httpsdoiorg101007s12155-016-9728-0
Xi Q (2000) Investigation on the distribution and potential of giant grassesin China PhD thesis Goettingen Germany Cuvillier Verlag
Xi Q amp Jezowkski S (2004) Plant resources of Triarrhena andMiscanthus species in China and its meaning for Europe PlantBreeding and Seed Science 49 63ndash77
Xue S Kalinina O amp Lewandowski I (2015) Present and future optionsfor the improvement of Miscanthus propagation techniques Renewableand Sustainable Energy Reviews 49 1233ndash1246
Yin K Q Gao C X amp Qiu J L (2017) Progress and prospectsin plant genome editing Nature Plants 3 httpsdoiorg101038nplants2017107
YookM (2016)Genetic diversity and transcriptome analysis for salt toler-ance inMiscanthus PhD thesis Seoul National University Korea
Zaidi S S E A Mahfouz M M amp Mansoor S (2017) CRISPR‐Cpf1 A new tool for plant genome editing Trends in PlantScience 22 550ndash553 httpsdoiorg101016jtplants201705001
Zalesny R S Stanturf J A Gardiner E S Perdue J H YoungT M Coyle D R hellip Hass A (2016) Ecosystem services ofwoody crop production systems BioEnergy Research 9 465ndash491 httpsdoiorg101007s12155-016-9737-z
Zhang Q X Sun Y Hu H K Chen B Hong C T Guo H Phellip Zheng B S (2012) Micropropagation and plant regenerationfrom embryogenic callus of Miscanthus sinensis Vitro Cellular ampDevelopmental Biology‐Plant 48 50ndash57 httpsdoiorg101007s11627-011-9387-y
Zhang Y Zalapa J E Jakubowski A R Price D L AcharyaA Wei Y hellip Casler M D (2011) Post‐glacial evolution of
Panicum virgatum Centers of diversity and gene pools revealedby SSR markers and cpDNA sequences Genetica 139 933ndash948httpsdoiorg101007s10709-011-9597-6
Zhao H Wang B He J Yang J Pan L Sun D amp Peng J(2013) Genetic diversity and population structure of Miscanthussinensis germplasm in China PloS One 8 e75672
Zhou X H Jacobs T B Xue L J Harding S A amp Tsai C J(2015) Exploiting SNPs for biallelic CRISPR mutations in theoutcrossing woody perennial Populus reveals 4‐coumarate CoAligase specificity and redundancy New Phytologist 208 298ndash301
Zhou R Macaya‐Sanz D Rodgers‐Melnick E Carlson C HGouker F E Evans L M hellip DiFazio S P (2018) Characteri-zation of a large sex determination region in Salix purpurea L(Salicaceae) Molecular Genetics and Genomics 1ndash16 httpsdoiorg101007s00438-018-1473-y
Zsuffa L Mosseler A amp Raj Y (1984) Prospects for interspecifichybridization in willow for biomass production In K L Perttu(Ed) Ecology and management of forest biomass production sys-tems Report 15 (pp 261ndash281) Uppsala Sweden SwedishUniversity of Agricultural Sciences
How to cite this article Clifton‐Brown J HarfoucheA Casler MD et al Breeding progress andpreparedness for mass‐scale deployment of perenniallignocellulosic biomass crops switchgrassmiscanthus willow and poplar GCB Bioenergy201800000ndash000 httpsdoiorg101111gcbb12566
34 | CLIFTON‐BROWN ET AL
Graphical AbstractThe contents of this page will be used as part of the graphical abstract of html only
It will not be published as part of main article
Plant breeding links the research effort with commercial mass upscaling The authorsrsquo assessment of development status ofthe four species is shown (poplar having two one for short rotation coppice (SRC) poplar and one for the more traditionalshort rotation forestry (SRF)) Mass scale deployment needs developments outside the breeding arenas to drive breedingactivities more rapidly and extensively
14Department of Crop Sciences amp Center for Advanced Bioenergy and Bioproducts Innovation 279 Edward R Madigan Laboratory University ofIllinois Urbana Illinois15Dipartimento di Agricoltura Alimentazione e AmbienteUniversitagrave degli Studi di CataniaCataniaItaly16Plant Breeding Wageningen University amp Research Wageningen The Netherlands17Battersea London UK18Julius Kuhn‐Institut (JKI) Bundesforschungsinstitut fur Kulturpflanzen Braunschweig Germany19Institute of Biological and Environmental Science University of Aberdeen Aberdeen UK20Taiwan Endemic Species Research Institute (TESRI) Nantou County Taiwan21Department of Agronomy amp The Key Laboratory of Crop Germplasm Resource of Zhejiang Province Zhejiang University Hangzhou China22Department of Agroecology Aarhus University Centre for Circular Bioeconomy Tjele Denmark23Department of Biobased Products and Energy Crops Institute of Crop Science University of Hohenheim Stuttgart Germany24Department of Plant Sciences Research Institute of Agriculture amp Life Sciences CALS Seoul National University Seoul Korea25Institute of Botany Jiangsu Province and Chinese Academy of Sciences Nanjing China26Natural Resources Research Institute University of Minnesota ndash Duluth Duluth Minnesota27Energene sp z oo Wrocław Poland28James Hutton Institute University of Dundee Dundee UK29GreenWood Resources Inc Portland Oregon30Hudson-Alpha Institute for Biotechnology Huntsville Alabama31Biological Sciences University of Southampton Southampton UK32The Center for Bioenergy Innovation Oak Ridge National Laboratory Oak Ridge Tennessee33Field Science Centre for the Northern Biosphere Hokkaido University Sapporo Japan34College of Agriculture and Life Sciences 2 Kangwon National University Chuncheon South Korea35USDA Forest Service Northern Research Station Rhinelander Wisconsin
CorrespondenceJohn Clifton‐Brown Institute ofBiological Environmental and RuralSciences Aberystwyth UniversityAberystwyth UKEmail jhcaberacuk
Funding informationFP7 Food Agriculture and FisheriesBiotechnology GrantAward NumberOPTIMISCFP7-289159 WATBIOFP7-311929 H2020 Environment GrantAward Number GRACE-745012Bio-Based Industries Joint Undertaking ItalianMinistry of Education Brain GainProgram (Rientro dei cervelli) USDepartment of Energy GrantAwardNumber DE-AC05-00OR22725 DE-SC0006634 DE-SC0012379 and DE-SC0018420 Department of Trade andIndustry (UK) GrantAward Number BW6005990000 Biotechnology andBiological Sciences Research CouncilGrantAward Number CSP17301 BBN0161491 N016149 LK0863 K01711X1 10963A01 G0162161 E0068331G00580X1 000I0410 Department forEnvironment Food and Rural AffairsGrantAward Number LK0863 NF0424NF0426 US Department of AgricultureNational Institute of Food and Agriculture
AbstractGenetic improvement through breeding is one of the key approaches to increasing
biomass supply This paper documents the breeding progress to date for four
perennial biomass crops (PBCs) that have high outputndashinput energy ratios namely
Panicum virgatum (switchgrass) species of the genera Miscanthus (miscanthus)
Salix (willow) and Populus (poplar) For each crop we report on the size of
germplasm collections the efforts to date to phenotype and genotype the diversity
available for breeding and on the scale of breeding work as indicated by number
of attempted crosses We also report on the development of faster and more pre-
cise breeding using molecular breeding techniques Poplar is the model tree for
genetic studies and is furthest ahead in terms of biological knowledge and genetic
resources Linkage maps transgenesis and genome editing methods are now being
used in commercially focused poplar breeding These are in development in
switchgrass miscanthus and willow generating large genetic and phenotypic data
sets requiring concomitant efforts in informatics to create summaries that can be
accessed and used by practical breeders Cultivars of switchgrass and miscanthus
can be seed‐based synthetic populations semihybrids or clones Willow and
poplar cultivars are commercially deployed as clones At local and regional level
the most advanced cultivars in each crop are at technology readiness levels which
could be scaled to planting rates of thousands of hectares per year in about
5 years with existing commercial developers Investment in further development
of better cultivars is subject to current market failure and the long breeding cycles
2 | CLIFTON‐BROWN ET AL
We conclude that sustained public investment in breeding plays a key role in
delivering future mass‐scale deployment of PBCs
KEYWORD S
bioenergy feedstocks lignocellulose M sacchariflorus M sinensis Miscanthus Panicum virgatum
perennial biomass crop Populus spp Salix spp
1 | INTRODUCTION
Increasing sustainable biomass production is an importantcomponent of the transition from a fossil fuel‐based econ-omy to renewables Taking the United Kingdom as an exam-ple Lovett Suumlnnenberg and Dockerty (2014) suggested that14 million ha of marginal agricultural land could be used forbiomass production without compromising food productionAssuming a biomass dry matter (DM) yield of 10 Mgha anda calorific value of 18 GJMg DM 14 million ha woulddeliver around 28 TWh of electricity (with 40 biomassconversion efficiency) which would be ~8 of primary UKelectricity generation (336 TWh in 2017 (DUKES 2017))To achieve this by 2050 planting rates of ~35000 hayearwould be needed from 2022 in line with calculations byEvans (2017) The current annual planting rates in the UnitedKingdom are orders of magnitude short of these levels atonly several hundred hectares per year Similar scenarioshave been generated for other countries (BMU 2009 Scar-lat Dallemand Monforti‐Ferrario amp Nita 2015)
If perennial biomass crops (PBCs) are to make a real con-tribution to sustainable development they should be grownon agricultural land which is less suitable for food crops(Lewandowski 2015) This economically ldquomarginalrdquo land istypically characterized by abiotic stresses (drought floodingstoniness steep slope exposure to wind and sub‐optimalaspect) low nutrients andor contaminated soils (Toacuteth et al2016) In these challenging environments PBCs need resili-ence traits They also need high outputinput ratios forenergy (typically 20ndash50) to deliver large carbon savingsLand may also be marginal due to environmental vulnerabil-ity Much of the value for society from the genetic improve-ment of these crops depends on positive effects arising fromhighly productive perennial systems In addition to produc-ing biomass as a carbon source to replace fossil carbon thesecrops reduce nitrate leaching (Pugesgaard Schelde LarsenLaeligrke amp Joslashrgensen 2015) making them good candidates tohelp fulfil Water Framework Directive (200060EC) and canincrease soil carbon storage during their production (McCal-mont et al 2017)
The objective of this paper was to report on the prepared-ness for wide deployment by summarizing the technicalstate of the art in breeding of four important PBCs namelyswitchgrass miscanthus willow and poplar These four
crops are the most promising and advanced PBCs for tem-perate regions and have therefore the focus here Switch-grass and miscanthus are both rhizomatous grasses with C4
photosynthesis while willow and poplar are trees with C3
photosynthesis Specifically this paper (a) reviews availablecrop trait genetic diversity information (b) assesses the pro-gress of conventional breeding technologies for yield resili-ence and biomass quality (c) reports on progress with newmolecular‐based breeding technologies to increase speedand precision of selection and (d) discusses the require-ments and next steps for breeding of PBCs including com-mercial considerations in order to sustainably meet thebiomass requirements of a growing worldwide bioeconomy
We summarize the crop‐specific attributes the location ofbreeding programmes the current availability of commercialcultivars and yield expectations in selected environments(Table 1) and the generalized breeding targets for all PBCs(Table 2) Economic information relating to the current mar-ket value of the biomass and the investment in breeding arepresented for different countriesregions in Table 3 We alsopresent a comparison of the prebreeding and conventionalbreeding efforts step‐by‐step starting with wild germplasmcollection and evaluation before wide crossing of wild rela-tives (Table 4) Hybridization is followed by at least 6 yearsof selection and evaluation before commercial upscaling canbegin (Figure 1) Recurrent selection often over decades isused within parent populations as part of an ongoing long‐term process to produce hybrid vigour (Brummer 1999) Inthe following sections the state of the art and new opportuni-ties of breeding switchgrass miscanthus willow and poplarare described The application of modern breeding technolo-gies is compared for the four crops in Table 5 It is mostadvanced in poplar and is therefore described in most detail
2 | SWITCHGRASS
Switchgrass is indigenous to the North American prairiesIt is grown from seed and harvested annually using tech-nology similar to that used for pastures Based on collec-tions from thousands of wild prairie remnants the geneticresources are roughly divided into lowland and upland eco-types and there are distinct clades within each ecotypewhich occur along both latitudinal and longitudinal gradi-ents (Evans et al 2018 Lu et al 2013 Zhang et al
CLIFTON‐BROWN ET AL | 3
TABLE1
Breeding‐relatedattributes
forfour
leadingperennialbiom
asscrops(PBCs)
Species
Switc
hgrass
Miscanthu
sW
illow
Pop
lar
Typ
eC4mdash
Grass
C4mdash
Grass
C3mdash
SRC
C3mdash
SRCSRF
Sourcesof
indigeno
usgerm
plasm
CAato
MXeastof
Rocky
Mou
ntains
Eastern
AsiaandOceania
Predom
inantly
Northernhemisph
ere
Northernhemisph
ere
Breedingsystem
Mon
oeciou
sou
tcrossing
Mon
oeciou
sou
tcrossing
Dioeciousou
tcrossing
Dioeciousou
tcrossing
Ploidy
4times8
times2times
3times
4times
2timesminus12
times2times
Specieswith
ingenu
s~4
50~1
4~4
00~3
0ndash32
Typ
esused
mainlyfor
breeding
US
Low
land
ecotyp
e(sub
trop
ical
clim
ates)andup
land
ecotyp
e(tem
perate
clim
ates)
EUb JPSK
andUS
MsinandMsac
EUS
viminalistimesschw
erinii
Sda
syclad
ostimesrehd
eriana
S
dasyclad
osS
viminalis
USandUKS
viminalistimesmiyab
eana
S
miyab
eana
US
Spu
rpurea
timesmiyab
eana
S
purpurea
Pop
ulus
tricho
carpa
Pdelto
ides
Pnigra
Psuaveolens
subsp
maximow
icziiPba
lsam
ifera
Palba
andhy
brids
Typ
ical
haploid
geno
mesize
(Mbp
)~1
500
Msin~5
400
andMsac~4
400
(Raybu
rnCrawfordRaybu
rnamp
Juvik
2009
)
~450
~485
plusmn10
Breedingprog
rammes
CA20
00(REAP
Quebec)
US 19
92(N
ebraska
19
92(O
klahom
a)
1992
(Georgia)
19
96(W
isconsin)
19
96(Sou
thDakota)
20
00(Tennessee)
20
02(M
ississippi)
20
07(O
klahom
a)
2008
(RutgersNew
Jersey)
20
12(Cornell
New
York)
20
12(U
rbana‐Champaign
Illin
ois)
DE19
90s(K
lein‐W
anzleben)
NL20
00s(W
ageningen)
UK20
04(A
berystwyth)
CN20
06(Chang
sha)
JP20
06(H
okkaido)
US 20
06(Californiamdash
CERESInc)
20
06(CaliforniaandIndianamdash
MBI)
20
08(U
rbana‐Champaign)
SK20
09(Suw
on‐SNU)and(M
uan
NICSof
RDA)
FR20
11(Estreacutees‐M
ons)
UK19
80s(Lon
gAshton
relocatedto
Rothamsted
Researchin
2002
)SE
19
80s(SvaloumlfWeibu
llSalix
energi
Europ
aAB)
US
1990
s(Cornell
New
York)
PL20
00s(O
lsztyn
University
ofWarmia
andMazury)
SE
19
39(M
ykingeEkebo
Research
Institu
te)
19
90s(U
ppsalaSL
U)
20
10(U
ppsalaST
T)
US 19
27(N
ewYork
OxfordPaper
Com
pany
andNew
YorkBotanical
Garden
Wheeler
etal20
15)
19
79(W
ashing
ton
UW
andOrego
nGWR)
19
80ndash1
995(M
ississippiMSS
tate
and
GWR)
19
96(M
innesotaUMD
NRRIand
GWR)
IT19
83(Piedm
ontAFV
)FR
20
01(O
rleacuteans
ampNancyIN
RA
Charrey‐sur‐Saocircne
ampPierroton
FCBA
Nog
ent‐sur‐V
ernisson
IRST
EA)
DE20
08(G
oumlttin
gen
NW‐FVA)
(Con
tinues)
4 | CLIFTON‐BROWN ET AL
TABLE
1(Contin
ued)
Species
Switc
hgrass
Miscanthu
sW
illow
Pop
lar
Current
commercial
varietieson
the
market
US
Nocommercial
hybrids
CA2
EU1(M
timesgfrom
different
origins)
+selected
Msinforthatching
inDK
US
3(but
2aregenetically
identical)
UK25
US
8EUDElt10
FR44
IT10ndash1
5SE
~1
4US
8ndash12
Southeast8ndash
14Upp
erMidwest10
PacificNorthwest
Precom
mercial
cultivars
expected
tobe
onthemarketin
3years
US
36registered
cultivars
(halfare
rand
omseed
increasesfrom
natural
prairiesandhalfarebred
varieties)
Mostarepu
blic
releasesfew
are
protected
patented
orlicensed
NL8
seeded
hybridsvanDinterSemo
MTA
UK4
seeded
hybridsCERES(Land
OLakes)andTerravestaLtdMTA
US
Non
eFR
Non
e
EU53
registered
with
CPV
OforPB
RUK20
registered
with
CPV
OforPB
RUS
18clon
esfrom
Cornellin
multisite
trials
CAun
quantified
UAlberta
andQuebec
FR8ndash
12SR
Fclon
eslicence‐based
IT5SR
Cand6SR
FAFV
MTA
orlicence
SE14
STTlicence‐based
andMTA
US
~19So
utheast5ndash7Upp
erMidwest
from
UMD
NRRIMTA6ndash12
North
Central
from
GWRMTAPacific
Northwest8clon
esGWRMTA24
inmultilocationyieldtrialsGWR
Com
mercial
yield
(tDM
haminus1year
minus1 )
US
3ndash18
EU8ndash
12CN20
ndash30
US
10ndash25
EU7ndash20
UK8ndash
14US
8ndash14
EU5ndash20
(SEandBaltic
Cou
ntries8ndash
12)
US
10ndash2
2(10ndash
12North
Central12
ndash16
Southeast15ndash2
2PacificNortheast)
Harvestrotatio
nand
commercial
stand
lifespan
Ann
ualfor10
ndash12years
Ann
ualfor10ndash2
5years(M
sac
hasbeen
used
for~3
0yearsin
China)
2‐to
4‐year
cyclefor22
ndash30years
SRC3‐
to7‐year
cyclefor20
years
SRF
10‐to12‐yearcycleforgt50
years
Adaptiverang
eOpen‐po
llinatedandsynthetic
cultivars
arelim
itedin
adaptatio
nby
temperature
andprecipitatio
n(~8breeding
zonesin
USandCA)
Standard
Mtimesgiswidelyadaptedin
EU
butislim
itedin
theUnitedStates
byinsufficient
winterh
ardiness
forN
orthern
Midwestand
heatintolerancein
the
southNovelMsactimesMsinhyb
rids
andMsintimesMsinhybrids
selected
incontinentalD
Ehave
show
nawide
adaptiv
erang
ein
EU(K
alininaetal
2017
)Ongoing
trialson
heavymetal
contam
inated
soils
indicatetoleranceby
exclusion(K
rzyżak
etal20
17)
Different
hybridsareneeded
fordifferent
zones
Besthy
bridsshow
someG
timesE(U
S)
andsomehy
bridslow
GtimesE(Fabio
etal20
17)
Different
hybridsareneeded
ford
ifferent
clim
aticzonesHyb
rids
thatareadapted
forg
rowingseason
sof
~6mon
thsand
relativ
elyshortd
aysin
Southern
Europ
earemaladaptedto
shortg
rowing
season
sof
~4mon
thsandrelativ
ely
long
days
inNorthernEU
The
mostb
roadly
adaptedvarietiescome
from
thePcana
densistaxo
n
Note
AFV
AlasiaFranco
VivaiC
ornell
CornellUniversity
CPV
OC
ommunity
PlantV
ariety
OfficeFC
BAF
orestCelluloseW
ood
ConstructionandFu
rnitu
reTechnologyInstitu
teG
timesEg
enotype‐by
‐environmentinterac-
tion
GWRG
reenWoodResourcesINRAF
renchNationalInstituteforA
griculturalR
esearch
IRST
EAN
ationalR
esearchUnito
fScience
andTechnologyforE
nvironmentand
AgricultureM
sacMsacchariflo
rusandM
timesg
(Mtimes
giganteus)M
sin
MiscanthussinensisM
BIMendelB
iotechnology
IncM
bpm
egabase
pairM
SStateM
ississippi
StateUniversity
MTAm
aterialtransferagreem
entNICS
NationalInstituteof
CropScience
NW‐
FVANorthwestGerman
ForestResearchInstitu
tePB
Rplantbreeders
rightRDARural
DevelopmentAdm
inistration
REAP
ResourceEfficient
Agriculture
Productio
nSL
USw
edishUniversity
ofAgriculturalSciences
SNUS
eoul
NationalU
niSR
Cshort‐ro
tatio
ncoppice
SRF
short‐rotationforestryS
TTS
weT
reeTech
tDM
haminus1year
minus1 tons
ofdrymatterperhectareperyearU
AlbertaUniversity
ofAlbertaU
MD
NRRIUniversity
ofMinnesotaDuluthsNaturalResources
ResearchInstitu
teU
MNU
niversity
ofMinnesotaU
WU
niversity
ofWashington
UWMU
niversity
ofWarmiaandMazury
a ISO
Alpha‐2
lettercountrycodes
b EU
isused
forEurope
CLIFTON‐BROWN ET AL | 5
2011) Genotype‐by‐environment interactions (G times E) arestrong and must be considered in breeding (Casler 2012Casler Mitchell amp Vogel 2012) Adaptation to environ-ment is regulated principally by responses to day‐lengthand temperature There are also strong genotype times environ-ment interactions between the drier western regions and thewetter eastern regions (Casler et al 2017)
The growing regions of North America are divided intofour adaptation zones for switchgrass each roughly corre-sponding to two official hardiness zones The lowland eco-types are generally late flowering high yielding andadapted to warmer climates but have lower drought andcold resistance than upland ecotypes (Casler 2012 Casleret al 2012)
In 2015 the US Department of Agriculture (USDA)National Plant Germplasm System GRIN (httpswwwars-gringovnpgs) had 181 switchgrass accessions of whichonly 96 were available for distribution due to limitationsassociated with seed multiplication (Casler Vogel amp Har-rison 2015) There are well over 2000 additional uncata-logued accessions (Table 1) held by various universities butthe USDA access to these is also constrained by the effortneeded in seed multiplication Switchgrass is a model herba-ceous species for conducting scientific research on biomass(Sanderson Adler Boateng Casler amp Sarath 2006) but lit-tle funding is available for the critical prebreeding work thatis necessary to link this biological research to commercialbreeding More than a million genotypes from ~2000 acces-sions (seed accessions contain many genotypes) have beenphenotypically screened in spaced plant nurseries and tenthousand of the most useful have been genotyped with differ-ent technologies depending on the technology available at
the time when these were performed From these character-ized genotypes parents are selected for exploratory pairwisecrosses to produce synthetic populations within ecotypesSwitchgrass like many grasses is outcrossing due to astrong genetically controlled self‐incompatibly (akin to theS‐Z‐locus system of other grasses (Martinez‐Reyna ampVogel 2002)) Thus the normal breeding approaches usedare F1 wide crosses and recurrent selection cycles within syn-thetic populations
The scale of these programmes varies from small‐scaleconventional breeding based solely on phenotypic selec-tion (eg REAP Canada Montreal Quebec) to large pro-grammes incorporating modern molecular breedingmethods (eg USDA‐ARS Madison Wisconsin) Earlyagronomic research and biomass production efforts werefocused on the seed‐based multiplication of promising wildaccessions from natural prairies Cultivars Alamo Kanlowand Cave‐in‐Rock were popular due to high yield andmoderate‐to‐wide adaptation Conventional breedingapproaches focussed on biomass production traits and haveled to the development of five cultivars particularly suitedto biomass production Cimarron EG1101 EG1102EG2101 and Liberty The first four of these represent thelowland ecotype and were developed either in Oklahomaor Georgia Liberty is a derivative of lowland times uplandhybrids developed in Nebraska following selection for lateflowering the high yield of the lowland ecotype and coldtolerance of the upland ecotype (Vogel et al 2014) Thesefive cultivars were all approximately 25ndash30 years in themaking counting from the initiation of these breeding pro-grammes Many more biomass‐type cultivars are expectedwithin the next few years as these and other breeding
TABLE 2 Generalized improvement targets for perennial biomass crops (PBCs)
Net energy yield per hectare
Increased yield
Reduced moisture content at harvest
Physical and chemical composition for different end‐use applications
Increased lignin content and decreased corrosive elements for thermal conversion
Reduced recalcitrance through decreased lignin content andor modified lignin monomer composition to reduce pretreatment requirementsfor next‐generation biofuels by saccharification and fermentation
Plant morphological differences which influence biomass harvest transport and storage (eg stem thickness)
Propagation costs
Improved cloning systems (trees and grasses)
Seed systems (grasses)
Optimizing agronomy for each new cultivar
Resilience through enhanced
Abiotic stress toleranceresistance (eg drought salinity and high and low temperature )
Biotic stress resistance (eg insects fungal bacterial and viral diseases)
Site adaptability especially to those of marginalcontaminated agricultural land
6 | CLIFTON‐BROWN ET AL
TABLE3
Preparedness
formassupscaling
currentmarketvalueandresearch
investmentin
four
perennialbiom
asscrops(PBCs)
Species
Switc
hgrass
Miscanthu
sW
illow
Pop
lar
Current
commercial
plantin
gcostsperha
USa20
0USD
bin
the
establishm
entyear
DE3
375
Eurob
(XueK
alininaamp
Lew
ando
wski20
15)
UK2
153
GBPb
(Evans201
6)redu
cedto
116
9GBPwith
Mxg
rhizom
ewhere
thefarm
erdo
estheland
preparation(w
wwterravestacom
)US
173
0ndash222
5USD
UK150
0ndash173
9GBPplus
land
preparation(Evans20
16)
US
197
6USD
(=80
0USD
acre)
IT110
0Euro
FRSE100
0ndash200
0Euro
US
863USD
ha(LazarusHeadleeamp
Zalesny
20
15)
Current
marketvalue
ofthebiom
asspert
DM
US
80ndash1
00USD
UK~8
0GBP(bales)(Terravesta
person
alcommun
ication)
US
94USD
(chipp
ed)
UK49
41GBP(chipp
ed)(Evans20
16)
US
55ndash7
0USD
IT10
0Euro
US
80ndash90USD
deliv
ered
basedon
40milesof
haulage
Scienceforg
enetic
improv
ementprojects
inthelast10
years
USgt50
projectsfund
edby
USDOEand
USD
ANIFA
CNgt
30projectsfund
edby
CN‐N
SFCandMBI
EUc 6projects(O
PTIM
ISCO
PTIM
A
WATBIO
SUNLIBBF
IBRAandGRACE)
FRB
FFSK
2projectsfund
edby
IPETandPM
BC
US
EBICABBImultip
leDOEFeedstock
Genom
icsandUSD
AAFR
Iprojects
UKB
SBECB
EGIN
(200
3ndash20
10)
US
USDOEJG
IGenom
eSequ
encing
(200
9ndash20
122
015ndash
2018
)USD
ANortheastSu
nGrant
(200
9ndash20
12)
USD
ANIFANEWBio
Con
sortium
(201
2ndash20
17)USD
ANIFAWillow
SkyC
AP(201
8ndash20
21)
EUF
P7(EnergyPo
plarN
ovelTreeWATBIO
Tree4Fu
ture)
SEC
limate‐adaptedpo
plarsandNanow
ood
SLU
andST
TUS
Bioenergy
Science(O
RNL)USD
Afeedstocks
geno
micsprog
rammeBRDIa
ndBRC‐CBI
Major
projects
supp
ortin
gcrossing
andselectioncycles
inthelast10
years
USandCA12
projects
NLRUEmiscanthu
sPP
P20
15ndash2
019
50ndash
100K
Euroyear
UKGIA
NT‐LIN
KPP
I20
11ndash2
016
13M
GBPyear
US
EBICABBI~0
5M
USD
year
UKBEGIN
2000
ndash201
0US
USD
ANortheastSu
nGrant
(200
9ndash20
12)USD
ANIFA
NEWBio
Con
sortium
(201
2ndash20
17)USD
ANIFA
Willow
SkyC
AP(201
8ndash20
21)
FR1
4region
alandnatio
nalp
rojects10
0ndash15
0k
Euroyear
SES
TTo
nebreeding
effortas
aninternalproject
in20
10with
aresulting
prog
enytrial
US
DOESu
nGrant
feedstockdevelopm
ent
partnershipprog
ramme(including
Willow
)
Current
annu
alinvestmentin
projects
fortranslationinto
commercial
hybrids
USgt20
MUSD
year
UKMUST
201
6ndash20
1905M
GBPyear
US
CABBIUSD
AAFR
I20
17ndash2
02205M
USD
year
US
USD
ANIFA
NEWBio
~04
MUSD
year
FRScienceforim
prov
ement~2
0kEuroyear
US
USD
Aun
derAFR
ICo‐ordinatedAg
Prod
ucer
projectsAnd
severaltranslational
geno
micsprojectswith
outpu
blic
fund
ing
Upscalin
gtim
e(years)
toproducesufficient
propagules
toplantgt10
0ha
US
Using
seed
2ndash3years
UKRhizomefor1ndash200
0hectares
canbe
ready
in6mon
ths
UKNL2
0ha
ofseed
hybridsplantedin
2018
In
2019
sufficientseedfor5
0ndash10
0ha
isexpected
UK3yearsusingconv
entio
nalcutting
sfaster
usingmicroprop
agation
FRIT3yearsby
vegetativ
eprop
agation
SEgt3yearsby
vegetativ
eprop
agation(cuttin
gs)
and3yearsby
microprop
agation
Note
AFR
IAgriculture
andFo
odResearchInitiativein
theUnitedStatesBEGIN
BiomassforEnergyGenetic
Improvem
entNetwork
BFF
BiomassFo
rtheFu
tureBRC‐CBI
Bioenergy
ResearchCentre‐Centrefor
Bioenergy
Innovatio
nBRDIBiomassresearch
developm
entinitiative
BSB
ECBBSR
CSu
stainableBioenergy
Centre
CABBICenterforadvanced
bioenergyandbioproductsinnovatio
nin
theUnitedStatesCN‐N
SFC
Natural
ScienceFo
undatio
nof
ChinaDOEDepartm
entof
Energyin
theUnitedStatesEBIEnergyBiosciences
Institu
teFIBRAFibrecropsas
sustainablesource
ofbiobased
materialforindustrial
productsin
Europeand
ChinaFP
7SeventhFram
eworkProgrammein
theEUGIA
NT‐LIN
KGenetic
improvem
entof
miscanthusas
asustainablefeedstockforbioenergyin
theUnitedKingdom
GRACEGRow
ingAdvancedindustrial
Crops
onmarginallandsforbiorEfineriesIPETKorea
Institu
teof
Planning
andEvaluationforTechnologyin
FoodA
gricultureFo
restry
andFisheriesJG
IJointGenom
eInstitu
teMBIMendelBiotechnology
IncMUST
Miscant-
husUpS
calin
gTechnology
NEWBioNortheast
WoodyW
arm‐seasonBiomassConsortiumNIFANationalInstitu
teof
Food
andAgriculture
intheUnitedStatesNovelTree
Novel
tree
breeding
strategiesOPT
IMAOpti-
mizationof
perennialgrassesforbiom
assproductio
nin
theMediterraneanareaOPT
IMISCOptim
izingbioenergyproductio
nfrom
Miscanthus
ORNLOak
Ridge
NationalLabPM
BCPlantMolecular
BreedingCenterof
theNextGenerationBiogreenResearchCenters
oftheRepublic
ofKoreaPP
Ipublicndashp
rivate
investmentPP
Ppublicndashp
rivate
partnership
RUEradiationuseefficiencySk
yCAP
Coordinated
AgriculturalProjectSL
U
SwedishUniversity
ofAgriculturalSciencesST
TSw
eTreeTech
SUNLIBBSu
stainableLiquidBiofuelsfrom
BiomassBiorefiningTree4Fu
tureDesigning
Trees
forthefutureUSD
AUSDepartm
entof
Agriculture
WATBIO
Developmentof
improved
perennialnonfoodbiom
assandbioproduct
cropsforwater
stressed
environm
ents
a ISO
Alpha‐2
lettercountrycodes
b Local
currencies
areused
asat
2018Euro
GBP(G
reat
Britain
Pound)USD
(USDollar)c EU
isused
forEurope
CLIFTON‐BROWN ET AL | 7
TABLE4
Prebreedingresearch
andthestatus
ofconventio
nalbreeding
infour
leadingperennialbiom
asscrops(PBCs)
Breeding
techno
logy
step
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sWillow
Poplar
1Collected
wild
accessions
or
secondarysources
availablefor
breeding
Provideabroadbase
of
useful
traits
Respect
forCBD
and
Nagoyaprotocol
on
collections
after2014
Not
allindigenous
genetic
resourcesareaccessible
under
CBD
forpolitical
reasons
USa~2
000
(181
inofficial
GRIN
gene
bank
ofwhich
96
wereavailable
others
in
variousunofficial
collections)
CNChangsha
~1000
andNanjin
g
2000from
China
DK~1
20from
JP
FRworking
collectionof
~100
(mainlyM
sin)
JP~1
500
from
JPS
KandCN
NLworking
collectionof
~300
Msin
SK~7
00from
SKC
NJP
andRU
UK~1
500
accessions
(from
~500
sites)
ofwhich
1000arein
field
nurseriesin
2018
US
14in
GRIN
global
US
Illin
ois~1
500
collections
made
inAsiawith
25
currently
available
inUnitedStates
forbreeding
UK1500with
about20
incommon
with
the
UnitedStates
US
350largelyunique
FR3370Populus
delto
ides
(650)Pnigra(2000)
Ptrichocarpa(600)and
Pmaximow
iczii(120)
IT30000P
delto
idsPnigra
Pmaximow
icziiand
Ptrichocarpa
SE13000
accessionsmainly
Ptrichocarpa
byST
TandSL
U
US
GWR150Ptrichocarpaand
300Pmaximow
icziiaccessions
forPacificNorthwestprogram
535Pdelto
ides
and~2
00
Pnigraaccessions
for
Southeastprogrammeand77
Psimoniifrom
Mongolia
UMD
NRRI550+
clonal
accessions
(primarily
Ptimescanadnesisa
long
with
Pdelto
ides
andPnigra
parents)
forUpper
Midwest
program
2Wild
accessions
which
have
undergone
phenotypic
screeningin
field
trials
Selectionof
parental
lines
with
useful
traits
Strong
partnerships
torun
multilocationfieldtrials
tophenotypeconsistently
theaccessionsgenotypes
indifferentenvironm
ents
anddatabases
Costof
runningmultilocation
US
gt500000genotypes
CN~1
250
inChangsha
~1700
in
HunanJiangsuS
handong
Hainan
NL~250genotypes
JP~1
200
SK~4
00
UK‐le
d~1
000
genotypesin
a
replicated
trial
US‐led
~1200
genotypesin
multi‐
locatio
nreplicated
trialsin
Asiaand
North
America
MsinSK
CNJP
CA
(Ontario)US(Illinoisand
Colorado)MsacSK
(Kangw
on)
CN
(Zhejiang)JP
(Sapporo)CA
(Onatario)U
S(Illinois)DK
(Aarhus)
UKgt400
US
180
FR2720
IT10000
SE~1
50
US
GWR~1
500
wild
Ptrichocarpaphenotypes
and
OPseedlots(total
of70)from
thePmaximow
icziispecies
complex
(suaveolens
cathayana
koreana
ussuriensismaximow
iczii)and
2000ndash3000Pdelto
ides
collections
(based
on104s‐
generatio
nfamilies)
UMD
NRRIscreened
seedlin
gsfrom
aPdelto
ides
(OPor
CP)
breeding
programme(1996ndash2016)
gt7200OPseedlin
gsof
PnigraunderMinnesota
clim
atic
conditions(improve
parental
populatio
n)
3Wild
germ
plasm
genotyping
Amanaged
living
collectionof
clonal
types
US
~20000genotypes
CNChangsha
~1000Nanjin
g37
FR44
bycpDNA
(Fenget
al
UK~4
00
US
225
(Con
tinues)
8 | CLIFTON‐BROWN ET AL
TABLE
4(Contin
ued)
Breeding
techno
logy
step
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sWillow
Poplar
Construct
phylogenetic
treesanddissim
ilarity
indices
The
type
ofmolecular
analysis
mdashAFL
PcpDNAR
ADseq
whole
genomesequencing
2014)
JP~1
200
NL250genotypesby
RADseq
UK~1
000
byRADseq
SK~3
00
US
Illin
oisRADseqon
allwild
accessions
FR2310
SE150
US
1500
4Exploratory
crossing
and
progenytests
Discovergood
parental
combinatio
nsmdashgeneral
combining
ability
Geographicseparatio
n
speciesor
phylogenetic
trees
Costly
long‐te
rmmultilocation
trialsforprogeny
US
gt10000
CN~1
0
FR~2
00
JP~3
0
NL3600
SK~2
0
UK~4
000
(~1500Msin)
US
500ndash1000
UK gt700since2003
gt800EWBP(1996ndash
2002)
US
800
FRCloned13
times13
factorial
matingdesign
with
3species[10
Pdelto
ides8Ptrichocarpa
8
Pnigra)
US
GWR1000exploratory
crossingsof
Pfrem
ontii
Psimoniiforim
proved
droughttolerance
UMD
NRRIcrossednativ
e
Pdelto
ides
with
non‐nativ
e
Ptrichocarpaand
Pmaximow
icziiMajority
of
progenypopulatio
nssuffered
from
clim
atic
andor
pathogenic
susceptib
ility
issues
5Wideintraspecies
hybridization
Discovergood
parental
combinatio
nsmdashgeneral
combining
ability
1to
4above
inform
atics
Flow
eringsynchronizationand
low
seed
set
US
~200
CN~1
0
NL1800
SK~1
0
UK~5
00
UK276
US
400
IT500
SE120
US
50ndash100
6With
inspecies
recurrentselection
Concentratio
nof
positiv
e
traits
Identificationof
theright
heterotic
groups
insteps
1ndash5
Difficultto
introducenew
germ
plasm
with
outdilutio
n
ofbesttraits
US
gt40
populatio
ns
undergoing
recurrentselection
NL2populatio
ns
UK6populatio
ns
US
~12populatio
nsto
beestablished
by2019
UK4
US
150
FRPdelto
ides
(gt30
FSfamilies
ndash900clones)Pnigra(gt40
FS
families
ndash1600clones)and
Ptrichocarpa(15FS
Families
minus350clones)
IT80
SEca50
parent
breeding
populatio
nswith
in
Ptrichocarpa
US
Goalsis~1
00parent
breeding
populatio
nswith
inPtrichocarpa
Pdelto
ides
PnigraandPmaximow
iczii
7Interspecies
hybrid
breeding
Com
bining
complem
entary
traitsto
producegood
morphotypes
and
heterosiseffects
Allthesteps1ndash6
Inearlystage
improvem
ents
areunpredictable
Awide
base
iscostly
tomanage
None
CNChangsha
3Nanjin
g
120MsintimesMflo
r
JP~5
with
Saccharum
spontaneum
SK5
UK420
US
250
FRPcanadensis(1800)
Pdelto
ides
timesPtrichocarpa
(800)and
PtrichocarpatimesPmaximow
iczii
(1200)
(Con
tinues)
CLIFTON‐BROWN ET AL | 9
TABLE
4(Contin
ued)
Breeding
techno
logy
step
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sWillow
Poplar
IT~1
20
US
525genotypesin
eastside
hybrid
programme
205westside
hybrid
programmeand~1
200
in
SoutheastPdelto
ides
program
8Chrom
osom
e
doublin
g
Arouteto
triploid
seeded
typesdoublin
gadiploid
parent
ofknow
n
breeding
valuefrom
recurrentselection
Needs
1ndash7to
help
identify
therightparental
lines
Doubled
plantsarenotorious
forrevertingto
diploid
US
~50plantstakenfrom
tetraploid
tooctaploid
which
arenow
infieldtrials
CN~1
0Msactriploids
DKSeveralattemptsin
mid
90s
FR5genotypesof
Msin
UK2Msin1Mflora
nd1Msac
US
~30
UKAttempted
butnot
routinelyused
US
Not
partof
theregular
breeding
becausethere
existnaturalpolyploids
NA
9DoubleHaploids
Arouteto
producing
homogeneous
progeny
andalso
forthe
introductio
nof
transgenes
orforgenome
editing
Needs
1ndash7to
help
identify
therightparental
lines
Fully
homozygousplantsare
weakandeasily
die
andmay
notflow
ersynchronously
US
Noneyetattempteddueto
poor
vigour
andviability
of
haploids
CN2genotypesof
Msin
andMflor
UKDihaploidsof
MsinMflor
and
Msacand2transgenic
Msinvia
anther
cultu
re
US
6createdandused
toidentify
miss‐identifying
paralogueloci
(causedby
recent
genome
duplicationin
miscanthus)Usedin
creatin
gthereferencegenomefor
MsinVeryweakplantsand
difficultto
retain
thelin
es
NA
NA
10Embryo
rescue
Anattractiv
etechniquefor
recovering
plantsfrom
sexual
crosses
The
majority
ofem
bryos
cannot
survivein
vivo
or
becomelong‐timedorm
ant
None
UKone3times
hybrid
UKEmbryosrescue
protocol
proved
robustat
8days
post‐pollin
ation
FRAnem
bryo
rescue
protocol
proved
robustandim
proved
hybridizationsuccessfor
Pdelto
ides
timesPtrichocarpa
crosses
11Po
llenstorage
Flexibility
tocross
interestingparents
with
outneed
of
flow
ering
synchronization
None
Nosuccessto
daten
otoriously
difficultwith
grassesincluding
sugarcane
Freshpollencommonly
used
incrossing
Pollenstoragein
some
interspecificcrosses
FRBothstored
(cryobank)
and
freshindividual
pollen
Note
AFL
Pam
plifiedfragmentlength
polymorphismCBDconventio
non
biological
diversity
CP
controlledpollinatio
nCpD
NAchloroplastDNAEWBP
Europeanwillow
breeding
programme
FSfull‐sib
GtimesE
genotype‐by‐environm
entGRIN
germ
plasm
resourcesinform
ationnetwork
GWRGreenWoodResourcesMflorMiscanthusflo
ridulusMsacMsacchariflo
rus
MsinMsinensisNAnotapplicableNRRINatural
Resources
ResearchInstitu
teOP
open
pollinatio
nRADseq
restrictionsite‐associatedDNA
sequencingSL
USw
edishUniversity
ofAgriculturalSciencesST
TSw
eTreeTech
UMDUniversity
ofMinnesota
Duluth
a ISO
Alpha‐2
lettercountrycodes
10 | CLIFTON‐BROWN ET AL
programmes mature The average rate of gain for biomassyield in long‐term switchgrass breeding programmes hasbeen 1ndash4 per year depending on ecotype populationand location of the breeding programme (Casler amp Vogel2014 Casler et al 2018) The hybrid derivative Libertyhas a biomass yield 43 higher than the better of its twoparents (Casler amp Vogel 2014 Vogel et al 2014) Thedevelopment of cold‐tolerant and late‐flowering lowland‐ecotype populations for the northern United States hasincreased biomass yields by 27 (Casler et al 2018)
Currently more than 20 recurrent selection populationsare being managed in the United States to select parentsfor improved yield yield resilience and compositional qual-ity of the biomass For the agronomic development andupscaling high seed multiplication rates need to be com-bined with lower seed dormancy to reduce both crop estab-lishment costs and risks Expresso is the first cultivar withsignificantly reduced seed dormancy which is the first steptowards development of domesticated populations (Casleret al 2015) Most phenotypic traits of interest to breeders
require a minimum of 2 years to be fully expressed whichresults in a breeding cycle that is at least two years Morecomplicated breeding programmes or traits that requiremore time to evaluate can extend the breeding cycle to 4ndash8 years per generation for example progeny testing forbiomass yield or field‐based selection for cold toleranceBreeding for a range of traits with such long cycles callsfor the development of molecular methods to reduce time-scales and improve breeding efficiency
Two association panels of switchgrass have been pheno-typically and genotypically characterized to identify quanti-tative trait loci (QTLs) that control important biomasstraits The northern panel consists of 60 populationsapproximately 65 from the upland ecotype The southernpanel consists of 48 populations approximately 65 fromthe lowland ecotype Numerous QTLs have been identifiedwithin the northern panel to date (Grabowski et al 2017)Both panels are the subject of additional studies focused onbiomass quality flowering and phenology and cold toler-ance Additionally numerous linkage maps have been
FIGURE 1 Cumulative minimum years needed for the conventional breeding cycle through the steps from wild germplasm to thecommercial hybrids in switchgrass miscanthus willow and poplar Information links between the steps are indicated by dotted arrows andhighlight the importance of long‐term informatics to maximize breeding gain
CLIFTON‐BROWN ET AL | 11
created by the pairwise crossing of individuals with diver-gent characteristics often to generate four‐way crosses thatare analysed as pseudo‐F2 crosses (Liu Wu Wang ampSamuels 2012 Okada et al 2010 Serba et al 2013Tornqvist et al 2018) Individual markers and QTLs iden-tified can be used to design marker‐assisted selection(MAS) programmes to accelerate breeding and increase itsefficiency Genomic prediction and selection (GS) holdseven more promise with the potential to double or triplethe rate of gain for biomass yield and other highly complexquantitative traits of switchgrass (Casler amp Ramstein 2018Ramstein et al 2016) The genome of switchgrass hasrecently been made public through the Joint Genome Insti-tute (httpsphytozomejgidoegov)
Transgenic approaches have been heavily relied upon togenerate unique genetic variants principally for traitsrelated to biomass quality (Merrick amp Fei 2015) Switch-grass is highly transformable using either Agrobacterium‐mediated transformation or biolistics bombardment butregeneration of plants is the bottleneck to these systemsTraditionally plants from the cultivar Alamo were the onlyregenerable genotypes but recent efforts have begun toidentify more genotypes from different populations that arecapable of both transformation and subsequent regeneration(King Bray Lafayette amp Parrott 2014 Li amp Qu 2011Ogawa et al 2014 Ogawa Honda Kondo amp Hara‐Nishi-mura 2016) Cell wall recalcitrance and improved sugarrelease are the most common targets for modification (Bis-wal et al 2018 Fu et al 2011) Transgenic approacheshave the potential to provide traits that cannot be bredusing natural genetic variability However they will stillrequire about 10ndash15 years and will cost $70ndash100 millionfor cultivar development and deployment (Harfouche Mei-lan amp Altman 2011) In addition there is commercialuncertainty due to the significant costs and unpredictabletimescales and outcomes of the regulatory approval processin the countries targeted for seed sales As seen in maizeone advantage of transgenic approaches is that they caneasily be incorporated into F1 hybrid cultivars (Casler2012 Casler et al 2012) but this does not decrease thetime required for cultivar development due to field evalua-tion and seed multiplication requirements
The potential impacts of unintentional gene flow andestablishment of non‐native transgene sequences in nativeprairie species via cross‐pollination are also major issues forthe environmental risk assessment These limit further thecommercialization of varieties made using these technolo-gies Although there is active research into switchgrass steril-ity mechanisms to curb unintended pollen‐mediated genetransfer it is likely that the first transgenic cultivars proposedfor release in the United States will be met with considerableopposition due to the potential for pollen flow to remainingwild prairie sites which account for lt1 of the original
prairie land area and are highly protected by various govern-mental and nongovernmental organizations (Casler et al2015) Evidence for landscape‐level pollen‐mediated geneflow from genetically modified Agrostis seed multiplicationfields (over a mountain range) to pollinate wild relatives(Watrud et al 2004) confirms the challenge of using trans-genic approaches Looking ahead genome editing technolo-gies hold considerable promise for creating targeted changesin phenotype (Burris Dlugosz Collins Stewart amp Lena-ghan 2016 Liu et al 2018) and at least in some jurisdic-tions it is likely that cultivars resulting from gene editingwill not need the same regulatory approval as GMOs (Jones2015a) However in July 2018 the European Court of Justice(ECJ) ruled that cultivars carrying mutations resulting fromgene editing should be regulated in the same way as GMOsThe ECJ ruled that such cultivars be distinguished from thosearising from untargeted mutation breeding which isexempted from regulation under Directive 200118EC
3 | MISCANTHUS
Miscanthus is indigenous to eastern Asia and Oceania whereit is traditionally used for forage thatching and papermaking(Xi 2000 Xi amp Jezowkski 2004) In the 1960s the highbiomass potential of a Japanese genotype introduced to Eur-ope by Danish nurseryman Aksel Olsen in 1935 was firstrecognized in Denmark (Linde‐Laursen 1993) Later thisaccession was characterized described and named asldquoM times giganteusrdquo (Greef amp Deuter 1993 Hodkinson ampRenvoize 2001) commonly abbreviated as Mxg It is a natu-rally occurring interspecies triploid hybrid between tetraploidM sacchariflorus (2n = 4x) and diploid M sinensis(2n = 2x) Despite its favourable agronomic characteristicsand ability to produce high yields in a wide range of environ-ments in Europe (Kalinina et al 2017) the risks of relianceon it as a single clone have been recognized Miscanthuslike switchgrass is outcrossing due to self‐incompatibility(Jiang et al 2017) Thus seeded hybrids are an option forcommercial breeding Miscanthus can also be vegetativelypropagated by rhizome or in vitro culture which allows thedevelopment of clones The breeding approaches are usuallybased on F1 crosses and recurrent selection cycles within thesynthetic populations There are several breeding pro-grammes that target improvement of miscanthus traitsincluding stress resilience targeted regional adaptation agro-nomic ldquoscalabilityrdquo through cheaper propagation fasterestablishment lower moisture and ash contents and greaterusable yield (Clifton‐Brown et al 2017)
Germplasm collections specifically to support breedingfor biomass started in the late 1980s and early 1990s inDenmark Germany and the United Kingdom (Clifton‐Brown Schwarz amp Hastings 2015) These collectionshave continued with successive expeditions from European
12 | CLIFTON‐BROWN ET AL
and US teams assembling diverse collections from a widegeographic range in eastern Asia including from ChinaJapan South Korea Russia and Taiwan (Hodkinson KlaasJones Prickett amp Barth 2015 Stewart et al 2009) Threekey miscanthus species for biomass production are M si-nensis M floridulus and M sacchariflorus M sinensis iswidely distributed throughout eastern Asia with an adap-tive range from the subtropics to southern Russia (Zhaoet al 2013) This species has small rhizomes and producesmany tightly packed shoots forming a ldquotuftrdquo M floridulushas a more southerly adaptive range with a rather similarmorphology to M sinensis but grows taller with thickerstems and is evergreen and less cold‐tolerant than the othermiscanthus species M sacchariflorus is the most northern‐adapted species ranging to 50 degN in eastern Russia (Clarket al 2016) Populations of diploid and tetraploid M sac-chariflorus are found in China (Xi 2000) and South Korea(Yook 2016) and eastern Russia but only tetraploids havebeen found in Japan (Clark Jin amp Petersen 2018)
Germplasm has been assembled from multiple collec-tions over the last century though some early collectionsare poorly documented This historical germplasm has beenused to initiate breeding programmes largely based on phe-notypic and genotypic characterization As many of theaccessions from these collections are ldquoorigin unknownrdquocrucial environmental envelope data are not available UK‐led expeditions started in 2006 and continued until 2011with European and Asian partners and have built up a com-prehensive collection of 1500 accessions from 500 sitesacross Eastern Asia including China Japan South Koreaand Taiwan These collections were guided using spatialclimatic data to identify variation in abiotic stress toleranceAccessions from these recent collections were planted fol-lowing quarantine in multilocation nursery trials at severallocations in Europe to examine trait expression in differentenvironments Based on the resulting phenotypic andmolecular marker data several studies (a) characterized pat-terns of population genetic structure (Slavov et al 20132014) (b) evaluated the statistical power of genomewideassociation studies (GWASs) and identified preliminarymarkerndashtrait associations (Slavov et al 2013 2014) and(c) assessed the potential of genomic prediction (Daveyet al 2017 Slavov et al 2014 2018b) Genomic indexselection in particular offers the possibility of exploringscenarios for different locations or industrial markets (Sla-vov et al 2018a 2018b)
Separately US‐led expeditions also collected about1500 accessions between 2010 and 2014 (Clark et al2014 2016 2018 2015) A comprehensive genetic analy-sis of the population structure has been produced by RAD-seq for M sinensis (Clark et al 2015 Van der WeijdeKamei et al 2017) and M sacchariflorus (Clark et al2018) Multilocation replicated field trials have also been
conducted on these materials in North America and inAsia GWAS has been conducted for both M sinensis anda subset of M sacchariflorus accessions (Clark et al2016) To date about 75 of these recent US‐led collec-tions are in nursery trials outside the United States Due tolengthy US quarantine procedures these are not yet avail-able for breeding in the United States However molecularanalyses have allowed us to identify and prioritize sets ofgenotypes that best encompass the genetic variation in eachspecies
While mostM sinensis accessions flower in northern Eur-ope very few M sacchariflorus accessions flower even inheated glasshouses For this reason the European pro-grammes in the United Kingdom the Netherlands and Francehave performed mainly M sinensis (intraspecies) hybridiza-tions (Table 4) Selected progeny become the parents of latergenerations (recurrent selection as in switchgrass) Seed setsof up to 400 seed per panicle occur inM sinensis In Aberyst-wyth and Illinois significant efforts to induce synchronousflowering in M sacchariflorus and M sinensis have beenmade because interspecies hybrids have proven higher yieldperformance and wide adaptability (Kalinina et al 2017) Ininterspecies pairwise crosses in glasshouses breathable bagsandor large crossing tubes or chambers in which two or morewhole plants fit are used for pollination control Seed sets arelower in bags than in the open air because bags restrict pollenmovement while increasing temperatures and reducinghumidity (Clifton‐Brown Senior amp Purdy 2018) About30 of attempted crosses produced 10 to 60 seeds per baggedpanicle The seed (thousand seed mass ranges from 05 to09 g) is threshed from the inflorescences and sown into mod-ular trays to produce plug plants which are then planted infield nurseries to identify key parental combinations
A breeding programme of this scale must serve theneeds of different environments accepting the commonpurpose is to optimize the interception of solar radiationAn ideal hybrid for a given environment combines adapta-tion to date of emergence with optimization of traits suchas height number of stems per plant flowering and senes-cence time to optimize solar interception to produce a highbiomass yield with low moisture content at harvest (Rob-son Farrar et al 2013 Robson Jensen et al 2013) By20132014 conventional breeding in Europe had producedintra‐ and interspecific fertile seeded hybrids When acohort (typically about 5) of outstanding crosses have beenidentified it is important to work on related upscaling mat-ters in parallel These are as follows
Assessment of the yield and critical traits in selectedhybrids using a network of field trials
Efficient cloning of the seed parents While in vitro andmacro‐cloning techniques are used some genotypes areamenable to neither technique
CLIFTON‐BROWN ET AL | 13
TABLE5
Status
ofmod
ernplantbreeding
techniqu
esin
four
leadingperennialbiom
asscrop
s(PBCs)
Breeding
techno
logy
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sW
illow
Pop
lar
MAS
Use
ofmarker
sequ
encesthat
correlatewith
atrait
allowingearly
prog
enyselectionor
rejectionlocatin
gkn
owngenesfor
useful
traits(suchas
height)from
other
speciesin
your
crop
Breeding
prog
ramme
relevant
biparental
crosses(Table
4)to
create
ldquomapping
popu
latio
nsrdquoand
QTLidentification
andestim
ationto
identifyrobu
stmarkers
alternatively
acomprehensive
panelof
unrelated
geno
typesfor
GWAS
Ineffectivewhere
traits
areaffected
bymany
geneswith
small
effects
US
Literally
dozens
ofstud
iesto
identify
SNPs
andmarkers
ofinterestNoattempts
atMASas
yetlargely
dueto
popu
latio
nspecificity
CNS
SRISS
Rmarkers
forM
sinMsacand
Mflor
FR2
Msin
mapping
popu
latio
ns(BFF
project)
NL2
Msin
mapping
popu
latio
ns(A
tienza
Ram
irezetal
2003
AtienzaSatovic
PetersenD
olstraamp
Martin
200
3Van
Der
Weijdeetal20
17a
Van
derW
eijde
Hux
ley
etal20
17
Van
derW
eijdeKam
ei
etal20
17V
ander
WeijdeKieseletal
2017
)SK
2GWASpanels(1
collectionof
Msin1
diploidbiparentalF 1
popu
latio
n)in
the
pipelin
eUK1
2families
tofind
QTLin
common
indifferentfam
ilies
ofthe
sameanddifferent
speciesandhy
brids
US
2largeGWAS
panels4
diploidbi‐
parentalF 1
popu
latio
ns
1F 2
popu
latio
n(add
ition
alin
pipelin
e)
UK16
families
for
QTLdiscov
ery
2crossesMASscreened
1po
tentialvariety
selected
US
9families
geno
typedby
GBS(8
areF 1o
neisF 2)a
numberof
prom
ising
QTLdevelopm
entof
markers
inprog
ress
FR(INRA)tested
inlargeFS
families
and
infactorialmating
design
low
efficiency
dueto
family
specificity
US
49families
GS
Metho
dto
accelerate
breeding
throug
hredu
cing
the
resourcesforcross
attemptsby
Aldquotraining
popu
latio
nrdquoof
individu
alsfrom
stages
abov
ethat
have
been
both
Risks
ofpo
orpredictio
nProg
eny
testingneedsto
becontinueddu
ring
the
timethetraining
setis
US
One
prog
rammeso
farmdash
USD
Ain
WisconsinThree
cycles
ofgeno
mic
selectioncompleted
CN~1
000
geno
types
NLprelim
inary
mod
elsforMsin
developed
UK~1
000
geno
types
Verysuitable
application
not
attempted
yet
Maybe
challeng
ingfor
interspecies
hybrids
FR(INRA)120
0geno
type
ofPop
ulus
nigraarebeingused
todevelop
intraspecificGS
(Con
tinues)
14 | CLIFTON‐BROWN ET AL
TABLE
5(Contin
ued)
Breeding
techno
logy
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sW
illow
Pop
lar
predictin
gthe
performance
ofprog
enyof
crosses
(and
inresearch
topredictbestparents
tousein
biparental
crosses)
geno
typedand
phenotyp
edto
developamod
elthattakesgeno
typic
datafrom
aldquocandidate
popu
latio
nrdquoof
untested
individu
als
andprod
uces
GEBVs(Jannink
Lorenzamp
Iwata
2010
)
beingph
enotyp
edand
geno
typed
Inthis
time
next‐generation
germ
plasm
andreadyto
beginthe
second
roun
dof
training
and
recalib
ratio
nof
geno
mic
predictio
nmod
els
US
Prelim
inary
mod
elsforMsac
and
Msin
developed
Work
isun
derw
ayto
deal
moreeffectivelywith
polyploidgenetics
calib
ratio
nsUS
125
0geno
types
arebeingused
todevelopGS
calib
ratio
ns
Traditio
nal
transgenesis
Efficient
introd
uctio
nof
ldquoforeign
rdquotraits
(possiblyfrom
other
genu
sspecies)
into
anelite
plant(ega
prov
enparent
ora
hybrid)that
needsa
simpletraitto
beim
prov
edValidate
cand
idategenesfrom
QTLstud
ies
MASand
know
ledg
eof
the
biolog
yof
thetrait
andsource
ofgenesto
confer
the
relevant
changesto
phenotyp
eWorking
transformation
protocol
IPissuescostof
regu
latory
approv
al
GMO
labelling
marketin
gissuestricky
tousetransgenes
inou
t‐breedersbecause
ofcomplexity
oftransformingandgene
flow
risks
US
Manyprog
ramsin
UnitedStates
are
creatin
gtransgenic
plantsusingAlamoas
asource
oftransformable
geno
typesManytraits
ofinterestNothing
commercial
yet
CNC
hang
shaBtg
ene
transformed
into
Msac
in20
04M
sin
transformed
with
MINAC2gene
and
Msactransform
edwith
Cry2A
ageneb
othwith
amarkerg
eneby
Agrob
acterium
JP1
Msintransgenic
forlow
temperature
tolerancewith
increased
expression
offructans
(unp
ublished)
NLImprov
ingprotocol
forM
sin
transformation
SK2
Msin
with
Arabido
psisand
Brachypod
ium
PhytochromeBgenes
(Hwang
Cho
etal
2014
Hwang
Lim
etal20
14)
UK2
Msin
and
1Mflorg
enotyp
es
UKRou
tine
transformationno
tyet
possibleResearchto
overcomerecalcitrance
ongo
ing
Currently
trying
differentspecies
andcond
ition
sPo
plar
transformationused
atpresent
US
Frequently
attempted
with
very
limitedsuccess
US
600transgenic
lines
have
been
characterized
mostly
performed
inthe
aspenhy
brids
reprod
uctiv
esterility
drou
ghttolerance
with
Kno
ckdo
wns
(Con
tinues)
CLIFTON‐BROWN ET AL | 15
TABLE
5(Contin
ued)
Breeding
techno
logy
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sW
illow
Pop
lar
variou
slytransformed
with
four
cellwall
genesiptand
uidA
genesusingtwo
selectionsystem
sby
biolisticsand
Agrob
acterium
Transform
edplantsare
beinganalysed
US
Prelim
inarywork
will
betakenforw
ardin
2018
ndash202
2in
CABBI
Genom
eediting
CRISPR
Refinem
entofexisting
traits
inusefulparents
orprom
isinghybrids
bygeneratingtargeted
mutations
ingenes
know
ntocontrolthe
traitof
interestF
irst
adouble‐stranded
breakismadeinthe
DNAwhich
isrepairedby
natural
DNArepair
machineryL
eads
tofram
eshiftSN
Por
can
usealdquorepair
templaterdquo
orcanbe
used
toinserta
transgene
intoaldquosafe
harbourlocusrdquoIt
couldbe
used
todeleterepressors(or
transcriptionfactors)
Identification
mapping
and
sequ
encing
oftarget
genes(from
DNA
sequ
ence)
avoiding
screening
outof
unintend
ededits
Manyregu
latory
authorities
have
not
decidedwhether
CRISPR
andother
geno
meediting
techno
logies
are
GMOsor
notIf
GMOsseecomment
onstage10
abov
eIf
noteditedcrop
swill
beregu
latedas
conv
entio
nalvarieties
CATechn
olog
yisstill
toonewand
switchg
rass
geno
meis
very
complexOther
labo
ratories
are
interestedbu
tno
tyet
mov
ingon
this
US
One
prog
rammeso
farmdash
USD
Ain
Albany
FRInitiated
in20
16(M
ISEDIT
project)
NLInitiatingin
2018
US
Initiatingin
2018
in
CABBI
UKNot
started
Not
possible
until
transformation
achieved
US
CRISPR
‐Cas9and
Cpf1have
been
successfully
developedin
Pop
ulus
andarehigh
lyefficientExamples
forlig
nin
biosyn
thesis
Note
BFF
BiomassFo
rtheFu
tureBtBacillus
thuringiensisCABBICenterforadvanced
bioenergyandbioproductsinnovatio
nCpf1
CRISPR
from
Prevotella
andFrancisella
1CRISPR
clusteredregularlyinterspaced
palin
drom
icrepeatsCRISPR
‐CasC
RISPR
‐associated
Cry2A
acrystaltoxins2A
asubfam
ilyproduced
byBtFS
full‐sibG
BS
genotyping
bysequencingG
EBVsgenomic‐estim
ated
breeding
valuesG
MOg
enetically
modi-
fied
organism
GS
genomicselection
GWAS
genomew
ideassociationstudy
INRAF
renchNationalInstituteforAgriculturalR
esearch
IPintellectualp
roperty
IPTisopentenyltransferase
ISSR
inter‐SSR
MflorMiscant-
husflo
ridulusMsacMsacchariflo
rusMsinMsinensisM
AS
marker‐a
ssistedselection
MISEDITm
iscanthusgene
editing
forseed‐propagatedtriploidsNACn
oapicalmeristemA
TAF1
2and
cup‐shaped
cotyledon2
‐lik
eQTLq
uantitativ
etraitlocusS
NP
singlenucleotid
epolymorphismS
SRsim
plesequence
repeatsUSD
AU
SDepartm
ento
fAgriculture
a ISO
Alpha‐2
lettercountrycodes
16 | CLIFTON‐BROWN ET AL
High seed production from field crossing trials con-ducted in locations where flowering in both seed andpollen parents is likely to happen synchronously
Scalable and adapted harvesting threshing and seed pro-cessing methods for producing high seed quality
The results of these parallel activities need to becombined to identify the upscaling pathway for eachhybrid if this cannot be achieved the hybrid will likelynot be commercially viable The UK‐led programme withpartners in Italy and Germany shows that seedbased mul-tiplication rates of 12000 are achievable several inter-specific hybrids (Clifton‐Brown et al 2017) Themultiplication rate of M sinensis is higher probably15000ndash10000 Conventional cloning from rhizome islimited to around 120 that is one ha could provide rhi-zomes for around 20 ha of new plantation
Multilocation field testing of wild and novel miscanthushybrids selected by breeding programmes in the Nether-lands and the United Kingdom was performed as part ofthe project Optimizing Miscanthus Biomass Production(OPTIMISC 2012ndash2016) These trials showed that com-mercial yields and biomass qualities (Kiesel et al 2017Van der Weijde et al 2017a Van der Weijde Kieselet al 2017) could be produced in a wide range of climatesand soil conditions from the temperate maritime climate ofwestern Wales to the continental climate of eastern Russiaand the Ukraine (Kalinina et al 2017) Extensive environ-mental measurements of soil and climate combined withgrowth monitoring are being used to understand abioticstresses (Nunn et al 2017 Van der Weijde Huxley et al2017) and develop genotype‐specific scenarios similar tothose reported earlier in Hastings et al (2009) Phenomicsexperiments on drought tolerance have been conducted onwild and improved germplasm (Malinowska Donnison ampRobson 2017 Van der Weijde Huxley et al 2017)Recently produced interspecific hybrids displaying excep-tional yield under drought (~30 greater than control Mxg)in field trials in Poland and Moldova are being furtherstudied in detail in the phenomics and genomics facility atAberystwyth to better understand genendashtrait associationswhich can be fed back into breeding
Intraspecific seeded hybrids of M sinensis produced inthe Netherlands and interspecific M sacchariflorus times M si-nensis hybrids produced by the UK‐led breeding programmehave entered yield testing in 2018 with the recently EU‐funded project ldquoGRowing Advanced industrial Crops on mar-ginal lands for biorEfineries (GRACE)rdquo (httpswwwgrace-bbieu) Substantial variation in biomass quality for sacchari-fication efficiency (glucose release as of dry matter) ashcontent and melting point has already been generated inintraspecific M sinensis hybrids (Van der Weijde Kiesel
et al 2017) across environments (Weijde Dolstra et al2017a) GRACE aims to establish more than 20 hectares ofnew inter‐ and intraspecific seeded hybrids across six Euro-pean countries This project is building the know‐how andagronomy needed to transition from small research plots tocommercial‐scale field sites and linking biomass productiondirectly to industrial applications The biomass produced byhybrids in different locations will be supplied to innovativeindustrial end‐users making a wide range of biobased prod-ucts both for chemicals and for energy In the United Statesmulti‐location yield were initiated in 2018 to evaluate new tri-ploid M times giganteus genotypes developed at Illinois Cur-rently infertile hybrids are favoured in the United Statesbecause this eliminates the risk of invasiveness from naturallydispersed viable seed The precautionary principle is appliedas fertile miscanthus has naturalized in several states (QuinnAllen amp Stewart 2010) North European multilocation fieldtrials in the EMI and OPTIMISC projects have shown thereis minimal risk of invasiveness even in years when fertileflowering hybrids produce viable seed Naturalized standshave not established here due perhaps to low dormancy pooroverwintering and low seedling competitive strength In addi-tion to breeding for nonshattering or sterile seeded hybridsQuinn et al (2010) suggest management strategies which canfurther minimize environmental opportunities to manage therisk of invasiveness
31 | Molecular breeding and biotechnology
In miscanthus new plant breeding techniques (Table 5)have focussed on developing molecular markers forbreeding in Europe the United States South Korea andJapan There are several publications on QTL mappingpopulations for key traits such as flowering (AtienzaRamirez amp Martin 2003) and compositional traits(Atienza Satovic Petersen Dolstra amp Martin 2003) Inthe United States and United Kingdom independent andinterconnected bi‐parental ldquomappingrdquo families have beenstudied (Dong et al 2018 Gifford Chae SwaminathanMoose amp Juvik 2015) alongside panels of diverse germ-plasm accessions for GWAS (Slavov et al 2013) Fur-ther developments calibrating GS with very large panelsof parents and cross progeny are underway (Daveyet al 2017) The recently completed first miscanthus ref-erence genome sequence is expected to improve the effi-ciency of MAS strategies and especially GWAS(httpsphytozomejgidoegovpzportalhtmlinfoalias=Org_Msinensis_er) For example without a referencegenome sequence Clark et al (2014) obtained 21207RADseq SNPs (single nucleotide polymorphisms) on apanel of 767 miscanthus genotypes (mostly M sinensis)but subsequent reanalysis of the RADseq data using the
CLIFTON‐BROWN ET AL | 17
new reference genome resulted in hundreds of thousandsof SNPs being called
Robust and effective in vitro regeneration systems havebeen developed for Miscanthus sinensis M times giganteusand M sacchariflorus (Dalton 2013 Guo et al 2013Hwang Cho et al 2014 Rambaud et al 2013 Ślusarkie-wicz‐Jarzina et al 2017 Wang et al 2011 Zhang et al2012) However there is still significant genotype speci-ficity and these methods need ldquoin‐houserdquo optimization anddevelopment to be used routinely They provide potentialroutes for rapid clonal propagation and also as a basis forgenetic transformation Stable transformation using bothbiolistics (Wang et al 2011) and Agrobacterium tumefa-ciens DNA delivery methods (Hwang Cho et al 2014Hwang Lim et al 2014) has been achieved in M sinen-sis The development of miscanthus transformation andgene editing to generate diplogametes for producing seed‐propagated triploid hybrids are performed as part of theFrench project MISEDIT (miscanthus gene editing forseed‐propagated triploids) There are no reports of genomeediting in any miscanthus species but new breeding inno-vations including genome editing are particularly relevantin this slow‐to‐breed nonfood bioenergy crop (Table 4)
4 | WILLOW
Willow (Salix spp) is a very diverse group of catkin‐bear-ing trees and shrubs Willow belongs to the family Sali-caceae which also includes the Populus genus There areapproximately 350 willow species (Argus 2007) foundmostly in temperate and arctic zones in the northern hemi-sphere A few are adapted to subtropical and tropicalzones The centre of diversity is believed to be in Asiawith over 200 species in China Around 120 species arefound in the former Soviet Union over 100 in NorthAmerica and around 65 species in Europe and one speciesis native to South America (Karp et al 2011) Willows aredioecious thus obligate outcrossers and highly heterozy-gous The haploid chromosome number is 19 (Hanley ampKarp 2014) Around 40 of species are polyploid (Sudaamp Argus 1968) ranging from triploids to the atypicaldodecaploid S maxxaliana with 2n=190 (Zsuffa et al1984)
Although almost exclusively native to the NorthernHemisphere willow has been grown around the globe formany thousands of years to support a wide range of appli-cations (Kuzovkina amp Quigley 2005 Stott 1992) How-ever it has been the focus of domestication for bioenergypurposes for only a relatively short period since the 1970sin North America and Europe For bioenergy breedershave focused their efforts on the shrub willows (subgenusVetix) because of their rapid juvenile growth rates as aresponse to coppicing on a 2‐ to 4‐year cycle that can be
accomplished using farm machinery rather than forestryequipment (Shield Macalpine Hanley amp Karp 2015Smart amp Cameron 2012)
Since shrub willow was not generally recognized as anagricultural crop until very recently there has been littlecommitment to building and maintaining germplasm reposi-tories of willow to support long‐term breeding One excep-tion is the United Kingdom where a large and well‐characterized Salix germplasm collection comprising over1500 accessions is held at Rothamsted Research (Stott1992 Trybush et al 2008) Originally initiated for use inbasketry in 1923 accessions have been added ever sinceIn the United States a germplasm collection of gt350accessions is located at Cornell University to support thebreeding programme there The UK and Cornell collectionshave a relatively small number of accessions in common(around 20) Taken together they represent much of thespecies diversity but only a small fraction of the overallgenetic diversity within the genus There are three activewillow breeding programmes in Europe RothamstedResearch (UK) Salixenergi Europa AB (SEE) and a pro-gramme at the University of Warmia and Mazury in Olsz-tyn (Poland) (abbreviations used in Table 1) There is oneactive US programme based at Cornell University Culti-vars are still being marketed by the European WillowBreeding Programme (EWBP) (UK) which was activelybreeding biomass varieties from 1996 to 2002 Cultivarsare protected by plant breedersrsquo rights (PBRs) in Europeand by plant patents in the United States The sharing ofgenetic resources in the willow community is generallyregulated by material transfer agreements (MTA) and tai-lored licensing agreements although the import of cuttingsinto North America is prohibited except under special quar-antine permit conditions
Efforts to augment breeding germplasm collection fromnature are continuing with phenotypic screening of wildgermplasm performed in field experiments with 177 S pur-purea genotypes in the United States (at sites in Genevaand Portland NY and Morgantown WV) that have beengenotyped using genotyping by sequencing (GBS) (Elshireet al 2011) In addition there are approximately 400accessions of S viminalis in Europe (near Pustnaumls Upp-sala Sweden and Woburn UK (Berlin et al 2014 Hal-lingbaumlck et al 2016) The S viminalis accessions wereinitially genotyped using 38 simple sequence repeats (SSR)markers to assess genetic diversity and screened with~1600 SNPs in genes of potential interest for phenologyand biomass traits Genetics and genomics combined withextensive phenotyping have substantially improved thegenetic basis of biomass‐related traits in willow and arenow being developed in targeted breeding via MAS Thisunderpinning work has been conducted on large specifi-cally developed biparental Salix mapping populations
18 | CLIFTON‐BROWN ET AL
(Hanley amp Karp 2014 Zhou et al 2018) as well asGWAS panels (Hallingbaumlck et al 2016)
Once promising parental combinations are identifiedcrosses are usually performed using fresh pollen frommaterial that has been subject to a phased removal fromcold storage (minus4degC) (Lindegaard amp Barker 1997 Macal-pine Shield Trybush Hayes amp Karp 2008 Mosseler1990) Pollen storage is useful in certain interspecific com-binations where flowering is not naturally synchronizedThis can be overcome by using pollen collection and stor-age protocol which involves extracting pollen using toluene(Kopp Maynard Niella Smart amp Abrahamson 2002)
The main breeding approach to improve willow yieldsrelies on species hybridization to capture hybrid vigour(Fabio et al 2017 Serapiglia Gouker amp Smart 2014) Inthe absence of genotypic models for heterosis breedershave extensively tested general and specific combiningability of parents to produce superior progeny The UKbreeding programmes (EWBP 1996ndash2002 and RothamstedResearch from 2003 on) have performed more than 1500exploratory cross‐pollinations The Cornell programme hassuccessfully completed about 550 crosses since 1998Investment into the characterization of genetic diversitycombined with progeny tests from exploratory crosses hasbeen used to produce hundreds of targeted intraspeciescrosses in the United Kingdom and United States respec-tively (see Table 1) To achieve long‐term gains beyond F1hybrids four intraspecific recurrent selection populationshave been created in the United Kingdom (for S dasycla-dos S viminalis and S miyabeana) and Cornell is pursu-ing recurrent selection of S purpurea Interspecifichybridizations with genotypes selected from the recurrentselection cycles are well advanced in willow with suchcrosses to date totalling 420 in the United Kingdom andover 100 in the United States
While species hybridization is common in Salix it isnot universal Of the crosses attempted about 50 hybri-dize and produce seed (Macalpine Shield amp Karp 2010)As the viability of seed from successful crosses is short (amatter of days at ambient temperatures) proper seed rear-ing and storage protocols are essential (Maroder PregoFacciuto amp Maldonado 2000)
Progeny from crosses are treated in different waysamong the breeding programmes at the seedling stage Inthe United States seedlings are planted into an irrigatedfield where plants are screened for two seasons beforebeing progressed to further field trials In the United King-dom seedlings are planted into trays of compost wherethey remain containerized in an irrigated nursery for theremainder of year one In the United Kingdom seedlingsare subject to two rounds of selection in the nursery yearThe first round takes place in September to select againstsusceptibility to rust infection (Melampsora spp) A second
round of selection in winter assesses tip damage from frostand giant willow aphid infestation In the United Stateswhere the rust pressure is lower screening for Melampsoraspp cannot be performed at the nursery stage Both pro-grammes monitor Melampsora spp pest susceptibilityyield and architecture over multiple years in field trialsSelected material is subject to two rounds of field trials fol-lowed by a final multilocation yield trial to identify vari-eties for commercialization
Promising selections (ie potential cultivars) need to beclonally propagated A rapid in vitro tissue culture propa-gation method has been developed (Palomo‐Riacuteos et al2015) This method can generate about 5000 viable trans-plantable clones from a single plant in just 24 weeks Anin vitro system can also accommodate early selection viamolecular or biochemical markers to increase selectionspeed Conventional breeding systems take 13 years viafour rounds of selection from crossing to selecting a variety(Figure 1) but this has the potential to be reduced to7 years if micropropagation and MAS selection are adopted(Hanley amp Karp 2014 Palomo‐Riacuteos et al 2015)
Willows are currently propagated commercially byplanting winter‐dormant stem cuttings in spring Commer-cial planting systems for willow use mechanical plantersthat cut and insert stem sections from whips into a well‐prepared soil One hectare of stock plants grown in specificmultiplication beds planted at 40000 plants per ha pro-duces planting material for 80 hectares of commercialshort‐rotation coppice willow annually (planted at 15000cuttings per hectare) (Whittaker et al 2016) When com-mercial plantations are established the industry standard isto plant intimate mixtures of ~5 diverse rust (Melampsoraspp)‐resistant varieties (McCracken amp Dawson 1997 VanDen Broek et al 2001)
The foundations for using new plant breeding tech-niques have been established with funding from both thepublic and the private sectors To establish QTL maps 16mapping populations from biparental crosses are understudy in the United Kingdom Nine are under study in theUnited States The average number of individuals in thesefamilies ranges from 150 to 947 (Hanley amp Karp 2014)GS is also being evaluated in S viminalis and preliminaryresults indicate that multiomic approaches combininggenomic and metabolomic data have great potential (Sla-vov amp Davey 2017) For both QTL and GS approachesthe field phenotyping demands are large as several thou-sand individuals need to be phenotyped for a wide rangeof traits These include the following dates of bud burstand growth succession stem height stem density wooddensity and disease resistance The greater the number ofindividuals the more precise the QTL marker maps andGS models are However the logistical and financial chal-lenges of phenotyping large numbers of individuals are
CLIFTON‐BROWN ET AL | 19
considerable because the willow crop is gt5 m tall in thesecond year There is tremendous potential to improve thethroughput of phenotyping using unmanned aerial systemswhich is being tested in the USDA National Institute ofFood and Agriculture (NIFA) Willow SkyCAP project atCornell Further investment in these approaches needs tobe sustained over many years fully realizes the potentialof a marker‐assisted selection programme for willow
To date despite considerable efforts in Europe and theUnited States to establish a routine transformation systemthere has not been a breakthrough in willow but attemptsare ongoing As some form of transformation is typically aprerequisite for genome editing techniques these have notyet been applied to willow
In Europe there are 53 short‐rotation coppice (SRC) bio-mass willow cultivars registered with the Community PlantVariety Office (CPVO) for PBRs of which ~25 are availablecommercially in the United Kingdom There are eightpatented cultivars commercially available in the UnitedStates In Sweden there are nine commercial cultivars regis-tered in Europe and two others which are unregistered(httpssalixenergiseplanting-material) Furthermore thereare about 20 precommercial hybrids in final yield trials inboth the United States and the United Kingdom It has beenestimated that it would take two years to produce the stockrequired to plant 50 ha commercially from the plant stock inthe final yield trials Breeding programmes have alreadydelivered rust‐resistant varieties and increases in yield to themarket The adoption of advanced breeding technologies willlikely lead to a step change in improving traits of interest
5 | POPLAR
Poplar a fast‐growing tree from the northern hemispherewith a small genome size has been adopted for commercialforestry and scientific purposes The genus Populus con-sists of about 29 species classified in six different sectionsPopulus (formerly Leuce) Tacamahaca Aigeiros AbasoTuranga and Leucoides (Eckenwalder 1996) The Populusspecies of most interest for breeding and testing in the Uni-ted States and Europe are P nigra P deltoides P maxi-mowiczii and P trichocarpa (Stanton 2014) Populusclones for biomass production are being developed byintra‐ and interspecies hybridization (DeWoody Trewin ampTaylor 2015 Richardson Isebrands amp Ball 2014 van derSchoot et al 2000) Recurrent selection approaches areused for gradual population improvement and to create eliteclonal lines for commercialization (Berguson McMahon ampRiemenschneider 2017 Neale amp Kremer 2011) Currentlypoplar breeding in the United States occurs in industrialand academic programmes located in the Southeast theMidwest and the Pacific Northwest These use six speciesand five interspecific taxa (Stanton 2014)
The southeastern programme historically focused onrecurrent selection of P deltoides from accessions made inthe lower Mississippi River alluvial plain (Robison Rous-seau amp Zhang 2006) More recently the genetic base hasbeen broadened to produce interspecific hybrids with resis-tance to the fungal infection Septoria musiva which causescankers
In the midwest of the United States populationimprovement efforts are focused on P deltoides selectionsfrom native provenances and hybrid crosses with acces-sions introduced from Europe Interspecific intercontinental(Europe and America) hybrid crosses between P nigra andP deltoides (P times canadensis) are behind many of theleading commercial hybrids which are the most advancedbreeding materials for many applications and regions InMinnesota previous breeding experience and efforts utiliz-ing P maximowiczii and P trichocarpa have been discon-tinued due to Septoria susceptibility and a lack of coldhardiness (Berguson et al 2017) Traits targeted forimprovement include yieldgrowth rate cold hardinessadventitious rooting resistance to Septoria and Melamp-sora leaf rust and stem form The Upper Midwest pro-gramme also carries out wide hybridizations within thesection Populus The P times wettsteinii (P trem-ula times P tremuloides) taxon is bred for gains in growthrate wood quality and resistance to the fungus Entoleucamammata which causes hypoxylon canker (David ampAnderson 2002)
In the Pacific Northwest GreenWood Resources Incleads poplar breeding that emphasizes interspecifichybrid improvement of P times generosa (P deltoides timesP trichocarpa and reciprocal) and P deltoides times P maxi-mowiczii taxa for coastal regions and the P times canadensistaxon for the drier continental regions Intraspecificimprovement of second‐generation breeding populations ofP deltoides P nigra P maximowiczii and P trichocarpaare also involved (Stanton et al 2010) The present focusof GreenWood Resourcesrsquo hybridization is bioenergy feed-stock improvement concentrating on coppice yield wood‐specific gravity and rate of sugar release
Industrial interest in poplar in the United States has his-torically come from the pulp and paper sector althoughveneer and dimensional lumber markets have been pursuedat times Currently the biomass market for liquid trans-portation fuels is being emphasized along with the use oftraditional and improved poplar genotypes for ecosystemservices such as phytoremediation (Tuskan amp Walsh 2001Zalesny et al 2016)
In Europe there are breeding programmes in FranceGermany Italy and Sweden These include the following(a) Alasia Franco Vivai (AFV) programme in northernItaly (b) the French programme led by the poplar Scien-tific Interest Group (GIS Peuplier) and carried out
20 | CLIFTON‐BROWN ET AL
collaboratively between the National Institute for Agricul-tural Research (INRA) the National Research Unit ofScience and Technology for Environment and Agriculture(IRSTEA) and the Forest Cellulose Wood Constructionand Furniture Technology Institute (FCBA) (c) the Ger-man programme at Northwest German Forest Research Sta-tion (NW‐FVA) at Hannoversch Muumlnden and (d) theSwedish programme at the Swedish University of Agricul-tural Sciences and SweTree Technologies AB (Table 1)
AFV leads an Italian poplar breeding programme usingextensive field‐grown germplasm collections of P albaP deltoides P nigra and P trichocarpa While interspeci-fic hybridization uses several taxa the focus is onP times canadensis The breeding programme addresses dis-ease resistance (Marssonina brunnea Melampsora larici‐populina Discosporium populeum and poplar mosaicvirus) growth rate and photoperiod adaptation AFV andGreenWood Resources collaborate in poplar improvementin Europe through the exchange of frozen pollen and seedfor reciprocal breeding projects Plantations in Poland andRomania are currently the focus of the collaboration
The ongoing French GIS Peuplier is developing a long‐term breeding programme based on intraspecific recurrentselection for the four parental species (P deltoides P tri-chocarpa P nigra and P maximowiczii) designed to betterbenefit from hybrid vigour demonstrated by the interspeci-fic crosses P canadensis P deltoides times P trichocarpaand P trichocarpa times P maximowiczii Current selectionpriorities are targeting adaptation to soil and climate condi-tions resistance and tolerance to the most economicallyimportant diseases and pests high volume production underSRC and traditional poplar cultivation regimes as well aswood quality of interest by different markets Currentlygenomic selection is under exploration to increase selectionaccuracy and selection intensity while maintaining geneticdiversity over generations
The German NW‐FVA programme is breeding intersec-tional AigeirosndashTacamahaca hybrids with a focus on resis-tance to Pollaccia elegans Xanthomonas populiDothichiza spp Marssonina brunnea and Melampsoraspp (Stanton 2014) Various cross combinations ofP maximowiczii P trichocarpa P nigra and P deltoideshave led to new cultivars suitable for deployment in vari-etal mixtures of five to ten genotypes of complementarystature high productivity and phenotypic stability (Weis-gerber 1993) The current priority is the selection of culti-vars for high‐yield short‐rotation biomass production Sixhundred P nigra genotypes are maintained in an ex situconservation programme An in situ P nigra conservationeffort involves an inventory of native stands which havebeen molecular fingerprinted for identity and diversity
The Swedish programme is concentrating on locallyadapted genotypes used for short‐rotation forestry (SRF)
because these meet the needs of the current pulping mar-kets Several field trials have shown that commercial poplarclones tested and deployed in Southern and Central Europeare not well adapted to photoperiods and low temperaturesin Sweden and in the Baltics Consequently SwedishUniversity of Agricultural Sciences and SweTree Technolo-gies AB started breeding in Sweden in 1990s to producepoplar clones better adapted to local climates and markets
51 | Molecular breeding technologies
Poplar genetic improvement cannot be rapidly achievedthrough traditional methods alone because of the longbreeding cycles outcrossing breeding systems and highheterozygosity Integrating modern genetic genomic andphenomics techniques with conventional breeding has thepotential to expedite poplar improvement
The genome of poplar has been sequenced (Tuskanet al 2006) It has an estimated genome size of485 plusmn 10 Mbp divided into 19 chromosomes This is smal-ler than other PBCs and makes poplar more amenable togenetic engineering (transgenesis) GS and genome editingPoplar has seen major investment in both the United Statesand Europe being the model system for woody perennialplant genetics and genomics research
52 | Targets for genetic modification
Traits targeted include wood properties (lignin content andcomposition) earlylate flowering male sterility to addressbiosafety regulation issues enhanced yield traits and herbi-cide tolerance These extensive transgenic experiments haveshown differences in recalcitrance to in vitro regenerationand genetic transformation in some of the most importantcommercial hybrid poplars (Alburquerque et al 2016)Further transgene expression stability is being studied Sofar China is the only country known to have commerciallyused transgenic insect‐resistant poplar A precommercialherbicide‐tolerant poplar was trialled for 8 years in the Uni-ted States (Li Meilan Ma Barish amp Strauss 2008) butcould not be released due to stringent environmental riskassessments required for regulatory approval This increasestranslation costs and delays reducing investor confidencefor commercial deployment (Harfouche et al 2011)
The first field trials of transgenic poplar were performedin France in 1987 (Fillatti Sellmer Mccown Haissig ampComai 1987) and in Belgium in 1988 (Deblock 1990)Although there have been a total of 28 research‐scale GMpoplar field trials approved in the European Union underCouncil Directive 90220EEC since October 1991 (inPoland Belgium Finland France Germany Spain Swe-den and in the United Kingdom (Pilate et al 2016) onlyauthorizations in Poland and Belgium are in place today In
CLIFTON‐BROWN ET AL | 21
the United States regulatory notifications and permits fornearly 20000 transgenic poplar trees derived from approxi-mately 600 different constructs have been issued since1995 by the USDAs Animal and Plant Health InspectionService (APHIS) (Strauss et al 2016)
53 | Genome editing CRISPR technologies
Clustered regularly interspaced palindromic repeats(CRISPR) and the CRISPR‐associated (CRISPR‐Cas)nucleases are a groundbreaking genome‐engineering toolthat complements classical plant breeding and transgenicmethods (Moreno‐Mateos et al 2017) Only two publishedstudies in poplar have applied the CRISPRCas9 technol-ogy One is in P tomentosa in which an endogenous phy-toene desaturase gene (PtoPDS) was successfully disruptedsite specifically in the first generation of transgenic plantsresulting in an albino and dwarf phenotype (Fan et al2015) The second was in P tremula times alba in whichhigh CRISPR‐Cas9 mutational efficiency was achieved forthree 4‐coumarateCoA ligase (4Cl) genes 4CL1 4CL2and 4CL5 associated with lignin and flavonoid biosynthe-sis (Zhou et al 2015) Due to its low cost precision andrapidness it is very probable that cultivars or clones pro-duced using CRISPR technology will be ready for market-ing in the near future (Yin et al 2017) Recently aCRISPR with a smaller associated endonuclease has beendiscovered from Prevotella and Francisella 1 (Cpf1) whichmay have advantages over Cas9 In addition there arereports of DNA‐free editing in plants using both CRISPRCpf1 and CRISPR Cas9 for example Ref (Kim et al2017 Mahfouz 2017 Zaidi et al 2017)
It remains unresolved whether plants modified by genomeediting will be regulated as genetically modified organisms(GMOs) by the relevant authorities in different countries(Lozano‐Juste amp Cutler 2014) Regulations to cover thesenew breeding techniques are still evolving but those coun-tries who have published specific guidance (including UnitedStates Argentina and Chile) are indicating that plants pos-sessing simple genome edits will not be regulated as conven-tional transgenesis (Jones 2015b) The first generation ofgenome‐edited crops will likely be phenocopy gene knock-outs that already exist to produce ldquonature identicalrdquo traits thatis traits that could also be derived by conventional breedingDespite this confidence in applying these new powerfulbreeding tools remains limited owing to the uncertain regula-tory environment in many parts of the world (Gao 2018)including the recent ECJ 2018 rulings mentioned earlier
54 | Genomics‐based breeding technologies
Poplar breeding programs are becoming well equipped withuseful genomics tools and resources that are critical to
explore genomewide variability and make use of the varia-tion for enhancing genetic gains Deep transcriptomesequencing resequencing of alternate genomes and GBStechnology for genomewide marker detection using next‐generation sequencing (NGS) are yielding valuable geno-mics tools GWAS with NGS‐based markers facilitatesmarker identification for MAS breeding by design and GS
GWAS approaches have provided a deeper understand-ing of genome function as well as allelic architectures ofcomplex traits (Huang et al 2010) and have been widelyimplemented in poplar for wood characteristics (Porthet al 2013) stomatal patterning carbon gain versus dis-ease resistance (McKown et al 2014) height and phenol-ogy (Evans et al 2014) cell wall chemistry (Mucheroet al 2015) growth and cell walls traits (Fahrenkroget al 2017) bark roughness (Bdeir et al 2017) andheight and diameter growth (Liu et al 2018) Using high‐throughput sequencing and genotyping platforms an enor-mous amount of SNP markers have been used to character-ize the linkage disequilibrium (LD) in poplar (eg Slavovet al 2012 discussed below)
The genetic architecture of photoperiodic traits in peren-nial trees is complex involving many loci However itshows high levels of conservation during evolution (Mau-rya amp Bhalerao 2017) These genomics tools can thereforebe used to address adaptation issues and fine‐tune themovement of elite lines into new environments For exam-ple poor timing of spring bud burst and autumn bud setcan result in frost damage resulting in yield losses (Ilstedt1996) These have been studied in P tremula genotypesalong a latitudinal cline in Sweden (~56ndash66oN) and haverevealed high nucleotide polymorphism in two nonsynony-mous SNPs within and around the phytochrome B2 locus(Ingvarsson Garcia Hall Luquez amp Jansson 2006 Ing-varsson Garcia Luquez Hall amp Jansson 2008) Rese-quencing 94 of these P tremula genotypes for GWASshowed that noncoding variation of a single genomicregion containing the PtFT2 gene described 65 ofobserved genetic variation in bud set along the latitudinalcline (Tan 2018)
Resequencing genomes is currently the most rapid andeffective method detecting genetic differences betweenvariants and for linking loci to complex and importantagronomical and biomass traits thus addressing breedingchallenges associated with long‐lived plants like poplars
To date whole genome resequencing initiatives havebeen launched for several poplar species and genotypes InPopulus LD studies based on genome resequencing sug-gested the feasibility of GWAS in undomesticated popula-tions (Slavov et al 2012) This plant population is beingused to inform breeding for bioenergy development Forexample the detection of reliable phenotypegenotype asso-ciations and molecular signatures of selection requires a
22 | CLIFTON‐BROWN ET AL
detailed knowledge about genomewide patterns of allelefrequency variation LD and recombination suggesting thatGWAS and GS in undomesticated populations may bemore feasible in Populus than previously assumed Slavovet al (2012) have resequenced 16 genomes of P tri-chocarpa and genotyped 120 trees from 10 subpopulationsusing 29213 SNPs (Geraldes et al 2013) The largest everSNP data set of genetic variations in poplar has recentlybeen released providing useful information for breedinghttpswwwbioenergycenterorgbescgwasindexcfmAlso deep sequencing of transcriptomes using RNA‐Seqhas been used for identification of functional genes andmolecular markers that is polymorphism markers andSSRs A multitissue and multiple experimental data set forP trichocarpa RNA‐Seq is publicly available (httpsjgidoegovdoe-jgi-plant-flagship-gene-atlas)
The availability of genomic information of DNA‐con-taining cell organelles (nucleus chloroplast and mitochon-dria) will also allow a holistic approach in poplarmolecular breeding in the near future (Kersten et al2016) Complete Populus genome sequences are availablefor nucleus (P trichocarpa section Tacamahaca) andchloroplasts (seven species and two clones from P trem-ula W52 and P tremula times P alba 717ndash1B4) A compara-tive approach revealed structural and functionalinformation broadening the knowledge base of PopuluscpDNA and stimulating future diagnostic marker develop-ment The availability of whole genome sequences of thesecellular compartments of P tremula holds promise forboosting marker‐assisted poplar breeding Other nucleargenome sequences from additional Populus species arenow available (eg P deltoides (httpsphytozomejgidoegovpz) and will become available in the forthcomingyears (eg P tremula and P tremuloidesmdashPopGenIE(Sjodin Street Sandberg Gustafsson amp Jansson 2009))Recently the characterization of the poplar pan‐genome bygenomewide identification of structural variation in threecrossable poplar species P nigra P deltoides and P tri-chocarpa revealed a deeper understanding of the role ofinter‐ and intraspecific structural variants in poplar pheno-type and may have important implications for breedingparticularly interspecific hybrids (Pinosio et al 2016)
GS has been proposed as an alternative to MAS in cropimprovement (Bernardo amp Yu 2007 Heffner Sorrells ampJannink 2009) GS is particularly well suited for specieswith long generation times for characteristics that displaymoderate‐to‐low heritability for traits that are expensive tomeasure and for selection of traits expressed late in the lifecycle as is the case for most traits of commercial value inforestry (Harfouche et al 2012) Current joint genomesequencing efforts to implement GS in poplar using geno-mic‐estimated breeding values for bioenergy conversiontraits from 49 P trichocarpa families and 20 full‐sib
progeny are taking place at the Oak Ridge National Labo-ratory and GreenWood Resources (Brian Stanton personalcommunication httpscbiornlgov) These data togetherwith the resequenced GWAS population data will be thebasis for developing GS algorithms Genomic breedingtools have been developed for the intraspecific programmetargeting yield resistance to Venturia shoot blightMelampsora leaf rust resistance to Cryptorhynchus lapathistem form wood‐specific gravity and wind firmness (Evanset al 2014 Guerra et al 2016) A newly developedldquobreeding with rare defective allelesrdquo (BRDA) technologyhas been developed to exploit natural variation of P nigraand identify defective variants of genes predicted by priortransgenic research to impact lignin properties Individualtrees carrying naturally defective alleles can then be incor-porated directly into breeding programs thereby bypassingthe need for transgenics (Vanholme et al 2013) Thisnovel breeding technology offers a reverse genetics com-plement to emerging GS for targeted improvement of quan-titative traits (Tsai 2013)
55 | Phenomics‐assisted breeding technology
Phenomics involves the characterization of phenomesmdashthefull set of phenotypes of given individual plants (HouleGovindaraju amp Omholt 2010) Traditional phenotypingtools which inefficiently measure a limited set of pheno-types have become a bottleneck in plant breeding studiesHigh‐throughput plant phenotyping facilities provide accu-rate screening of thousands of plant breeding lines clones orpopulations over time (Fu 2015) are critical for acceleratinggenomics‐based breeding Automated image collection andanalysis phenomics technologies allow accurate and nonde-structive measurements of a diversity of phenotypic traits inlarge breeding populations (Gegas Gay Camargo amp Doo-nan 2014 Goggin Lorence amp Topp 2015 Ludovisi et al2017 Shakoor Lee amp Mockler 2017) One important con-sideration is the identification of relevant and quantifiabletarget traits that are early diagnostic indicators of biomassyield Good progress has been made in elucidating theseunderpinning morpho‐physiological traits that are amenableto remote sensing in Populus (Harfouche Meilan amp Altman2014 Rae Robinson Street amp Taylor 2004) Morerecently Ludovisi et al (2017) developed a novel methodol-ogy for field phenomics of drought stress in a P nigra F2partially inbred population using thermal infrared imagesrecorded from an unmanned aerial vehicle‐based platform
Energy is the current main market for poplar biomass butthe market return provided is not sufficient to support pro-duction expansion even with added demand for environmen-tal and land management ldquoecosystem servicesrdquo such as thetreatment of effluent phytoremediation riparian buffer zonesand agro‐forestry plantings Aviation fuel is a significant
CLIFTON‐BROWN ET AL | 23
target market (Crawford et al 2016) To serve this marketand to reduce current carbon costs of production (Budsberget al 2016) key improvement traits in addition to yield(eg coppice regeneration pestdisease resistance water‐and nutrient‐use efficiencies) will be trace greenhouse gas(GHG) emissions (eg isoprene volatiles) site adaptabilityand biomass conversion efficiency Efforts are underway tohave national environmental protection agenciesrsquo approvalfor poplar hybrids qualifying for renewable energy credits
6 | REFLECTIONS ON THECOMMERCIALIZATIONCHALLENGE
The research and innovation activities reviewed in thispaper aim to advance the genetic improvement of speciesthat can provide feedstocks for bioenergy applicationsshould those markets eventually develop These marketsneed to generate sufficient revenue and adequately dis-tribute it to the actors along the value chain The work onall four crops shares one thing in common long‐termefforts to integrate fundamental knowledge into breedingand crop development along a research and development(RampD) pipeline The development of miscanthus led byAberystwyth University exemplifies the concerted researcheffort that has integrated the RampD activities from eight pro-jects over 14 years with background core research fundingalong an emerging innovation chain (Figure 2) This pro-gramme has produced a first range of conventionally bredseeded interspecies hybrids which are now in upscaling tri-als (Table 3) The application of molecular approaches(Table 5) with further conventional breeding (Table 4)offers the prospect of a second range of improved seededhybrids This example shows that research‐based support ofthe development of new crops or crop types requires along‐term commitment that goes beyond that normallyavailable from project‐based funding (Figure 1) Innovationin this sector requires continuous resourcing of conven-tional breeding operations and capability to minimize timeand investment losses caused by funding discontinuities
This challenge is increased further by the well‐knownmarket failure in the breeding of many agricultural cropspecies The UK Department for Environment Food andRural Affairs (Defra) examined the role of geneticimprovement in relation to nonmarket outcomes such asenvironmental protection and concluded that public invest-ment in breeding was required if profound market failure isto be addressed (Defra 2002) With the exception ofwidely grown hybrid crops such as maize and some high‐value horticultural crops royalties arising from plant breed-ersrsquo rights or other returns to breeders fail to adequatelycompensate for the full cost for research‐based plant breed-ing The result even for major crops such as wheat is sub‐
optimal investment and suboptimal returns for society Thismarket failure is especially acute for perennial crops devel-oped for improved sustainability rather than consumerappeal (Tracy et al 2016) Figure 3 illustrates the underly-ing challenge of capturing value for the breeding effortThe ldquovalley of deathrdquo that results from the low and delayedreturns to investment applies generally to the research‐to‐product innovation pipeline (Beard Ford Koutsky amp Spi-wak 2009) and certainly to most agricultural crop speciesHowever this schematic is particularly relevant to PBCsMost of the value for society from the improved breedingof these crops comes from changes in how agricultural landis used that is it depends on the increased production ofthese crops The value for society includes many ecosys-tems benefits the effects of a return to seminatural peren-nial crop cover that protects soils the increase in soilcarbon storage the protection of vulnerable land or the cul-tivation of polluted soils and the reductions in GHG emis-sions (Lewandowski 2016) By its very nature theproduction of biomass on agricultural land marginal forfood production challenges farm‐level profitability Thecosts of planting material and one‐time nature of cropestablishment are major early‐stage costs and therefore theopportunities for conventional royalty capture by breedersthat are manifold for annual crops are limited for PBCs(Hastings et al 2017) Public investment in developingPBCs for the nonfood biobased sector needs to providemore long‐term support for this critical foundation to a sus-tainable bioeconomy
7 | CONCLUSIONS
This paper provides an overview of research‐based plantbreeding in four leading PBCs For all four PBC generasignificant progress has been made in genetic improvementthrough collaboration between research scientists and thoseoperating ongoing breeding programmes Compared withthe main food crops most PBC breeding programmes dateback only a few decades (Table 1) This breeding efforthas thus co‐evolved with molecular biology and the result-ing ‐omics technologies that can support breeding Thedevelopment of all four PBCs has depended strongly onpublic investment in research and innovation The natureand driver of the investment varied In close associationwith public research organizations or universities all theseprogrammes started with germplasm collection and charac-terization which underpin the selection of parents forexploratory wide crosses for progeny testing (Figure 1Table 4)
Public support for switchgrass in North America wasexplicitly linked to plant breeding with 12 breeding pro-grammes supported in the United States and CanadaSwitchgrass breeding efforts to date using conventional
24 | CLIFTON‐BROWN ET AL
breeding have resulted in over 36 registered cultivars inthe United States (Table 1) with the development of dedi-cated biomass‐type cultivars coming within the past fewyears While ‐omics technologies have been incorporatedinto several of these breeding programmes they have notyet led to commercial deployment in either conventional orhybrid cultivars
Willow genetic improvement was led by the researchcommunity closely linked to plant breeding programmesWillow and poplar have the longest record of public invest-ment in genetic improvement that can be traced back to1920s in the United Kingdom and United States respec-tively Like switchgrass breeding programmes for willoware connected to public research efforts The United King-dom in partnership with the programme based at CornellUniversity remains the European leader in willowimprovement with a long‐term breeding effort closelylinked to supporting biological research at Rothamsted Inwillow F1 hybrids have produced impressive yield gainsover parental germplasm by capturing hybrid vigour Over30 willow clones are commercially available in the UnitedStates and Europe and a further ~90 are under precommer-cial testing (Table 1)
Compared with willow and poplar miscanthus is a rela-tive newcomer with all the current breeding programmesstarting in the 2000s Clonal M times giganteus propagated byrhizomes is expected to be replaced by more readily scal-able seeded hybrids from intra‐ (M sinensis) and inter‐(M sacchariflorus times M sinensis) species crosses with highseed multiplication rates (of gt2000) The first group ofhybrid cultivars is expected to be market‐ready around2022
Of the four genera used as PBCs Populus is the mostadvanced in terms of achievements in biological researchas a result of its use as a model for basic research of treesMuch of this biological research is not directly connectedto plant breeding Nevertheless reflecting the fact thatpoplar is widely grown as a single‐stem tree in SRF thereare about 60 commercially available clones and an addi-tional 80 clones in commercial pipelines (Table 1) Trans-genic poplar hybrids have moved beyond proof of conceptto commercial reality in China
Many PBC programmes have initiated long‐term con-ventional recurrent selection breeding cycles for populationimprovement which is a key process in increasing yieldthrough hybrid vigour As this approach requires manyyears most programmes are experimenting with molecularbreeding methods as these have the potential to accelerateprecision breeding For all four PBCs investments in basicgenetic and genomic resources including the developmentof mapping populations for QTLs and whole genomesequences are available to support long‐term advancesMore recently association genetics with panels of diversegermplasm are being used as training populations for GSmodels (Table 5) These efforts are benefitting from pub-licly available DNA sequences and whole genome assem-blies in crop databases Key to these accelerated breedingtechnologies are developments in novel phenomics tech-nologies to bridge the genotypephenotype gap In poplarnovel remote sensing field phenotyping is now beingdeployed to assist breeders These advances are being com-bined with in vitro and in planta modern molecular breed-ing techniques such as CRISPR (Table 5) CRISPRtechnology for genome editing has been proven in poplar
FIGURE 2 A schematic developmentpathway for miscanthus in the UnitedKingdom related to the investment in RampDprojects at Aberystwyth (top coloured areasfor projects in the three categories basicresearch breeding and commercialupscaling) leading to a projected croppingarea of 350000 ha by 2030 with clonal andsuccessive ranges of improved seed‐basedhybrids Purple represents theBiotechnology and Biological SciencesResearch Council (BBSRC) and brown theDepartment for Environment Food andRural Affairs (Defra) (UK Nationalfunding) blue bars represent EU fundingand green private sector funding (Terravestaand CERES) and GIANT‐LINK andMiscanthus Upscaling Technology (MUST)are publicndashprivate‐initiatives (PPI)
CLIFTON‐BROWN ET AL | 25
This technology is also being applied in switchgrass andmiscanthus (Table 5) but the future of CRISPR in com-mercial breeding for the European market is uncertain inthe light of recent ECJ 2018 rulings
There is integration of research and plant breeding itselfin all four PBCs Therefore estimating the ongoing costs ofmaintaining these breeding programmes is difficult Invest-ment in research also seeks wider benefits associated withtechnological advances in plant science rather than cultivardevelopment per se However in all cases the conventionalbreeding cycle shown in Figure 1 is the basic ldquoenginerdquo withmolecular technologies (‐omics) serving to accelerate thisengine The history of the development shows that the exis-tence of these breeding programmes is essential to gain ben-efits from the biological research Despite this it is thisessential step that is at most risk from reductions in invest-ment A conventional breeding programme typicallyrequires a breeder and several technicians who are supported
over the long term (20ndash30 years Figure 1) at costs of about05 to 10 million Euro per year (as of 2018) The analysisreported here shows that the time needed to perform onecycle of conventional breeding bringing germplasm fromthe wild to a commercial hybrid ranged from 11 years inswitchgrass to 26 years in poplar (Figure 1) In a maturecrop grown on over 100000 ha with effective cultivar pro-tection and a suitable business model this level of revenuecould come from royalties Until such levels are reachedPBCs lie in the innovation valley of death (Figure 3) andneed public support
Applying industrial ldquotechnology readiness levelsrdquo (TRL)originally developed for aerospace (Heacuteder 2017) to ourplant breeding efforts we estimate many promising hybridscultivars are at TRL levels of 3ndash4 In Table 3 experts ineach crop estimate that it would take 3 years from now toupscale planting material from leading cultivars in plot tri-als to 100 ha
FIGURE 3 A schematic relating some of the steps in the innovation chain from relatively basic crop science research through to thedeployment in commercial cropping systems and value chains The shape of the funnel above the expanding development and deploymentrepresents the availability of investment along the development chain from relatively basic research at the top to the upscaled deployment at thebottom Plant breeding links the research effort with the development of cropping systems The constriction represents the constrained fundingfor breeding that links conventional public research investment and the potential returns from commercial development The handover pointsbetween publicly funded work to develop the germplasm resources (often known as prebreeding) the breeding and the subsequent cropdevelopment are shown on the left The constriction point is aggravated by the lack academic rewards for this essential breeding activity Theoutcome is such that this innovation system is constrained by the precarious resourcing of plant breeding The authorsrsquo assessment ofdevelopment status of the four species is shown (poplar having two one for short‐rotation coppice (SRC) poplar and one for the more traditionalshort‐rotation forestry (SRF)) The four new perennial biomass crops (PBCs) are now in the critical phase of depending of plant breedingprogress without the income stream from a large crop production base
26 | CLIFTON‐BROWN ET AL
Taking the UK example mentioned in the introductionplanting rates of ~35000 ha per year from 2022 onwardsare needed to reach over 1 m ha by 2050 Ongoing workin the UK‐funded Miscanthus Upscaling Technology(MUST) project shows that ramping annual hybrid seedproduction from the current level of sufficient seed for10 ha in 2018 to 35000 ha would take about 5 yearsassuming no setbacks If current hybrids of any of the fourPBCs in the upscaling pipeline fail on any step for exam-ple lower than expected multiplication rates or unforeseenagronomic barriers then further selections from ongoingbreeding are needed to replace earlier candidates
In conclusion the breeding foundations have been laidwell for switchgrass miscanthus willow and poplar owingto public funding over the long time periods necessaryImproved cultivars or genotypes are available that could bescaled up over a few years if real sustained market oppor-tunities emerged in response to sustained favourable poli-cies and industrial market pull The potential contributionsof growing and using these PBCs for socioeconomic andenvironmental benefits are clear but how farmers andothers in commercial value chains are rewarded for mass‐scale deployment as is necessary is not obvious at presentTherefore mass‐scale deployment of these lignocellulosecrops needs developments outside the breeding arenas todrive breeding activities more rapidly and extensively
8 | UNCITED REFERENCE
Huang et al (2019)
ACKNOWLEDGEMENTS
UK The UK‐led miscanthus research and breeding wasmainly supported by the Biotechnology and BiologicalSciences Research Council (BBSRC) Department for Envi-ronment Food and Rural Affairs (Defra) the BBSRC CSPstrategic funding grant BBCSP17301 Innovate UKBBSRC ldquoMUSTrdquo BBN0161491 CERES Inc and Ter-ravesta Ltd through the GIANT‐LINK project (LK0863)Genomic selection and genomewide association study activi-ties were supported by BBSRC grant BBK01711X1 theBBSRC strategic programme grant on Energy Grasses ampBio‐refining BBSEW10963A01 The UK‐led willow RampDwork reported here was supported by BBSRC (BBSEC00005199 BBSEC00005201 BBG0162161 BBE0068331 BBG00580X1 and BBSEC000I0410) Defra(NF0424 NF0426) and the Department of Trade and Indus-try (DTI) (BW6005990000) IT The Brain Gain Program(Rientro dei cervelli) of the Italian Ministry of EducationUniversity and Research supports Antoine Harfouche USContributions by Gerald Tuskan to this manuscript were sup-ported by the Center for Bioenergy Innovation a US
Department of Energy Bioenergy Research Center supportedby the Office of Biological and Environmental Research inthe DOE Office of Science under contract number DE‐AC05‐00OR22725 Willow breeding efforts at CornellUniversity have been supported by grants from the USDepartment of Agriculture National Institute of Food andAgriculture Contributions by the University of Illinois weresupported primarily by the DOE Office of Science Office ofBiological and Environmental Research (BER) grant nosDE‐SC0006634 DE‐SC0012379 and DE‐SC0018420 (Cen-ter for Advanced Bioenergy and Bioproducts Innovation)and the Energy Biosciences Institute EU We would like tofurther acknowledge contributions from the EU projectsldquoOPTIMISCrdquo FP7‐289159 on miscanthus and ldquoWATBIOrdquoFP7‐311929 on poplar and miscanthus as well as ldquoGRACErdquoH2020‐EU326 Biobased Industries Joint Technology Ini-tiative (BBI‐JTI) Project ID 745012 on miscanthus
CONFLICT OF INTEREST
The authors declare that progress reported in this paperwhich includes input from industrial partners is not biasedby their business interests
ORCID
John Clifton‐Brown httporcidorg0000-0001-6477-5452Antoine Harfouche httporcidorg0000-0003-3998-0694Lawrence B Smart httporcidorg0000-0002-7812-7736Danny Awty‐Carroll httporcidorg0000-0001-5855-0775VasileBotnari httpsorcidorg0000-0002-0470-0384Lindsay V Clark httporcidorg0000-0002-3881-9252SalvatoreCosentino httporcidorg0000-0001-7076-8777Lin SHuang httporcidorg0000-0003-3079-326XJon P McCalmont httporcidorg0000-0002-5978-9574Paul Robson httporcidorg0000-0003-1841-3594AnatoliiSandu httporcidorg0000-0002-7900-448XDaniloScordia httporcidorg0000-0002-3822-788XRezaShafiei httporcidorg0000-0003-0891-0730Gail Taylor httporcidorg0000-0001-8470-6390IrisLewandowski httporcidorg0000-0002-0388-4521
REFERENCES
Alburquerque N Baldacci‐Cresp F Baucher M Casacuberta JM Collonnier C ElJaziri M hellip Burgos L (2016) New trans-formation technologies for trees In C Vettori F Gallardo H
CLIFTON‐BROWN ET AL | 27
Haumlggman V Kazana F Migliacci G Pilateamp M Fladung(Eds) Biosafety of forest transgenic trees (pp 31ndash66) DordrechtNetherlands Springer
Argus G W (2007) Salix (Salicaceae) distribution maps and a syn-opsis of their classification in North America north of MexicoHarvard Papers in Botany 12 335ndash368 httpsdoiorg1031001043-4534(2007)12[335SSDMAA]20CO2
Atienza S G Ramirez M C amp Martin A (2003) Mapping‐QTLscontrolling flowering date in Miscanthus sinensis Anderss CerealResearch Communications 31 265ndash271
Atienza S G Satovic Z Petersen K K Dolstra O amp Martin A(2003) Identification of QTLs influencing agronomic traits inMiscanthus sinensis Anderss I Total height flag‐leaf height andstem diameter Theoretical and Applied Genetics 107 123ndash129
Atienza S G Satovic Z Petersen K K Dolstra O amp Martin A(2003) Influencing combustion quality in Miscanthus sinensisAnderss Identification of QTLs for calcium phosphorus and sul-phur content Plant Breeding 122 141ndash145
Bdeir R Muchero W Yordanov Y Tuskan G A Busov V ampGailing O (2017) Quantitative trait locus mapping of Populusbark features and stem diameter BMC Plant Biology 17 224httpsdoiorg101186s12870-017-1166-4
Beard T R Ford G S Koutsky T M amp Spiwak L J (2009) AValley of Death in the innovation sequence An economic investi-gation Research Evaluation 18 343ndash356 httpsdoiorg103152095820209X481057
Berguson W McMahon B amp Riemenschneider D (2017) Addi-tive and non‐additive genetic variances for tree growth in severalhybrid poplar populations and implications regarding breedingstrategy Silvae Genetica 66(1) 33ndash39 httpsdoiorg101515sg-2017-0005
Berlin S Trybush S O Fogelqvist J Gyllenstrand N Halling-baumlck H R Aringhman I hellip Hanley S J (2014) Genetic diversitypopulation structure and phenotypic variation in European Salixviminalis L (Salicaceae) Tree Genetics amp Genomes 10 1595ndash1610 httpsdoiorg101007s11295-014-0782-5
Bernardo R amp Yu J (2007) Prospects for genomewide selectionfor quantitative traits in maize Crop Science 47 1082ndash1090httpsdoiorg102135cropsci2006110690
Biswal A K Atmodjo M A Li M Baxter H L Yoo C GPu Y hellip Mohnen D (2018) Sugar release and growth of bio-fuel crops are improved by downregulation of pectin biosynthesisNature Biotechnology 36 249ndash257
Bmu (2009) National biomass action plan for Germany (p 17) Bun-desministerium fuumlr Umwelt Naturschutz und ReaktorsicherheitBundesministerium fuumlr Ernaumlhrung (BMU) amp Landwirtschaft undVerbraucherschutz (BMELV) 11055 Berlin | Germany Retrievedfrom httpwwwbmeldeSharedDocsDownloadsENPublicationsBiomassActionPlanpdf__blob=publicationFile
Brummer E C (1999) Capturing heterosis in forage crop cultivardevelopment Crop Science 39 943ndash954 httpsdoiorg102135cropsci19990011183X003900040001x
Budsberg E Crawford J T Morgan H Chin W S Bura R ampGustafson R (2016) Hydrocarbon bio‐jet fuel from bioconver-sion of poplar biomass Life cycle assessment Biotechnology forBiofuels 9 170 httpsdoiorg101186s13068-016-0582-2
Burris K P Dlugosz E M Collins A G Stewart C N amp Lena-ghan S C (2016) Development of a rapid low‐cost protoplasttransfection system for switchgrass (Panicum virgatum L) Plant
Cell Reports 35 693ndash704 httpsdoiorg101007s00299-015-1913-7
Casler M (2012) Switchgrass breeding genetics and genomics InA Monti (Ed) Switchgrass (pp 29ndash54) London UK Springer
Casler M Mitchell R amp Vogel K (2012) Switchgrass In CKole C P Joshi amp D R Shonnard (Eds) Handbook of bioen-ergy crop plants Vol 2 New York NY Taylor amp Francis
Casler M D amp Ramstein G P (2018) Breeding for biomass yieldin switchgrass using surrogate measures of yield BioEnergyResearch 11 6ndash12
Casler M D Sosa S Hoffman L Mayton H Ernst C Adler PR hellip Bonos S A (2017) Biomass Yield of Switchgrass Culti-vars under High‐ versus Low‐Input Conditions Crop Science 57821ndash832 httpsdoiorg102135cropsci2016080698
Casler M D amp Vogel K P (2014) Selection for biomass yield inupland lowland and hybrid switchgrass Crop Science 54 626ndash636 httpsdoiorg102135cropsci2013040239
Casler M D Vogel K P amp Harrison M (2015) Switchgrassgermplasm resources Crop Science 55 2463ndash2478 httpsdoiorg102135cropsci2015020076
Casler M D Vogel K P Lee D K Mitchell R B Adler P RSulc R M hellip Moore K J (2018) 30 Years of progress towardincreasing biomass yield of switchgrass and big bluestem CropScience 58 1242
Clark L V Brummer J E Głowacka K Hall M C Heo K PengJ hellip Sacks E J (2014) A footprint of past climate change on thediversity and population structure of Miscanthus sinensis Annals ofBotany 114 97ndash107 httpsdoiorg101093aobmcu084
Clark L V Dzyubenko E Dzyubenko N Bagmet L SabitovA Chebukin P hellip Sacks E J (2016) Ecological characteristicsand in situ genetic associations for yield‐component traits of wildMiscanthus from eastern Russia Annals of Botany 118 941ndash955
Clark L V Jin X Petersen K K Anzoua K G Bagmet LChebukin P hellip Sacks E J(2018) Population structure of Mis-canthus sacchariflorus reveals two major polyploidization eventstetraploid‐mediated unidirectional introgression from diploid Msinensis and diversity centred around the Yellow Sea Annals ofBotany 20 1ndash18 httpsdoiorg101093aobmcy1161
Clark L V Stewart J R Nishiwaki A Toma Y Kjeldsen J BJoslashrgensen U hellip Sacks E J (2015) Genetic structure ofMiscanthus sinensis and Miscanthus sacchariflorus in Japan indi-cates a gradient of bidirectional but asymmetric introgressionJournal of Experimental Botany 66 4213ndash4225
Clifton‐Brown J Hastings A Mos M McCalmont J P Ashman CAwty‐Carroll DhellipCracroft‐EleyW (2017) Progress in upscalingMis-canthus biomass production for the European bio‐economy with seed‐based hybridsGlobal Change Biology Bioenergy 9 6ndash17
Clifton‐Brown J Schwarz K U amp Hastings A (2015) History ofthe development of Miscanthus as a bioenergy crop From smallbeginnings to potential realisation Biology and Environment‐Pro-ceedings of the Royal Irish Academy 115B 45ndash57
Clifton‐Brown J Senior H Purdy S Horsnell R Lankamp BMuumlennekhoff A-K hellip Bentley A R (2018) Investigating thepotential of novel non‐woven fabrics to increase pollination effi-ciency in plant breeding PloS One (Accepted)
Crawford J T Shan CW Budsberg E Morgan H Bura R ampGus-tafson R (2016) Hydrocarbon bio‐jet fuel from bioconversion ofpoplar biomass Techno‐economic assessment Biotechnology forBiofuels 9 141 httpsdoiorg101186s13068-016-0545-7
28 | CLIFTON‐BROWN ET AL
Dalton S (2013) Biotechnology of Miscanthus In S M Jain amp SD Gupta (Eds) Biotechnology of neglected and underutilizedcrops (pp 243ndash294) Dordrecht Netherlands Springer
Davey C L Robson P Hawkins S Farrar K Clifton‐Brown J CDonnison I S amp Slavov G T (2017) Genetic relationshipsbetween spring emergence canopy phenology and biomass yieldincrease the accuracy of genomic prediction in Miscanthus Journalof Experimental Botany 68 5093ndash5102 httpsdoiorg101093jxberx339
David X amp Anderson X (2002) Aspen and larch genetics cooper-ative annual report 13 St Paul MN Department of ForestResources University of Minnesota
Deblock M (1990) Factors influencing the tissue‐culture and theAgrobacterium‐Tumefaciens‐mediated transformation of hybridaspen and poplar clones Plant Physiology 93 1110ndash1116httpsdoiorg101104pp9331110
Defra (2002) The role of future public research investment in thegenetic improvement of UK‐grown crops (p 220) LondonRetrieved from sciencesearchdefragovukDefaultaspxMenu=MenuampModule=MoreampLocation=NoneampCompleted=220ampProjectID=10412
Dewoody J Trewin H amp Taylor G (2015) Genetic and morpho-logical differentiation in Populus nigra L Isolation by coloniza-tion or isolation by adaptation Molecular Ecology 24 2641ndash2655
Dong H Liu S Clark L V Sharma S Gifford J M Juvik JA hellip Sacks E J (2018) Genetic mapping of biomass yield inthree interconnected Miscanthus populations Global Change Biol-ogy Bioenergy 10 165ndash185
Dukes (2017) Digest of UK Energy Statistics (DUKES) (p 264) Lon-don UK Department for Business Energy amp Industrial StrategyRetrieved from wwwgovukgovernmentcollectionsdigest-of-uk-energy-statistics-dukes
Eckenwalder J (1996) Systematics and evolution of Populus In RStettler H Bradshaw J Heilman amp T Hinckley (Eds) Biologyof Populus and its implications for management and conservation(pp 7‐32) Ottawa ON National Research Council of Canada
Elshire R J Glaubitz J C Sun Q Poland J A Kawamoto KBuckler E S amp Mitchell S E (2011) A robust simple geno-typing‐by‐sequencing (GBS) approach for high diversity speciesPloS One 6 e19379
Evans H (2016) Bioenergy crops in the UK Case studies of suc-cessful whole farm integration (p 23) Loughborough UKEnergy Technologies Institute Retrieved from httpswwweticouklibrarybioenergy-crops-in-the-uk-case-studies-on-successful-whole-farm-integration-evidence-pack
Evans H (2017) Increasing UK biomass production through moreproductive use of land (p 7) Loughborough UK Energy Tech-nologies Institute Retrieved from httpswwweticouklibraryan-eti-perspective-increasing-uk-biomass-production-through-more-productive-use-of-land
Evans J Sanciangco M D Lau K H Crisovan E Barry KDaum C hellip Buell C R (2018) Extensive genetic diversity ispresent within North American switchgrass germplasm The PlantGenome 11 1ndash16 httpsdoiorg103835plantgenome2017060055
Evans L M Slavov G T Rodgers‐Melnick E Martin J RanjanP Muchero W hellip DiFazio S P (2014) Population genomicsof Populus trichocarpa identifies signatures of selection and
adaptive trait associations Nature Genetics 46 1089ndash1096httpsdoiorg101038ng3075
Fabio E S Volk T A Miller R O Serapiglia M J Gauch HG Van Rees K C J hellip Smart L B (2017) Genotypetimes envi-ronment interaction analysis of North American shrub willowyield trials confirms superior performance of triploid hybrids Glo-bal Change Biology Bioenergy 9 445ndash459 httpsdoiorg101111gcbb12344
Fahrenkrog A M Neves L G Resende M F R Vazquez A Ide los Campos G Dervinis C hellip Kirst M (2017) Genome‐wide association study reveals putative regulators of bioenergytraits in Populus deltoides New Phytologist 213 799ndash811
Fan D Liu T T Li C F Jiao B Li S Hou Y S amp Luo KM (2015) Efficient CRISPRCas9‐mediated targeted mutagenesisin Populus in the first generation Scientific Reports 5 12217
Feng X P Lourgant K Castric V Saumitou-Laprade P ZhengB S Jiang D M amp Brancourt-Hulmel M (2014) The discov-ery of natural accessions related to Miscanthus x giganteus usingchloroplast DNA Crop Science 54 1645ndash1655
Fillatti J J Sellmer J Mccown B Haissig B amp Comai L(1987) Agrobacterium-mediated transformation and regenerationof Populus Molecular amp General Genetics 206 192ndash199
Fu Y B (2015) Understanding crop genetic diversity under modernplant breeding Theoretical and Applied Genetics 128 2131ndash2142 httpsdoiorg101007s00122-015-2585-y
Fu C Mielenz J R Xiao X Ge Y Hamilton C Y RodriguezM hellip Wang Z‐Y (2011) Genetic manipulation of ligninreduces recalcitrance and improves ethanol production fromswitchgrass Proceedings of the National Academy of Sciences ofthe United States of America 108 3803ndash3808 httpsdoiorg101073pnas1100310108
Gao C (2018) The future of CRISPR technologies in agricultureNature Reviews Molecular Cell Biology 19(5) 275ndash276
Gegas V Gay A Camargo A amp Doonan J (2014) Challengesof Crop Phenomics in the Post-genomic Era In J Hancock (Ed)Phenomics (pp 142ndash171) Boca Raton FL CRC Press
Geraldes A DiFazio S P Slavov G T Ranjan P Muchero WHannemann J hellip Tuskan G A (2013) A 34K SNP genotypingarray for Populus trichocarpa Design application to the study ofnatural populations and transferability to other Populus speciesMolecular Ecology Resources 13 306ndash323
Gifford J M Chae W B Swaminathan K Moose S P amp JuvikJ A (2015) Mapping the genome of Miscanthus sinensis forQTL associated with biomass productivity Global Change Biol-ogy Bioenergy 7 797ndash810
Goggin F L Lorence A amp Topp C N (2015) Applying high‐throughput phenotyping to plant‐insect interactions Picturingmore resistant crops Current Opinion in Insect Science 9 69ndash76httpsdoiorg101016jcois201503002
Grabowski P P Evans J Daum C Deshpande S Barry K WKennedy M hellip Casler M D (2017) Genome‐wide associationswith flowering time in switchgrass using exome‐capture sequenc-ing data New Phytologist 213 154ndash169 httpsdoiorg101111nph14101
Greef J M amp Deuter M (1993) Syntaxonomy of Miscanthus xgiganteus GREEF et DEU Angewandte Botanik 67 87ndash90
Guerra F P Richards J H Fiehn O Famula R Stanton B JShuren R hellip Neale D B (2016) Analysis of the genetic varia-tion in growth ecophysiology and chemical and metabolomic
CLIFTON‐BROWN ET AL | 29
composition of wood of Populus trichocarpa provenances TreeGenetics amp Genomes 12 6 httpsdoiorg101007s11295-015-0965-8
Guo H P Shao R X Hong C T Hu H K Zheng B S ampZhang Q X (2013) Rapid in vitro propagation of bioenergy cropMiscanthus sacchariflorus Applied Mechanics and Materials 260181ndash186
Hallingbaumlck H R Fogelqvist J Powers S J Turrion‐Gomez JRossiter R Amey J hellip Roumlnnberg‐Waumlstljung A‐C (2016)Association mapping in Salix viminalis L (Salicaceae)ndashIdentifica-tion of candidate genes associated with growth and phenologyGlobal Change Biology Bioenergy 8 670ndash685
Hanley S J amp Karp A (2014) Genetic strategies for dissectingcomplex traits in biomass willows (Salix spp) Tree Physiology34 1167ndash1180 httpsdoiorg101093treephystpt089
Harfouche A Meilan R amp Altman A (2011) Tree genetic engi-neering and applications to sustainable forestry and biomass pro-duction Trends in Biotechnology 29 9ndash17 httpsdoiorg101016jtibtech201009003
Harfouche A Meilan R amp Altman A (2014) Molecular and phys-iological responses to abiotic stress in forest trees and their rele-vance to tree improvement Tree Physiology 34 1181ndash1198httpsdoiorg101093treephystpu012
Harfouche A Meilan R Kirst M Morgante M Boerjan WSabatti M amp Mugnozza G S (2012) Accelerating the domesti-cation of forest trees in a changing world Trends in PlantScience 17 64ndash72 httpsdoiorg101016jtplants201111005
Hastings A Clifton‐Brown J Wattenbach M Mitchell C PStampfl P amp Smith P (2009) Future energy potential of Mis-canthus in Europe Global Change Biology Bioenergy 1 180ndash196
Hastings A Mos M Yesufu J AMccalmont J Schwarz KShafei RhellipClifton-Brown J (2017) Economic and environmen-tal assessment of seed and rhizome propagated Miscanthus in theUK Frontiers in Plant Science 8 16 httpsdoiorg103389fpls201701058
Heacuteder M (2017) From NASA to EU The evolution of the TRLscale in Public Sector Innovation The Innovation Journal 22 1ndash23 httpsdoiorg103389fpls201701058
Heffner E L Sorrells M E amp Jannink J L (2009) Genomicselection for crop improvement Crop Science 49 1ndash12 httpsdoiorg102135cropsci2008080512
Hodkinson T R Klaas M Jones M B Prickett R amp Barth S(2015) Miscanthus A case study for the utilization of naturalgenetic variation Plant Genetic Resources‐Characterization andUtilization 13 219ndash237 httpsdoiorg101017S147926211400094X
Hodkinson T R amp Renvoize S (2001) Nomenclature of Miscant-hus xgiganteus (Poaceae) Kew Bulletin 56 759ndash760 httpsdoiorg1023074117709
Houle D Govindaraju D R amp Omholt S (2010) Phenomics Thenext challenge Nature Reviews Genetics 11 855ndash866 httpsdoiorg101038nrg2897
Huang X H Wei X H Sang T Zhao Q Feng Q Zhao Yhellip Han B (2010) Genome‐wide association studies of 14 agro-nomic traits in rice landraces Nature Genetics 42 961ndashU976
Hwang O‐J Cho M‐A Han Y‐J Kim Y‐M Lim S‐H KimD‐S hellip Kim J‐I (2014) Agrobacterium‐mediated genetic trans-formation of Miscanthus sinensis Plant Cell Tissue and OrganCulture (PCTOC) 117 51ndash63
Hwang O‐J Lim S‐H Han Y‐J Shin A‐Y Kim D‐S ampKim J‐I (2014) Phenotypic characterization of transgenic Mis-canthus sinensis plants overexpressing Arabidopsis phytochromeB International Journal of Photoenergy 2014501016
Ilstedt B (1996) Genetics and performance of Belgian poplar clonestested in Sweden International Journal of Forest Genetics 3183ndash195
Ingvarsson P K Garcia M V Hall D Luquez V amp Jansson S(2006) Clinal variation in phyB2 a candidate gene for day‐length‐induced growth cessation and bud set across a latitudinalgradient in European aspen (Populus tremula) Genetics 1721845ndash1853
Ingvarsson P K Garcia M V Luquez V Hall D amp Jansson S(2008) Nucleotide polymorphism and phenotypic associationswithin and around the phytochrome B2 locus in European aspen(Populus tremula Salicaceae) Genetics 178 2217ndash2226 httpsdoiorg101534genetics107082354
Jannink J‐L Lorenz A J amp Iwata H (2010) Genomic selectionin plant breeding From theory to practice Briefings in FunctionalGenomics 9 166ndash177 httpsdoiorg101093bfgpelq001
Jiang J Guan Y Mccormick S Juvik J Lubberstedt T amp FeiS‐Z (2017) Gametophytic self‐incompatibility is operative inMiscanthus sinensis (Poaceae) and is affected by pistil age CropScience 57 1948ndash1956
Jones H D (2015a) Future of breeding by genome editing is in thehands of regulators GM Crops amp Food 6 223ndash232
Jones H D (2015b) Regulatory uncertainty over genome editingNature Plants 1 14011
Kalinina O Nunn C Sanderson R Hastings A F S van der Weijde TOumlzguumlven MhellipClifton‐Brown J C (2017) ExtendingMiscanthus cul-tivation with novel germplasm at six contrasting sites Frontiers in PlantScience 8563 httpsdoiorg103389fpls201700563
Karp A Hanley S J Trybush S O Macalpine W Pei M ampShield I (2011) Genetic improvement of willow for bioenergyand biofuels free access Journal of Integrative Plant Biology 53151ndash165 httpsdoiorg101111j1744-7909201001015x
Kersten B Faivre Rampant P Mader M Le Paslier M‐C Bou-non R Berard A hellip Fladung M (2016) Genome sequences ofPopulus tremula chloroplast and mitochondrion Implications forholistic poplar breeding PloS One 11 e0147209 httpsdoiorg101371journalpone0147209
Kiesel A Nunn C Iqbal Y Van der Weijde T Wagner MOumlzguumlven M hellip Lewandowski I (2017) Site‐specific manage-ment of Miscanthus genotypes for combustion and anaerobicdigestion A comparison of energy yields Frontiers in PlantScience 8 347 httpsdoiorg103389fpls201700347
Kim H Kim S T Ryu J Kang B C Kim J S amp Kim S G(2017) CRISPRCpf1‐mediated DNA‐free plant genome editingNature Communications 8 14406 httpsdoiorg101038ncomms14406
King Z R Bray A L Lafayette P R amp Parrott W A (2014)Biolistic transformation of elite genotypes of switchgrass (Pan-icum virgatum L) Plant Cell Reports 33 313ndash322 httpsdoiorg101007s00299-013-1531-1
Kopp R F Maynard C A De Niella P R Smart L B amp Abra-hamson L P (2002) Collection and storage of pollen from Salix(Salicaceae) American Journal of Botany 89 248ndash252
Krzyżak J Pogrzeba M Rusinowski S Clifton‐Brown J McCal-mont J P Kiesel A hellip Mos M (2017) Heavy metal uptake
30 | CLIFTON‐BROWN ET AL
by novel miscanthus seed‐based hybrids cultivated in heavy metalcontaminated soil Civil and Environmental Engineering Reports26 121ndash132 httpsdoiorg101515ceer-2017-0040
Kuzovkina Y A amp Quigley M F (2005) Willows beyond wet-lands Uses of Salix L species for environmental projects WaterAir and Soil Pollution 162 183ndash204 httpsdoiorg101007s11270-005-6272-5
Lazarus W Headlee W L amp Zalesny R S (2015) Impacts of sup-plyshed‐level differences in productivity and land costs on the eco-nomics of hybrid poplar production in Minnesota USA BioEnergyResearch 8 231ndash248 httpsdoiorg101007s12155-014-9520-y
Lewandowski I (2015) Securing a sustainable biomass supply in agrowing bioeconomy Global Food Security 6 34ndash42 httpsdoiorg101016jgfs201510001
Lewandowski I (2016) The role of perennial biomass crops in agrowing bioeconomy In S Barth D Murphy‐Bokern O Kalin-ina G Taylor amp M Jones (Eds) Perennial biomass crops for aresource‐constrained world (pp 3ndash13) Cham SwitzerlandSpringer Switzerland AG
Li J L Meilan R Ma C Barish M amp Strauss S H (2008)Stability of herbicide resistance over 8 years of coppice in field‐grown genetically engineered poplars Western Journal of AppliedForestry 23 89ndash93
Li R Y amp Qu R D (2011) High throughput Agrobacterium‐medi-ated switchgrass transformation Biomass amp Bioenergy 35 1046ndash1054 httpsdoiorg101016jbiombioe201011025
Lindegaard K N amp Barker J H A (1997) Breeding willows forbiomass Aspects of Applied Biology 49 155ndash162
Linde‐Laursen I B (1993) Cytogenetic analysis of MiscanthusGiganteus an interspecific hybrid Hereditas 119 297ndash300httpsdoiorg101111j1601-5223199300297x
Liu Y Merrick P Zhang Z Z Ji C H Yang B amp Fei S Z(2018) Targeted mutagenesis in tetraploid switchgrass (Panicumvirgatum L) using CRISPRCas9 Plant Biotechnology Journal16 381ndash393
Liu L L Wu Y Q Wang Y W amp Samuels T (2012) A high‐density simple sequence repeat‐based genetic linkage map ofswitchgrass G3 (Bethesda Md) 2 357ndash370
Lovett A Suumlnnenberg G amp Dockerty T (2014) The availabilityof land for perennial energy crops in Great Britain GlobalChange Biology Bioenergy 6 99ndash107 httpsdoiorg101111gcbb12147
Lozano‐Juste J amp Cutler S R (2014) Plant genome engineering infull bloom Trends in Plant Science 19 284ndash287 httpsdoiorg101016jtplants201402014
Lu F Lipka A E Glaubitz J Elshire R Cherney J H CaslerM D hellip Costich D E (2013) Switchgrass Genomic DiversityPloidy and Evolution Novel Insights from a Network‐Based SNPDiscovery Protocol Plos Genetics 9 e1003215
Ludovisi R Tauro F Salvati R Khoury S Mugnozza G S ampHarfouche A (2017) UAV‐based thermal imaging for high‐throughput field phenotyping of black poplar response to droughtFrontiers in Plant Science 81681
Macalpine W Shield I amp Karp A (2010) Seed to near market vari-ety the BEGIN willow breeding pipeline 2003ndash2010 and beyond InBioten conference proceedings Birmingham pp 21ndash23
Macalpine W J Shield I F Trybush S O Hayes C M ampKarp A (2008) Overcoming barriers to crossing in willow (Salixspp) breeding Aspects of Applied Biology 90 173ndash180
Mahfouz M M (2017) The efficient tool CRISPR‐Cpf1 NaturePlants 3 17028
Malinowska M Donnison I S amp Robson P R (2017) Phenomicsanalysis of drought responses in Miscanthus collected from differ-ent geographical locations Global Change Biology Bioenergy 978ndash91
Maroder H L Prego I A Facciuto G R amp Maldonado S B(2000) Storage behaviour of Salix alba and Salix matsudanaseeds Annals of Botany 86 1017ndash1021 httpsdoiorg101006anbo20001265
Martinez‐Reyna J amp Vogel K (2002) Incompatibility systems inswitchgrass Crop Science 42 1800ndash1805 httpsdoiorg102135cropsci20021800
Maurya J P amp Bhalerao R P (2017) Photoperiod‐and tempera-ture‐mediated control of growth cessation and dormancy in treesA molecular perspective Annals of Botany 120 351ndash360httpsdoiorg101093aobmcx061
McCalmont J P Hastings A Mcnamara N P Richter G MRobson P Donnison I S amp Clifton‐Brown J (2017) Environ-mental costs and benefits of growing Miscanthus for bioenergy inthe UK Global Change Biology Bioenergy 9 489ndash507
McCracken A R amp Dawson W M (1997) Using mixtures of wil-low clones as a means of controlling rust disease Aspects ofApplied Biology 49 97ndash103
McKown A D Klaacutepště J Guy R D Geraldes A Porth I Hanne-mann JhellipDouglas C J (2014) Genome‐wide association implicatesnumerous genes underlying ecological trait variation in natural popula-tions ofPopulus trichocarpaNew Phytologist 203 535ndash553
Merrick P amp Fei S Z (2015) Plant regeneration and genetic transforma-tion in switchgrass ndash A review Journal of Integrative Agriculture 14483ndash493 httpsdoiorg101016S2095-3119(14)60921-7
Moreno‐Mateos M A Fernandez J P Rouet R Lane M AVejnar C E Mis E hellip Giraldez A J (2017) CRISPR‐Cpf1mediates efficient homology‐directed repair and temperature‐con-trolled genome editing Nature Communications 8 2024 httpsdoiorg101038s41467-017-01836-2
Mosseler A (1990) Hybrid performance and species crossabilityrelationships in willows (Salix) Canadian Journal of Botany 682329ndash2338
Muchero W Guo J DiFazio S P Chen J‐G Ranjan P Slavov GT hellip Tuskan G A (2015) High‐resolution genetic mapping ofallelic variants associated with cell wall chemistry in Populus BMCGenomics 16 24 httpsdoiorg101186s12864-015-1215-z
Neale D B amp Kremer A (2011) Forest tree genomics Growingresources and applications Nature Reviews Genetics 12 111ndash122 httpsdoiorg101038nrg2931
Nunn C Hastings A F S J Kalinina O Oumlzguumlven M SchuumlleH Tarakanov I G hellip Clifton‐Brown J C (2017) Environmen-tal influences on the growing season duration and ripening ofdiverse Miscanthus germplasm grown in six countries Frontiersin Plant Science 8 907 httpsdoiorg103389fpls201700907
Ogawa Y Honda M Kondo Y amp Hara‐Nishimura I (2016) Anefficient Agrobacterium‐mediated transformation method forswitchgrass genotypes using Type I callus Plant Biotechnology33 19ndash26
Ogawa Y Shirakawa M Koumoto Y Honda M Asami Y KondoY ampHara‐Nishimura I (2014) A simple and reliable multi‐gene trans-formation method for switchgrass Plant Cell Reports 33 1161ndash1172httpsdoiorg101007s00299-014-1605-8
CLIFTON‐BROWN ET AL | 31
Okada M Lanzatella C Saha M C Bouton J Wu R L amp Tobias CM (2010) Complete switchgrass genetic maps reveal subgenomecollinearity preferential pairing and multilocus interactions Genetics185 745ndash760 httpsdoiorg101534genetics110113910
Palomo‐Riacuteos E Macalpine W Shield I Amey J Karaoğlu C WestJ hellip Jones H D (2015) Efficient method for rapid multiplication ofclean and healthy willow clones via in vitro propagation with broadgenotype applicability Canadian Journal of Forest Research 451662ndash1667 httpsdoiorg101139cjfr-2015-0055
Pilate G Allona I Boerjan W Deacutejardin A Fladung M Gal-lardo F hellip Halpin C (2016) Lessons from 25 years of GM treefield trials in Europe and prospects for the future In C Vettori FGallardo H Haumlggman V Kazana F Migliacci G Pilateamp MFladung (Eds) Biosafety of forest transgenic trees (pp 67ndash100)Dordrecht Netherlands Springer Switzerland AG
Pinosio S Giacomello S Faivre-Rampant P Taylor G Jorge VLe Paslier M C amp Marroni F (2016) Characterization of thepoplar pan-genome by genome-wide identification of structuralvariation Molecular Biology and Evolution 33 2706ndash2719
Porth I Klaacutepště J Skyba O Friedmann M C Hannemann JEhlting J hellip Douglas C J (2013) Network analysis reveals therelationship among wood properties gene expression levels andgenotypes of natural Populus trichocarpa accessions New Phytol-ogist 200 727ndash742
Pugesgaard S Schelde K Larsen S U Laeligrke P E amp JoslashrgensenU (2015) Comparing annual and perennial crops for bioenergyproductionndashinfluence on nitrate leaching and energy balance Glo-bal Change Biology Bioenergy 7 1136ndash1149 httpsdoiorg101111gcbb12215
Quinn L D Allen D J amp Stewart J R (2010) Invasivenesspotential of Miscanthus sinensis Implications for bioenergy pro-duction in the United States Global Change Biology Bioenergy2 310ndash320 httpsdoiorg101111j1757-1707201001062x
Rae A M Robinson K M Street N R amp Taylor G (2004)Morphological and physiological traits influencing biomass pro-ductivity in short‐rotation coppice poplar Canadian Journal ofForest Research‐Revue Canadienne De Recherche Forestiere 341488ndash1498 httpsdoiorg101139x04-033
Rambaud C Arnoult S Bluteau A Mansard M C Blassiau Camp Brancourt‐Hulmel M (2013) Shoot organogenesis in threeMiscanthus species and evaluation for genetic uniformity usingAFLP analysis Plant Cell Tissue and Organ Culture (PCTOC)113 437ndash448 httpsdoiorg101007s11240-012-0284-9
Ramstein G P Evans J Kaeppler S M Mitchell R B Vogel K PBuell C R amp Casler M D (2016) Accuracy of genomic prediction inswitchgrass (Panicum virgatum L) improved by accounting for linkagedisequilibriumG3 (Bethesda Md) 6 1049ndash1062
Rayburn A L Crawford J Rayburn C M amp Juvik J A (2009)Genome size of three Miscanthus species Plant Molecular Biol-ogy Reporter 27 184 httpsdoiorg101007s11105-008-0070-3
Richardson J Isebrands J amp Ball J (2014) Ecology and physiol-ogy of poplars and willows In J Isebrands amp J Richardson(Eds) Poplars and willows Trees for society and the environ-ment (pp 92‐123) Boston MA and Rome CABI and FAO
Robison T L Rousseau R J amp Zhang J (2006) Biomass produc-tivity improvement for eastern cottonwood Biomass amp Bioenergy30 735ndash739 httpsdoiorg101016jbiombioe200601012
Robson P R Farrar K Gay A P Jensen E F Clifton‐Brown JC amp Donnison I S (2013) Variation in canopy duration in the
perennial biofuel crop Miscanthus reveals complex associationswith yield Journal of Experimental Botany 64 2373ndash2383
Robson P Jensen E Hawkins S White S R Kenobi K Clif-ton‐Brown J hellip Farrar K (2013) Accelerating the domestica-tion of a bioenergy crop Identifying and modelling morphologicaltargets for sustainable yield increase in Miscanthus Journal ofExperimental Botany 64 4143ndash4155
Sanderson M A Adler P R Boateng A A Casler M D ampSarath G (2006) Switchgrass as a biofuels feedstock in theUSA Canadian Journal of Plant Science 86 1315ndash1325httpsdoiorg104141P06-136
Scarlat N Dallemand J F Monforti‐Ferrario F amp Nita V(2015) The role of biomass and bioenergy in a future bioecon-omy Policies and facts Environmental Development 15 3ndash34httpsdoiorg101016jenvdev201503006
Serapiglia M J Gouker F E amp Smart L B (2014) Early selec-tion of novel triploid hybrids of shrub willow with improved bio-mass yield relative to diploids BMC Plant Biology 14 74httpsdoiorg1011861471-2229-14-74
Serba D Wu L Daverdin G Bahri B A Wang X Kilian Ahellip Devos K M (2013) Linkage maps of lowland and upland tet-raploid switchgrass ecotypes BioEnergy Research 6 953ndash965httpsdoiorg101007s12155-013-9315-6
Shakoor N Lee S amp Mockler T C (2017) High throughput phe-notyping to accelerate crop breeding and monitoring of diseases inthe field Current Opinion in Plant Biology 38 184ndash192 httpsdoiorg101016jpbi201705006
Shield I Macalpine W Hanley S amp Karp A (2015) Breedingwillow for short rotation coppice energy cropping In Industrialcrops Breeding for BioEnergy and Bioproducts (pp 67ndash80) NewYork NY Springer
Sjodin A Street N R Sandberg G Gustafsson P amp Jansson S (2009)The Populus Genome Integrative Explorer (PopGenIE) A new resourcefor exploring the Populus genomeNewPhytologist 182 1013ndash1025
Slavov G amp Davey C (2017) Integrated genomic prediction inbioenergy crops In Abstracts from the International Conferenceon Developing biomass crops for future climates pp 34 24ndash27September 2017 Oriel College Oxford wwwwatbioeu
Slavov G Davey C Bosch M Robson P Donnison I ampMackay I (2018a) Genomic index selection provides a pragmaticframework for setting and refining multi‐objective breeding targetsin Miscanthus Annals of Botany (in press)
Slavov G T DiFazio S P Martin J Schackwitz W MucheroW Rodgers‐Melnick E hellip Tuskan G A (2012) Genome rese-quencing reveals multiscale geographic structure and extensivelinkage disequilibrium in the forest tree Populus trichocarpa NewPhytologist 196 713ndash725
Slavov G Davey C Robson P Donnison I amp Mackay I(2018b) Domestication of Miscanthus through genomic indexselection In Plant and Animal Genome Conference 13 January2018 San Diego CA Retrieved from httpspagconfexcompagxxvimeetingappcgiPaper30622
Slavov G T Nipper R Robson P Farrar K Allison G GBosch M hellip Jensen E (2013) Genome‐wide association studiesand prediction of 17 traits related to phenology biomass and cellwall composition in the energy grass Miscanthus sinensis NewPhytologist 201 1227ndash1239
Slavov G Robson P Jensen E Hodgson E Farrar K AllisonG hellip Donnison I (2014) Contrasting geographic patterns of
32 | CLIFTON‐BROWN ET AL
genetic variation for molecular markers vs phenotypic traits in theenergy grass Miscanthus sinensis Global Change Biology Bioen-ergy 5 562ndash571
Ślusarkiewicz‐Jarzina A Ponitka A Cerazy‐Waliszewska J Woj-ciechowicz M K Sobańska K Jeżowski S amp Pniewski T(2017) Effective and simple in vitro regeneration system of Mis-canthus sinensis Mtimes giganteus and M sacchariflorus for plant-ing and biotechnology purposes Biomass and Bioenergy 107219ndash226 httpsdoiorg101016jbiombioe201710012
Smart L amp Cameron K (2012) Shrub willow In C Kole S Joshiamp D Shonnard (Eds) Handbook of bioenergy crop plants (pp687ndash708) Boca Raton FL Taylor and Francis Group
Stanton B J (2014) The domestication and conservation of Populusand Salix genetic resources (Chapter 4) In Isebrands J ampRichardson J (Eds) Poplars and willows Trees for society andthe environment (pp 124ndash200) Rome Italy CAB InternationalFood and Agricultural Organization of the United Nations
Stanton B J Neale D B amp Li S (2010) Populus breeding Fromthe classical to the genomic approach In S Jansson R Bha-lerao amp A T Groover (Eds) Genetics and genomics of popu-lus (pp 309ndash348) New York NY Springer
Stewart J R Toma Y Fernandez F G Nishiwaki A Yamada T ampBollero G (2009) The ecology and agronomy ofMiscanthus sinensisa species important to bioenergy crop development in its native range inJapan A reviewGlobal Change Biology Bioenergy 1 126ndash153
Stott K G (1992) Willows in the service of man Proceedings of the RoyalSociety of Edinburgh Section B Biological Sciences 98 169ndash182
Strauss S H Ma C Ault K amp Klocko A L (2016) Lessonsfrom two decades of field trials with genetically modified trees inthe USA Biology and regulatory compliance In C Vettori FGallardo H Haumlggman V Kazana F Migliacci G Pilate amp MFladung (Eds) Biosafety of forest transgenic trees (pp 101ndash124)Dordrecht Netherlands Springer
Suda Y amp Argus G W (1968) Chromosome numbers of some NorthAmerican Salix Brittonia 20 191ndash197 httpsdoiorg1023072805440
Tan B (2018) Genomic selection and genome‐wide association studies todissect quantitative traits in forest trees Umearing Sweden University
Tornqvist C Taylor M Jiang Y Evans J Buell C KaepplerS amp Casler M (2018) Quantitative trait locus mapping for flow-ering time in a lowland x upland switchgrass pseudo‐F2 popula-tion The Plant Genome 11 170093
Toacuteth G Hermann T Da Silva M amp Montanarella L (2016)Heavy metals in agricultural soils of the European Union withimplications for food safety Environment International 88 299ndash309 httpsdoiorg101016jenvint201512017
Tracy W Dawson J Moore V amp Fisch J (2016) Intellectual propertyrights and public plant breeding Recommendations and proceedings ofa conference on best practices for intellectual property protection of pub-lically developed plant germplasm (p 70) University of Wisconsin‐Madison Retrieved from httpshostcalswisceduagronomywp-contentuploadssites1620162005Proceedings-IPR-Finalpdf
Trybush S Jahodovaacute Š Macalpine W amp Karp A (2008) A geneticstudy of a Salix germplasm resource reveals new insights into rela-tionships among subgenera sections and species BioEnergyResearch 1 67ndash79 httpsdoiorg101007s12155-008-9007-9
Tsai C J (2013) Next‐generation sequencing for next‐generationbreeding and more New Phytologist 198 635ndash637 httpsdoiorg101111nph12245
Tuskan G A Difazio S Jansson S Bohlmann J Grigoriev I HellstenU hellip Rokhsar D (2006) The genome of black cottonwood Populustrichocarpa (Torr ampGray) Science 313 1596ndash1604
Tuskan G A amp Walsh M E (2001) Short‐rotation woody cropsystems atmospheric carbon dioxide and carbon management AUS case study The Forestry Chronicle 77 259ndash264 httpsdoiorg105558tfc77259-2
Van Den Broek R Treffers D‐J Meeusen M Van Wijk ANieuwlaar E amp Turkenburg W (2001) Green energy or organicfood A life‐cycle assessment comparing two uses of set‐asideland Journal of Industrial Ecology 5 65ndash87 httpsdoiorg101162108819801760049477
Van Der Schoot J Pospiskova M Vosman B amp Smulders M J M(2000) Development and characterization of microsatellite markers inblack poplar (Populus nigra L) Theoretical and Applied Genetics 101317ndash322 httpsdoiorg101007s001220051485
Van Der Weijde T Dolstra O Visser R G amp Trindade L M(2017a) Stability of cell wall composition and saccharificationefficiency in Miscanthus across diverse environments Frontiers inPlant Science 7 2004
Van der Weijde T Huxley L M Hawkins S Sembiring E HFarrar K Dolstra O hellip Trindade L M (2017) Impact ofdrought stress on growth and quality of miscanthus for biofuelproduction Global Change Biology Bioenergy 9 770ndash782
Van der Weijde T Kamei C L A Severing E I Torres A FGomez L D Dolstra O hellip Trindade L M (2017) Geneticcomplexity of miscanthus cell wall composition and biomass qual-ity for biofuels BMC Genomics 18 406
Van der Weijde T Kiesel A Iqbal Y Muylle H Dolstra O Visser RG FhellipTrindade LM (2017) Evaluation ofMiscanthus sinensis bio-mass quality as feedstock for conversion into different bioenergy prod-uctsGlobal Change Biology Bioenergy 9 176ndash190
Vanholme B Cesarino I Goeminne G Kim H Marroni F VanAcker R hellip Boerjan W (2013) Breeding with rare defectivealleles (BRDA) A natural Populus nigra HCT mutant with modi-fied lignin as a case study New Phytologist 198 765ndash776
Vogel K P Mitchell R B Casler M D amp Sarath G (2014)Registration of Liberty Switchgrass Journal of Plant Registra-tions 8 242ndash247 httpsdoiorg103198jpr2013120076crc
Wang X Yamada T Kong F‐J Abe Y Hoshino Y Sato Hhellip Yamada T (2011) Establishment of an efficient in vitro cul-ture and particle bombardment‐mediated transformation systems inMiscanthus sinensis Anderss a potential bioenergy crop GlobalChange Biology Bioenergy 3 322ndash332 httpsdoiorg101111j1757-1707201101090x
Watrud L S Lee E H Fairbrother A Burdick C Reichman JR Bollman M hellip Van de Water P K (2004) Evidence forlandscape‐level pollen‐mediated gene flow from genetically modi-fied creeping bentgrass with CP4 EPSPS as a marker Proceedingsof the National Academy of Sciences of the United States of Amer-ica 101 14533ndash14538
Weisgerber H (1993) Poplar breeding for the purpose of biomassproduction in short‐rotation periods in Germany ndash Problems andfirst findings Forestry Chronicle 69 727ndash729 httpsdoiorg105558tfc69727-6
Wheeler N C Steiner K C Schlarbaum S E amp Neale D B(2015) The evolution of forest genetics and tree improvementresearch in the United States Journal of Forestry 113 500ndash510httpsdoiorg105849jof14-120
CLIFTON‐BROWN ET AL | 33
Whittaker C Macalpine W J Yates N E amp Shield I (2016)Dry matter losses and methane emissions during wood chip stor-age The impact on full life cycle greenhouse gas savings of shortrotation coppice willow for heat BioEnergy Research 9 820ndash835 httpsdoiorg101007s12155-016-9728-0
Xi Q (2000) Investigation on the distribution and potential of giant grassesin China PhD thesis Goettingen Germany Cuvillier Verlag
Xi Q amp Jezowkski S (2004) Plant resources of Triarrhena andMiscanthus species in China and its meaning for Europe PlantBreeding and Seed Science 49 63ndash77
Xue S Kalinina O amp Lewandowski I (2015) Present and future optionsfor the improvement of Miscanthus propagation techniques Renewableand Sustainable Energy Reviews 49 1233ndash1246
Yin K Q Gao C X amp Qiu J L (2017) Progress and prospectsin plant genome editing Nature Plants 3 httpsdoiorg101038nplants2017107
YookM (2016)Genetic diversity and transcriptome analysis for salt toler-ance inMiscanthus PhD thesis Seoul National University Korea
Zaidi S S E A Mahfouz M M amp Mansoor S (2017) CRISPR‐Cpf1 A new tool for plant genome editing Trends in PlantScience 22 550ndash553 httpsdoiorg101016jtplants201705001
Zalesny R S Stanturf J A Gardiner E S Perdue J H YoungT M Coyle D R hellip Hass A (2016) Ecosystem services ofwoody crop production systems BioEnergy Research 9 465ndash491 httpsdoiorg101007s12155-016-9737-z
Zhang Q X Sun Y Hu H K Chen B Hong C T Guo H Phellip Zheng B S (2012) Micropropagation and plant regenerationfrom embryogenic callus of Miscanthus sinensis Vitro Cellular ampDevelopmental Biology‐Plant 48 50ndash57 httpsdoiorg101007s11627-011-9387-y
Zhang Y Zalapa J E Jakubowski A R Price D L AcharyaA Wei Y hellip Casler M D (2011) Post‐glacial evolution of
Panicum virgatum Centers of diversity and gene pools revealedby SSR markers and cpDNA sequences Genetica 139 933ndash948httpsdoiorg101007s10709-011-9597-6
Zhao H Wang B He J Yang J Pan L Sun D amp Peng J(2013) Genetic diversity and population structure of Miscanthussinensis germplasm in China PloS One 8 e75672
Zhou X H Jacobs T B Xue L J Harding S A amp Tsai C J(2015) Exploiting SNPs for biallelic CRISPR mutations in theoutcrossing woody perennial Populus reveals 4‐coumarate CoAligase specificity and redundancy New Phytologist 208 298ndash301
Zhou R Macaya‐Sanz D Rodgers‐Melnick E Carlson C HGouker F E Evans L M hellip DiFazio S P (2018) Characteri-zation of a large sex determination region in Salix purpurea L(Salicaceae) Molecular Genetics and Genomics 1ndash16 httpsdoiorg101007s00438-018-1473-y
Zsuffa L Mosseler A amp Raj Y (1984) Prospects for interspecifichybridization in willow for biomass production In K L Perttu(Ed) Ecology and management of forest biomass production sys-tems Report 15 (pp 261ndash281) Uppsala Sweden SwedishUniversity of Agricultural Sciences
How to cite this article Clifton‐Brown J HarfoucheA Casler MD et al Breeding progress andpreparedness for mass‐scale deployment of perenniallignocellulosic biomass crops switchgrassmiscanthus willow and poplar GCB Bioenergy201800000ndash000 httpsdoiorg101111gcbb12566
34 | CLIFTON‐BROWN ET AL
Graphical AbstractThe contents of this page will be used as part of the graphical abstract of html only
It will not be published as part of main article
Plant breeding links the research effort with commercial mass upscaling The authorsrsquo assessment of development status ofthe four species is shown (poplar having two one for short rotation coppice (SRC) poplar and one for the more traditionalshort rotation forestry (SRF)) Mass scale deployment needs developments outside the breeding arenas to drive breedingactivities more rapidly and extensively
We conclude that sustained public investment in breeding plays a key role in
delivering future mass‐scale deployment of PBCs
KEYWORD S
bioenergy feedstocks lignocellulose M sacchariflorus M sinensis Miscanthus Panicum virgatum
perennial biomass crop Populus spp Salix spp
1 | INTRODUCTION
Increasing sustainable biomass production is an importantcomponent of the transition from a fossil fuel‐based econ-omy to renewables Taking the United Kingdom as an exam-ple Lovett Suumlnnenberg and Dockerty (2014) suggested that14 million ha of marginal agricultural land could be used forbiomass production without compromising food productionAssuming a biomass dry matter (DM) yield of 10 Mgha anda calorific value of 18 GJMg DM 14 million ha woulddeliver around 28 TWh of electricity (with 40 biomassconversion efficiency) which would be ~8 of primary UKelectricity generation (336 TWh in 2017 (DUKES 2017))To achieve this by 2050 planting rates of ~35000 hayearwould be needed from 2022 in line with calculations byEvans (2017) The current annual planting rates in the UnitedKingdom are orders of magnitude short of these levels atonly several hundred hectares per year Similar scenarioshave been generated for other countries (BMU 2009 Scar-lat Dallemand Monforti‐Ferrario amp Nita 2015)
If perennial biomass crops (PBCs) are to make a real con-tribution to sustainable development they should be grownon agricultural land which is less suitable for food crops(Lewandowski 2015) This economically ldquomarginalrdquo land istypically characterized by abiotic stresses (drought floodingstoniness steep slope exposure to wind and sub‐optimalaspect) low nutrients andor contaminated soils (Toacuteth et al2016) In these challenging environments PBCs need resili-ence traits They also need high outputinput ratios forenergy (typically 20ndash50) to deliver large carbon savingsLand may also be marginal due to environmental vulnerabil-ity Much of the value for society from the genetic improve-ment of these crops depends on positive effects arising fromhighly productive perennial systems In addition to produc-ing biomass as a carbon source to replace fossil carbon thesecrops reduce nitrate leaching (Pugesgaard Schelde LarsenLaeligrke amp Joslashrgensen 2015) making them good candidates tohelp fulfil Water Framework Directive (200060EC) and canincrease soil carbon storage during their production (McCal-mont et al 2017)
The objective of this paper was to report on the prepared-ness for wide deployment by summarizing the technicalstate of the art in breeding of four important PBCs namelyswitchgrass miscanthus willow and poplar These four
crops are the most promising and advanced PBCs for tem-perate regions and have therefore the focus here Switch-grass and miscanthus are both rhizomatous grasses with C4
photosynthesis while willow and poplar are trees with C3
photosynthesis Specifically this paper (a) reviews availablecrop trait genetic diversity information (b) assesses the pro-gress of conventional breeding technologies for yield resili-ence and biomass quality (c) reports on progress with newmolecular‐based breeding technologies to increase speedand precision of selection and (d) discusses the require-ments and next steps for breeding of PBCs including com-mercial considerations in order to sustainably meet thebiomass requirements of a growing worldwide bioeconomy
We summarize the crop‐specific attributes the location ofbreeding programmes the current availability of commercialcultivars and yield expectations in selected environments(Table 1) and the generalized breeding targets for all PBCs(Table 2) Economic information relating to the current mar-ket value of the biomass and the investment in breeding arepresented for different countriesregions in Table 3 We alsopresent a comparison of the prebreeding and conventionalbreeding efforts step‐by‐step starting with wild germplasmcollection and evaluation before wide crossing of wild rela-tives (Table 4) Hybridization is followed by at least 6 yearsof selection and evaluation before commercial upscaling canbegin (Figure 1) Recurrent selection often over decades isused within parent populations as part of an ongoing long‐term process to produce hybrid vigour (Brummer 1999) Inthe following sections the state of the art and new opportuni-ties of breeding switchgrass miscanthus willow and poplarare described The application of modern breeding technolo-gies is compared for the four crops in Table 5 It is mostadvanced in poplar and is therefore described in most detail
2 | SWITCHGRASS
Switchgrass is indigenous to the North American prairiesIt is grown from seed and harvested annually using tech-nology similar to that used for pastures Based on collec-tions from thousands of wild prairie remnants the geneticresources are roughly divided into lowland and upland eco-types and there are distinct clades within each ecotypewhich occur along both latitudinal and longitudinal gradi-ents (Evans et al 2018 Lu et al 2013 Zhang et al
CLIFTON‐BROWN ET AL | 3
TABLE1
Breeding‐relatedattributes
forfour
leadingperennialbiom
asscrops(PBCs)
Species
Switc
hgrass
Miscanthu
sW
illow
Pop
lar
Typ
eC4mdash
Grass
C4mdash
Grass
C3mdash
SRC
C3mdash
SRCSRF
Sourcesof
indigeno
usgerm
plasm
CAato
MXeastof
Rocky
Mou
ntains
Eastern
AsiaandOceania
Predom
inantly
Northernhemisph
ere
Northernhemisph
ere
Breedingsystem
Mon
oeciou
sou
tcrossing
Mon
oeciou
sou
tcrossing
Dioeciousou
tcrossing
Dioeciousou
tcrossing
Ploidy
4times8
times2times
3times
4times
2timesminus12
times2times
Specieswith
ingenu
s~4
50~1
4~4
00~3
0ndash32
Typ
esused
mainlyfor
breeding
US
Low
land
ecotyp
e(sub
trop
ical
clim
ates)andup
land
ecotyp
e(tem
perate
clim
ates)
EUb JPSK
andUS
MsinandMsac
EUS
viminalistimesschw
erinii
Sda
syclad
ostimesrehd
eriana
S
dasyclad
osS
viminalis
USandUKS
viminalistimesmiyab
eana
S
miyab
eana
US
Spu
rpurea
timesmiyab
eana
S
purpurea
Pop
ulus
tricho
carpa
Pdelto
ides
Pnigra
Psuaveolens
subsp
maximow
icziiPba
lsam
ifera
Palba
andhy
brids
Typ
ical
haploid
geno
mesize
(Mbp
)~1
500
Msin~5
400
andMsac~4
400
(Raybu
rnCrawfordRaybu
rnamp
Juvik
2009
)
~450
~485
plusmn10
Breedingprog
rammes
CA20
00(REAP
Quebec)
US 19
92(N
ebraska
19
92(O
klahom
a)
1992
(Georgia)
19
96(W
isconsin)
19
96(Sou
thDakota)
20
00(Tennessee)
20
02(M
ississippi)
20
07(O
klahom
a)
2008
(RutgersNew
Jersey)
20
12(Cornell
New
York)
20
12(U
rbana‐Champaign
Illin
ois)
DE19
90s(K
lein‐W
anzleben)
NL20
00s(W
ageningen)
UK20
04(A
berystwyth)
CN20
06(Chang
sha)
JP20
06(H
okkaido)
US 20
06(Californiamdash
CERESInc)
20
06(CaliforniaandIndianamdash
MBI)
20
08(U
rbana‐Champaign)
SK20
09(Suw
on‐SNU)and(M
uan
NICSof
RDA)
FR20
11(Estreacutees‐M
ons)
UK19
80s(Lon
gAshton
relocatedto
Rothamsted
Researchin
2002
)SE
19
80s(SvaloumlfWeibu
llSalix
energi
Europ
aAB)
US
1990
s(Cornell
New
York)
PL20
00s(O
lsztyn
University
ofWarmia
andMazury)
SE
19
39(M
ykingeEkebo
Research
Institu
te)
19
90s(U
ppsalaSL
U)
20
10(U
ppsalaST
T)
US 19
27(N
ewYork
OxfordPaper
Com
pany
andNew
YorkBotanical
Garden
Wheeler
etal20
15)
19
79(W
ashing
ton
UW
andOrego
nGWR)
19
80ndash1
995(M
ississippiMSS
tate
and
GWR)
19
96(M
innesotaUMD
NRRIand
GWR)
IT19
83(Piedm
ontAFV
)FR
20
01(O
rleacuteans
ampNancyIN
RA
Charrey‐sur‐Saocircne
ampPierroton
FCBA
Nog
ent‐sur‐V
ernisson
IRST
EA)
DE20
08(G
oumlttin
gen
NW‐FVA)
(Con
tinues)
4 | CLIFTON‐BROWN ET AL
TABLE
1(Contin
ued)
Species
Switc
hgrass
Miscanthu
sW
illow
Pop
lar
Current
commercial
varietieson
the
market
US
Nocommercial
hybrids
CA2
EU1(M
timesgfrom
different
origins)
+selected
Msinforthatching
inDK
US
3(but
2aregenetically
identical)
UK25
US
8EUDElt10
FR44
IT10ndash1
5SE
~1
4US
8ndash12
Southeast8ndash
14Upp
erMidwest10
PacificNorthwest
Precom
mercial
cultivars
expected
tobe
onthemarketin
3years
US
36registered
cultivars
(halfare
rand
omseed
increasesfrom
natural
prairiesandhalfarebred
varieties)
Mostarepu
blic
releasesfew
are
protected
patented
orlicensed
NL8
seeded
hybridsvanDinterSemo
MTA
UK4
seeded
hybridsCERES(Land
OLakes)andTerravestaLtdMTA
US
Non
eFR
Non
e
EU53
registered
with
CPV
OforPB
RUK20
registered
with
CPV
OforPB
RUS
18clon
esfrom
Cornellin
multisite
trials
CAun
quantified
UAlberta
andQuebec
FR8ndash
12SR
Fclon
eslicence‐based
IT5SR
Cand6SR
FAFV
MTA
orlicence
SE14
STTlicence‐based
andMTA
US
~19So
utheast5ndash7Upp
erMidwest
from
UMD
NRRIMTA6ndash12
North
Central
from
GWRMTAPacific
Northwest8clon
esGWRMTA24
inmultilocationyieldtrialsGWR
Com
mercial
yield
(tDM
haminus1year
minus1 )
US
3ndash18
EU8ndash
12CN20
ndash30
US
10ndash25
EU7ndash20
UK8ndash
14US
8ndash14
EU5ndash20
(SEandBaltic
Cou
ntries8ndash
12)
US
10ndash2
2(10ndash
12North
Central12
ndash16
Southeast15ndash2
2PacificNortheast)
Harvestrotatio
nand
commercial
stand
lifespan
Ann
ualfor10
ndash12years
Ann
ualfor10ndash2
5years(M
sac
hasbeen
used
for~3
0yearsin
China)
2‐to
4‐year
cyclefor22
ndash30years
SRC3‐
to7‐year
cyclefor20
years
SRF
10‐to12‐yearcycleforgt50
years
Adaptiverang
eOpen‐po
llinatedandsynthetic
cultivars
arelim
itedin
adaptatio
nby
temperature
andprecipitatio
n(~8breeding
zonesin
USandCA)
Standard
Mtimesgiswidelyadaptedin
EU
butislim
itedin
theUnitedStates
byinsufficient
winterh
ardiness
forN
orthern
Midwestand
heatintolerancein
the
southNovelMsactimesMsinhyb
rids
andMsintimesMsinhybrids
selected
incontinentalD
Ehave
show
nawide
adaptiv
erang
ein
EU(K
alininaetal
2017
)Ongoing
trialson
heavymetal
contam
inated
soils
indicatetoleranceby
exclusion(K
rzyżak
etal20
17)
Different
hybridsareneeded
fordifferent
zones
Besthy
bridsshow
someG
timesE(U
S)
andsomehy
bridslow
GtimesE(Fabio
etal20
17)
Different
hybridsareneeded
ford
ifferent
clim
aticzonesHyb
rids
thatareadapted
forg
rowingseason
sof
~6mon
thsand
relativ
elyshortd
aysin
Southern
Europ
earemaladaptedto
shortg
rowing
season
sof
~4mon
thsandrelativ
ely
long
days
inNorthernEU
The
mostb
roadly
adaptedvarietiescome
from
thePcana
densistaxo
n
Note
AFV
AlasiaFranco
VivaiC
ornell
CornellUniversity
CPV
OC
ommunity
PlantV
ariety
OfficeFC
BAF
orestCelluloseW
ood
ConstructionandFu
rnitu
reTechnologyInstitu
teG
timesEg
enotype‐by
‐environmentinterac-
tion
GWRG
reenWoodResourcesINRAF
renchNationalInstituteforA
griculturalR
esearch
IRST
EAN
ationalR
esearchUnito
fScience
andTechnologyforE
nvironmentand
AgricultureM
sacMsacchariflo
rusandM
timesg
(Mtimes
giganteus)M
sin
MiscanthussinensisM
BIMendelB
iotechnology
IncM
bpm
egabase
pairM
SStateM
ississippi
StateUniversity
MTAm
aterialtransferagreem
entNICS
NationalInstituteof
CropScience
NW‐
FVANorthwestGerman
ForestResearchInstitu
tePB
Rplantbreeders
rightRDARural
DevelopmentAdm
inistration
REAP
ResourceEfficient
Agriculture
Productio
nSL
USw
edishUniversity
ofAgriculturalSciences
SNUS
eoul
NationalU
niSR
Cshort‐ro
tatio
ncoppice
SRF
short‐rotationforestryS
TTS
weT
reeTech
tDM
haminus1year
minus1 tons
ofdrymatterperhectareperyearU
AlbertaUniversity
ofAlbertaU
MD
NRRIUniversity
ofMinnesotaDuluthsNaturalResources
ResearchInstitu
teU
MNU
niversity
ofMinnesotaU
WU
niversity
ofWashington
UWMU
niversity
ofWarmiaandMazury
a ISO
Alpha‐2
lettercountrycodes
b EU
isused
forEurope
CLIFTON‐BROWN ET AL | 5
2011) Genotype‐by‐environment interactions (G times E) arestrong and must be considered in breeding (Casler 2012Casler Mitchell amp Vogel 2012) Adaptation to environ-ment is regulated principally by responses to day‐lengthand temperature There are also strong genotype times environ-ment interactions between the drier western regions and thewetter eastern regions (Casler et al 2017)
The growing regions of North America are divided intofour adaptation zones for switchgrass each roughly corre-sponding to two official hardiness zones The lowland eco-types are generally late flowering high yielding andadapted to warmer climates but have lower drought andcold resistance than upland ecotypes (Casler 2012 Casleret al 2012)
In 2015 the US Department of Agriculture (USDA)National Plant Germplasm System GRIN (httpswwwars-gringovnpgs) had 181 switchgrass accessions of whichonly 96 were available for distribution due to limitationsassociated with seed multiplication (Casler Vogel amp Har-rison 2015) There are well over 2000 additional uncata-logued accessions (Table 1) held by various universities butthe USDA access to these is also constrained by the effortneeded in seed multiplication Switchgrass is a model herba-ceous species for conducting scientific research on biomass(Sanderson Adler Boateng Casler amp Sarath 2006) but lit-tle funding is available for the critical prebreeding work thatis necessary to link this biological research to commercialbreeding More than a million genotypes from ~2000 acces-sions (seed accessions contain many genotypes) have beenphenotypically screened in spaced plant nurseries and tenthousand of the most useful have been genotyped with differ-ent technologies depending on the technology available at
the time when these were performed From these character-ized genotypes parents are selected for exploratory pairwisecrosses to produce synthetic populations within ecotypesSwitchgrass like many grasses is outcrossing due to astrong genetically controlled self‐incompatibly (akin to theS‐Z‐locus system of other grasses (Martinez‐Reyna ampVogel 2002)) Thus the normal breeding approaches usedare F1 wide crosses and recurrent selection cycles within syn-thetic populations
The scale of these programmes varies from small‐scaleconventional breeding based solely on phenotypic selec-tion (eg REAP Canada Montreal Quebec) to large pro-grammes incorporating modern molecular breedingmethods (eg USDA‐ARS Madison Wisconsin) Earlyagronomic research and biomass production efforts werefocused on the seed‐based multiplication of promising wildaccessions from natural prairies Cultivars Alamo Kanlowand Cave‐in‐Rock were popular due to high yield andmoderate‐to‐wide adaptation Conventional breedingapproaches focussed on biomass production traits and haveled to the development of five cultivars particularly suitedto biomass production Cimarron EG1101 EG1102EG2101 and Liberty The first four of these represent thelowland ecotype and were developed either in Oklahomaor Georgia Liberty is a derivative of lowland times uplandhybrids developed in Nebraska following selection for lateflowering the high yield of the lowland ecotype and coldtolerance of the upland ecotype (Vogel et al 2014) Thesefive cultivars were all approximately 25ndash30 years in themaking counting from the initiation of these breeding pro-grammes Many more biomass‐type cultivars are expectedwithin the next few years as these and other breeding
TABLE 2 Generalized improvement targets for perennial biomass crops (PBCs)
Net energy yield per hectare
Increased yield
Reduced moisture content at harvest
Physical and chemical composition for different end‐use applications
Increased lignin content and decreased corrosive elements for thermal conversion
Reduced recalcitrance through decreased lignin content andor modified lignin monomer composition to reduce pretreatment requirementsfor next‐generation biofuels by saccharification and fermentation
Plant morphological differences which influence biomass harvest transport and storage (eg stem thickness)
Propagation costs
Improved cloning systems (trees and grasses)
Seed systems (grasses)
Optimizing agronomy for each new cultivar
Resilience through enhanced
Abiotic stress toleranceresistance (eg drought salinity and high and low temperature )
Biotic stress resistance (eg insects fungal bacterial and viral diseases)
Site adaptability especially to those of marginalcontaminated agricultural land
6 | CLIFTON‐BROWN ET AL
TABLE3
Preparedness
formassupscaling
currentmarketvalueandresearch
investmentin
four
perennialbiom
asscrops(PBCs)
Species
Switc
hgrass
Miscanthu
sW
illow
Pop
lar
Current
commercial
plantin
gcostsperha
USa20
0USD
bin
the
establishm
entyear
DE3
375
Eurob
(XueK
alininaamp
Lew
ando
wski20
15)
UK2
153
GBPb
(Evans201
6)redu
cedto
116
9GBPwith
Mxg
rhizom
ewhere
thefarm
erdo
estheland
preparation(w
wwterravestacom
)US
173
0ndash222
5USD
UK150
0ndash173
9GBPplus
land
preparation(Evans20
16)
US
197
6USD
(=80
0USD
acre)
IT110
0Euro
FRSE100
0ndash200
0Euro
US
863USD
ha(LazarusHeadleeamp
Zalesny
20
15)
Current
marketvalue
ofthebiom
asspert
DM
US
80ndash1
00USD
UK~8
0GBP(bales)(Terravesta
person
alcommun
ication)
US
94USD
(chipp
ed)
UK49
41GBP(chipp
ed)(Evans20
16)
US
55ndash7
0USD
IT10
0Euro
US
80ndash90USD
deliv
ered
basedon
40milesof
haulage
Scienceforg
enetic
improv
ementprojects
inthelast10
years
USgt50
projectsfund
edby
USDOEand
USD
ANIFA
CNgt
30projectsfund
edby
CN‐N
SFCandMBI
EUc 6projects(O
PTIM
ISCO
PTIM
A
WATBIO
SUNLIBBF
IBRAandGRACE)
FRB
FFSK
2projectsfund
edby
IPETandPM
BC
US
EBICABBImultip
leDOEFeedstock
Genom
icsandUSD
AAFR
Iprojects
UKB
SBECB
EGIN
(200
3ndash20
10)
US
USDOEJG
IGenom
eSequ
encing
(200
9ndash20
122
015ndash
2018
)USD
ANortheastSu
nGrant
(200
9ndash20
12)
USD
ANIFANEWBio
Con
sortium
(201
2ndash20
17)USD
ANIFAWillow
SkyC
AP(201
8ndash20
21)
EUF
P7(EnergyPo
plarN
ovelTreeWATBIO
Tree4Fu
ture)
SEC
limate‐adaptedpo
plarsandNanow
ood
SLU
andST
TUS
Bioenergy
Science(O
RNL)USD
Afeedstocks
geno
micsprog
rammeBRDIa
ndBRC‐CBI
Major
projects
supp
ortin
gcrossing
andselectioncycles
inthelast10
years
USandCA12
projects
NLRUEmiscanthu
sPP
P20
15ndash2
019
50ndash
100K
Euroyear
UKGIA
NT‐LIN
KPP
I20
11ndash2
016
13M
GBPyear
US
EBICABBI~0
5M
USD
year
UKBEGIN
2000
ndash201
0US
USD
ANortheastSu
nGrant
(200
9ndash20
12)USD
ANIFA
NEWBio
Con
sortium
(201
2ndash20
17)USD
ANIFA
Willow
SkyC
AP(201
8ndash20
21)
FR1
4region
alandnatio
nalp
rojects10
0ndash15
0k
Euroyear
SES
TTo
nebreeding
effortas
aninternalproject
in20
10with
aresulting
prog
enytrial
US
DOESu
nGrant
feedstockdevelopm
ent
partnershipprog
ramme(including
Willow
)
Current
annu
alinvestmentin
projects
fortranslationinto
commercial
hybrids
USgt20
MUSD
year
UKMUST
201
6ndash20
1905M
GBPyear
US
CABBIUSD
AAFR
I20
17ndash2
02205M
USD
year
US
USD
ANIFA
NEWBio
~04
MUSD
year
FRScienceforim
prov
ement~2
0kEuroyear
US
USD
Aun
derAFR
ICo‐ordinatedAg
Prod
ucer
projectsAnd
severaltranslational
geno
micsprojectswith
outpu
blic
fund
ing
Upscalin
gtim
e(years)
toproducesufficient
propagules
toplantgt10
0ha
US
Using
seed
2ndash3years
UKRhizomefor1ndash200
0hectares
canbe
ready
in6mon
ths
UKNL2
0ha
ofseed
hybridsplantedin
2018
In
2019
sufficientseedfor5
0ndash10
0ha
isexpected
UK3yearsusingconv
entio
nalcutting
sfaster
usingmicroprop
agation
FRIT3yearsby
vegetativ
eprop
agation
SEgt3yearsby
vegetativ
eprop
agation(cuttin
gs)
and3yearsby
microprop
agation
Note
AFR
IAgriculture
andFo
odResearchInitiativein
theUnitedStatesBEGIN
BiomassforEnergyGenetic
Improvem
entNetwork
BFF
BiomassFo
rtheFu
tureBRC‐CBI
Bioenergy
ResearchCentre‐Centrefor
Bioenergy
Innovatio
nBRDIBiomassresearch
developm
entinitiative
BSB
ECBBSR
CSu
stainableBioenergy
Centre
CABBICenterforadvanced
bioenergyandbioproductsinnovatio
nin
theUnitedStatesCN‐N
SFC
Natural
ScienceFo
undatio
nof
ChinaDOEDepartm
entof
Energyin
theUnitedStatesEBIEnergyBiosciences
Institu
teFIBRAFibrecropsas
sustainablesource
ofbiobased
materialforindustrial
productsin
Europeand
ChinaFP
7SeventhFram
eworkProgrammein
theEUGIA
NT‐LIN
KGenetic
improvem
entof
miscanthusas
asustainablefeedstockforbioenergyin
theUnitedKingdom
GRACEGRow
ingAdvancedindustrial
Crops
onmarginallandsforbiorEfineriesIPETKorea
Institu
teof
Planning
andEvaluationforTechnologyin
FoodA
gricultureFo
restry
andFisheriesJG
IJointGenom
eInstitu
teMBIMendelBiotechnology
IncMUST
Miscant-
husUpS
calin
gTechnology
NEWBioNortheast
WoodyW
arm‐seasonBiomassConsortiumNIFANationalInstitu
teof
Food
andAgriculture
intheUnitedStatesNovelTree
Novel
tree
breeding
strategiesOPT
IMAOpti-
mizationof
perennialgrassesforbiom
assproductio
nin
theMediterraneanareaOPT
IMISCOptim
izingbioenergyproductio
nfrom
Miscanthus
ORNLOak
Ridge
NationalLabPM
BCPlantMolecular
BreedingCenterof
theNextGenerationBiogreenResearchCenters
oftheRepublic
ofKoreaPP
Ipublicndashp
rivate
investmentPP
Ppublicndashp
rivate
partnership
RUEradiationuseefficiencySk
yCAP
Coordinated
AgriculturalProjectSL
U
SwedishUniversity
ofAgriculturalSciencesST
TSw
eTreeTech
SUNLIBBSu
stainableLiquidBiofuelsfrom
BiomassBiorefiningTree4Fu
tureDesigning
Trees
forthefutureUSD
AUSDepartm
entof
Agriculture
WATBIO
Developmentof
improved
perennialnonfoodbiom
assandbioproduct
cropsforwater
stressed
environm
ents
a ISO
Alpha‐2
lettercountrycodes
b Local
currencies
areused
asat
2018Euro
GBP(G
reat
Britain
Pound)USD
(USDollar)c EU
isused
forEurope
CLIFTON‐BROWN ET AL | 7
TABLE4
Prebreedingresearch
andthestatus
ofconventio
nalbreeding
infour
leadingperennialbiom
asscrops(PBCs)
Breeding
techno
logy
step
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sWillow
Poplar
1Collected
wild
accessions
or
secondarysources
availablefor
breeding
Provideabroadbase
of
useful
traits
Respect
forCBD
and
Nagoyaprotocol
on
collections
after2014
Not
allindigenous
genetic
resourcesareaccessible
under
CBD
forpolitical
reasons
USa~2
000
(181
inofficial
GRIN
gene
bank
ofwhich
96
wereavailable
others
in
variousunofficial
collections)
CNChangsha
~1000
andNanjin
g
2000from
China
DK~1
20from
JP
FRworking
collectionof
~100
(mainlyM
sin)
JP~1
500
from
JPS
KandCN
NLworking
collectionof
~300
Msin
SK~7
00from
SKC
NJP
andRU
UK~1
500
accessions
(from
~500
sites)
ofwhich
1000arein
field
nurseriesin
2018
US
14in
GRIN
global
US
Illin
ois~1
500
collections
made
inAsiawith
25
currently
available
inUnitedStates
forbreeding
UK1500with
about20
incommon
with
the
UnitedStates
US
350largelyunique
FR3370Populus
delto
ides
(650)Pnigra(2000)
Ptrichocarpa(600)and
Pmaximow
iczii(120)
IT30000P
delto
idsPnigra
Pmaximow
icziiand
Ptrichocarpa
SE13000
accessionsmainly
Ptrichocarpa
byST
TandSL
U
US
GWR150Ptrichocarpaand
300Pmaximow
icziiaccessions
forPacificNorthwestprogram
535Pdelto
ides
and~2
00
Pnigraaccessions
for
Southeastprogrammeand77
Psimoniifrom
Mongolia
UMD
NRRI550+
clonal
accessions
(primarily
Ptimescanadnesisa
long
with
Pdelto
ides
andPnigra
parents)
forUpper
Midwest
program
2Wild
accessions
which
have
undergone
phenotypic
screeningin
field
trials
Selectionof
parental
lines
with
useful
traits
Strong
partnerships
torun
multilocationfieldtrials
tophenotypeconsistently
theaccessionsgenotypes
indifferentenvironm
ents
anddatabases
Costof
runningmultilocation
US
gt500000genotypes
CN~1
250
inChangsha
~1700
in
HunanJiangsuS
handong
Hainan
NL~250genotypes
JP~1
200
SK~4
00
UK‐le
d~1
000
genotypesin
a
replicated
trial
US‐led
~1200
genotypesin
multi‐
locatio
nreplicated
trialsin
Asiaand
North
America
MsinSK
CNJP
CA
(Ontario)US(Illinoisand
Colorado)MsacSK
(Kangw
on)
CN
(Zhejiang)JP
(Sapporo)CA
(Onatario)U
S(Illinois)DK
(Aarhus)
UKgt400
US
180
FR2720
IT10000
SE~1
50
US
GWR~1
500
wild
Ptrichocarpaphenotypes
and
OPseedlots(total
of70)from
thePmaximow
icziispecies
complex
(suaveolens
cathayana
koreana
ussuriensismaximow
iczii)and
2000ndash3000Pdelto
ides
collections
(based
on104s‐
generatio
nfamilies)
UMD
NRRIscreened
seedlin
gsfrom
aPdelto
ides
(OPor
CP)
breeding
programme(1996ndash2016)
gt7200OPseedlin
gsof
PnigraunderMinnesota
clim
atic
conditions(improve
parental
populatio
n)
3Wild
germ
plasm
genotyping
Amanaged
living
collectionof
clonal
types
US
~20000genotypes
CNChangsha
~1000Nanjin
g37
FR44
bycpDNA
(Fenget
al
UK~4
00
US
225
(Con
tinues)
8 | CLIFTON‐BROWN ET AL
TABLE
4(Contin
ued)
Breeding
techno
logy
step
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sWillow
Poplar
Construct
phylogenetic
treesanddissim
ilarity
indices
The
type
ofmolecular
analysis
mdashAFL
PcpDNAR
ADseq
whole
genomesequencing
2014)
JP~1
200
NL250genotypesby
RADseq
UK~1
000
byRADseq
SK~3
00
US
Illin
oisRADseqon
allwild
accessions
FR2310
SE150
US
1500
4Exploratory
crossing
and
progenytests
Discovergood
parental
combinatio
nsmdashgeneral
combining
ability
Geographicseparatio
n
speciesor
phylogenetic
trees
Costly
long‐te
rmmultilocation
trialsforprogeny
US
gt10000
CN~1
0
FR~2
00
JP~3
0
NL3600
SK~2
0
UK~4
000
(~1500Msin)
US
500ndash1000
UK gt700since2003
gt800EWBP(1996ndash
2002)
US
800
FRCloned13
times13
factorial
matingdesign
with
3species[10
Pdelto
ides8Ptrichocarpa
8
Pnigra)
US
GWR1000exploratory
crossingsof
Pfrem
ontii
Psimoniiforim
proved
droughttolerance
UMD
NRRIcrossednativ
e
Pdelto
ides
with
non‐nativ
e
Ptrichocarpaand
Pmaximow
icziiMajority
of
progenypopulatio
nssuffered
from
clim
atic
andor
pathogenic
susceptib
ility
issues
5Wideintraspecies
hybridization
Discovergood
parental
combinatio
nsmdashgeneral
combining
ability
1to
4above
inform
atics
Flow
eringsynchronizationand
low
seed
set
US
~200
CN~1
0
NL1800
SK~1
0
UK~5
00
UK276
US
400
IT500
SE120
US
50ndash100
6With
inspecies
recurrentselection
Concentratio
nof
positiv
e
traits
Identificationof
theright
heterotic
groups
insteps
1ndash5
Difficultto
introducenew
germ
plasm
with
outdilutio
n
ofbesttraits
US
gt40
populatio
ns
undergoing
recurrentselection
NL2populatio
ns
UK6populatio
ns
US
~12populatio
nsto
beestablished
by2019
UK4
US
150
FRPdelto
ides
(gt30
FSfamilies
ndash900clones)Pnigra(gt40
FS
families
ndash1600clones)and
Ptrichocarpa(15FS
Families
minus350clones)
IT80
SEca50
parent
breeding
populatio
nswith
in
Ptrichocarpa
US
Goalsis~1
00parent
breeding
populatio
nswith
inPtrichocarpa
Pdelto
ides
PnigraandPmaximow
iczii
7Interspecies
hybrid
breeding
Com
bining
complem
entary
traitsto
producegood
morphotypes
and
heterosiseffects
Allthesteps1ndash6
Inearlystage
improvem
ents
areunpredictable
Awide
base
iscostly
tomanage
None
CNChangsha
3Nanjin
g
120MsintimesMflo
r
JP~5
with
Saccharum
spontaneum
SK5
UK420
US
250
FRPcanadensis(1800)
Pdelto
ides
timesPtrichocarpa
(800)and
PtrichocarpatimesPmaximow
iczii
(1200)
(Con
tinues)
CLIFTON‐BROWN ET AL | 9
TABLE
4(Contin
ued)
Breeding
techno
logy
step
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sWillow
Poplar
IT~1
20
US
525genotypesin
eastside
hybrid
programme
205westside
hybrid
programmeand~1
200
in
SoutheastPdelto
ides
program
8Chrom
osom
e
doublin
g
Arouteto
triploid
seeded
typesdoublin
gadiploid
parent
ofknow
n
breeding
valuefrom
recurrentselection
Needs
1ndash7to
help
identify
therightparental
lines
Doubled
plantsarenotorious
forrevertingto
diploid
US
~50plantstakenfrom
tetraploid
tooctaploid
which
arenow
infieldtrials
CN~1
0Msactriploids
DKSeveralattemptsin
mid
90s
FR5genotypesof
Msin
UK2Msin1Mflora
nd1Msac
US
~30
UKAttempted
butnot
routinelyused
US
Not
partof
theregular
breeding
becausethere
existnaturalpolyploids
NA
9DoubleHaploids
Arouteto
producing
homogeneous
progeny
andalso
forthe
introductio
nof
transgenes
orforgenome
editing
Needs
1ndash7to
help
identify
therightparental
lines
Fully
homozygousplantsare
weakandeasily
die
andmay
notflow
ersynchronously
US
Noneyetattempteddueto
poor
vigour
andviability
of
haploids
CN2genotypesof
Msin
andMflor
UKDihaploidsof
MsinMflor
and
Msacand2transgenic
Msinvia
anther
cultu
re
US
6createdandused
toidentify
miss‐identifying
paralogueloci
(causedby
recent
genome
duplicationin
miscanthus)Usedin
creatin
gthereferencegenomefor
MsinVeryweakplantsand
difficultto
retain
thelin
es
NA
NA
10Embryo
rescue
Anattractiv
etechniquefor
recovering
plantsfrom
sexual
crosses
The
majority
ofem
bryos
cannot
survivein
vivo
or
becomelong‐timedorm
ant
None
UKone3times
hybrid
UKEmbryosrescue
protocol
proved
robustat
8days
post‐pollin
ation
FRAnem
bryo
rescue
protocol
proved
robustandim
proved
hybridizationsuccessfor
Pdelto
ides
timesPtrichocarpa
crosses
11Po
llenstorage
Flexibility
tocross
interestingparents
with
outneed
of
flow
ering
synchronization
None
Nosuccessto
daten
otoriously
difficultwith
grassesincluding
sugarcane
Freshpollencommonly
used
incrossing
Pollenstoragein
some
interspecificcrosses
FRBothstored
(cryobank)
and
freshindividual
pollen
Note
AFL
Pam
plifiedfragmentlength
polymorphismCBDconventio
non
biological
diversity
CP
controlledpollinatio
nCpD
NAchloroplastDNAEWBP
Europeanwillow
breeding
programme
FSfull‐sib
GtimesE
genotype‐by‐environm
entGRIN
germ
plasm
resourcesinform
ationnetwork
GWRGreenWoodResourcesMflorMiscanthusflo
ridulusMsacMsacchariflo
rus
MsinMsinensisNAnotapplicableNRRINatural
Resources
ResearchInstitu
teOP
open
pollinatio
nRADseq
restrictionsite‐associatedDNA
sequencingSL
USw
edishUniversity
ofAgriculturalSciencesST
TSw
eTreeTech
UMDUniversity
ofMinnesota
Duluth
a ISO
Alpha‐2
lettercountrycodes
10 | CLIFTON‐BROWN ET AL
programmes mature The average rate of gain for biomassyield in long‐term switchgrass breeding programmes hasbeen 1ndash4 per year depending on ecotype populationand location of the breeding programme (Casler amp Vogel2014 Casler et al 2018) The hybrid derivative Libertyhas a biomass yield 43 higher than the better of its twoparents (Casler amp Vogel 2014 Vogel et al 2014) Thedevelopment of cold‐tolerant and late‐flowering lowland‐ecotype populations for the northern United States hasincreased biomass yields by 27 (Casler et al 2018)
Currently more than 20 recurrent selection populationsare being managed in the United States to select parentsfor improved yield yield resilience and compositional qual-ity of the biomass For the agronomic development andupscaling high seed multiplication rates need to be com-bined with lower seed dormancy to reduce both crop estab-lishment costs and risks Expresso is the first cultivar withsignificantly reduced seed dormancy which is the first steptowards development of domesticated populations (Casleret al 2015) Most phenotypic traits of interest to breeders
require a minimum of 2 years to be fully expressed whichresults in a breeding cycle that is at least two years Morecomplicated breeding programmes or traits that requiremore time to evaluate can extend the breeding cycle to 4ndash8 years per generation for example progeny testing forbiomass yield or field‐based selection for cold toleranceBreeding for a range of traits with such long cycles callsfor the development of molecular methods to reduce time-scales and improve breeding efficiency
Two association panels of switchgrass have been pheno-typically and genotypically characterized to identify quanti-tative trait loci (QTLs) that control important biomasstraits The northern panel consists of 60 populationsapproximately 65 from the upland ecotype The southernpanel consists of 48 populations approximately 65 fromthe lowland ecotype Numerous QTLs have been identifiedwithin the northern panel to date (Grabowski et al 2017)Both panels are the subject of additional studies focused onbiomass quality flowering and phenology and cold toler-ance Additionally numerous linkage maps have been
FIGURE 1 Cumulative minimum years needed for the conventional breeding cycle through the steps from wild germplasm to thecommercial hybrids in switchgrass miscanthus willow and poplar Information links between the steps are indicated by dotted arrows andhighlight the importance of long‐term informatics to maximize breeding gain
CLIFTON‐BROWN ET AL | 11
created by the pairwise crossing of individuals with diver-gent characteristics often to generate four‐way crosses thatare analysed as pseudo‐F2 crosses (Liu Wu Wang ampSamuels 2012 Okada et al 2010 Serba et al 2013Tornqvist et al 2018) Individual markers and QTLs iden-tified can be used to design marker‐assisted selection(MAS) programmes to accelerate breeding and increase itsefficiency Genomic prediction and selection (GS) holdseven more promise with the potential to double or triplethe rate of gain for biomass yield and other highly complexquantitative traits of switchgrass (Casler amp Ramstein 2018Ramstein et al 2016) The genome of switchgrass hasrecently been made public through the Joint Genome Insti-tute (httpsphytozomejgidoegov)
Transgenic approaches have been heavily relied upon togenerate unique genetic variants principally for traitsrelated to biomass quality (Merrick amp Fei 2015) Switch-grass is highly transformable using either Agrobacterium‐mediated transformation or biolistics bombardment butregeneration of plants is the bottleneck to these systemsTraditionally plants from the cultivar Alamo were the onlyregenerable genotypes but recent efforts have begun toidentify more genotypes from different populations that arecapable of both transformation and subsequent regeneration(King Bray Lafayette amp Parrott 2014 Li amp Qu 2011Ogawa et al 2014 Ogawa Honda Kondo amp Hara‐Nishi-mura 2016) Cell wall recalcitrance and improved sugarrelease are the most common targets for modification (Bis-wal et al 2018 Fu et al 2011) Transgenic approacheshave the potential to provide traits that cannot be bredusing natural genetic variability However they will stillrequire about 10ndash15 years and will cost $70ndash100 millionfor cultivar development and deployment (Harfouche Mei-lan amp Altman 2011) In addition there is commercialuncertainty due to the significant costs and unpredictabletimescales and outcomes of the regulatory approval processin the countries targeted for seed sales As seen in maizeone advantage of transgenic approaches is that they caneasily be incorporated into F1 hybrid cultivars (Casler2012 Casler et al 2012) but this does not decrease thetime required for cultivar development due to field evalua-tion and seed multiplication requirements
The potential impacts of unintentional gene flow andestablishment of non‐native transgene sequences in nativeprairie species via cross‐pollination are also major issues forthe environmental risk assessment These limit further thecommercialization of varieties made using these technolo-gies Although there is active research into switchgrass steril-ity mechanisms to curb unintended pollen‐mediated genetransfer it is likely that the first transgenic cultivars proposedfor release in the United States will be met with considerableopposition due to the potential for pollen flow to remainingwild prairie sites which account for lt1 of the original
prairie land area and are highly protected by various govern-mental and nongovernmental organizations (Casler et al2015) Evidence for landscape‐level pollen‐mediated geneflow from genetically modified Agrostis seed multiplicationfields (over a mountain range) to pollinate wild relatives(Watrud et al 2004) confirms the challenge of using trans-genic approaches Looking ahead genome editing technolo-gies hold considerable promise for creating targeted changesin phenotype (Burris Dlugosz Collins Stewart amp Lena-ghan 2016 Liu et al 2018) and at least in some jurisdic-tions it is likely that cultivars resulting from gene editingwill not need the same regulatory approval as GMOs (Jones2015a) However in July 2018 the European Court of Justice(ECJ) ruled that cultivars carrying mutations resulting fromgene editing should be regulated in the same way as GMOsThe ECJ ruled that such cultivars be distinguished from thosearising from untargeted mutation breeding which isexempted from regulation under Directive 200118EC
3 | MISCANTHUS
Miscanthus is indigenous to eastern Asia and Oceania whereit is traditionally used for forage thatching and papermaking(Xi 2000 Xi amp Jezowkski 2004) In the 1960s the highbiomass potential of a Japanese genotype introduced to Eur-ope by Danish nurseryman Aksel Olsen in 1935 was firstrecognized in Denmark (Linde‐Laursen 1993) Later thisaccession was characterized described and named asldquoM times giganteusrdquo (Greef amp Deuter 1993 Hodkinson ampRenvoize 2001) commonly abbreviated as Mxg It is a natu-rally occurring interspecies triploid hybrid between tetraploidM sacchariflorus (2n = 4x) and diploid M sinensis(2n = 2x) Despite its favourable agronomic characteristicsand ability to produce high yields in a wide range of environ-ments in Europe (Kalinina et al 2017) the risks of relianceon it as a single clone have been recognized Miscanthuslike switchgrass is outcrossing due to self‐incompatibility(Jiang et al 2017) Thus seeded hybrids are an option forcommercial breeding Miscanthus can also be vegetativelypropagated by rhizome or in vitro culture which allows thedevelopment of clones The breeding approaches are usuallybased on F1 crosses and recurrent selection cycles within thesynthetic populations There are several breeding pro-grammes that target improvement of miscanthus traitsincluding stress resilience targeted regional adaptation agro-nomic ldquoscalabilityrdquo through cheaper propagation fasterestablishment lower moisture and ash contents and greaterusable yield (Clifton‐Brown et al 2017)
Germplasm collections specifically to support breedingfor biomass started in the late 1980s and early 1990s inDenmark Germany and the United Kingdom (Clifton‐Brown Schwarz amp Hastings 2015) These collectionshave continued with successive expeditions from European
12 | CLIFTON‐BROWN ET AL
and US teams assembling diverse collections from a widegeographic range in eastern Asia including from ChinaJapan South Korea Russia and Taiwan (Hodkinson KlaasJones Prickett amp Barth 2015 Stewart et al 2009) Threekey miscanthus species for biomass production are M si-nensis M floridulus and M sacchariflorus M sinensis iswidely distributed throughout eastern Asia with an adap-tive range from the subtropics to southern Russia (Zhaoet al 2013) This species has small rhizomes and producesmany tightly packed shoots forming a ldquotuftrdquo M floridulushas a more southerly adaptive range with a rather similarmorphology to M sinensis but grows taller with thickerstems and is evergreen and less cold‐tolerant than the othermiscanthus species M sacchariflorus is the most northern‐adapted species ranging to 50 degN in eastern Russia (Clarket al 2016) Populations of diploid and tetraploid M sac-chariflorus are found in China (Xi 2000) and South Korea(Yook 2016) and eastern Russia but only tetraploids havebeen found in Japan (Clark Jin amp Petersen 2018)
Germplasm has been assembled from multiple collec-tions over the last century though some early collectionsare poorly documented This historical germplasm has beenused to initiate breeding programmes largely based on phe-notypic and genotypic characterization As many of theaccessions from these collections are ldquoorigin unknownrdquocrucial environmental envelope data are not available UK‐led expeditions started in 2006 and continued until 2011with European and Asian partners and have built up a com-prehensive collection of 1500 accessions from 500 sitesacross Eastern Asia including China Japan South Koreaand Taiwan These collections were guided using spatialclimatic data to identify variation in abiotic stress toleranceAccessions from these recent collections were planted fol-lowing quarantine in multilocation nursery trials at severallocations in Europe to examine trait expression in differentenvironments Based on the resulting phenotypic andmolecular marker data several studies (a) characterized pat-terns of population genetic structure (Slavov et al 20132014) (b) evaluated the statistical power of genomewideassociation studies (GWASs) and identified preliminarymarkerndashtrait associations (Slavov et al 2013 2014) and(c) assessed the potential of genomic prediction (Daveyet al 2017 Slavov et al 2014 2018b) Genomic indexselection in particular offers the possibility of exploringscenarios for different locations or industrial markets (Sla-vov et al 2018a 2018b)
Separately US‐led expeditions also collected about1500 accessions between 2010 and 2014 (Clark et al2014 2016 2018 2015) A comprehensive genetic analy-sis of the population structure has been produced by RAD-seq for M sinensis (Clark et al 2015 Van der WeijdeKamei et al 2017) and M sacchariflorus (Clark et al2018) Multilocation replicated field trials have also been
conducted on these materials in North America and inAsia GWAS has been conducted for both M sinensis anda subset of M sacchariflorus accessions (Clark et al2016) To date about 75 of these recent US‐led collec-tions are in nursery trials outside the United States Due tolengthy US quarantine procedures these are not yet avail-able for breeding in the United States However molecularanalyses have allowed us to identify and prioritize sets ofgenotypes that best encompass the genetic variation in eachspecies
While mostM sinensis accessions flower in northern Eur-ope very few M sacchariflorus accessions flower even inheated glasshouses For this reason the European pro-grammes in the United Kingdom the Netherlands and Francehave performed mainly M sinensis (intraspecies) hybridiza-tions (Table 4) Selected progeny become the parents of latergenerations (recurrent selection as in switchgrass) Seed setsof up to 400 seed per panicle occur inM sinensis In Aberyst-wyth and Illinois significant efforts to induce synchronousflowering in M sacchariflorus and M sinensis have beenmade because interspecies hybrids have proven higher yieldperformance and wide adaptability (Kalinina et al 2017) Ininterspecies pairwise crosses in glasshouses breathable bagsandor large crossing tubes or chambers in which two or morewhole plants fit are used for pollination control Seed sets arelower in bags than in the open air because bags restrict pollenmovement while increasing temperatures and reducinghumidity (Clifton‐Brown Senior amp Purdy 2018) About30 of attempted crosses produced 10 to 60 seeds per baggedpanicle The seed (thousand seed mass ranges from 05 to09 g) is threshed from the inflorescences and sown into mod-ular trays to produce plug plants which are then planted infield nurseries to identify key parental combinations
A breeding programme of this scale must serve theneeds of different environments accepting the commonpurpose is to optimize the interception of solar radiationAn ideal hybrid for a given environment combines adapta-tion to date of emergence with optimization of traits suchas height number of stems per plant flowering and senes-cence time to optimize solar interception to produce a highbiomass yield with low moisture content at harvest (Rob-son Farrar et al 2013 Robson Jensen et al 2013) By20132014 conventional breeding in Europe had producedintra‐ and interspecific fertile seeded hybrids When acohort (typically about 5) of outstanding crosses have beenidentified it is important to work on related upscaling mat-ters in parallel These are as follows
Assessment of the yield and critical traits in selectedhybrids using a network of field trials
Efficient cloning of the seed parents While in vitro andmacro‐cloning techniques are used some genotypes areamenable to neither technique
CLIFTON‐BROWN ET AL | 13
TABLE5
Status
ofmod
ernplantbreeding
techniqu
esin
four
leadingperennialbiom
asscrop
s(PBCs)
Breeding
techno
logy
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sW
illow
Pop
lar
MAS
Use
ofmarker
sequ
encesthat
correlatewith
atrait
allowingearly
prog
enyselectionor
rejectionlocatin
gkn
owngenesfor
useful
traits(suchas
height)from
other
speciesin
your
crop
Breeding
prog
ramme
relevant
biparental
crosses(Table
4)to
create
ldquomapping
popu
latio
nsrdquoand
QTLidentification
andestim
ationto
identifyrobu
stmarkers
alternatively
acomprehensive
panelof
unrelated
geno
typesfor
GWAS
Ineffectivewhere
traits
areaffected
bymany
geneswith
small
effects
US
Literally
dozens
ofstud
iesto
identify
SNPs
andmarkers
ofinterestNoattempts
atMASas
yetlargely
dueto
popu
latio
nspecificity
CNS
SRISS
Rmarkers
forM
sinMsacand
Mflor
FR2
Msin
mapping
popu
latio
ns(BFF
project)
NL2
Msin
mapping
popu
latio
ns(A
tienza
Ram
irezetal
2003
AtienzaSatovic
PetersenD
olstraamp
Martin
200
3Van
Der
Weijdeetal20
17a
Van
derW
eijde
Hux
ley
etal20
17
Van
derW
eijdeKam
ei
etal20
17V
ander
WeijdeKieseletal
2017
)SK
2GWASpanels(1
collectionof
Msin1
diploidbiparentalF 1
popu
latio
n)in
the
pipelin
eUK1
2families
tofind
QTLin
common
indifferentfam
ilies
ofthe
sameanddifferent
speciesandhy
brids
US
2largeGWAS
panels4
diploidbi‐
parentalF 1
popu
latio
ns
1F 2
popu
latio
n(add
ition
alin
pipelin
e)
UK16
families
for
QTLdiscov
ery
2crossesMASscreened
1po
tentialvariety
selected
US
9families
geno
typedby
GBS(8
areF 1o
neisF 2)a
numberof
prom
ising
QTLdevelopm
entof
markers
inprog
ress
FR(INRA)tested
inlargeFS
families
and
infactorialmating
design
low
efficiency
dueto
family
specificity
US
49families
GS
Metho
dto
accelerate
breeding
throug
hredu
cing
the
resourcesforcross
attemptsby
Aldquotraining
popu
latio
nrdquoof
individu
alsfrom
stages
abov
ethat
have
been
both
Risks
ofpo
orpredictio
nProg
eny
testingneedsto
becontinueddu
ring
the
timethetraining
setis
US
One
prog
rammeso
farmdash
USD
Ain
WisconsinThree
cycles
ofgeno
mic
selectioncompleted
CN~1
000
geno
types
NLprelim
inary
mod
elsforMsin
developed
UK~1
000
geno
types
Verysuitable
application
not
attempted
yet
Maybe
challeng
ingfor
interspecies
hybrids
FR(INRA)120
0geno
type
ofPop
ulus
nigraarebeingused
todevelop
intraspecificGS
(Con
tinues)
14 | CLIFTON‐BROWN ET AL
TABLE
5(Contin
ued)
Breeding
techno
logy
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sW
illow
Pop
lar
predictin
gthe
performance
ofprog
enyof
crosses
(and
inresearch
topredictbestparents
tousein
biparental
crosses)
geno
typedand
phenotyp
edto
developamod
elthattakesgeno
typic
datafrom
aldquocandidate
popu
latio
nrdquoof
untested
individu
als
andprod
uces
GEBVs(Jannink
Lorenzamp
Iwata
2010
)
beingph
enotyp
edand
geno
typed
Inthis
time
next‐generation
germ
plasm
andreadyto
beginthe
second
roun
dof
training
and
recalib
ratio
nof
geno
mic
predictio
nmod
els
US
Prelim
inary
mod
elsforMsac
and
Msin
developed
Work
isun
derw
ayto
deal
moreeffectivelywith
polyploidgenetics
calib
ratio
nsUS
125
0geno
types
arebeingused
todevelopGS
calib
ratio
ns
Traditio
nal
transgenesis
Efficient
introd
uctio
nof
ldquoforeign
rdquotraits
(possiblyfrom
other
genu
sspecies)
into
anelite
plant(ega
prov
enparent
ora
hybrid)that
needsa
simpletraitto
beim
prov
edValidate
cand
idategenesfrom
QTLstud
ies
MASand
know
ledg
eof
the
biolog
yof
thetrait
andsource
ofgenesto
confer
the
relevant
changesto
phenotyp
eWorking
transformation
protocol
IPissuescostof
regu
latory
approv
al
GMO
labelling
marketin
gissuestricky
tousetransgenes
inou
t‐breedersbecause
ofcomplexity
oftransformingandgene
flow
risks
US
Manyprog
ramsin
UnitedStates
are
creatin
gtransgenic
plantsusingAlamoas
asource
oftransformable
geno
typesManytraits
ofinterestNothing
commercial
yet
CNC
hang
shaBtg
ene
transformed
into
Msac
in20
04M
sin
transformed
with
MINAC2gene
and
Msactransform
edwith
Cry2A
ageneb
othwith
amarkerg
eneby
Agrob
acterium
JP1
Msintransgenic
forlow
temperature
tolerancewith
increased
expression
offructans
(unp
ublished)
NLImprov
ingprotocol
forM
sin
transformation
SK2
Msin
with
Arabido
psisand
Brachypod
ium
PhytochromeBgenes
(Hwang
Cho
etal
2014
Hwang
Lim
etal20
14)
UK2
Msin
and
1Mflorg
enotyp
es
UKRou
tine
transformationno
tyet
possibleResearchto
overcomerecalcitrance
ongo
ing
Currently
trying
differentspecies
andcond
ition
sPo
plar
transformationused
atpresent
US
Frequently
attempted
with
very
limitedsuccess
US
600transgenic
lines
have
been
characterized
mostly
performed
inthe
aspenhy
brids
reprod
uctiv
esterility
drou
ghttolerance
with
Kno
ckdo
wns
(Con
tinues)
CLIFTON‐BROWN ET AL | 15
TABLE
5(Contin
ued)
Breeding
techno
logy
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sW
illow
Pop
lar
variou
slytransformed
with
four
cellwall
genesiptand
uidA
genesusingtwo
selectionsystem
sby
biolisticsand
Agrob
acterium
Transform
edplantsare
beinganalysed
US
Prelim
inarywork
will
betakenforw
ardin
2018
ndash202
2in
CABBI
Genom
eediting
CRISPR
Refinem
entofexisting
traits
inusefulparents
orprom
isinghybrids
bygeneratingtargeted
mutations
ingenes
know
ntocontrolthe
traitof
interestF
irst
adouble‐stranded
breakismadeinthe
DNAwhich
isrepairedby
natural
DNArepair
machineryL
eads
tofram
eshiftSN
Por
can
usealdquorepair
templaterdquo
orcanbe
used
toinserta
transgene
intoaldquosafe
harbourlocusrdquoIt
couldbe
used
todeleterepressors(or
transcriptionfactors)
Identification
mapping
and
sequ
encing
oftarget
genes(from
DNA
sequ
ence)
avoiding
screening
outof
unintend
ededits
Manyregu
latory
authorities
have
not
decidedwhether
CRISPR
andother
geno
meediting
techno
logies
are
GMOsor
notIf
GMOsseecomment
onstage10
abov
eIf
noteditedcrop
swill
beregu
latedas
conv
entio
nalvarieties
CATechn
olog
yisstill
toonewand
switchg
rass
geno
meis
very
complexOther
labo
ratories
are
interestedbu
tno
tyet
mov
ingon
this
US
One
prog
rammeso
farmdash
USD
Ain
Albany
FRInitiated
in20
16(M
ISEDIT
project)
NLInitiatingin
2018
US
Initiatingin
2018
in
CABBI
UKNot
started
Not
possible
until
transformation
achieved
US
CRISPR
‐Cas9and
Cpf1have
been
successfully
developedin
Pop
ulus
andarehigh
lyefficientExamples
forlig
nin
biosyn
thesis
Note
BFF
BiomassFo
rtheFu
tureBtBacillus
thuringiensisCABBICenterforadvanced
bioenergyandbioproductsinnovatio
nCpf1
CRISPR
from
Prevotella
andFrancisella
1CRISPR
clusteredregularlyinterspaced
palin
drom
icrepeatsCRISPR
‐CasC
RISPR
‐associated
Cry2A
acrystaltoxins2A
asubfam
ilyproduced
byBtFS
full‐sibG
BS
genotyping
bysequencingG
EBVsgenomic‐estim
ated
breeding
valuesG
MOg
enetically
modi-
fied
organism
GS
genomicselection
GWAS
genomew
ideassociationstudy
INRAF
renchNationalInstituteforAgriculturalR
esearch
IPintellectualp
roperty
IPTisopentenyltransferase
ISSR
inter‐SSR
MflorMiscant-
husflo
ridulusMsacMsacchariflo
rusMsinMsinensisM
AS
marker‐a
ssistedselection
MISEDITm
iscanthusgene
editing
forseed‐propagatedtriploidsNACn
oapicalmeristemA
TAF1
2and
cup‐shaped
cotyledon2
‐lik
eQTLq
uantitativ
etraitlocusS
NP
singlenucleotid
epolymorphismS
SRsim
plesequence
repeatsUSD
AU
SDepartm
ento
fAgriculture
a ISO
Alpha‐2
lettercountrycodes
16 | CLIFTON‐BROWN ET AL
High seed production from field crossing trials con-ducted in locations where flowering in both seed andpollen parents is likely to happen synchronously
Scalable and adapted harvesting threshing and seed pro-cessing methods for producing high seed quality
The results of these parallel activities need to becombined to identify the upscaling pathway for eachhybrid if this cannot be achieved the hybrid will likelynot be commercially viable The UK‐led programme withpartners in Italy and Germany shows that seedbased mul-tiplication rates of 12000 are achievable several inter-specific hybrids (Clifton‐Brown et al 2017) Themultiplication rate of M sinensis is higher probably15000ndash10000 Conventional cloning from rhizome islimited to around 120 that is one ha could provide rhi-zomes for around 20 ha of new plantation
Multilocation field testing of wild and novel miscanthushybrids selected by breeding programmes in the Nether-lands and the United Kingdom was performed as part ofthe project Optimizing Miscanthus Biomass Production(OPTIMISC 2012ndash2016) These trials showed that com-mercial yields and biomass qualities (Kiesel et al 2017Van der Weijde et al 2017a Van der Weijde Kieselet al 2017) could be produced in a wide range of climatesand soil conditions from the temperate maritime climate ofwestern Wales to the continental climate of eastern Russiaand the Ukraine (Kalinina et al 2017) Extensive environ-mental measurements of soil and climate combined withgrowth monitoring are being used to understand abioticstresses (Nunn et al 2017 Van der Weijde Huxley et al2017) and develop genotype‐specific scenarios similar tothose reported earlier in Hastings et al (2009) Phenomicsexperiments on drought tolerance have been conducted onwild and improved germplasm (Malinowska Donnison ampRobson 2017 Van der Weijde Huxley et al 2017)Recently produced interspecific hybrids displaying excep-tional yield under drought (~30 greater than control Mxg)in field trials in Poland and Moldova are being furtherstudied in detail in the phenomics and genomics facility atAberystwyth to better understand genendashtrait associationswhich can be fed back into breeding
Intraspecific seeded hybrids of M sinensis produced inthe Netherlands and interspecific M sacchariflorus times M si-nensis hybrids produced by the UK‐led breeding programmehave entered yield testing in 2018 with the recently EU‐funded project ldquoGRowing Advanced industrial Crops on mar-ginal lands for biorEfineries (GRACE)rdquo (httpswwwgrace-bbieu) Substantial variation in biomass quality for sacchari-fication efficiency (glucose release as of dry matter) ashcontent and melting point has already been generated inintraspecific M sinensis hybrids (Van der Weijde Kiesel
et al 2017) across environments (Weijde Dolstra et al2017a) GRACE aims to establish more than 20 hectares ofnew inter‐ and intraspecific seeded hybrids across six Euro-pean countries This project is building the know‐how andagronomy needed to transition from small research plots tocommercial‐scale field sites and linking biomass productiondirectly to industrial applications The biomass produced byhybrids in different locations will be supplied to innovativeindustrial end‐users making a wide range of biobased prod-ucts both for chemicals and for energy In the United Statesmulti‐location yield were initiated in 2018 to evaluate new tri-ploid M times giganteus genotypes developed at Illinois Cur-rently infertile hybrids are favoured in the United Statesbecause this eliminates the risk of invasiveness from naturallydispersed viable seed The precautionary principle is appliedas fertile miscanthus has naturalized in several states (QuinnAllen amp Stewart 2010) North European multilocation fieldtrials in the EMI and OPTIMISC projects have shown thereis minimal risk of invasiveness even in years when fertileflowering hybrids produce viable seed Naturalized standshave not established here due perhaps to low dormancy pooroverwintering and low seedling competitive strength In addi-tion to breeding for nonshattering or sterile seeded hybridsQuinn et al (2010) suggest management strategies which canfurther minimize environmental opportunities to manage therisk of invasiveness
31 | Molecular breeding and biotechnology
In miscanthus new plant breeding techniques (Table 5)have focussed on developing molecular markers forbreeding in Europe the United States South Korea andJapan There are several publications on QTL mappingpopulations for key traits such as flowering (AtienzaRamirez amp Martin 2003) and compositional traits(Atienza Satovic Petersen Dolstra amp Martin 2003) Inthe United States and United Kingdom independent andinterconnected bi‐parental ldquomappingrdquo families have beenstudied (Dong et al 2018 Gifford Chae SwaminathanMoose amp Juvik 2015) alongside panels of diverse germ-plasm accessions for GWAS (Slavov et al 2013) Fur-ther developments calibrating GS with very large panelsof parents and cross progeny are underway (Daveyet al 2017) The recently completed first miscanthus ref-erence genome sequence is expected to improve the effi-ciency of MAS strategies and especially GWAS(httpsphytozomejgidoegovpzportalhtmlinfoalias=Org_Msinensis_er) For example without a referencegenome sequence Clark et al (2014) obtained 21207RADseq SNPs (single nucleotide polymorphisms) on apanel of 767 miscanthus genotypes (mostly M sinensis)but subsequent reanalysis of the RADseq data using the
CLIFTON‐BROWN ET AL | 17
new reference genome resulted in hundreds of thousandsof SNPs being called
Robust and effective in vitro regeneration systems havebeen developed for Miscanthus sinensis M times giganteusand M sacchariflorus (Dalton 2013 Guo et al 2013Hwang Cho et al 2014 Rambaud et al 2013 Ślusarkie-wicz‐Jarzina et al 2017 Wang et al 2011 Zhang et al2012) However there is still significant genotype speci-ficity and these methods need ldquoin‐houserdquo optimization anddevelopment to be used routinely They provide potentialroutes for rapid clonal propagation and also as a basis forgenetic transformation Stable transformation using bothbiolistics (Wang et al 2011) and Agrobacterium tumefa-ciens DNA delivery methods (Hwang Cho et al 2014Hwang Lim et al 2014) has been achieved in M sinen-sis The development of miscanthus transformation andgene editing to generate diplogametes for producing seed‐propagated triploid hybrids are performed as part of theFrench project MISEDIT (miscanthus gene editing forseed‐propagated triploids) There are no reports of genomeediting in any miscanthus species but new breeding inno-vations including genome editing are particularly relevantin this slow‐to‐breed nonfood bioenergy crop (Table 4)
4 | WILLOW
Willow (Salix spp) is a very diverse group of catkin‐bear-ing trees and shrubs Willow belongs to the family Sali-caceae which also includes the Populus genus There areapproximately 350 willow species (Argus 2007) foundmostly in temperate and arctic zones in the northern hemi-sphere A few are adapted to subtropical and tropicalzones The centre of diversity is believed to be in Asiawith over 200 species in China Around 120 species arefound in the former Soviet Union over 100 in NorthAmerica and around 65 species in Europe and one speciesis native to South America (Karp et al 2011) Willows aredioecious thus obligate outcrossers and highly heterozy-gous The haploid chromosome number is 19 (Hanley ampKarp 2014) Around 40 of species are polyploid (Sudaamp Argus 1968) ranging from triploids to the atypicaldodecaploid S maxxaliana with 2n=190 (Zsuffa et al1984)
Although almost exclusively native to the NorthernHemisphere willow has been grown around the globe formany thousands of years to support a wide range of appli-cations (Kuzovkina amp Quigley 2005 Stott 1992) How-ever it has been the focus of domestication for bioenergypurposes for only a relatively short period since the 1970sin North America and Europe For bioenergy breedershave focused their efforts on the shrub willows (subgenusVetix) because of their rapid juvenile growth rates as aresponse to coppicing on a 2‐ to 4‐year cycle that can be
accomplished using farm machinery rather than forestryequipment (Shield Macalpine Hanley amp Karp 2015Smart amp Cameron 2012)
Since shrub willow was not generally recognized as anagricultural crop until very recently there has been littlecommitment to building and maintaining germplasm reposi-tories of willow to support long‐term breeding One excep-tion is the United Kingdom where a large and well‐characterized Salix germplasm collection comprising over1500 accessions is held at Rothamsted Research (Stott1992 Trybush et al 2008) Originally initiated for use inbasketry in 1923 accessions have been added ever sinceIn the United States a germplasm collection of gt350accessions is located at Cornell University to support thebreeding programme there The UK and Cornell collectionshave a relatively small number of accessions in common(around 20) Taken together they represent much of thespecies diversity but only a small fraction of the overallgenetic diversity within the genus There are three activewillow breeding programmes in Europe RothamstedResearch (UK) Salixenergi Europa AB (SEE) and a pro-gramme at the University of Warmia and Mazury in Olsz-tyn (Poland) (abbreviations used in Table 1) There is oneactive US programme based at Cornell University Culti-vars are still being marketed by the European WillowBreeding Programme (EWBP) (UK) which was activelybreeding biomass varieties from 1996 to 2002 Cultivarsare protected by plant breedersrsquo rights (PBRs) in Europeand by plant patents in the United States The sharing ofgenetic resources in the willow community is generallyregulated by material transfer agreements (MTA) and tai-lored licensing agreements although the import of cuttingsinto North America is prohibited except under special quar-antine permit conditions
Efforts to augment breeding germplasm collection fromnature are continuing with phenotypic screening of wildgermplasm performed in field experiments with 177 S pur-purea genotypes in the United States (at sites in Genevaand Portland NY and Morgantown WV) that have beengenotyped using genotyping by sequencing (GBS) (Elshireet al 2011) In addition there are approximately 400accessions of S viminalis in Europe (near Pustnaumls Upp-sala Sweden and Woburn UK (Berlin et al 2014 Hal-lingbaumlck et al 2016) The S viminalis accessions wereinitially genotyped using 38 simple sequence repeats (SSR)markers to assess genetic diversity and screened with~1600 SNPs in genes of potential interest for phenologyand biomass traits Genetics and genomics combined withextensive phenotyping have substantially improved thegenetic basis of biomass‐related traits in willow and arenow being developed in targeted breeding via MAS Thisunderpinning work has been conducted on large specifi-cally developed biparental Salix mapping populations
18 | CLIFTON‐BROWN ET AL
(Hanley amp Karp 2014 Zhou et al 2018) as well asGWAS panels (Hallingbaumlck et al 2016)
Once promising parental combinations are identifiedcrosses are usually performed using fresh pollen frommaterial that has been subject to a phased removal fromcold storage (minus4degC) (Lindegaard amp Barker 1997 Macal-pine Shield Trybush Hayes amp Karp 2008 Mosseler1990) Pollen storage is useful in certain interspecific com-binations where flowering is not naturally synchronizedThis can be overcome by using pollen collection and stor-age protocol which involves extracting pollen using toluene(Kopp Maynard Niella Smart amp Abrahamson 2002)
The main breeding approach to improve willow yieldsrelies on species hybridization to capture hybrid vigour(Fabio et al 2017 Serapiglia Gouker amp Smart 2014) Inthe absence of genotypic models for heterosis breedershave extensively tested general and specific combiningability of parents to produce superior progeny The UKbreeding programmes (EWBP 1996ndash2002 and RothamstedResearch from 2003 on) have performed more than 1500exploratory cross‐pollinations The Cornell programme hassuccessfully completed about 550 crosses since 1998Investment into the characterization of genetic diversitycombined with progeny tests from exploratory crosses hasbeen used to produce hundreds of targeted intraspeciescrosses in the United Kingdom and United States respec-tively (see Table 1) To achieve long‐term gains beyond F1hybrids four intraspecific recurrent selection populationshave been created in the United Kingdom (for S dasycla-dos S viminalis and S miyabeana) and Cornell is pursu-ing recurrent selection of S purpurea Interspecifichybridizations with genotypes selected from the recurrentselection cycles are well advanced in willow with suchcrosses to date totalling 420 in the United Kingdom andover 100 in the United States
While species hybridization is common in Salix it isnot universal Of the crosses attempted about 50 hybri-dize and produce seed (Macalpine Shield amp Karp 2010)As the viability of seed from successful crosses is short (amatter of days at ambient temperatures) proper seed rear-ing and storage protocols are essential (Maroder PregoFacciuto amp Maldonado 2000)
Progeny from crosses are treated in different waysamong the breeding programmes at the seedling stage Inthe United States seedlings are planted into an irrigatedfield where plants are screened for two seasons beforebeing progressed to further field trials In the United King-dom seedlings are planted into trays of compost wherethey remain containerized in an irrigated nursery for theremainder of year one In the United Kingdom seedlingsare subject to two rounds of selection in the nursery yearThe first round takes place in September to select againstsusceptibility to rust infection (Melampsora spp) A second
round of selection in winter assesses tip damage from frostand giant willow aphid infestation In the United Stateswhere the rust pressure is lower screening for Melampsoraspp cannot be performed at the nursery stage Both pro-grammes monitor Melampsora spp pest susceptibilityyield and architecture over multiple years in field trialsSelected material is subject to two rounds of field trials fol-lowed by a final multilocation yield trial to identify vari-eties for commercialization
Promising selections (ie potential cultivars) need to beclonally propagated A rapid in vitro tissue culture propa-gation method has been developed (Palomo‐Riacuteos et al2015) This method can generate about 5000 viable trans-plantable clones from a single plant in just 24 weeks Anin vitro system can also accommodate early selection viamolecular or biochemical markers to increase selectionspeed Conventional breeding systems take 13 years viafour rounds of selection from crossing to selecting a variety(Figure 1) but this has the potential to be reduced to7 years if micropropagation and MAS selection are adopted(Hanley amp Karp 2014 Palomo‐Riacuteos et al 2015)
Willows are currently propagated commercially byplanting winter‐dormant stem cuttings in spring Commer-cial planting systems for willow use mechanical plantersthat cut and insert stem sections from whips into a well‐prepared soil One hectare of stock plants grown in specificmultiplication beds planted at 40000 plants per ha pro-duces planting material for 80 hectares of commercialshort‐rotation coppice willow annually (planted at 15000cuttings per hectare) (Whittaker et al 2016) When com-mercial plantations are established the industry standard isto plant intimate mixtures of ~5 diverse rust (Melampsoraspp)‐resistant varieties (McCracken amp Dawson 1997 VanDen Broek et al 2001)
The foundations for using new plant breeding tech-niques have been established with funding from both thepublic and the private sectors To establish QTL maps 16mapping populations from biparental crosses are understudy in the United Kingdom Nine are under study in theUnited States The average number of individuals in thesefamilies ranges from 150 to 947 (Hanley amp Karp 2014)GS is also being evaluated in S viminalis and preliminaryresults indicate that multiomic approaches combininggenomic and metabolomic data have great potential (Sla-vov amp Davey 2017) For both QTL and GS approachesthe field phenotyping demands are large as several thou-sand individuals need to be phenotyped for a wide rangeof traits These include the following dates of bud burstand growth succession stem height stem density wooddensity and disease resistance The greater the number ofindividuals the more precise the QTL marker maps andGS models are However the logistical and financial chal-lenges of phenotyping large numbers of individuals are
CLIFTON‐BROWN ET AL | 19
considerable because the willow crop is gt5 m tall in thesecond year There is tremendous potential to improve thethroughput of phenotyping using unmanned aerial systemswhich is being tested in the USDA National Institute ofFood and Agriculture (NIFA) Willow SkyCAP project atCornell Further investment in these approaches needs tobe sustained over many years fully realizes the potentialof a marker‐assisted selection programme for willow
To date despite considerable efforts in Europe and theUnited States to establish a routine transformation systemthere has not been a breakthrough in willow but attemptsare ongoing As some form of transformation is typically aprerequisite for genome editing techniques these have notyet been applied to willow
In Europe there are 53 short‐rotation coppice (SRC) bio-mass willow cultivars registered with the Community PlantVariety Office (CPVO) for PBRs of which ~25 are availablecommercially in the United Kingdom There are eightpatented cultivars commercially available in the UnitedStates In Sweden there are nine commercial cultivars regis-tered in Europe and two others which are unregistered(httpssalixenergiseplanting-material) Furthermore thereare about 20 precommercial hybrids in final yield trials inboth the United States and the United Kingdom It has beenestimated that it would take two years to produce the stockrequired to plant 50 ha commercially from the plant stock inthe final yield trials Breeding programmes have alreadydelivered rust‐resistant varieties and increases in yield to themarket The adoption of advanced breeding technologies willlikely lead to a step change in improving traits of interest
5 | POPLAR
Poplar a fast‐growing tree from the northern hemispherewith a small genome size has been adopted for commercialforestry and scientific purposes The genus Populus con-sists of about 29 species classified in six different sectionsPopulus (formerly Leuce) Tacamahaca Aigeiros AbasoTuranga and Leucoides (Eckenwalder 1996) The Populusspecies of most interest for breeding and testing in the Uni-ted States and Europe are P nigra P deltoides P maxi-mowiczii and P trichocarpa (Stanton 2014) Populusclones for biomass production are being developed byintra‐ and interspecies hybridization (DeWoody Trewin ampTaylor 2015 Richardson Isebrands amp Ball 2014 van derSchoot et al 2000) Recurrent selection approaches areused for gradual population improvement and to create eliteclonal lines for commercialization (Berguson McMahon ampRiemenschneider 2017 Neale amp Kremer 2011) Currentlypoplar breeding in the United States occurs in industrialand academic programmes located in the Southeast theMidwest and the Pacific Northwest These use six speciesand five interspecific taxa (Stanton 2014)
The southeastern programme historically focused onrecurrent selection of P deltoides from accessions made inthe lower Mississippi River alluvial plain (Robison Rous-seau amp Zhang 2006) More recently the genetic base hasbeen broadened to produce interspecific hybrids with resis-tance to the fungal infection Septoria musiva which causescankers
In the midwest of the United States populationimprovement efforts are focused on P deltoides selectionsfrom native provenances and hybrid crosses with acces-sions introduced from Europe Interspecific intercontinental(Europe and America) hybrid crosses between P nigra andP deltoides (P times canadensis) are behind many of theleading commercial hybrids which are the most advancedbreeding materials for many applications and regions InMinnesota previous breeding experience and efforts utiliz-ing P maximowiczii and P trichocarpa have been discon-tinued due to Septoria susceptibility and a lack of coldhardiness (Berguson et al 2017) Traits targeted forimprovement include yieldgrowth rate cold hardinessadventitious rooting resistance to Septoria and Melamp-sora leaf rust and stem form The Upper Midwest pro-gramme also carries out wide hybridizations within thesection Populus The P times wettsteinii (P trem-ula times P tremuloides) taxon is bred for gains in growthrate wood quality and resistance to the fungus Entoleucamammata which causes hypoxylon canker (David ampAnderson 2002)
In the Pacific Northwest GreenWood Resources Incleads poplar breeding that emphasizes interspecifichybrid improvement of P times generosa (P deltoides timesP trichocarpa and reciprocal) and P deltoides times P maxi-mowiczii taxa for coastal regions and the P times canadensistaxon for the drier continental regions Intraspecificimprovement of second‐generation breeding populations ofP deltoides P nigra P maximowiczii and P trichocarpaare also involved (Stanton et al 2010) The present focusof GreenWood Resourcesrsquo hybridization is bioenergy feed-stock improvement concentrating on coppice yield wood‐specific gravity and rate of sugar release
Industrial interest in poplar in the United States has his-torically come from the pulp and paper sector althoughveneer and dimensional lumber markets have been pursuedat times Currently the biomass market for liquid trans-portation fuels is being emphasized along with the use oftraditional and improved poplar genotypes for ecosystemservices such as phytoremediation (Tuskan amp Walsh 2001Zalesny et al 2016)
In Europe there are breeding programmes in FranceGermany Italy and Sweden These include the following(a) Alasia Franco Vivai (AFV) programme in northernItaly (b) the French programme led by the poplar Scien-tific Interest Group (GIS Peuplier) and carried out
20 | CLIFTON‐BROWN ET AL
collaboratively between the National Institute for Agricul-tural Research (INRA) the National Research Unit ofScience and Technology for Environment and Agriculture(IRSTEA) and the Forest Cellulose Wood Constructionand Furniture Technology Institute (FCBA) (c) the Ger-man programme at Northwest German Forest Research Sta-tion (NW‐FVA) at Hannoversch Muumlnden and (d) theSwedish programme at the Swedish University of Agricul-tural Sciences and SweTree Technologies AB (Table 1)
AFV leads an Italian poplar breeding programme usingextensive field‐grown germplasm collections of P albaP deltoides P nigra and P trichocarpa While interspeci-fic hybridization uses several taxa the focus is onP times canadensis The breeding programme addresses dis-ease resistance (Marssonina brunnea Melampsora larici‐populina Discosporium populeum and poplar mosaicvirus) growth rate and photoperiod adaptation AFV andGreenWood Resources collaborate in poplar improvementin Europe through the exchange of frozen pollen and seedfor reciprocal breeding projects Plantations in Poland andRomania are currently the focus of the collaboration
The ongoing French GIS Peuplier is developing a long‐term breeding programme based on intraspecific recurrentselection for the four parental species (P deltoides P tri-chocarpa P nigra and P maximowiczii) designed to betterbenefit from hybrid vigour demonstrated by the interspeci-fic crosses P canadensis P deltoides times P trichocarpaand P trichocarpa times P maximowiczii Current selectionpriorities are targeting adaptation to soil and climate condi-tions resistance and tolerance to the most economicallyimportant diseases and pests high volume production underSRC and traditional poplar cultivation regimes as well aswood quality of interest by different markets Currentlygenomic selection is under exploration to increase selectionaccuracy and selection intensity while maintaining geneticdiversity over generations
The German NW‐FVA programme is breeding intersec-tional AigeirosndashTacamahaca hybrids with a focus on resis-tance to Pollaccia elegans Xanthomonas populiDothichiza spp Marssonina brunnea and Melampsoraspp (Stanton 2014) Various cross combinations ofP maximowiczii P trichocarpa P nigra and P deltoideshave led to new cultivars suitable for deployment in vari-etal mixtures of five to ten genotypes of complementarystature high productivity and phenotypic stability (Weis-gerber 1993) The current priority is the selection of culti-vars for high‐yield short‐rotation biomass production Sixhundred P nigra genotypes are maintained in an ex situconservation programme An in situ P nigra conservationeffort involves an inventory of native stands which havebeen molecular fingerprinted for identity and diversity
The Swedish programme is concentrating on locallyadapted genotypes used for short‐rotation forestry (SRF)
because these meet the needs of the current pulping mar-kets Several field trials have shown that commercial poplarclones tested and deployed in Southern and Central Europeare not well adapted to photoperiods and low temperaturesin Sweden and in the Baltics Consequently SwedishUniversity of Agricultural Sciences and SweTree Technolo-gies AB started breeding in Sweden in 1990s to producepoplar clones better adapted to local climates and markets
51 | Molecular breeding technologies
Poplar genetic improvement cannot be rapidly achievedthrough traditional methods alone because of the longbreeding cycles outcrossing breeding systems and highheterozygosity Integrating modern genetic genomic andphenomics techniques with conventional breeding has thepotential to expedite poplar improvement
The genome of poplar has been sequenced (Tuskanet al 2006) It has an estimated genome size of485 plusmn 10 Mbp divided into 19 chromosomes This is smal-ler than other PBCs and makes poplar more amenable togenetic engineering (transgenesis) GS and genome editingPoplar has seen major investment in both the United Statesand Europe being the model system for woody perennialplant genetics and genomics research
52 | Targets for genetic modification
Traits targeted include wood properties (lignin content andcomposition) earlylate flowering male sterility to addressbiosafety regulation issues enhanced yield traits and herbi-cide tolerance These extensive transgenic experiments haveshown differences in recalcitrance to in vitro regenerationand genetic transformation in some of the most importantcommercial hybrid poplars (Alburquerque et al 2016)Further transgene expression stability is being studied Sofar China is the only country known to have commerciallyused transgenic insect‐resistant poplar A precommercialherbicide‐tolerant poplar was trialled for 8 years in the Uni-ted States (Li Meilan Ma Barish amp Strauss 2008) butcould not be released due to stringent environmental riskassessments required for regulatory approval This increasestranslation costs and delays reducing investor confidencefor commercial deployment (Harfouche et al 2011)
The first field trials of transgenic poplar were performedin France in 1987 (Fillatti Sellmer Mccown Haissig ampComai 1987) and in Belgium in 1988 (Deblock 1990)Although there have been a total of 28 research‐scale GMpoplar field trials approved in the European Union underCouncil Directive 90220EEC since October 1991 (inPoland Belgium Finland France Germany Spain Swe-den and in the United Kingdom (Pilate et al 2016) onlyauthorizations in Poland and Belgium are in place today In
CLIFTON‐BROWN ET AL | 21
the United States regulatory notifications and permits fornearly 20000 transgenic poplar trees derived from approxi-mately 600 different constructs have been issued since1995 by the USDAs Animal and Plant Health InspectionService (APHIS) (Strauss et al 2016)
53 | Genome editing CRISPR technologies
Clustered regularly interspaced palindromic repeats(CRISPR) and the CRISPR‐associated (CRISPR‐Cas)nucleases are a groundbreaking genome‐engineering toolthat complements classical plant breeding and transgenicmethods (Moreno‐Mateos et al 2017) Only two publishedstudies in poplar have applied the CRISPRCas9 technol-ogy One is in P tomentosa in which an endogenous phy-toene desaturase gene (PtoPDS) was successfully disruptedsite specifically in the first generation of transgenic plantsresulting in an albino and dwarf phenotype (Fan et al2015) The second was in P tremula times alba in whichhigh CRISPR‐Cas9 mutational efficiency was achieved forthree 4‐coumarateCoA ligase (4Cl) genes 4CL1 4CL2and 4CL5 associated with lignin and flavonoid biosynthe-sis (Zhou et al 2015) Due to its low cost precision andrapidness it is very probable that cultivars or clones pro-duced using CRISPR technology will be ready for market-ing in the near future (Yin et al 2017) Recently aCRISPR with a smaller associated endonuclease has beendiscovered from Prevotella and Francisella 1 (Cpf1) whichmay have advantages over Cas9 In addition there arereports of DNA‐free editing in plants using both CRISPRCpf1 and CRISPR Cas9 for example Ref (Kim et al2017 Mahfouz 2017 Zaidi et al 2017)
It remains unresolved whether plants modified by genomeediting will be regulated as genetically modified organisms(GMOs) by the relevant authorities in different countries(Lozano‐Juste amp Cutler 2014) Regulations to cover thesenew breeding techniques are still evolving but those coun-tries who have published specific guidance (including UnitedStates Argentina and Chile) are indicating that plants pos-sessing simple genome edits will not be regulated as conven-tional transgenesis (Jones 2015b) The first generation ofgenome‐edited crops will likely be phenocopy gene knock-outs that already exist to produce ldquonature identicalrdquo traits thatis traits that could also be derived by conventional breedingDespite this confidence in applying these new powerfulbreeding tools remains limited owing to the uncertain regula-tory environment in many parts of the world (Gao 2018)including the recent ECJ 2018 rulings mentioned earlier
54 | Genomics‐based breeding technologies
Poplar breeding programs are becoming well equipped withuseful genomics tools and resources that are critical to
explore genomewide variability and make use of the varia-tion for enhancing genetic gains Deep transcriptomesequencing resequencing of alternate genomes and GBStechnology for genomewide marker detection using next‐generation sequencing (NGS) are yielding valuable geno-mics tools GWAS with NGS‐based markers facilitatesmarker identification for MAS breeding by design and GS
GWAS approaches have provided a deeper understand-ing of genome function as well as allelic architectures ofcomplex traits (Huang et al 2010) and have been widelyimplemented in poplar for wood characteristics (Porthet al 2013) stomatal patterning carbon gain versus dis-ease resistance (McKown et al 2014) height and phenol-ogy (Evans et al 2014) cell wall chemistry (Mucheroet al 2015) growth and cell walls traits (Fahrenkroget al 2017) bark roughness (Bdeir et al 2017) andheight and diameter growth (Liu et al 2018) Using high‐throughput sequencing and genotyping platforms an enor-mous amount of SNP markers have been used to character-ize the linkage disequilibrium (LD) in poplar (eg Slavovet al 2012 discussed below)
The genetic architecture of photoperiodic traits in peren-nial trees is complex involving many loci However itshows high levels of conservation during evolution (Mau-rya amp Bhalerao 2017) These genomics tools can thereforebe used to address adaptation issues and fine‐tune themovement of elite lines into new environments For exam-ple poor timing of spring bud burst and autumn bud setcan result in frost damage resulting in yield losses (Ilstedt1996) These have been studied in P tremula genotypesalong a latitudinal cline in Sweden (~56ndash66oN) and haverevealed high nucleotide polymorphism in two nonsynony-mous SNPs within and around the phytochrome B2 locus(Ingvarsson Garcia Hall Luquez amp Jansson 2006 Ing-varsson Garcia Luquez Hall amp Jansson 2008) Rese-quencing 94 of these P tremula genotypes for GWASshowed that noncoding variation of a single genomicregion containing the PtFT2 gene described 65 ofobserved genetic variation in bud set along the latitudinalcline (Tan 2018)
Resequencing genomes is currently the most rapid andeffective method detecting genetic differences betweenvariants and for linking loci to complex and importantagronomical and biomass traits thus addressing breedingchallenges associated with long‐lived plants like poplars
To date whole genome resequencing initiatives havebeen launched for several poplar species and genotypes InPopulus LD studies based on genome resequencing sug-gested the feasibility of GWAS in undomesticated popula-tions (Slavov et al 2012) This plant population is beingused to inform breeding for bioenergy development Forexample the detection of reliable phenotypegenotype asso-ciations and molecular signatures of selection requires a
22 | CLIFTON‐BROWN ET AL
detailed knowledge about genomewide patterns of allelefrequency variation LD and recombination suggesting thatGWAS and GS in undomesticated populations may bemore feasible in Populus than previously assumed Slavovet al (2012) have resequenced 16 genomes of P tri-chocarpa and genotyped 120 trees from 10 subpopulationsusing 29213 SNPs (Geraldes et al 2013) The largest everSNP data set of genetic variations in poplar has recentlybeen released providing useful information for breedinghttpswwwbioenergycenterorgbescgwasindexcfmAlso deep sequencing of transcriptomes using RNA‐Seqhas been used for identification of functional genes andmolecular markers that is polymorphism markers andSSRs A multitissue and multiple experimental data set forP trichocarpa RNA‐Seq is publicly available (httpsjgidoegovdoe-jgi-plant-flagship-gene-atlas)
The availability of genomic information of DNA‐con-taining cell organelles (nucleus chloroplast and mitochon-dria) will also allow a holistic approach in poplarmolecular breeding in the near future (Kersten et al2016) Complete Populus genome sequences are availablefor nucleus (P trichocarpa section Tacamahaca) andchloroplasts (seven species and two clones from P trem-ula W52 and P tremula times P alba 717ndash1B4) A compara-tive approach revealed structural and functionalinformation broadening the knowledge base of PopuluscpDNA and stimulating future diagnostic marker develop-ment The availability of whole genome sequences of thesecellular compartments of P tremula holds promise forboosting marker‐assisted poplar breeding Other nucleargenome sequences from additional Populus species arenow available (eg P deltoides (httpsphytozomejgidoegovpz) and will become available in the forthcomingyears (eg P tremula and P tremuloidesmdashPopGenIE(Sjodin Street Sandberg Gustafsson amp Jansson 2009))Recently the characterization of the poplar pan‐genome bygenomewide identification of structural variation in threecrossable poplar species P nigra P deltoides and P tri-chocarpa revealed a deeper understanding of the role ofinter‐ and intraspecific structural variants in poplar pheno-type and may have important implications for breedingparticularly interspecific hybrids (Pinosio et al 2016)
GS has been proposed as an alternative to MAS in cropimprovement (Bernardo amp Yu 2007 Heffner Sorrells ampJannink 2009) GS is particularly well suited for specieswith long generation times for characteristics that displaymoderate‐to‐low heritability for traits that are expensive tomeasure and for selection of traits expressed late in the lifecycle as is the case for most traits of commercial value inforestry (Harfouche et al 2012) Current joint genomesequencing efforts to implement GS in poplar using geno-mic‐estimated breeding values for bioenergy conversiontraits from 49 P trichocarpa families and 20 full‐sib
progeny are taking place at the Oak Ridge National Labo-ratory and GreenWood Resources (Brian Stanton personalcommunication httpscbiornlgov) These data togetherwith the resequenced GWAS population data will be thebasis for developing GS algorithms Genomic breedingtools have been developed for the intraspecific programmetargeting yield resistance to Venturia shoot blightMelampsora leaf rust resistance to Cryptorhynchus lapathistem form wood‐specific gravity and wind firmness (Evanset al 2014 Guerra et al 2016) A newly developedldquobreeding with rare defective allelesrdquo (BRDA) technologyhas been developed to exploit natural variation of P nigraand identify defective variants of genes predicted by priortransgenic research to impact lignin properties Individualtrees carrying naturally defective alleles can then be incor-porated directly into breeding programs thereby bypassingthe need for transgenics (Vanholme et al 2013) Thisnovel breeding technology offers a reverse genetics com-plement to emerging GS for targeted improvement of quan-titative traits (Tsai 2013)
55 | Phenomics‐assisted breeding technology
Phenomics involves the characterization of phenomesmdashthefull set of phenotypes of given individual plants (HouleGovindaraju amp Omholt 2010) Traditional phenotypingtools which inefficiently measure a limited set of pheno-types have become a bottleneck in plant breeding studiesHigh‐throughput plant phenotyping facilities provide accu-rate screening of thousands of plant breeding lines clones orpopulations over time (Fu 2015) are critical for acceleratinggenomics‐based breeding Automated image collection andanalysis phenomics technologies allow accurate and nonde-structive measurements of a diversity of phenotypic traits inlarge breeding populations (Gegas Gay Camargo amp Doo-nan 2014 Goggin Lorence amp Topp 2015 Ludovisi et al2017 Shakoor Lee amp Mockler 2017) One important con-sideration is the identification of relevant and quantifiabletarget traits that are early diagnostic indicators of biomassyield Good progress has been made in elucidating theseunderpinning morpho‐physiological traits that are amenableto remote sensing in Populus (Harfouche Meilan amp Altman2014 Rae Robinson Street amp Taylor 2004) Morerecently Ludovisi et al (2017) developed a novel methodol-ogy for field phenomics of drought stress in a P nigra F2partially inbred population using thermal infrared imagesrecorded from an unmanned aerial vehicle‐based platform
Energy is the current main market for poplar biomass butthe market return provided is not sufficient to support pro-duction expansion even with added demand for environmen-tal and land management ldquoecosystem servicesrdquo such as thetreatment of effluent phytoremediation riparian buffer zonesand agro‐forestry plantings Aviation fuel is a significant
CLIFTON‐BROWN ET AL | 23
target market (Crawford et al 2016) To serve this marketand to reduce current carbon costs of production (Budsberget al 2016) key improvement traits in addition to yield(eg coppice regeneration pestdisease resistance water‐and nutrient‐use efficiencies) will be trace greenhouse gas(GHG) emissions (eg isoprene volatiles) site adaptabilityand biomass conversion efficiency Efforts are underway tohave national environmental protection agenciesrsquo approvalfor poplar hybrids qualifying for renewable energy credits
6 | REFLECTIONS ON THECOMMERCIALIZATIONCHALLENGE
The research and innovation activities reviewed in thispaper aim to advance the genetic improvement of speciesthat can provide feedstocks for bioenergy applicationsshould those markets eventually develop These marketsneed to generate sufficient revenue and adequately dis-tribute it to the actors along the value chain The work onall four crops shares one thing in common long‐termefforts to integrate fundamental knowledge into breedingand crop development along a research and development(RampD) pipeline The development of miscanthus led byAberystwyth University exemplifies the concerted researcheffort that has integrated the RampD activities from eight pro-jects over 14 years with background core research fundingalong an emerging innovation chain (Figure 2) This pro-gramme has produced a first range of conventionally bredseeded interspecies hybrids which are now in upscaling tri-als (Table 3) The application of molecular approaches(Table 5) with further conventional breeding (Table 4)offers the prospect of a second range of improved seededhybrids This example shows that research‐based support ofthe development of new crops or crop types requires along‐term commitment that goes beyond that normallyavailable from project‐based funding (Figure 1) Innovationin this sector requires continuous resourcing of conven-tional breeding operations and capability to minimize timeand investment losses caused by funding discontinuities
This challenge is increased further by the well‐knownmarket failure in the breeding of many agricultural cropspecies The UK Department for Environment Food andRural Affairs (Defra) examined the role of geneticimprovement in relation to nonmarket outcomes such asenvironmental protection and concluded that public invest-ment in breeding was required if profound market failure isto be addressed (Defra 2002) With the exception ofwidely grown hybrid crops such as maize and some high‐value horticultural crops royalties arising from plant breed-ersrsquo rights or other returns to breeders fail to adequatelycompensate for the full cost for research‐based plant breed-ing The result even for major crops such as wheat is sub‐
optimal investment and suboptimal returns for society Thismarket failure is especially acute for perennial crops devel-oped for improved sustainability rather than consumerappeal (Tracy et al 2016) Figure 3 illustrates the underly-ing challenge of capturing value for the breeding effortThe ldquovalley of deathrdquo that results from the low and delayedreturns to investment applies generally to the research‐to‐product innovation pipeline (Beard Ford Koutsky amp Spi-wak 2009) and certainly to most agricultural crop speciesHowever this schematic is particularly relevant to PBCsMost of the value for society from the improved breedingof these crops comes from changes in how agricultural landis used that is it depends on the increased production ofthese crops The value for society includes many ecosys-tems benefits the effects of a return to seminatural peren-nial crop cover that protects soils the increase in soilcarbon storage the protection of vulnerable land or the cul-tivation of polluted soils and the reductions in GHG emis-sions (Lewandowski 2016) By its very nature theproduction of biomass on agricultural land marginal forfood production challenges farm‐level profitability Thecosts of planting material and one‐time nature of cropestablishment are major early‐stage costs and therefore theopportunities for conventional royalty capture by breedersthat are manifold for annual crops are limited for PBCs(Hastings et al 2017) Public investment in developingPBCs for the nonfood biobased sector needs to providemore long‐term support for this critical foundation to a sus-tainable bioeconomy
7 | CONCLUSIONS
This paper provides an overview of research‐based plantbreeding in four leading PBCs For all four PBC generasignificant progress has been made in genetic improvementthrough collaboration between research scientists and thoseoperating ongoing breeding programmes Compared withthe main food crops most PBC breeding programmes dateback only a few decades (Table 1) This breeding efforthas thus co‐evolved with molecular biology and the result-ing ‐omics technologies that can support breeding Thedevelopment of all four PBCs has depended strongly onpublic investment in research and innovation The natureand driver of the investment varied In close associationwith public research organizations or universities all theseprogrammes started with germplasm collection and charac-terization which underpin the selection of parents forexploratory wide crosses for progeny testing (Figure 1Table 4)
Public support for switchgrass in North America wasexplicitly linked to plant breeding with 12 breeding pro-grammes supported in the United States and CanadaSwitchgrass breeding efforts to date using conventional
24 | CLIFTON‐BROWN ET AL
breeding have resulted in over 36 registered cultivars inthe United States (Table 1) with the development of dedi-cated biomass‐type cultivars coming within the past fewyears While ‐omics technologies have been incorporatedinto several of these breeding programmes they have notyet led to commercial deployment in either conventional orhybrid cultivars
Willow genetic improvement was led by the researchcommunity closely linked to plant breeding programmesWillow and poplar have the longest record of public invest-ment in genetic improvement that can be traced back to1920s in the United Kingdom and United States respec-tively Like switchgrass breeding programmes for willoware connected to public research efforts The United King-dom in partnership with the programme based at CornellUniversity remains the European leader in willowimprovement with a long‐term breeding effort closelylinked to supporting biological research at Rothamsted Inwillow F1 hybrids have produced impressive yield gainsover parental germplasm by capturing hybrid vigour Over30 willow clones are commercially available in the UnitedStates and Europe and a further ~90 are under precommer-cial testing (Table 1)
Compared with willow and poplar miscanthus is a rela-tive newcomer with all the current breeding programmesstarting in the 2000s Clonal M times giganteus propagated byrhizomes is expected to be replaced by more readily scal-able seeded hybrids from intra‐ (M sinensis) and inter‐(M sacchariflorus times M sinensis) species crosses with highseed multiplication rates (of gt2000) The first group ofhybrid cultivars is expected to be market‐ready around2022
Of the four genera used as PBCs Populus is the mostadvanced in terms of achievements in biological researchas a result of its use as a model for basic research of treesMuch of this biological research is not directly connectedto plant breeding Nevertheless reflecting the fact thatpoplar is widely grown as a single‐stem tree in SRF thereare about 60 commercially available clones and an addi-tional 80 clones in commercial pipelines (Table 1) Trans-genic poplar hybrids have moved beyond proof of conceptto commercial reality in China
Many PBC programmes have initiated long‐term con-ventional recurrent selection breeding cycles for populationimprovement which is a key process in increasing yieldthrough hybrid vigour As this approach requires manyyears most programmes are experimenting with molecularbreeding methods as these have the potential to accelerateprecision breeding For all four PBCs investments in basicgenetic and genomic resources including the developmentof mapping populations for QTLs and whole genomesequences are available to support long‐term advancesMore recently association genetics with panels of diversegermplasm are being used as training populations for GSmodels (Table 5) These efforts are benefitting from pub-licly available DNA sequences and whole genome assem-blies in crop databases Key to these accelerated breedingtechnologies are developments in novel phenomics tech-nologies to bridge the genotypephenotype gap In poplarnovel remote sensing field phenotyping is now beingdeployed to assist breeders These advances are being com-bined with in vitro and in planta modern molecular breed-ing techniques such as CRISPR (Table 5) CRISPRtechnology for genome editing has been proven in poplar
FIGURE 2 A schematic developmentpathway for miscanthus in the UnitedKingdom related to the investment in RampDprojects at Aberystwyth (top coloured areasfor projects in the three categories basicresearch breeding and commercialupscaling) leading to a projected croppingarea of 350000 ha by 2030 with clonal andsuccessive ranges of improved seed‐basedhybrids Purple represents theBiotechnology and Biological SciencesResearch Council (BBSRC) and brown theDepartment for Environment Food andRural Affairs (Defra) (UK Nationalfunding) blue bars represent EU fundingand green private sector funding (Terravestaand CERES) and GIANT‐LINK andMiscanthus Upscaling Technology (MUST)are publicndashprivate‐initiatives (PPI)
CLIFTON‐BROWN ET AL | 25
This technology is also being applied in switchgrass andmiscanthus (Table 5) but the future of CRISPR in com-mercial breeding for the European market is uncertain inthe light of recent ECJ 2018 rulings
There is integration of research and plant breeding itselfin all four PBCs Therefore estimating the ongoing costs ofmaintaining these breeding programmes is difficult Invest-ment in research also seeks wider benefits associated withtechnological advances in plant science rather than cultivardevelopment per se However in all cases the conventionalbreeding cycle shown in Figure 1 is the basic ldquoenginerdquo withmolecular technologies (‐omics) serving to accelerate thisengine The history of the development shows that the exis-tence of these breeding programmes is essential to gain ben-efits from the biological research Despite this it is thisessential step that is at most risk from reductions in invest-ment A conventional breeding programme typicallyrequires a breeder and several technicians who are supported
over the long term (20ndash30 years Figure 1) at costs of about05 to 10 million Euro per year (as of 2018) The analysisreported here shows that the time needed to perform onecycle of conventional breeding bringing germplasm fromthe wild to a commercial hybrid ranged from 11 years inswitchgrass to 26 years in poplar (Figure 1) In a maturecrop grown on over 100000 ha with effective cultivar pro-tection and a suitable business model this level of revenuecould come from royalties Until such levels are reachedPBCs lie in the innovation valley of death (Figure 3) andneed public support
Applying industrial ldquotechnology readiness levelsrdquo (TRL)originally developed for aerospace (Heacuteder 2017) to ourplant breeding efforts we estimate many promising hybridscultivars are at TRL levels of 3ndash4 In Table 3 experts ineach crop estimate that it would take 3 years from now toupscale planting material from leading cultivars in plot tri-als to 100 ha
FIGURE 3 A schematic relating some of the steps in the innovation chain from relatively basic crop science research through to thedeployment in commercial cropping systems and value chains The shape of the funnel above the expanding development and deploymentrepresents the availability of investment along the development chain from relatively basic research at the top to the upscaled deployment at thebottom Plant breeding links the research effort with the development of cropping systems The constriction represents the constrained fundingfor breeding that links conventional public research investment and the potential returns from commercial development The handover pointsbetween publicly funded work to develop the germplasm resources (often known as prebreeding) the breeding and the subsequent cropdevelopment are shown on the left The constriction point is aggravated by the lack academic rewards for this essential breeding activity Theoutcome is such that this innovation system is constrained by the precarious resourcing of plant breeding The authorsrsquo assessment ofdevelopment status of the four species is shown (poplar having two one for short‐rotation coppice (SRC) poplar and one for the more traditionalshort‐rotation forestry (SRF)) The four new perennial biomass crops (PBCs) are now in the critical phase of depending of plant breedingprogress without the income stream from a large crop production base
26 | CLIFTON‐BROWN ET AL
Taking the UK example mentioned in the introductionplanting rates of ~35000 ha per year from 2022 onwardsare needed to reach over 1 m ha by 2050 Ongoing workin the UK‐funded Miscanthus Upscaling Technology(MUST) project shows that ramping annual hybrid seedproduction from the current level of sufficient seed for10 ha in 2018 to 35000 ha would take about 5 yearsassuming no setbacks If current hybrids of any of the fourPBCs in the upscaling pipeline fail on any step for exam-ple lower than expected multiplication rates or unforeseenagronomic barriers then further selections from ongoingbreeding are needed to replace earlier candidates
In conclusion the breeding foundations have been laidwell for switchgrass miscanthus willow and poplar owingto public funding over the long time periods necessaryImproved cultivars or genotypes are available that could bescaled up over a few years if real sustained market oppor-tunities emerged in response to sustained favourable poli-cies and industrial market pull The potential contributionsof growing and using these PBCs for socioeconomic andenvironmental benefits are clear but how farmers andothers in commercial value chains are rewarded for mass‐scale deployment as is necessary is not obvious at presentTherefore mass‐scale deployment of these lignocellulosecrops needs developments outside the breeding arenas todrive breeding activities more rapidly and extensively
8 | UNCITED REFERENCE
Huang et al (2019)
ACKNOWLEDGEMENTS
UK The UK‐led miscanthus research and breeding wasmainly supported by the Biotechnology and BiologicalSciences Research Council (BBSRC) Department for Envi-ronment Food and Rural Affairs (Defra) the BBSRC CSPstrategic funding grant BBCSP17301 Innovate UKBBSRC ldquoMUSTrdquo BBN0161491 CERES Inc and Ter-ravesta Ltd through the GIANT‐LINK project (LK0863)Genomic selection and genomewide association study activi-ties were supported by BBSRC grant BBK01711X1 theBBSRC strategic programme grant on Energy Grasses ampBio‐refining BBSEW10963A01 The UK‐led willow RampDwork reported here was supported by BBSRC (BBSEC00005199 BBSEC00005201 BBG0162161 BBE0068331 BBG00580X1 and BBSEC000I0410) Defra(NF0424 NF0426) and the Department of Trade and Indus-try (DTI) (BW6005990000) IT The Brain Gain Program(Rientro dei cervelli) of the Italian Ministry of EducationUniversity and Research supports Antoine Harfouche USContributions by Gerald Tuskan to this manuscript were sup-ported by the Center for Bioenergy Innovation a US
Department of Energy Bioenergy Research Center supportedby the Office of Biological and Environmental Research inthe DOE Office of Science under contract number DE‐AC05‐00OR22725 Willow breeding efforts at CornellUniversity have been supported by grants from the USDepartment of Agriculture National Institute of Food andAgriculture Contributions by the University of Illinois weresupported primarily by the DOE Office of Science Office ofBiological and Environmental Research (BER) grant nosDE‐SC0006634 DE‐SC0012379 and DE‐SC0018420 (Cen-ter for Advanced Bioenergy and Bioproducts Innovation)and the Energy Biosciences Institute EU We would like tofurther acknowledge contributions from the EU projectsldquoOPTIMISCrdquo FP7‐289159 on miscanthus and ldquoWATBIOrdquoFP7‐311929 on poplar and miscanthus as well as ldquoGRACErdquoH2020‐EU326 Biobased Industries Joint Technology Ini-tiative (BBI‐JTI) Project ID 745012 on miscanthus
CONFLICT OF INTEREST
The authors declare that progress reported in this paperwhich includes input from industrial partners is not biasedby their business interests
ORCID
John Clifton‐Brown httporcidorg0000-0001-6477-5452Antoine Harfouche httporcidorg0000-0003-3998-0694Lawrence B Smart httporcidorg0000-0002-7812-7736Danny Awty‐Carroll httporcidorg0000-0001-5855-0775VasileBotnari httpsorcidorg0000-0002-0470-0384Lindsay V Clark httporcidorg0000-0002-3881-9252SalvatoreCosentino httporcidorg0000-0001-7076-8777Lin SHuang httporcidorg0000-0003-3079-326XJon P McCalmont httporcidorg0000-0002-5978-9574Paul Robson httporcidorg0000-0003-1841-3594AnatoliiSandu httporcidorg0000-0002-7900-448XDaniloScordia httporcidorg0000-0002-3822-788XRezaShafiei httporcidorg0000-0003-0891-0730Gail Taylor httporcidorg0000-0001-8470-6390IrisLewandowski httporcidorg0000-0002-0388-4521
REFERENCES
Alburquerque N Baldacci‐Cresp F Baucher M Casacuberta JM Collonnier C ElJaziri M hellip Burgos L (2016) New trans-formation technologies for trees In C Vettori F Gallardo H
CLIFTON‐BROWN ET AL | 27
Haumlggman V Kazana F Migliacci G Pilateamp M Fladung(Eds) Biosafety of forest transgenic trees (pp 31ndash66) DordrechtNetherlands Springer
Argus G W (2007) Salix (Salicaceae) distribution maps and a syn-opsis of their classification in North America north of MexicoHarvard Papers in Botany 12 335ndash368 httpsdoiorg1031001043-4534(2007)12[335SSDMAA]20CO2
Atienza S G Ramirez M C amp Martin A (2003) Mapping‐QTLscontrolling flowering date in Miscanthus sinensis Anderss CerealResearch Communications 31 265ndash271
Atienza S G Satovic Z Petersen K K Dolstra O amp Martin A(2003) Identification of QTLs influencing agronomic traits inMiscanthus sinensis Anderss I Total height flag‐leaf height andstem diameter Theoretical and Applied Genetics 107 123ndash129
Atienza S G Satovic Z Petersen K K Dolstra O amp Martin A(2003) Influencing combustion quality in Miscanthus sinensisAnderss Identification of QTLs for calcium phosphorus and sul-phur content Plant Breeding 122 141ndash145
Bdeir R Muchero W Yordanov Y Tuskan G A Busov V ampGailing O (2017) Quantitative trait locus mapping of Populusbark features and stem diameter BMC Plant Biology 17 224httpsdoiorg101186s12870-017-1166-4
Beard T R Ford G S Koutsky T M amp Spiwak L J (2009) AValley of Death in the innovation sequence An economic investi-gation Research Evaluation 18 343ndash356 httpsdoiorg103152095820209X481057
Berguson W McMahon B amp Riemenschneider D (2017) Addi-tive and non‐additive genetic variances for tree growth in severalhybrid poplar populations and implications regarding breedingstrategy Silvae Genetica 66(1) 33ndash39 httpsdoiorg101515sg-2017-0005
Berlin S Trybush S O Fogelqvist J Gyllenstrand N Halling-baumlck H R Aringhman I hellip Hanley S J (2014) Genetic diversitypopulation structure and phenotypic variation in European Salixviminalis L (Salicaceae) Tree Genetics amp Genomes 10 1595ndash1610 httpsdoiorg101007s11295-014-0782-5
Bernardo R amp Yu J (2007) Prospects for genomewide selectionfor quantitative traits in maize Crop Science 47 1082ndash1090httpsdoiorg102135cropsci2006110690
Biswal A K Atmodjo M A Li M Baxter H L Yoo C GPu Y hellip Mohnen D (2018) Sugar release and growth of bio-fuel crops are improved by downregulation of pectin biosynthesisNature Biotechnology 36 249ndash257
Bmu (2009) National biomass action plan for Germany (p 17) Bun-desministerium fuumlr Umwelt Naturschutz und ReaktorsicherheitBundesministerium fuumlr Ernaumlhrung (BMU) amp Landwirtschaft undVerbraucherschutz (BMELV) 11055 Berlin | Germany Retrievedfrom httpwwwbmeldeSharedDocsDownloadsENPublicationsBiomassActionPlanpdf__blob=publicationFile
Brummer E C (1999) Capturing heterosis in forage crop cultivardevelopment Crop Science 39 943ndash954 httpsdoiorg102135cropsci19990011183X003900040001x
Budsberg E Crawford J T Morgan H Chin W S Bura R ampGustafson R (2016) Hydrocarbon bio‐jet fuel from bioconver-sion of poplar biomass Life cycle assessment Biotechnology forBiofuels 9 170 httpsdoiorg101186s13068-016-0582-2
Burris K P Dlugosz E M Collins A G Stewart C N amp Lena-ghan S C (2016) Development of a rapid low‐cost protoplasttransfection system for switchgrass (Panicum virgatum L) Plant
Cell Reports 35 693ndash704 httpsdoiorg101007s00299-015-1913-7
Casler M (2012) Switchgrass breeding genetics and genomics InA Monti (Ed) Switchgrass (pp 29ndash54) London UK Springer
Casler M Mitchell R amp Vogel K (2012) Switchgrass In CKole C P Joshi amp D R Shonnard (Eds) Handbook of bioen-ergy crop plants Vol 2 New York NY Taylor amp Francis
Casler M D amp Ramstein G P (2018) Breeding for biomass yieldin switchgrass using surrogate measures of yield BioEnergyResearch 11 6ndash12
Casler M D Sosa S Hoffman L Mayton H Ernst C Adler PR hellip Bonos S A (2017) Biomass Yield of Switchgrass Culti-vars under High‐ versus Low‐Input Conditions Crop Science 57821ndash832 httpsdoiorg102135cropsci2016080698
Casler M D amp Vogel K P (2014) Selection for biomass yield inupland lowland and hybrid switchgrass Crop Science 54 626ndash636 httpsdoiorg102135cropsci2013040239
Casler M D Vogel K P amp Harrison M (2015) Switchgrassgermplasm resources Crop Science 55 2463ndash2478 httpsdoiorg102135cropsci2015020076
Casler M D Vogel K P Lee D K Mitchell R B Adler P RSulc R M hellip Moore K J (2018) 30 Years of progress towardincreasing biomass yield of switchgrass and big bluestem CropScience 58 1242
Clark L V Brummer J E Głowacka K Hall M C Heo K PengJ hellip Sacks E J (2014) A footprint of past climate change on thediversity and population structure of Miscanthus sinensis Annals ofBotany 114 97ndash107 httpsdoiorg101093aobmcu084
Clark L V Dzyubenko E Dzyubenko N Bagmet L SabitovA Chebukin P hellip Sacks E J (2016) Ecological characteristicsand in situ genetic associations for yield‐component traits of wildMiscanthus from eastern Russia Annals of Botany 118 941ndash955
Clark L V Jin X Petersen K K Anzoua K G Bagmet LChebukin P hellip Sacks E J(2018) Population structure of Mis-canthus sacchariflorus reveals two major polyploidization eventstetraploid‐mediated unidirectional introgression from diploid Msinensis and diversity centred around the Yellow Sea Annals ofBotany 20 1ndash18 httpsdoiorg101093aobmcy1161
Clark L V Stewart J R Nishiwaki A Toma Y Kjeldsen J BJoslashrgensen U hellip Sacks E J (2015) Genetic structure ofMiscanthus sinensis and Miscanthus sacchariflorus in Japan indi-cates a gradient of bidirectional but asymmetric introgressionJournal of Experimental Botany 66 4213ndash4225
Clifton‐Brown J Hastings A Mos M McCalmont J P Ashman CAwty‐Carroll DhellipCracroft‐EleyW (2017) Progress in upscalingMis-canthus biomass production for the European bio‐economy with seed‐based hybridsGlobal Change Biology Bioenergy 9 6ndash17
Clifton‐Brown J Schwarz K U amp Hastings A (2015) History ofthe development of Miscanthus as a bioenergy crop From smallbeginnings to potential realisation Biology and Environment‐Pro-ceedings of the Royal Irish Academy 115B 45ndash57
Clifton‐Brown J Senior H Purdy S Horsnell R Lankamp BMuumlennekhoff A-K hellip Bentley A R (2018) Investigating thepotential of novel non‐woven fabrics to increase pollination effi-ciency in plant breeding PloS One (Accepted)
Crawford J T Shan CW Budsberg E Morgan H Bura R ampGus-tafson R (2016) Hydrocarbon bio‐jet fuel from bioconversion ofpoplar biomass Techno‐economic assessment Biotechnology forBiofuels 9 141 httpsdoiorg101186s13068-016-0545-7
28 | CLIFTON‐BROWN ET AL
Dalton S (2013) Biotechnology of Miscanthus In S M Jain amp SD Gupta (Eds) Biotechnology of neglected and underutilizedcrops (pp 243ndash294) Dordrecht Netherlands Springer
Davey C L Robson P Hawkins S Farrar K Clifton‐Brown J CDonnison I S amp Slavov G T (2017) Genetic relationshipsbetween spring emergence canopy phenology and biomass yieldincrease the accuracy of genomic prediction in Miscanthus Journalof Experimental Botany 68 5093ndash5102 httpsdoiorg101093jxberx339
David X amp Anderson X (2002) Aspen and larch genetics cooper-ative annual report 13 St Paul MN Department of ForestResources University of Minnesota
Deblock M (1990) Factors influencing the tissue‐culture and theAgrobacterium‐Tumefaciens‐mediated transformation of hybridaspen and poplar clones Plant Physiology 93 1110ndash1116httpsdoiorg101104pp9331110
Defra (2002) The role of future public research investment in thegenetic improvement of UK‐grown crops (p 220) LondonRetrieved from sciencesearchdefragovukDefaultaspxMenu=MenuampModule=MoreampLocation=NoneampCompleted=220ampProjectID=10412
Dewoody J Trewin H amp Taylor G (2015) Genetic and morpho-logical differentiation in Populus nigra L Isolation by coloniza-tion or isolation by adaptation Molecular Ecology 24 2641ndash2655
Dong H Liu S Clark L V Sharma S Gifford J M Juvik JA hellip Sacks E J (2018) Genetic mapping of biomass yield inthree interconnected Miscanthus populations Global Change Biol-ogy Bioenergy 10 165ndash185
Dukes (2017) Digest of UK Energy Statistics (DUKES) (p 264) Lon-don UK Department for Business Energy amp Industrial StrategyRetrieved from wwwgovukgovernmentcollectionsdigest-of-uk-energy-statistics-dukes
Eckenwalder J (1996) Systematics and evolution of Populus In RStettler H Bradshaw J Heilman amp T Hinckley (Eds) Biologyof Populus and its implications for management and conservation(pp 7‐32) Ottawa ON National Research Council of Canada
Elshire R J Glaubitz J C Sun Q Poland J A Kawamoto KBuckler E S amp Mitchell S E (2011) A robust simple geno-typing‐by‐sequencing (GBS) approach for high diversity speciesPloS One 6 e19379
Evans H (2016) Bioenergy crops in the UK Case studies of suc-cessful whole farm integration (p 23) Loughborough UKEnergy Technologies Institute Retrieved from httpswwweticouklibrarybioenergy-crops-in-the-uk-case-studies-on-successful-whole-farm-integration-evidence-pack
Evans H (2017) Increasing UK biomass production through moreproductive use of land (p 7) Loughborough UK Energy Tech-nologies Institute Retrieved from httpswwweticouklibraryan-eti-perspective-increasing-uk-biomass-production-through-more-productive-use-of-land
Evans J Sanciangco M D Lau K H Crisovan E Barry KDaum C hellip Buell C R (2018) Extensive genetic diversity ispresent within North American switchgrass germplasm The PlantGenome 11 1ndash16 httpsdoiorg103835plantgenome2017060055
Evans L M Slavov G T Rodgers‐Melnick E Martin J RanjanP Muchero W hellip DiFazio S P (2014) Population genomicsof Populus trichocarpa identifies signatures of selection and
adaptive trait associations Nature Genetics 46 1089ndash1096httpsdoiorg101038ng3075
Fabio E S Volk T A Miller R O Serapiglia M J Gauch HG Van Rees K C J hellip Smart L B (2017) Genotypetimes envi-ronment interaction analysis of North American shrub willowyield trials confirms superior performance of triploid hybrids Glo-bal Change Biology Bioenergy 9 445ndash459 httpsdoiorg101111gcbb12344
Fahrenkrog A M Neves L G Resende M F R Vazquez A Ide los Campos G Dervinis C hellip Kirst M (2017) Genome‐wide association study reveals putative regulators of bioenergytraits in Populus deltoides New Phytologist 213 799ndash811
Fan D Liu T T Li C F Jiao B Li S Hou Y S amp Luo KM (2015) Efficient CRISPRCas9‐mediated targeted mutagenesisin Populus in the first generation Scientific Reports 5 12217
Feng X P Lourgant K Castric V Saumitou-Laprade P ZhengB S Jiang D M amp Brancourt-Hulmel M (2014) The discov-ery of natural accessions related to Miscanthus x giganteus usingchloroplast DNA Crop Science 54 1645ndash1655
Fillatti J J Sellmer J Mccown B Haissig B amp Comai L(1987) Agrobacterium-mediated transformation and regenerationof Populus Molecular amp General Genetics 206 192ndash199
Fu Y B (2015) Understanding crop genetic diversity under modernplant breeding Theoretical and Applied Genetics 128 2131ndash2142 httpsdoiorg101007s00122-015-2585-y
Fu C Mielenz J R Xiao X Ge Y Hamilton C Y RodriguezM hellip Wang Z‐Y (2011) Genetic manipulation of ligninreduces recalcitrance and improves ethanol production fromswitchgrass Proceedings of the National Academy of Sciences ofthe United States of America 108 3803ndash3808 httpsdoiorg101073pnas1100310108
Gao C (2018) The future of CRISPR technologies in agricultureNature Reviews Molecular Cell Biology 19(5) 275ndash276
Gegas V Gay A Camargo A amp Doonan J (2014) Challengesof Crop Phenomics in the Post-genomic Era In J Hancock (Ed)Phenomics (pp 142ndash171) Boca Raton FL CRC Press
Geraldes A DiFazio S P Slavov G T Ranjan P Muchero WHannemann J hellip Tuskan G A (2013) A 34K SNP genotypingarray for Populus trichocarpa Design application to the study ofnatural populations and transferability to other Populus speciesMolecular Ecology Resources 13 306ndash323
Gifford J M Chae W B Swaminathan K Moose S P amp JuvikJ A (2015) Mapping the genome of Miscanthus sinensis forQTL associated with biomass productivity Global Change Biol-ogy Bioenergy 7 797ndash810
Goggin F L Lorence A amp Topp C N (2015) Applying high‐throughput phenotyping to plant‐insect interactions Picturingmore resistant crops Current Opinion in Insect Science 9 69ndash76httpsdoiorg101016jcois201503002
Grabowski P P Evans J Daum C Deshpande S Barry K WKennedy M hellip Casler M D (2017) Genome‐wide associationswith flowering time in switchgrass using exome‐capture sequenc-ing data New Phytologist 213 154ndash169 httpsdoiorg101111nph14101
Greef J M amp Deuter M (1993) Syntaxonomy of Miscanthus xgiganteus GREEF et DEU Angewandte Botanik 67 87ndash90
Guerra F P Richards J H Fiehn O Famula R Stanton B JShuren R hellip Neale D B (2016) Analysis of the genetic varia-tion in growth ecophysiology and chemical and metabolomic
CLIFTON‐BROWN ET AL | 29
composition of wood of Populus trichocarpa provenances TreeGenetics amp Genomes 12 6 httpsdoiorg101007s11295-015-0965-8
Guo H P Shao R X Hong C T Hu H K Zheng B S ampZhang Q X (2013) Rapid in vitro propagation of bioenergy cropMiscanthus sacchariflorus Applied Mechanics and Materials 260181ndash186
Hallingbaumlck H R Fogelqvist J Powers S J Turrion‐Gomez JRossiter R Amey J hellip Roumlnnberg‐Waumlstljung A‐C (2016)Association mapping in Salix viminalis L (Salicaceae)ndashIdentifica-tion of candidate genes associated with growth and phenologyGlobal Change Biology Bioenergy 8 670ndash685
Hanley S J amp Karp A (2014) Genetic strategies for dissectingcomplex traits in biomass willows (Salix spp) Tree Physiology34 1167ndash1180 httpsdoiorg101093treephystpt089
Harfouche A Meilan R amp Altman A (2011) Tree genetic engi-neering and applications to sustainable forestry and biomass pro-duction Trends in Biotechnology 29 9ndash17 httpsdoiorg101016jtibtech201009003
Harfouche A Meilan R amp Altman A (2014) Molecular and phys-iological responses to abiotic stress in forest trees and their rele-vance to tree improvement Tree Physiology 34 1181ndash1198httpsdoiorg101093treephystpu012
Harfouche A Meilan R Kirst M Morgante M Boerjan WSabatti M amp Mugnozza G S (2012) Accelerating the domesti-cation of forest trees in a changing world Trends in PlantScience 17 64ndash72 httpsdoiorg101016jtplants201111005
Hastings A Clifton‐Brown J Wattenbach M Mitchell C PStampfl P amp Smith P (2009) Future energy potential of Mis-canthus in Europe Global Change Biology Bioenergy 1 180ndash196
Hastings A Mos M Yesufu J AMccalmont J Schwarz KShafei RhellipClifton-Brown J (2017) Economic and environmen-tal assessment of seed and rhizome propagated Miscanthus in theUK Frontiers in Plant Science 8 16 httpsdoiorg103389fpls201701058
Heacuteder M (2017) From NASA to EU The evolution of the TRLscale in Public Sector Innovation The Innovation Journal 22 1ndash23 httpsdoiorg103389fpls201701058
Heffner E L Sorrells M E amp Jannink J L (2009) Genomicselection for crop improvement Crop Science 49 1ndash12 httpsdoiorg102135cropsci2008080512
Hodkinson T R Klaas M Jones M B Prickett R amp Barth S(2015) Miscanthus A case study for the utilization of naturalgenetic variation Plant Genetic Resources‐Characterization andUtilization 13 219ndash237 httpsdoiorg101017S147926211400094X
Hodkinson T R amp Renvoize S (2001) Nomenclature of Miscant-hus xgiganteus (Poaceae) Kew Bulletin 56 759ndash760 httpsdoiorg1023074117709
Houle D Govindaraju D R amp Omholt S (2010) Phenomics Thenext challenge Nature Reviews Genetics 11 855ndash866 httpsdoiorg101038nrg2897
Huang X H Wei X H Sang T Zhao Q Feng Q Zhao Yhellip Han B (2010) Genome‐wide association studies of 14 agro-nomic traits in rice landraces Nature Genetics 42 961ndashU976
Hwang O‐J Cho M‐A Han Y‐J Kim Y‐M Lim S‐H KimD‐S hellip Kim J‐I (2014) Agrobacterium‐mediated genetic trans-formation of Miscanthus sinensis Plant Cell Tissue and OrganCulture (PCTOC) 117 51ndash63
Hwang O‐J Lim S‐H Han Y‐J Shin A‐Y Kim D‐S ampKim J‐I (2014) Phenotypic characterization of transgenic Mis-canthus sinensis plants overexpressing Arabidopsis phytochromeB International Journal of Photoenergy 2014501016
Ilstedt B (1996) Genetics and performance of Belgian poplar clonestested in Sweden International Journal of Forest Genetics 3183ndash195
Ingvarsson P K Garcia M V Hall D Luquez V amp Jansson S(2006) Clinal variation in phyB2 a candidate gene for day‐length‐induced growth cessation and bud set across a latitudinalgradient in European aspen (Populus tremula) Genetics 1721845ndash1853
Ingvarsson P K Garcia M V Luquez V Hall D amp Jansson S(2008) Nucleotide polymorphism and phenotypic associationswithin and around the phytochrome B2 locus in European aspen(Populus tremula Salicaceae) Genetics 178 2217ndash2226 httpsdoiorg101534genetics107082354
Jannink J‐L Lorenz A J amp Iwata H (2010) Genomic selectionin plant breeding From theory to practice Briefings in FunctionalGenomics 9 166ndash177 httpsdoiorg101093bfgpelq001
Jiang J Guan Y Mccormick S Juvik J Lubberstedt T amp FeiS‐Z (2017) Gametophytic self‐incompatibility is operative inMiscanthus sinensis (Poaceae) and is affected by pistil age CropScience 57 1948ndash1956
Jones H D (2015a) Future of breeding by genome editing is in thehands of regulators GM Crops amp Food 6 223ndash232
Jones H D (2015b) Regulatory uncertainty over genome editingNature Plants 1 14011
Kalinina O Nunn C Sanderson R Hastings A F S van der Weijde TOumlzguumlven MhellipClifton‐Brown J C (2017) ExtendingMiscanthus cul-tivation with novel germplasm at six contrasting sites Frontiers in PlantScience 8563 httpsdoiorg103389fpls201700563
Karp A Hanley S J Trybush S O Macalpine W Pei M ampShield I (2011) Genetic improvement of willow for bioenergyand biofuels free access Journal of Integrative Plant Biology 53151ndash165 httpsdoiorg101111j1744-7909201001015x
Kersten B Faivre Rampant P Mader M Le Paslier M‐C Bou-non R Berard A hellip Fladung M (2016) Genome sequences ofPopulus tremula chloroplast and mitochondrion Implications forholistic poplar breeding PloS One 11 e0147209 httpsdoiorg101371journalpone0147209
Kiesel A Nunn C Iqbal Y Van der Weijde T Wagner MOumlzguumlven M hellip Lewandowski I (2017) Site‐specific manage-ment of Miscanthus genotypes for combustion and anaerobicdigestion A comparison of energy yields Frontiers in PlantScience 8 347 httpsdoiorg103389fpls201700347
Kim H Kim S T Ryu J Kang B C Kim J S amp Kim S G(2017) CRISPRCpf1‐mediated DNA‐free plant genome editingNature Communications 8 14406 httpsdoiorg101038ncomms14406
King Z R Bray A L Lafayette P R amp Parrott W A (2014)Biolistic transformation of elite genotypes of switchgrass (Pan-icum virgatum L) Plant Cell Reports 33 313ndash322 httpsdoiorg101007s00299-013-1531-1
Kopp R F Maynard C A De Niella P R Smart L B amp Abra-hamson L P (2002) Collection and storage of pollen from Salix(Salicaceae) American Journal of Botany 89 248ndash252
Krzyżak J Pogrzeba M Rusinowski S Clifton‐Brown J McCal-mont J P Kiesel A hellip Mos M (2017) Heavy metal uptake
30 | CLIFTON‐BROWN ET AL
by novel miscanthus seed‐based hybrids cultivated in heavy metalcontaminated soil Civil and Environmental Engineering Reports26 121ndash132 httpsdoiorg101515ceer-2017-0040
Kuzovkina Y A amp Quigley M F (2005) Willows beyond wet-lands Uses of Salix L species for environmental projects WaterAir and Soil Pollution 162 183ndash204 httpsdoiorg101007s11270-005-6272-5
Lazarus W Headlee W L amp Zalesny R S (2015) Impacts of sup-plyshed‐level differences in productivity and land costs on the eco-nomics of hybrid poplar production in Minnesota USA BioEnergyResearch 8 231ndash248 httpsdoiorg101007s12155-014-9520-y
Lewandowski I (2015) Securing a sustainable biomass supply in agrowing bioeconomy Global Food Security 6 34ndash42 httpsdoiorg101016jgfs201510001
Lewandowski I (2016) The role of perennial biomass crops in agrowing bioeconomy In S Barth D Murphy‐Bokern O Kalin-ina G Taylor amp M Jones (Eds) Perennial biomass crops for aresource‐constrained world (pp 3ndash13) Cham SwitzerlandSpringer Switzerland AG
Li J L Meilan R Ma C Barish M amp Strauss S H (2008)Stability of herbicide resistance over 8 years of coppice in field‐grown genetically engineered poplars Western Journal of AppliedForestry 23 89ndash93
Li R Y amp Qu R D (2011) High throughput Agrobacterium‐medi-ated switchgrass transformation Biomass amp Bioenergy 35 1046ndash1054 httpsdoiorg101016jbiombioe201011025
Lindegaard K N amp Barker J H A (1997) Breeding willows forbiomass Aspects of Applied Biology 49 155ndash162
Linde‐Laursen I B (1993) Cytogenetic analysis of MiscanthusGiganteus an interspecific hybrid Hereditas 119 297ndash300httpsdoiorg101111j1601-5223199300297x
Liu Y Merrick P Zhang Z Z Ji C H Yang B amp Fei S Z(2018) Targeted mutagenesis in tetraploid switchgrass (Panicumvirgatum L) using CRISPRCas9 Plant Biotechnology Journal16 381ndash393
Liu L L Wu Y Q Wang Y W amp Samuels T (2012) A high‐density simple sequence repeat‐based genetic linkage map ofswitchgrass G3 (Bethesda Md) 2 357ndash370
Lovett A Suumlnnenberg G amp Dockerty T (2014) The availabilityof land for perennial energy crops in Great Britain GlobalChange Biology Bioenergy 6 99ndash107 httpsdoiorg101111gcbb12147
Lozano‐Juste J amp Cutler S R (2014) Plant genome engineering infull bloom Trends in Plant Science 19 284ndash287 httpsdoiorg101016jtplants201402014
Lu F Lipka A E Glaubitz J Elshire R Cherney J H CaslerM D hellip Costich D E (2013) Switchgrass Genomic DiversityPloidy and Evolution Novel Insights from a Network‐Based SNPDiscovery Protocol Plos Genetics 9 e1003215
Ludovisi R Tauro F Salvati R Khoury S Mugnozza G S ampHarfouche A (2017) UAV‐based thermal imaging for high‐throughput field phenotyping of black poplar response to droughtFrontiers in Plant Science 81681
Macalpine W Shield I amp Karp A (2010) Seed to near market vari-ety the BEGIN willow breeding pipeline 2003ndash2010 and beyond InBioten conference proceedings Birmingham pp 21ndash23
Macalpine W J Shield I F Trybush S O Hayes C M ampKarp A (2008) Overcoming barriers to crossing in willow (Salixspp) breeding Aspects of Applied Biology 90 173ndash180
Mahfouz M M (2017) The efficient tool CRISPR‐Cpf1 NaturePlants 3 17028
Malinowska M Donnison I S amp Robson P R (2017) Phenomicsanalysis of drought responses in Miscanthus collected from differ-ent geographical locations Global Change Biology Bioenergy 978ndash91
Maroder H L Prego I A Facciuto G R amp Maldonado S B(2000) Storage behaviour of Salix alba and Salix matsudanaseeds Annals of Botany 86 1017ndash1021 httpsdoiorg101006anbo20001265
Martinez‐Reyna J amp Vogel K (2002) Incompatibility systems inswitchgrass Crop Science 42 1800ndash1805 httpsdoiorg102135cropsci20021800
Maurya J P amp Bhalerao R P (2017) Photoperiod‐and tempera-ture‐mediated control of growth cessation and dormancy in treesA molecular perspective Annals of Botany 120 351ndash360httpsdoiorg101093aobmcx061
McCalmont J P Hastings A Mcnamara N P Richter G MRobson P Donnison I S amp Clifton‐Brown J (2017) Environ-mental costs and benefits of growing Miscanthus for bioenergy inthe UK Global Change Biology Bioenergy 9 489ndash507
McCracken A R amp Dawson W M (1997) Using mixtures of wil-low clones as a means of controlling rust disease Aspects ofApplied Biology 49 97ndash103
McKown A D Klaacutepště J Guy R D Geraldes A Porth I Hanne-mann JhellipDouglas C J (2014) Genome‐wide association implicatesnumerous genes underlying ecological trait variation in natural popula-tions ofPopulus trichocarpaNew Phytologist 203 535ndash553
Merrick P amp Fei S Z (2015) Plant regeneration and genetic transforma-tion in switchgrass ndash A review Journal of Integrative Agriculture 14483ndash493 httpsdoiorg101016S2095-3119(14)60921-7
Moreno‐Mateos M A Fernandez J P Rouet R Lane M AVejnar C E Mis E hellip Giraldez A J (2017) CRISPR‐Cpf1mediates efficient homology‐directed repair and temperature‐con-trolled genome editing Nature Communications 8 2024 httpsdoiorg101038s41467-017-01836-2
Mosseler A (1990) Hybrid performance and species crossabilityrelationships in willows (Salix) Canadian Journal of Botany 682329ndash2338
Muchero W Guo J DiFazio S P Chen J‐G Ranjan P Slavov GT hellip Tuskan G A (2015) High‐resolution genetic mapping ofallelic variants associated with cell wall chemistry in Populus BMCGenomics 16 24 httpsdoiorg101186s12864-015-1215-z
Neale D B amp Kremer A (2011) Forest tree genomics Growingresources and applications Nature Reviews Genetics 12 111ndash122 httpsdoiorg101038nrg2931
Nunn C Hastings A F S J Kalinina O Oumlzguumlven M SchuumlleH Tarakanov I G hellip Clifton‐Brown J C (2017) Environmen-tal influences on the growing season duration and ripening ofdiverse Miscanthus germplasm grown in six countries Frontiersin Plant Science 8 907 httpsdoiorg103389fpls201700907
Ogawa Y Honda M Kondo Y amp Hara‐Nishimura I (2016) Anefficient Agrobacterium‐mediated transformation method forswitchgrass genotypes using Type I callus Plant Biotechnology33 19ndash26
Ogawa Y Shirakawa M Koumoto Y Honda M Asami Y KondoY ampHara‐Nishimura I (2014) A simple and reliable multi‐gene trans-formation method for switchgrass Plant Cell Reports 33 1161ndash1172httpsdoiorg101007s00299-014-1605-8
CLIFTON‐BROWN ET AL | 31
Okada M Lanzatella C Saha M C Bouton J Wu R L amp Tobias CM (2010) Complete switchgrass genetic maps reveal subgenomecollinearity preferential pairing and multilocus interactions Genetics185 745ndash760 httpsdoiorg101534genetics110113910
Palomo‐Riacuteos E Macalpine W Shield I Amey J Karaoğlu C WestJ hellip Jones H D (2015) Efficient method for rapid multiplication ofclean and healthy willow clones via in vitro propagation with broadgenotype applicability Canadian Journal of Forest Research 451662ndash1667 httpsdoiorg101139cjfr-2015-0055
Pilate G Allona I Boerjan W Deacutejardin A Fladung M Gal-lardo F hellip Halpin C (2016) Lessons from 25 years of GM treefield trials in Europe and prospects for the future In C Vettori FGallardo H Haumlggman V Kazana F Migliacci G Pilateamp MFladung (Eds) Biosafety of forest transgenic trees (pp 67ndash100)Dordrecht Netherlands Springer Switzerland AG
Pinosio S Giacomello S Faivre-Rampant P Taylor G Jorge VLe Paslier M C amp Marroni F (2016) Characterization of thepoplar pan-genome by genome-wide identification of structuralvariation Molecular Biology and Evolution 33 2706ndash2719
Porth I Klaacutepště J Skyba O Friedmann M C Hannemann JEhlting J hellip Douglas C J (2013) Network analysis reveals therelationship among wood properties gene expression levels andgenotypes of natural Populus trichocarpa accessions New Phytol-ogist 200 727ndash742
Pugesgaard S Schelde K Larsen S U Laeligrke P E amp JoslashrgensenU (2015) Comparing annual and perennial crops for bioenergyproductionndashinfluence on nitrate leaching and energy balance Glo-bal Change Biology Bioenergy 7 1136ndash1149 httpsdoiorg101111gcbb12215
Quinn L D Allen D J amp Stewart J R (2010) Invasivenesspotential of Miscanthus sinensis Implications for bioenergy pro-duction in the United States Global Change Biology Bioenergy2 310ndash320 httpsdoiorg101111j1757-1707201001062x
Rae A M Robinson K M Street N R amp Taylor G (2004)Morphological and physiological traits influencing biomass pro-ductivity in short‐rotation coppice poplar Canadian Journal ofForest Research‐Revue Canadienne De Recherche Forestiere 341488ndash1498 httpsdoiorg101139x04-033
Rambaud C Arnoult S Bluteau A Mansard M C Blassiau Camp Brancourt‐Hulmel M (2013) Shoot organogenesis in threeMiscanthus species and evaluation for genetic uniformity usingAFLP analysis Plant Cell Tissue and Organ Culture (PCTOC)113 437ndash448 httpsdoiorg101007s11240-012-0284-9
Ramstein G P Evans J Kaeppler S M Mitchell R B Vogel K PBuell C R amp Casler M D (2016) Accuracy of genomic prediction inswitchgrass (Panicum virgatum L) improved by accounting for linkagedisequilibriumG3 (Bethesda Md) 6 1049ndash1062
Rayburn A L Crawford J Rayburn C M amp Juvik J A (2009)Genome size of three Miscanthus species Plant Molecular Biol-ogy Reporter 27 184 httpsdoiorg101007s11105-008-0070-3
Richardson J Isebrands J amp Ball J (2014) Ecology and physiol-ogy of poplars and willows In J Isebrands amp J Richardson(Eds) Poplars and willows Trees for society and the environ-ment (pp 92‐123) Boston MA and Rome CABI and FAO
Robison T L Rousseau R J amp Zhang J (2006) Biomass produc-tivity improvement for eastern cottonwood Biomass amp Bioenergy30 735ndash739 httpsdoiorg101016jbiombioe200601012
Robson P R Farrar K Gay A P Jensen E F Clifton‐Brown JC amp Donnison I S (2013) Variation in canopy duration in the
perennial biofuel crop Miscanthus reveals complex associationswith yield Journal of Experimental Botany 64 2373ndash2383
Robson P Jensen E Hawkins S White S R Kenobi K Clif-ton‐Brown J hellip Farrar K (2013) Accelerating the domestica-tion of a bioenergy crop Identifying and modelling morphologicaltargets for sustainable yield increase in Miscanthus Journal ofExperimental Botany 64 4143ndash4155
Sanderson M A Adler P R Boateng A A Casler M D ampSarath G (2006) Switchgrass as a biofuels feedstock in theUSA Canadian Journal of Plant Science 86 1315ndash1325httpsdoiorg104141P06-136
Scarlat N Dallemand J F Monforti‐Ferrario F amp Nita V(2015) The role of biomass and bioenergy in a future bioecon-omy Policies and facts Environmental Development 15 3ndash34httpsdoiorg101016jenvdev201503006
Serapiglia M J Gouker F E amp Smart L B (2014) Early selec-tion of novel triploid hybrids of shrub willow with improved bio-mass yield relative to diploids BMC Plant Biology 14 74httpsdoiorg1011861471-2229-14-74
Serba D Wu L Daverdin G Bahri B A Wang X Kilian Ahellip Devos K M (2013) Linkage maps of lowland and upland tet-raploid switchgrass ecotypes BioEnergy Research 6 953ndash965httpsdoiorg101007s12155-013-9315-6
Shakoor N Lee S amp Mockler T C (2017) High throughput phe-notyping to accelerate crop breeding and monitoring of diseases inthe field Current Opinion in Plant Biology 38 184ndash192 httpsdoiorg101016jpbi201705006
Shield I Macalpine W Hanley S amp Karp A (2015) Breedingwillow for short rotation coppice energy cropping In Industrialcrops Breeding for BioEnergy and Bioproducts (pp 67ndash80) NewYork NY Springer
Sjodin A Street N R Sandberg G Gustafsson P amp Jansson S (2009)The Populus Genome Integrative Explorer (PopGenIE) A new resourcefor exploring the Populus genomeNewPhytologist 182 1013ndash1025
Slavov G amp Davey C (2017) Integrated genomic prediction inbioenergy crops In Abstracts from the International Conferenceon Developing biomass crops for future climates pp 34 24ndash27September 2017 Oriel College Oxford wwwwatbioeu
Slavov G Davey C Bosch M Robson P Donnison I ampMackay I (2018a) Genomic index selection provides a pragmaticframework for setting and refining multi‐objective breeding targetsin Miscanthus Annals of Botany (in press)
Slavov G T DiFazio S P Martin J Schackwitz W MucheroW Rodgers‐Melnick E hellip Tuskan G A (2012) Genome rese-quencing reveals multiscale geographic structure and extensivelinkage disequilibrium in the forest tree Populus trichocarpa NewPhytologist 196 713ndash725
Slavov G Davey C Robson P Donnison I amp Mackay I(2018b) Domestication of Miscanthus through genomic indexselection In Plant and Animal Genome Conference 13 January2018 San Diego CA Retrieved from httpspagconfexcompagxxvimeetingappcgiPaper30622
Slavov G T Nipper R Robson P Farrar K Allison G GBosch M hellip Jensen E (2013) Genome‐wide association studiesand prediction of 17 traits related to phenology biomass and cellwall composition in the energy grass Miscanthus sinensis NewPhytologist 201 1227ndash1239
Slavov G Robson P Jensen E Hodgson E Farrar K AllisonG hellip Donnison I (2014) Contrasting geographic patterns of
32 | CLIFTON‐BROWN ET AL
genetic variation for molecular markers vs phenotypic traits in theenergy grass Miscanthus sinensis Global Change Biology Bioen-ergy 5 562ndash571
Ślusarkiewicz‐Jarzina A Ponitka A Cerazy‐Waliszewska J Woj-ciechowicz M K Sobańska K Jeżowski S amp Pniewski T(2017) Effective and simple in vitro regeneration system of Mis-canthus sinensis Mtimes giganteus and M sacchariflorus for plant-ing and biotechnology purposes Biomass and Bioenergy 107219ndash226 httpsdoiorg101016jbiombioe201710012
Smart L amp Cameron K (2012) Shrub willow In C Kole S Joshiamp D Shonnard (Eds) Handbook of bioenergy crop plants (pp687ndash708) Boca Raton FL Taylor and Francis Group
Stanton B J (2014) The domestication and conservation of Populusand Salix genetic resources (Chapter 4) In Isebrands J ampRichardson J (Eds) Poplars and willows Trees for society andthe environment (pp 124ndash200) Rome Italy CAB InternationalFood and Agricultural Organization of the United Nations
Stanton B J Neale D B amp Li S (2010) Populus breeding Fromthe classical to the genomic approach In S Jansson R Bha-lerao amp A T Groover (Eds) Genetics and genomics of popu-lus (pp 309ndash348) New York NY Springer
Stewart J R Toma Y Fernandez F G Nishiwaki A Yamada T ampBollero G (2009) The ecology and agronomy ofMiscanthus sinensisa species important to bioenergy crop development in its native range inJapan A reviewGlobal Change Biology Bioenergy 1 126ndash153
Stott K G (1992) Willows in the service of man Proceedings of the RoyalSociety of Edinburgh Section B Biological Sciences 98 169ndash182
Strauss S H Ma C Ault K amp Klocko A L (2016) Lessonsfrom two decades of field trials with genetically modified trees inthe USA Biology and regulatory compliance In C Vettori FGallardo H Haumlggman V Kazana F Migliacci G Pilate amp MFladung (Eds) Biosafety of forest transgenic trees (pp 101ndash124)Dordrecht Netherlands Springer
Suda Y amp Argus G W (1968) Chromosome numbers of some NorthAmerican Salix Brittonia 20 191ndash197 httpsdoiorg1023072805440
Tan B (2018) Genomic selection and genome‐wide association studies todissect quantitative traits in forest trees Umearing Sweden University
Tornqvist C Taylor M Jiang Y Evans J Buell C KaepplerS amp Casler M (2018) Quantitative trait locus mapping for flow-ering time in a lowland x upland switchgrass pseudo‐F2 popula-tion The Plant Genome 11 170093
Toacuteth G Hermann T Da Silva M amp Montanarella L (2016)Heavy metals in agricultural soils of the European Union withimplications for food safety Environment International 88 299ndash309 httpsdoiorg101016jenvint201512017
Tracy W Dawson J Moore V amp Fisch J (2016) Intellectual propertyrights and public plant breeding Recommendations and proceedings ofa conference on best practices for intellectual property protection of pub-lically developed plant germplasm (p 70) University of Wisconsin‐Madison Retrieved from httpshostcalswisceduagronomywp-contentuploadssites1620162005Proceedings-IPR-Finalpdf
Trybush S Jahodovaacute Š Macalpine W amp Karp A (2008) A geneticstudy of a Salix germplasm resource reveals new insights into rela-tionships among subgenera sections and species BioEnergyResearch 1 67ndash79 httpsdoiorg101007s12155-008-9007-9
Tsai C J (2013) Next‐generation sequencing for next‐generationbreeding and more New Phytologist 198 635ndash637 httpsdoiorg101111nph12245
Tuskan G A Difazio S Jansson S Bohlmann J Grigoriev I HellstenU hellip Rokhsar D (2006) The genome of black cottonwood Populustrichocarpa (Torr ampGray) Science 313 1596ndash1604
Tuskan G A amp Walsh M E (2001) Short‐rotation woody cropsystems atmospheric carbon dioxide and carbon management AUS case study The Forestry Chronicle 77 259ndash264 httpsdoiorg105558tfc77259-2
Van Den Broek R Treffers D‐J Meeusen M Van Wijk ANieuwlaar E amp Turkenburg W (2001) Green energy or organicfood A life‐cycle assessment comparing two uses of set‐asideland Journal of Industrial Ecology 5 65ndash87 httpsdoiorg101162108819801760049477
Van Der Schoot J Pospiskova M Vosman B amp Smulders M J M(2000) Development and characterization of microsatellite markers inblack poplar (Populus nigra L) Theoretical and Applied Genetics 101317ndash322 httpsdoiorg101007s001220051485
Van Der Weijde T Dolstra O Visser R G amp Trindade L M(2017a) Stability of cell wall composition and saccharificationefficiency in Miscanthus across diverse environments Frontiers inPlant Science 7 2004
Van der Weijde T Huxley L M Hawkins S Sembiring E HFarrar K Dolstra O hellip Trindade L M (2017) Impact ofdrought stress on growth and quality of miscanthus for biofuelproduction Global Change Biology Bioenergy 9 770ndash782
Van der Weijde T Kamei C L A Severing E I Torres A FGomez L D Dolstra O hellip Trindade L M (2017) Geneticcomplexity of miscanthus cell wall composition and biomass qual-ity for biofuels BMC Genomics 18 406
Van der Weijde T Kiesel A Iqbal Y Muylle H Dolstra O Visser RG FhellipTrindade LM (2017) Evaluation ofMiscanthus sinensis bio-mass quality as feedstock for conversion into different bioenergy prod-uctsGlobal Change Biology Bioenergy 9 176ndash190
Vanholme B Cesarino I Goeminne G Kim H Marroni F VanAcker R hellip Boerjan W (2013) Breeding with rare defectivealleles (BRDA) A natural Populus nigra HCT mutant with modi-fied lignin as a case study New Phytologist 198 765ndash776
Vogel K P Mitchell R B Casler M D amp Sarath G (2014)Registration of Liberty Switchgrass Journal of Plant Registra-tions 8 242ndash247 httpsdoiorg103198jpr2013120076crc
Wang X Yamada T Kong F‐J Abe Y Hoshino Y Sato Hhellip Yamada T (2011) Establishment of an efficient in vitro cul-ture and particle bombardment‐mediated transformation systems inMiscanthus sinensis Anderss a potential bioenergy crop GlobalChange Biology Bioenergy 3 322ndash332 httpsdoiorg101111j1757-1707201101090x
Watrud L S Lee E H Fairbrother A Burdick C Reichman JR Bollman M hellip Van de Water P K (2004) Evidence forlandscape‐level pollen‐mediated gene flow from genetically modi-fied creeping bentgrass with CP4 EPSPS as a marker Proceedingsof the National Academy of Sciences of the United States of Amer-ica 101 14533ndash14538
Weisgerber H (1993) Poplar breeding for the purpose of biomassproduction in short‐rotation periods in Germany ndash Problems andfirst findings Forestry Chronicle 69 727ndash729 httpsdoiorg105558tfc69727-6
Wheeler N C Steiner K C Schlarbaum S E amp Neale D B(2015) The evolution of forest genetics and tree improvementresearch in the United States Journal of Forestry 113 500ndash510httpsdoiorg105849jof14-120
CLIFTON‐BROWN ET AL | 33
Whittaker C Macalpine W J Yates N E amp Shield I (2016)Dry matter losses and methane emissions during wood chip stor-age The impact on full life cycle greenhouse gas savings of shortrotation coppice willow for heat BioEnergy Research 9 820ndash835 httpsdoiorg101007s12155-016-9728-0
Xi Q (2000) Investigation on the distribution and potential of giant grassesin China PhD thesis Goettingen Germany Cuvillier Verlag
Xi Q amp Jezowkski S (2004) Plant resources of Triarrhena andMiscanthus species in China and its meaning for Europe PlantBreeding and Seed Science 49 63ndash77
Xue S Kalinina O amp Lewandowski I (2015) Present and future optionsfor the improvement of Miscanthus propagation techniques Renewableand Sustainable Energy Reviews 49 1233ndash1246
Yin K Q Gao C X amp Qiu J L (2017) Progress and prospectsin plant genome editing Nature Plants 3 httpsdoiorg101038nplants2017107
YookM (2016)Genetic diversity and transcriptome analysis for salt toler-ance inMiscanthus PhD thesis Seoul National University Korea
Zaidi S S E A Mahfouz M M amp Mansoor S (2017) CRISPR‐Cpf1 A new tool for plant genome editing Trends in PlantScience 22 550ndash553 httpsdoiorg101016jtplants201705001
Zalesny R S Stanturf J A Gardiner E S Perdue J H YoungT M Coyle D R hellip Hass A (2016) Ecosystem services ofwoody crop production systems BioEnergy Research 9 465ndash491 httpsdoiorg101007s12155-016-9737-z
Zhang Q X Sun Y Hu H K Chen B Hong C T Guo H Phellip Zheng B S (2012) Micropropagation and plant regenerationfrom embryogenic callus of Miscanthus sinensis Vitro Cellular ampDevelopmental Biology‐Plant 48 50ndash57 httpsdoiorg101007s11627-011-9387-y
Zhang Y Zalapa J E Jakubowski A R Price D L AcharyaA Wei Y hellip Casler M D (2011) Post‐glacial evolution of
Panicum virgatum Centers of diversity and gene pools revealedby SSR markers and cpDNA sequences Genetica 139 933ndash948httpsdoiorg101007s10709-011-9597-6
Zhao H Wang B He J Yang J Pan L Sun D amp Peng J(2013) Genetic diversity and population structure of Miscanthussinensis germplasm in China PloS One 8 e75672
Zhou X H Jacobs T B Xue L J Harding S A amp Tsai C J(2015) Exploiting SNPs for biallelic CRISPR mutations in theoutcrossing woody perennial Populus reveals 4‐coumarate CoAligase specificity and redundancy New Phytologist 208 298ndash301
Zhou R Macaya‐Sanz D Rodgers‐Melnick E Carlson C HGouker F E Evans L M hellip DiFazio S P (2018) Characteri-zation of a large sex determination region in Salix purpurea L(Salicaceae) Molecular Genetics and Genomics 1ndash16 httpsdoiorg101007s00438-018-1473-y
Zsuffa L Mosseler A amp Raj Y (1984) Prospects for interspecifichybridization in willow for biomass production In K L Perttu(Ed) Ecology and management of forest biomass production sys-tems Report 15 (pp 261ndash281) Uppsala Sweden SwedishUniversity of Agricultural Sciences
How to cite this article Clifton‐Brown J HarfoucheA Casler MD et al Breeding progress andpreparedness for mass‐scale deployment of perenniallignocellulosic biomass crops switchgrassmiscanthus willow and poplar GCB Bioenergy201800000ndash000 httpsdoiorg101111gcbb12566
34 | CLIFTON‐BROWN ET AL
Graphical AbstractThe contents of this page will be used as part of the graphical abstract of html only
It will not be published as part of main article
Plant breeding links the research effort with commercial mass upscaling The authorsrsquo assessment of development status ofthe four species is shown (poplar having two one for short rotation coppice (SRC) poplar and one for the more traditionalshort rotation forestry (SRF)) Mass scale deployment needs developments outside the breeding arenas to drive breedingactivities more rapidly and extensively
TABLE1
Breeding‐relatedattributes
forfour
leadingperennialbiom
asscrops(PBCs)
Species
Switc
hgrass
Miscanthu
sW
illow
Pop
lar
Typ
eC4mdash
Grass
C4mdash
Grass
C3mdash
SRC
C3mdash
SRCSRF
Sourcesof
indigeno
usgerm
plasm
CAato
MXeastof
Rocky
Mou
ntains
Eastern
AsiaandOceania
Predom
inantly
Northernhemisph
ere
Northernhemisph
ere
Breedingsystem
Mon
oeciou
sou
tcrossing
Mon
oeciou
sou
tcrossing
Dioeciousou
tcrossing
Dioeciousou
tcrossing
Ploidy
4times8
times2times
3times
4times
2timesminus12
times2times
Specieswith
ingenu
s~4
50~1
4~4
00~3
0ndash32
Typ
esused
mainlyfor
breeding
US
Low
land
ecotyp
e(sub
trop
ical
clim
ates)andup
land
ecotyp
e(tem
perate
clim
ates)
EUb JPSK
andUS
MsinandMsac
EUS
viminalistimesschw
erinii
Sda
syclad
ostimesrehd
eriana
S
dasyclad
osS
viminalis
USandUKS
viminalistimesmiyab
eana
S
miyab
eana
US
Spu
rpurea
timesmiyab
eana
S
purpurea
Pop
ulus
tricho
carpa
Pdelto
ides
Pnigra
Psuaveolens
subsp
maximow
icziiPba
lsam
ifera
Palba
andhy
brids
Typ
ical
haploid
geno
mesize
(Mbp
)~1
500
Msin~5
400
andMsac~4
400
(Raybu
rnCrawfordRaybu
rnamp
Juvik
2009
)
~450
~485
plusmn10
Breedingprog
rammes
CA20
00(REAP
Quebec)
US 19
92(N
ebraska
19
92(O
klahom
a)
1992
(Georgia)
19
96(W
isconsin)
19
96(Sou
thDakota)
20
00(Tennessee)
20
02(M
ississippi)
20
07(O
klahom
a)
2008
(RutgersNew
Jersey)
20
12(Cornell
New
York)
20
12(U
rbana‐Champaign
Illin
ois)
DE19
90s(K
lein‐W
anzleben)
NL20
00s(W
ageningen)
UK20
04(A
berystwyth)
CN20
06(Chang
sha)
JP20
06(H
okkaido)
US 20
06(Californiamdash
CERESInc)
20
06(CaliforniaandIndianamdash
MBI)
20
08(U
rbana‐Champaign)
SK20
09(Suw
on‐SNU)and(M
uan
NICSof
RDA)
FR20
11(Estreacutees‐M
ons)
UK19
80s(Lon
gAshton
relocatedto
Rothamsted
Researchin
2002
)SE
19
80s(SvaloumlfWeibu
llSalix
energi
Europ
aAB)
US
1990
s(Cornell
New
York)
PL20
00s(O
lsztyn
University
ofWarmia
andMazury)
SE
19
39(M
ykingeEkebo
Research
Institu
te)
19
90s(U
ppsalaSL
U)
20
10(U
ppsalaST
T)
US 19
27(N
ewYork
OxfordPaper
Com
pany
andNew
YorkBotanical
Garden
Wheeler
etal20
15)
19
79(W
ashing
ton
UW
andOrego
nGWR)
19
80ndash1
995(M
ississippiMSS
tate
and
GWR)
19
96(M
innesotaUMD
NRRIand
GWR)
IT19
83(Piedm
ontAFV
)FR
20
01(O
rleacuteans
ampNancyIN
RA
Charrey‐sur‐Saocircne
ampPierroton
FCBA
Nog
ent‐sur‐V
ernisson
IRST
EA)
DE20
08(G
oumlttin
gen
NW‐FVA)
(Con
tinues)
4 | CLIFTON‐BROWN ET AL
TABLE
1(Contin
ued)
Species
Switc
hgrass
Miscanthu
sW
illow
Pop
lar
Current
commercial
varietieson
the
market
US
Nocommercial
hybrids
CA2
EU1(M
timesgfrom
different
origins)
+selected
Msinforthatching
inDK
US
3(but
2aregenetically
identical)
UK25
US
8EUDElt10
FR44
IT10ndash1
5SE
~1
4US
8ndash12
Southeast8ndash
14Upp
erMidwest10
PacificNorthwest
Precom
mercial
cultivars
expected
tobe
onthemarketin
3years
US
36registered
cultivars
(halfare
rand
omseed
increasesfrom
natural
prairiesandhalfarebred
varieties)
Mostarepu
blic
releasesfew
are
protected
patented
orlicensed
NL8
seeded
hybridsvanDinterSemo
MTA
UK4
seeded
hybridsCERES(Land
OLakes)andTerravestaLtdMTA
US
Non
eFR
Non
e
EU53
registered
with
CPV
OforPB
RUK20
registered
with
CPV
OforPB
RUS
18clon
esfrom
Cornellin
multisite
trials
CAun
quantified
UAlberta
andQuebec
FR8ndash
12SR
Fclon
eslicence‐based
IT5SR
Cand6SR
FAFV
MTA
orlicence
SE14
STTlicence‐based
andMTA
US
~19So
utheast5ndash7Upp
erMidwest
from
UMD
NRRIMTA6ndash12
North
Central
from
GWRMTAPacific
Northwest8clon
esGWRMTA24
inmultilocationyieldtrialsGWR
Com
mercial
yield
(tDM
haminus1year
minus1 )
US
3ndash18
EU8ndash
12CN20
ndash30
US
10ndash25
EU7ndash20
UK8ndash
14US
8ndash14
EU5ndash20
(SEandBaltic
Cou
ntries8ndash
12)
US
10ndash2
2(10ndash
12North
Central12
ndash16
Southeast15ndash2
2PacificNortheast)
Harvestrotatio
nand
commercial
stand
lifespan
Ann
ualfor10
ndash12years
Ann
ualfor10ndash2
5years(M
sac
hasbeen
used
for~3
0yearsin
China)
2‐to
4‐year
cyclefor22
ndash30years
SRC3‐
to7‐year
cyclefor20
years
SRF
10‐to12‐yearcycleforgt50
years
Adaptiverang
eOpen‐po
llinatedandsynthetic
cultivars
arelim
itedin
adaptatio
nby
temperature
andprecipitatio
n(~8breeding
zonesin
USandCA)
Standard
Mtimesgiswidelyadaptedin
EU
butislim
itedin
theUnitedStates
byinsufficient
winterh
ardiness
forN
orthern
Midwestand
heatintolerancein
the
southNovelMsactimesMsinhyb
rids
andMsintimesMsinhybrids
selected
incontinentalD
Ehave
show
nawide
adaptiv
erang
ein
EU(K
alininaetal
2017
)Ongoing
trialson
heavymetal
contam
inated
soils
indicatetoleranceby
exclusion(K
rzyżak
etal20
17)
Different
hybridsareneeded
fordifferent
zones
Besthy
bridsshow
someG
timesE(U
S)
andsomehy
bridslow
GtimesE(Fabio
etal20
17)
Different
hybridsareneeded
ford
ifferent
clim
aticzonesHyb
rids
thatareadapted
forg
rowingseason
sof
~6mon
thsand
relativ
elyshortd
aysin
Southern
Europ
earemaladaptedto
shortg
rowing
season
sof
~4mon
thsandrelativ
ely
long
days
inNorthernEU
The
mostb
roadly
adaptedvarietiescome
from
thePcana
densistaxo
n
Note
AFV
AlasiaFranco
VivaiC
ornell
CornellUniversity
CPV
OC
ommunity
PlantV
ariety
OfficeFC
BAF
orestCelluloseW
ood
ConstructionandFu
rnitu
reTechnologyInstitu
teG
timesEg
enotype‐by
‐environmentinterac-
tion
GWRG
reenWoodResourcesINRAF
renchNationalInstituteforA
griculturalR
esearch
IRST
EAN
ationalR
esearchUnito
fScience
andTechnologyforE
nvironmentand
AgricultureM
sacMsacchariflo
rusandM
timesg
(Mtimes
giganteus)M
sin
MiscanthussinensisM
BIMendelB
iotechnology
IncM
bpm
egabase
pairM
SStateM
ississippi
StateUniversity
MTAm
aterialtransferagreem
entNICS
NationalInstituteof
CropScience
NW‐
FVANorthwestGerman
ForestResearchInstitu
tePB
Rplantbreeders
rightRDARural
DevelopmentAdm
inistration
REAP
ResourceEfficient
Agriculture
Productio
nSL
USw
edishUniversity
ofAgriculturalSciences
SNUS
eoul
NationalU
niSR
Cshort‐ro
tatio
ncoppice
SRF
short‐rotationforestryS
TTS
weT
reeTech
tDM
haminus1year
minus1 tons
ofdrymatterperhectareperyearU
AlbertaUniversity
ofAlbertaU
MD
NRRIUniversity
ofMinnesotaDuluthsNaturalResources
ResearchInstitu
teU
MNU
niversity
ofMinnesotaU
WU
niversity
ofWashington
UWMU
niversity
ofWarmiaandMazury
a ISO
Alpha‐2
lettercountrycodes
b EU
isused
forEurope
CLIFTON‐BROWN ET AL | 5
2011) Genotype‐by‐environment interactions (G times E) arestrong and must be considered in breeding (Casler 2012Casler Mitchell amp Vogel 2012) Adaptation to environ-ment is regulated principally by responses to day‐lengthand temperature There are also strong genotype times environ-ment interactions between the drier western regions and thewetter eastern regions (Casler et al 2017)
The growing regions of North America are divided intofour adaptation zones for switchgrass each roughly corre-sponding to two official hardiness zones The lowland eco-types are generally late flowering high yielding andadapted to warmer climates but have lower drought andcold resistance than upland ecotypes (Casler 2012 Casleret al 2012)
In 2015 the US Department of Agriculture (USDA)National Plant Germplasm System GRIN (httpswwwars-gringovnpgs) had 181 switchgrass accessions of whichonly 96 were available for distribution due to limitationsassociated with seed multiplication (Casler Vogel amp Har-rison 2015) There are well over 2000 additional uncata-logued accessions (Table 1) held by various universities butthe USDA access to these is also constrained by the effortneeded in seed multiplication Switchgrass is a model herba-ceous species for conducting scientific research on biomass(Sanderson Adler Boateng Casler amp Sarath 2006) but lit-tle funding is available for the critical prebreeding work thatis necessary to link this biological research to commercialbreeding More than a million genotypes from ~2000 acces-sions (seed accessions contain many genotypes) have beenphenotypically screened in spaced plant nurseries and tenthousand of the most useful have been genotyped with differ-ent technologies depending on the technology available at
the time when these were performed From these character-ized genotypes parents are selected for exploratory pairwisecrosses to produce synthetic populations within ecotypesSwitchgrass like many grasses is outcrossing due to astrong genetically controlled self‐incompatibly (akin to theS‐Z‐locus system of other grasses (Martinez‐Reyna ampVogel 2002)) Thus the normal breeding approaches usedare F1 wide crosses and recurrent selection cycles within syn-thetic populations
The scale of these programmes varies from small‐scaleconventional breeding based solely on phenotypic selec-tion (eg REAP Canada Montreal Quebec) to large pro-grammes incorporating modern molecular breedingmethods (eg USDA‐ARS Madison Wisconsin) Earlyagronomic research and biomass production efforts werefocused on the seed‐based multiplication of promising wildaccessions from natural prairies Cultivars Alamo Kanlowand Cave‐in‐Rock were popular due to high yield andmoderate‐to‐wide adaptation Conventional breedingapproaches focussed on biomass production traits and haveled to the development of five cultivars particularly suitedto biomass production Cimarron EG1101 EG1102EG2101 and Liberty The first four of these represent thelowland ecotype and were developed either in Oklahomaor Georgia Liberty is a derivative of lowland times uplandhybrids developed in Nebraska following selection for lateflowering the high yield of the lowland ecotype and coldtolerance of the upland ecotype (Vogel et al 2014) Thesefive cultivars were all approximately 25ndash30 years in themaking counting from the initiation of these breeding pro-grammes Many more biomass‐type cultivars are expectedwithin the next few years as these and other breeding
TABLE 2 Generalized improvement targets for perennial biomass crops (PBCs)
Net energy yield per hectare
Increased yield
Reduced moisture content at harvest
Physical and chemical composition for different end‐use applications
Increased lignin content and decreased corrosive elements for thermal conversion
Reduced recalcitrance through decreased lignin content andor modified lignin monomer composition to reduce pretreatment requirementsfor next‐generation biofuels by saccharification and fermentation
Plant morphological differences which influence biomass harvest transport and storage (eg stem thickness)
Propagation costs
Improved cloning systems (trees and grasses)
Seed systems (grasses)
Optimizing agronomy for each new cultivar
Resilience through enhanced
Abiotic stress toleranceresistance (eg drought salinity and high and low temperature )
Biotic stress resistance (eg insects fungal bacterial and viral diseases)
Site adaptability especially to those of marginalcontaminated agricultural land
6 | CLIFTON‐BROWN ET AL
TABLE3
Preparedness
formassupscaling
currentmarketvalueandresearch
investmentin
four
perennialbiom
asscrops(PBCs)
Species
Switc
hgrass
Miscanthu
sW
illow
Pop
lar
Current
commercial
plantin
gcostsperha
USa20
0USD
bin
the
establishm
entyear
DE3
375
Eurob
(XueK
alininaamp
Lew
ando
wski20
15)
UK2
153
GBPb
(Evans201
6)redu
cedto
116
9GBPwith
Mxg
rhizom
ewhere
thefarm
erdo
estheland
preparation(w
wwterravestacom
)US
173
0ndash222
5USD
UK150
0ndash173
9GBPplus
land
preparation(Evans20
16)
US
197
6USD
(=80
0USD
acre)
IT110
0Euro
FRSE100
0ndash200
0Euro
US
863USD
ha(LazarusHeadleeamp
Zalesny
20
15)
Current
marketvalue
ofthebiom
asspert
DM
US
80ndash1
00USD
UK~8
0GBP(bales)(Terravesta
person
alcommun
ication)
US
94USD
(chipp
ed)
UK49
41GBP(chipp
ed)(Evans20
16)
US
55ndash7
0USD
IT10
0Euro
US
80ndash90USD
deliv
ered
basedon
40milesof
haulage
Scienceforg
enetic
improv
ementprojects
inthelast10
years
USgt50
projectsfund
edby
USDOEand
USD
ANIFA
CNgt
30projectsfund
edby
CN‐N
SFCandMBI
EUc 6projects(O
PTIM
ISCO
PTIM
A
WATBIO
SUNLIBBF
IBRAandGRACE)
FRB
FFSK
2projectsfund
edby
IPETandPM
BC
US
EBICABBImultip
leDOEFeedstock
Genom
icsandUSD
AAFR
Iprojects
UKB
SBECB
EGIN
(200
3ndash20
10)
US
USDOEJG
IGenom
eSequ
encing
(200
9ndash20
122
015ndash
2018
)USD
ANortheastSu
nGrant
(200
9ndash20
12)
USD
ANIFANEWBio
Con
sortium
(201
2ndash20
17)USD
ANIFAWillow
SkyC
AP(201
8ndash20
21)
EUF
P7(EnergyPo
plarN
ovelTreeWATBIO
Tree4Fu
ture)
SEC
limate‐adaptedpo
plarsandNanow
ood
SLU
andST
TUS
Bioenergy
Science(O
RNL)USD
Afeedstocks
geno
micsprog
rammeBRDIa
ndBRC‐CBI
Major
projects
supp
ortin
gcrossing
andselectioncycles
inthelast10
years
USandCA12
projects
NLRUEmiscanthu
sPP
P20
15ndash2
019
50ndash
100K
Euroyear
UKGIA
NT‐LIN
KPP
I20
11ndash2
016
13M
GBPyear
US
EBICABBI~0
5M
USD
year
UKBEGIN
2000
ndash201
0US
USD
ANortheastSu
nGrant
(200
9ndash20
12)USD
ANIFA
NEWBio
Con
sortium
(201
2ndash20
17)USD
ANIFA
Willow
SkyC
AP(201
8ndash20
21)
FR1
4region
alandnatio
nalp
rojects10
0ndash15
0k
Euroyear
SES
TTo
nebreeding
effortas
aninternalproject
in20
10with
aresulting
prog
enytrial
US
DOESu
nGrant
feedstockdevelopm
ent
partnershipprog
ramme(including
Willow
)
Current
annu
alinvestmentin
projects
fortranslationinto
commercial
hybrids
USgt20
MUSD
year
UKMUST
201
6ndash20
1905M
GBPyear
US
CABBIUSD
AAFR
I20
17ndash2
02205M
USD
year
US
USD
ANIFA
NEWBio
~04
MUSD
year
FRScienceforim
prov
ement~2
0kEuroyear
US
USD
Aun
derAFR
ICo‐ordinatedAg
Prod
ucer
projectsAnd
severaltranslational
geno
micsprojectswith
outpu
blic
fund
ing
Upscalin
gtim
e(years)
toproducesufficient
propagules
toplantgt10
0ha
US
Using
seed
2ndash3years
UKRhizomefor1ndash200
0hectares
canbe
ready
in6mon
ths
UKNL2
0ha
ofseed
hybridsplantedin
2018
In
2019
sufficientseedfor5
0ndash10
0ha
isexpected
UK3yearsusingconv
entio
nalcutting
sfaster
usingmicroprop
agation
FRIT3yearsby
vegetativ
eprop
agation
SEgt3yearsby
vegetativ
eprop
agation(cuttin
gs)
and3yearsby
microprop
agation
Note
AFR
IAgriculture
andFo
odResearchInitiativein
theUnitedStatesBEGIN
BiomassforEnergyGenetic
Improvem
entNetwork
BFF
BiomassFo
rtheFu
tureBRC‐CBI
Bioenergy
ResearchCentre‐Centrefor
Bioenergy
Innovatio
nBRDIBiomassresearch
developm
entinitiative
BSB
ECBBSR
CSu
stainableBioenergy
Centre
CABBICenterforadvanced
bioenergyandbioproductsinnovatio
nin
theUnitedStatesCN‐N
SFC
Natural
ScienceFo
undatio
nof
ChinaDOEDepartm
entof
Energyin
theUnitedStatesEBIEnergyBiosciences
Institu
teFIBRAFibrecropsas
sustainablesource
ofbiobased
materialforindustrial
productsin
Europeand
ChinaFP
7SeventhFram
eworkProgrammein
theEUGIA
NT‐LIN
KGenetic
improvem
entof
miscanthusas
asustainablefeedstockforbioenergyin
theUnitedKingdom
GRACEGRow
ingAdvancedindustrial
Crops
onmarginallandsforbiorEfineriesIPETKorea
Institu
teof
Planning
andEvaluationforTechnologyin
FoodA
gricultureFo
restry
andFisheriesJG
IJointGenom
eInstitu
teMBIMendelBiotechnology
IncMUST
Miscant-
husUpS
calin
gTechnology
NEWBioNortheast
WoodyW
arm‐seasonBiomassConsortiumNIFANationalInstitu
teof
Food
andAgriculture
intheUnitedStatesNovelTree
Novel
tree
breeding
strategiesOPT
IMAOpti-
mizationof
perennialgrassesforbiom
assproductio
nin
theMediterraneanareaOPT
IMISCOptim
izingbioenergyproductio
nfrom
Miscanthus
ORNLOak
Ridge
NationalLabPM
BCPlantMolecular
BreedingCenterof
theNextGenerationBiogreenResearchCenters
oftheRepublic
ofKoreaPP
Ipublicndashp
rivate
investmentPP
Ppublicndashp
rivate
partnership
RUEradiationuseefficiencySk
yCAP
Coordinated
AgriculturalProjectSL
U
SwedishUniversity
ofAgriculturalSciencesST
TSw
eTreeTech
SUNLIBBSu
stainableLiquidBiofuelsfrom
BiomassBiorefiningTree4Fu
tureDesigning
Trees
forthefutureUSD
AUSDepartm
entof
Agriculture
WATBIO
Developmentof
improved
perennialnonfoodbiom
assandbioproduct
cropsforwater
stressed
environm
ents
a ISO
Alpha‐2
lettercountrycodes
b Local
currencies
areused
asat
2018Euro
GBP(G
reat
Britain
Pound)USD
(USDollar)c EU
isused
forEurope
CLIFTON‐BROWN ET AL | 7
TABLE4
Prebreedingresearch
andthestatus
ofconventio
nalbreeding
infour
leadingperennialbiom
asscrops(PBCs)
Breeding
techno
logy
step
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sWillow
Poplar
1Collected
wild
accessions
or
secondarysources
availablefor
breeding
Provideabroadbase
of
useful
traits
Respect
forCBD
and
Nagoyaprotocol
on
collections
after2014
Not
allindigenous
genetic
resourcesareaccessible
under
CBD
forpolitical
reasons
USa~2
000
(181
inofficial
GRIN
gene
bank
ofwhich
96
wereavailable
others
in
variousunofficial
collections)
CNChangsha
~1000
andNanjin
g
2000from
China
DK~1
20from
JP
FRworking
collectionof
~100
(mainlyM
sin)
JP~1
500
from
JPS
KandCN
NLworking
collectionof
~300
Msin
SK~7
00from
SKC
NJP
andRU
UK~1
500
accessions
(from
~500
sites)
ofwhich
1000arein
field
nurseriesin
2018
US
14in
GRIN
global
US
Illin
ois~1
500
collections
made
inAsiawith
25
currently
available
inUnitedStates
forbreeding
UK1500with
about20
incommon
with
the
UnitedStates
US
350largelyunique
FR3370Populus
delto
ides
(650)Pnigra(2000)
Ptrichocarpa(600)and
Pmaximow
iczii(120)
IT30000P
delto
idsPnigra
Pmaximow
icziiand
Ptrichocarpa
SE13000
accessionsmainly
Ptrichocarpa
byST
TandSL
U
US
GWR150Ptrichocarpaand
300Pmaximow
icziiaccessions
forPacificNorthwestprogram
535Pdelto
ides
and~2
00
Pnigraaccessions
for
Southeastprogrammeand77
Psimoniifrom
Mongolia
UMD
NRRI550+
clonal
accessions
(primarily
Ptimescanadnesisa
long
with
Pdelto
ides
andPnigra
parents)
forUpper
Midwest
program
2Wild
accessions
which
have
undergone
phenotypic
screeningin
field
trials
Selectionof
parental
lines
with
useful
traits
Strong
partnerships
torun
multilocationfieldtrials
tophenotypeconsistently
theaccessionsgenotypes
indifferentenvironm
ents
anddatabases
Costof
runningmultilocation
US
gt500000genotypes
CN~1
250
inChangsha
~1700
in
HunanJiangsuS
handong
Hainan
NL~250genotypes
JP~1
200
SK~4
00
UK‐le
d~1
000
genotypesin
a
replicated
trial
US‐led
~1200
genotypesin
multi‐
locatio
nreplicated
trialsin
Asiaand
North
America
MsinSK
CNJP
CA
(Ontario)US(Illinoisand
Colorado)MsacSK
(Kangw
on)
CN
(Zhejiang)JP
(Sapporo)CA
(Onatario)U
S(Illinois)DK
(Aarhus)
UKgt400
US
180
FR2720
IT10000
SE~1
50
US
GWR~1
500
wild
Ptrichocarpaphenotypes
and
OPseedlots(total
of70)from
thePmaximow
icziispecies
complex
(suaveolens
cathayana
koreana
ussuriensismaximow
iczii)and
2000ndash3000Pdelto
ides
collections
(based
on104s‐
generatio
nfamilies)
UMD
NRRIscreened
seedlin
gsfrom
aPdelto
ides
(OPor
CP)
breeding
programme(1996ndash2016)
gt7200OPseedlin
gsof
PnigraunderMinnesota
clim
atic
conditions(improve
parental
populatio
n)
3Wild
germ
plasm
genotyping
Amanaged
living
collectionof
clonal
types
US
~20000genotypes
CNChangsha
~1000Nanjin
g37
FR44
bycpDNA
(Fenget
al
UK~4
00
US
225
(Con
tinues)
8 | CLIFTON‐BROWN ET AL
TABLE
4(Contin
ued)
Breeding
techno
logy
step
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sWillow
Poplar
Construct
phylogenetic
treesanddissim
ilarity
indices
The
type
ofmolecular
analysis
mdashAFL
PcpDNAR
ADseq
whole
genomesequencing
2014)
JP~1
200
NL250genotypesby
RADseq
UK~1
000
byRADseq
SK~3
00
US
Illin
oisRADseqon
allwild
accessions
FR2310
SE150
US
1500
4Exploratory
crossing
and
progenytests
Discovergood
parental
combinatio
nsmdashgeneral
combining
ability
Geographicseparatio
n
speciesor
phylogenetic
trees
Costly
long‐te
rmmultilocation
trialsforprogeny
US
gt10000
CN~1
0
FR~2
00
JP~3
0
NL3600
SK~2
0
UK~4
000
(~1500Msin)
US
500ndash1000
UK gt700since2003
gt800EWBP(1996ndash
2002)
US
800
FRCloned13
times13
factorial
matingdesign
with
3species[10
Pdelto
ides8Ptrichocarpa
8
Pnigra)
US
GWR1000exploratory
crossingsof
Pfrem
ontii
Psimoniiforim
proved
droughttolerance
UMD
NRRIcrossednativ
e
Pdelto
ides
with
non‐nativ
e
Ptrichocarpaand
Pmaximow
icziiMajority
of
progenypopulatio
nssuffered
from
clim
atic
andor
pathogenic
susceptib
ility
issues
5Wideintraspecies
hybridization
Discovergood
parental
combinatio
nsmdashgeneral
combining
ability
1to
4above
inform
atics
Flow
eringsynchronizationand
low
seed
set
US
~200
CN~1
0
NL1800
SK~1
0
UK~5
00
UK276
US
400
IT500
SE120
US
50ndash100
6With
inspecies
recurrentselection
Concentratio
nof
positiv
e
traits
Identificationof
theright
heterotic
groups
insteps
1ndash5
Difficultto
introducenew
germ
plasm
with
outdilutio
n
ofbesttraits
US
gt40
populatio
ns
undergoing
recurrentselection
NL2populatio
ns
UK6populatio
ns
US
~12populatio
nsto
beestablished
by2019
UK4
US
150
FRPdelto
ides
(gt30
FSfamilies
ndash900clones)Pnigra(gt40
FS
families
ndash1600clones)and
Ptrichocarpa(15FS
Families
minus350clones)
IT80
SEca50
parent
breeding
populatio
nswith
in
Ptrichocarpa
US
Goalsis~1
00parent
breeding
populatio
nswith
inPtrichocarpa
Pdelto
ides
PnigraandPmaximow
iczii
7Interspecies
hybrid
breeding
Com
bining
complem
entary
traitsto
producegood
morphotypes
and
heterosiseffects
Allthesteps1ndash6
Inearlystage
improvem
ents
areunpredictable
Awide
base
iscostly
tomanage
None
CNChangsha
3Nanjin
g
120MsintimesMflo
r
JP~5
with
Saccharum
spontaneum
SK5
UK420
US
250
FRPcanadensis(1800)
Pdelto
ides
timesPtrichocarpa
(800)and
PtrichocarpatimesPmaximow
iczii
(1200)
(Con
tinues)
CLIFTON‐BROWN ET AL | 9
TABLE
4(Contin
ued)
Breeding
techno
logy
step
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sWillow
Poplar
IT~1
20
US
525genotypesin
eastside
hybrid
programme
205westside
hybrid
programmeand~1
200
in
SoutheastPdelto
ides
program
8Chrom
osom
e
doublin
g
Arouteto
triploid
seeded
typesdoublin
gadiploid
parent
ofknow
n
breeding
valuefrom
recurrentselection
Needs
1ndash7to
help
identify
therightparental
lines
Doubled
plantsarenotorious
forrevertingto
diploid
US
~50plantstakenfrom
tetraploid
tooctaploid
which
arenow
infieldtrials
CN~1
0Msactriploids
DKSeveralattemptsin
mid
90s
FR5genotypesof
Msin
UK2Msin1Mflora
nd1Msac
US
~30
UKAttempted
butnot
routinelyused
US
Not
partof
theregular
breeding
becausethere
existnaturalpolyploids
NA
9DoubleHaploids
Arouteto
producing
homogeneous
progeny
andalso
forthe
introductio
nof
transgenes
orforgenome
editing
Needs
1ndash7to
help
identify
therightparental
lines
Fully
homozygousplantsare
weakandeasily
die
andmay
notflow
ersynchronously
US
Noneyetattempteddueto
poor
vigour
andviability
of
haploids
CN2genotypesof
Msin
andMflor
UKDihaploidsof
MsinMflor
and
Msacand2transgenic
Msinvia
anther
cultu
re
US
6createdandused
toidentify
miss‐identifying
paralogueloci
(causedby
recent
genome
duplicationin
miscanthus)Usedin
creatin
gthereferencegenomefor
MsinVeryweakplantsand
difficultto
retain
thelin
es
NA
NA
10Embryo
rescue
Anattractiv
etechniquefor
recovering
plantsfrom
sexual
crosses
The
majority
ofem
bryos
cannot
survivein
vivo
or
becomelong‐timedorm
ant
None
UKone3times
hybrid
UKEmbryosrescue
protocol
proved
robustat
8days
post‐pollin
ation
FRAnem
bryo
rescue
protocol
proved
robustandim
proved
hybridizationsuccessfor
Pdelto
ides
timesPtrichocarpa
crosses
11Po
llenstorage
Flexibility
tocross
interestingparents
with
outneed
of
flow
ering
synchronization
None
Nosuccessto
daten
otoriously
difficultwith
grassesincluding
sugarcane
Freshpollencommonly
used
incrossing
Pollenstoragein
some
interspecificcrosses
FRBothstored
(cryobank)
and
freshindividual
pollen
Note
AFL
Pam
plifiedfragmentlength
polymorphismCBDconventio
non
biological
diversity
CP
controlledpollinatio
nCpD
NAchloroplastDNAEWBP
Europeanwillow
breeding
programme
FSfull‐sib
GtimesE
genotype‐by‐environm
entGRIN
germ
plasm
resourcesinform
ationnetwork
GWRGreenWoodResourcesMflorMiscanthusflo
ridulusMsacMsacchariflo
rus
MsinMsinensisNAnotapplicableNRRINatural
Resources
ResearchInstitu
teOP
open
pollinatio
nRADseq
restrictionsite‐associatedDNA
sequencingSL
USw
edishUniversity
ofAgriculturalSciencesST
TSw
eTreeTech
UMDUniversity
ofMinnesota
Duluth
a ISO
Alpha‐2
lettercountrycodes
10 | CLIFTON‐BROWN ET AL
programmes mature The average rate of gain for biomassyield in long‐term switchgrass breeding programmes hasbeen 1ndash4 per year depending on ecotype populationand location of the breeding programme (Casler amp Vogel2014 Casler et al 2018) The hybrid derivative Libertyhas a biomass yield 43 higher than the better of its twoparents (Casler amp Vogel 2014 Vogel et al 2014) Thedevelopment of cold‐tolerant and late‐flowering lowland‐ecotype populations for the northern United States hasincreased biomass yields by 27 (Casler et al 2018)
Currently more than 20 recurrent selection populationsare being managed in the United States to select parentsfor improved yield yield resilience and compositional qual-ity of the biomass For the agronomic development andupscaling high seed multiplication rates need to be com-bined with lower seed dormancy to reduce both crop estab-lishment costs and risks Expresso is the first cultivar withsignificantly reduced seed dormancy which is the first steptowards development of domesticated populations (Casleret al 2015) Most phenotypic traits of interest to breeders
require a minimum of 2 years to be fully expressed whichresults in a breeding cycle that is at least two years Morecomplicated breeding programmes or traits that requiremore time to evaluate can extend the breeding cycle to 4ndash8 years per generation for example progeny testing forbiomass yield or field‐based selection for cold toleranceBreeding for a range of traits with such long cycles callsfor the development of molecular methods to reduce time-scales and improve breeding efficiency
Two association panels of switchgrass have been pheno-typically and genotypically characterized to identify quanti-tative trait loci (QTLs) that control important biomasstraits The northern panel consists of 60 populationsapproximately 65 from the upland ecotype The southernpanel consists of 48 populations approximately 65 fromthe lowland ecotype Numerous QTLs have been identifiedwithin the northern panel to date (Grabowski et al 2017)Both panels are the subject of additional studies focused onbiomass quality flowering and phenology and cold toler-ance Additionally numerous linkage maps have been
FIGURE 1 Cumulative minimum years needed for the conventional breeding cycle through the steps from wild germplasm to thecommercial hybrids in switchgrass miscanthus willow and poplar Information links between the steps are indicated by dotted arrows andhighlight the importance of long‐term informatics to maximize breeding gain
CLIFTON‐BROWN ET AL | 11
created by the pairwise crossing of individuals with diver-gent characteristics often to generate four‐way crosses thatare analysed as pseudo‐F2 crosses (Liu Wu Wang ampSamuels 2012 Okada et al 2010 Serba et al 2013Tornqvist et al 2018) Individual markers and QTLs iden-tified can be used to design marker‐assisted selection(MAS) programmes to accelerate breeding and increase itsefficiency Genomic prediction and selection (GS) holdseven more promise with the potential to double or triplethe rate of gain for biomass yield and other highly complexquantitative traits of switchgrass (Casler amp Ramstein 2018Ramstein et al 2016) The genome of switchgrass hasrecently been made public through the Joint Genome Insti-tute (httpsphytozomejgidoegov)
Transgenic approaches have been heavily relied upon togenerate unique genetic variants principally for traitsrelated to biomass quality (Merrick amp Fei 2015) Switch-grass is highly transformable using either Agrobacterium‐mediated transformation or biolistics bombardment butregeneration of plants is the bottleneck to these systemsTraditionally plants from the cultivar Alamo were the onlyregenerable genotypes but recent efforts have begun toidentify more genotypes from different populations that arecapable of both transformation and subsequent regeneration(King Bray Lafayette amp Parrott 2014 Li amp Qu 2011Ogawa et al 2014 Ogawa Honda Kondo amp Hara‐Nishi-mura 2016) Cell wall recalcitrance and improved sugarrelease are the most common targets for modification (Bis-wal et al 2018 Fu et al 2011) Transgenic approacheshave the potential to provide traits that cannot be bredusing natural genetic variability However they will stillrequire about 10ndash15 years and will cost $70ndash100 millionfor cultivar development and deployment (Harfouche Mei-lan amp Altman 2011) In addition there is commercialuncertainty due to the significant costs and unpredictabletimescales and outcomes of the regulatory approval processin the countries targeted for seed sales As seen in maizeone advantage of transgenic approaches is that they caneasily be incorporated into F1 hybrid cultivars (Casler2012 Casler et al 2012) but this does not decrease thetime required for cultivar development due to field evalua-tion and seed multiplication requirements
The potential impacts of unintentional gene flow andestablishment of non‐native transgene sequences in nativeprairie species via cross‐pollination are also major issues forthe environmental risk assessment These limit further thecommercialization of varieties made using these technolo-gies Although there is active research into switchgrass steril-ity mechanisms to curb unintended pollen‐mediated genetransfer it is likely that the first transgenic cultivars proposedfor release in the United States will be met with considerableopposition due to the potential for pollen flow to remainingwild prairie sites which account for lt1 of the original
prairie land area and are highly protected by various govern-mental and nongovernmental organizations (Casler et al2015) Evidence for landscape‐level pollen‐mediated geneflow from genetically modified Agrostis seed multiplicationfields (over a mountain range) to pollinate wild relatives(Watrud et al 2004) confirms the challenge of using trans-genic approaches Looking ahead genome editing technolo-gies hold considerable promise for creating targeted changesin phenotype (Burris Dlugosz Collins Stewart amp Lena-ghan 2016 Liu et al 2018) and at least in some jurisdic-tions it is likely that cultivars resulting from gene editingwill not need the same regulatory approval as GMOs (Jones2015a) However in July 2018 the European Court of Justice(ECJ) ruled that cultivars carrying mutations resulting fromgene editing should be regulated in the same way as GMOsThe ECJ ruled that such cultivars be distinguished from thosearising from untargeted mutation breeding which isexempted from regulation under Directive 200118EC
3 | MISCANTHUS
Miscanthus is indigenous to eastern Asia and Oceania whereit is traditionally used for forage thatching and papermaking(Xi 2000 Xi amp Jezowkski 2004) In the 1960s the highbiomass potential of a Japanese genotype introduced to Eur-ope by Danish nurseryman Aksel Olsen in 1935 was firstrecognized in Denmark (Linde‐Laursen 1993) Later thisaccession was characterized described and named asldquoM times giganteusrdquo (Greef amp Deuter 1993 Hodkinson ampRenvoize 2001) commonly abbreviated as Mxg It is a natu-rally occurring interspecies triploid hybrid between tetraploidM sacchariflorus (2n = 4x) and diploid M sinensis(2n = 2x) Despite its favourable agronomic characteristicsand ability to produce high yields in a wide range of environ-ments in Europe (Kalinina et al 2017) the risks of relianceon it as a single clone have been recognized Miscanthuslike switchgrass is outcrossing due to self‐incompatibility(Jiang et al 2017) Thus seeded hybrids are an option forcommercial breeding Miscanthus can also be vegetativelypropagated by rhizome or in vitro culture which allows thedevelopment of clones The breeding approaches are usuallybased on F1 crosses and recurrent selection cycles within thesynthetic populations There are several breeding pro-grammes that target improvement of miscanthus traitsincluding stress resilience targeted regional adaptation agro-nomic ldquoscalabilityrdquo through cheaper propagation fasterestablishment lower moisture and ash contents and greaterusable yield (Clifton‐Brown et al 2017)
Germplasm collections specifically to support breedingfor biomass started in the late 1980s and early 1990s inDenmark Germany and the United Kingdom (Clifton‐Brown Schwarz amp Hastings 2015) These collectionshave continued with successive expeditions from European
12 | CLIFTON‐BROWN ET AL
and US teams assembling diverse collections from a widegeographic range in eastern Asia including from ChinaJapan South Korea Russia and Taiwan (Hodkinson KlaasJones Prickett amp Barth 2015 Stewart et al 2009) Threekey miscanthus species for biomass production are M si-nensis M floridulus and M sacchariflorus M sinensis iswidely distributed throughout eastern Asia with an adap-tive range from the subtropics to southern Russia (Zhaoet al 2013) This species has small rhizomes and producesmany tightly packed shoots forming a ldquotuftrdquo M floridulushas a more southerly adaptive range with a rather similarmorphology to M sinensis but grows taller with thickerstems and is evergreen and less cold‐tolerant than the othermiscanthus species M sacchariflorus is the most northern‐adapted species ranging to 50 degN in eastern Russia (Clarket al 2016) Populations of diploid and tetraploid M sac-chariflorus are found in China (Xi 2000) and South Korea(Yook 2016) and eastern Russia but only tetraploids havebeen found in Japan (Clark Jin amp Petersen 2018)
Germplasm has been assembled from multiple collec-tions over the last century though some early collectionsare poorly documented This historical germplasm has beenused to initiate breeding programmes largely based on phe-notypic and genotypic characterization As many of theaccessions from these collections are ldquoorigin unknownrdquocrucial environmental envelope data are not available UK‐led expeditions started in 2006 and continued until 2011with European and Asian partners and have built up a com-prehensive collection of 1500 accessions from 500 sitesacross Eastern Asia including China Japan South Koreaand Taiwan These collections were guided using spatialclimatic data to identify variation in abiotic stress toleranceAccessions from these recent collections were planted fol-lowing quarantine in multilocation nursery trials at severallocations in Europe to examine trait expression in differentenvironments Based on the resulting phenotypic andmolecular marker data several studies (a) characterized pat-terns of population genetic structure (Slavov et al 20132014) (b) evaluated the statistical power of genomewideassociation studies (GWASs) and identified preliminarymarkerndashtrait associations (Slavov et al 2013 2014) and(c) assessed the potential of genomic prediction (Daveyet al 2017 Slavov et al 2014 2018b) Genomic indexselection in particular offers the possibility of exploringscenarios for different locations or industrial markets (Sla-vov et al 2018a 2018b)
Separately US‐led expeditions also collected about1500 accessions between 2010 and 2014 (Clark et al2014 2016 2018 2015) A comprehensive genetic analy-sis of the population structure has been produced by RAD-seq for M sinensis (Clark et al 2015 Van der WeijdeKamei et al 2017) and M sacchariflorus (Clark et al2018) Multilocation replicated field trials have also been
conducted on these materials in North America and inAsia GWAS has been conducted for both M sinensis anda subset of M sacchariflorus accessions (Clark et al2016) To date about 75 of these recent US‐led collec-tions are in nursery trials outside the United States Due tolengthy US quarantine procedures these are not yet avail-able for breeding in the United States However molecularanalyses have allowed us to identify and prioritize sets ofgenotypes that best encompass the genetic variation in eachspecies
While mostM sinensis accessions flower in northern Eur-ope very few M sacchariflorus accessions flower even inheated glasshouses For this reason the European pro-grammes in the United Kingdom the Netherlands and Francehave performed mainly M sinensis (intraspecies) hybridiza-tions (Table 4) Selected progeny become the parents of latergenerations (recurrent selection as in switchgrass) Seed setsof up to 400 seed per panicle occur inM sinensis In Aberyst-wyth and Illinois significant efforts to induce synchronousflowering in M sacchariflorus and M sinensis have beenmade because interspecies hybrids have proven higher yieldperformance and wide adaptability (Kalinina et al 2017) Ininterspecies pairwise crosses in glasshouses breathable bagsandor large crossing tubes or chambers in which two or morewhole plants fit are used for pollination control Seed sets arelower in bags than in the open air because bags restrict pollenmovement while increasing temperatures and reducinghumidity (Clifton‐Brown Senior amp Purdy 2018) About30 of attempted crosses produced 10 to 60 seeds per baggedpanicle The seed (thousand seed mass ranges from 05 to09 g) is threshed from the inflorescences and sown into mod-ular trays to produce plug plants which are then planted infield nurseries to identify key parental combinations
A breeding programme of this scale must serve theneeds of different environments accepting the commonpurpose is to optimize the interception of solar radiationAn ideal hybrid for a given environment combines adapta-tion to date of emergence with optimization of traits suchas height number of stems per plant flowering and senes-cence time to optimize solar interception to produce a highbiomass yield with low moisture content at harvest (Rob-son Farrar et al 2013 Robson Jensen et al 2013) By20132014 conventional breeding in Europe had producedintra‐ and interspecific fertile seeded hybrids When acohort (typically about 5) of outstanding crosses have beenidentified it is important to work on related upscaling mat-ters in parallel These are as follows
Assessment of the yield and critical traits in selectedhybrids using a network of field trials
Efficient cloning of the seed parents While in vitro andmacro‐cloning techniques are used some genotypes areamenable to neither technique
CLIFTON‐BROWN ET AL | 13
TABLE5
Status
ofmod
ernplantbreeding
techniqu
esin
four
leadingperennialbiom
asscrop
s(PBCs)
Breeding
techno
logy
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sW
illow
Pop
lar
MAS
Use
ofmarker
sequ
encesthat
correlatewith
atrait
allowingearly
prog
enyselectionor
rejectionlocatin
gkn
owngenesfor
useful
traits(suchas
height)from
other
speciesin
your
crop
Breeding
prog
ramme
relevant
biparental
crosses(Table
4)to
create
ldquomapping
popu
latio
nsrdquoand
QTLidentification
andestim
ationto
identifyrobu
stmarkers
alternatively
acomprehensive
panelof
unrelated
geno
typesfor
GWAS
Ineffectivewhere
traits
areaffected
bymany
geneswith
small
effects
US
Literally
dozens
ofstud
iesto
identify
SNPs
andmarkers
ofinterestNoattempts
atMASas
yetlargely
dueto
popu
latio
nspecificity
CNS
SRISS
Rmarkers
forM
sinMsacand
Mflor
FR2
Msin
mapping
popu
latio
ns(BFF
project)
NL2
Msin
mapping
popu
latio
ns(A
tienza
Ram
irezetal
2003
AtienzaSatovic
PetersenD
olstraamp
Martin
200
3Van
Der
Weijdeetal20
17a
Van
derW
eijde
Hux
ley
etal20
17
Van
derW
eijdeKam
ei
etal20
17V
ander
WeijdeKieseletal
2017
)SK
2GWASpanels(1
collectionof
Msin1
diploidbiparentalF 1
popu
latio
n)in
the
pipelin
eUK1
2families
tofind
QTLin
common
indifferentfam
ilies
ofthe
sameanddifferent
speciesandhy
brids
US
2largeGWAS
panels4
diploidbi‐
parentalF 1
popu
latio
ns
1F 2
popu
latio
n(add
ition
alin
pipelin
e)
UK16
families
for
QTLdiscov
ery
2crossesMASscreened
1po
tentialvariety
selected
US
9families
geno
typedby
GBS(8
areF 1o
neisF 2)a
numberof
prom
ising
QTLdevelopm
entof
markers
inprog
ress
FR(INRA)tested
inlargeFS
families
and
infactorialmating
design
low
efficiency
dueto
family
specificity
US
49families
GS
Metho
dto
accelerate
breeding
throug
hredu
cing
the
resourcesforcross
attemptsby
Aldquotraining
popu
latio
nrdquoof
individu
alsfrom
stages
abov
ethat
have
been
both
Risks
ofpo
orpredictio
nProg
eny
testingneedsto
becontinueddu
ring
the
timethetraining
setis
US
One
prog
rammeso
farmdash
USD
Ain
WisconsinThree
cycles
ofgeno
mic
selectioncompleted
CN~1
000
geno
types
NLprelim
inary
mod
elsforMsin
developed
UK~1
000
geno
types
Verysuitable
application
not
attempted
yet
Maybe
challeng
ingfor
interspecies
hybrids
FR(INRA)120
0geno
type
ofPop
ulus
nigraarebeingused
todevelop
intraspecificGS
(Con
tinues)
14 | CLIFTON‐BROWN ET AL
TABLE
5(Contin
ued)
Breeding
techno
logy
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sW
illow
Pop
lar
predictin
gthe
performance
ofprog
enyof
crosses
(and
inresearch
topredictbestparents
tousein
biparental
crosses)
geno
typedand
phenotyp
edto
developamod
elthattakesgeno
typic
datafrom
aldquocandidate
popu
latio
nrdquoof
untested
individu
als
andprod
uces
GEBVs(Jannink
Lorenzamp
Iwata
2010
)
beingph
enotyp
edand
geno
typed
Inthis
time
next‐generation
germ
plasm
andreadyto
beginthe
second
roun
dof
training
and
recalib
ratio
nof
geno
mic
predictio
nmod
els
US
Prelim
inary
mod
elsforMsac
and
Msin
developed
Work
isun
derw
ayto
deal
moreeffectivelywith
polyploidgenetics
calib
ratio
nsUS
125
0geno
types
arebeingused
todevelopGS
calib
ratio
ns
Traditio
nal
transgenesis
Efficient
introd
uctio
nof
ldquoforeign
rdquotraits
(possiblyfrom
other
genu
sspecies)
into
anelite
plant(ega
prov
enparent
ora
hybrid)that
needsa
simpletraitto
beim
prov
edValidate
cand
idategenesfrom
QTLstud
ies
MASand
know
ledg
eof
the
biolog
yof
thetrait
andsource
ofgenesto
confer
the
relevant
changesto
phenotyp
eWorking
transformation
protocol
IPissuescostof
regu
latory
approv
al
GMO
labelling
marketin
gissuestricky
tousetransgenes
inou
t‐breedersbecause
ofcomplexity
oftransformingandgene
flow
risks
US
Manyprog
ramsin
UnitedStates
are
creatin
gtransgenic
plantsusingAlamoas
asource
oftransformable
geno
typesManytraits
ofinterestNothing
commercial
yet
CNC
hang
shaBtg
ene
transformed
into
Msac
in20
04M
sin
transformed
with
MINAC2gene
and
Msactransform
edwith
Cry2A
ageneb
othwith
amarkerg
eneby
Agrob
acterium
JP1
Msintransgenic
forlow
temperature
tolerancewith
increased
expression
offructans
(unp
ublished)
NLImprov
ingprotocol
forM
sin
transformation
SK2
Msin
with
Arabido
psisand
Brachypod
ium
PhytochromeBgenes
(Hwang
Cho
etal
2014
Hwang
Lim
etal20
14)
UK2
Msin
and
1Mflorg
enotyp
es
UKRou
tine
transformationno
tyet
possibleResearchto
overcomerecalcitrance
ongo
ing
Currently
trying
differentspecies
andcond
ition
sPo
plar
transformationused
atpresent
US
Frequently
attempted
with
very
limitedsuccess
US
600transgenic
lines
have
been
characterized
mostly
performed
inthe
aspenhy
brids
reprod
uctiv
esterility
drou
ghttolerance
with
Kno
ckdo
wns
(Con
tinues)
CLIFTON‐BROWN ET AL | 15
TABLE
5(Contin
ued)
Breeding
techno
logy
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sW
illow
Pop
lar
variou
slytransformed
with
four
cellwall
genesiptand
uidA
genesusingtwo
selectionsystem
sby
biolisticsand
Agrob
acterium
Transform
edplantsare
beinganalysed
US
Prelim
inarywork
will
betakenforw
ardin
2018
ndash202
2in
CABBI
Genom
eediting
CRISPR
Refinem
entofexisting
traits
inusefulparents
orprom
isinghybrids
bygeneratingtargeted
mutations
ingenes
know
ntocontrolthe
traitof
interestF
irst
adouble‐stranded
breakismadeinthe
DNAwhich
isrepairedby
natural
DNArepair
machineryL
eads
tofram
eshiftSN
Por
can
usealdquorepair
templaterdquo
orcanbe
used
toinserta
transgene
intoaldquosafe
harbourlocusrdquoIt
couldbe
used
todeleterepressors(or
transcriptionfactors)
Identification
mapping
and
sequ
encing
oftarget
genes(from
DNA
sequ
ence)
avoiding
screening
outof
unintend
ededits
Manyregu
latory
authorities
have
not
decidedwhether
CRISPR
andother
geno
meediting
techno
logies
are
GMOsor
notIf
GMOsseecomment
onstage10
abov
eIf
noteditedcrop
swill
beregu
latedas
conv
entio
nalvarieties
CATechn
olog
yisstill
toonewand
switchg
rass
geno
meis
very
complexOther
labo
ratories
are
interestedbu
tno
tyet
mov
ingon
this
US
One
prog
rammeso
farmdash
USD
Ain
Albany
FRInitiated
in20
16(M
ISEDIT
project)
NLInitiatingin
2018
US
Initiatingin
2018
in
CABBI
UKNot
started
Not
possible
until
transformation
achieved
US
CRISPR
‐Cas9and
Cpf1have
been
successfully
developedin
Pop
ulus
andarehigh
lyefficientExamples
forlig
nin
biosyn
thesis
Note
BFF
BiomassFo
rtheFu
tureBtBacillus
thuringiensisCABBICenterforadvanced
bioenergyandbioproductsinnovatio
nCpf1
CRISPR
from
Prevotella
andFrancisella
1CRISPR
clusteredregularlyinterspaced
palin
drom
icrepeatsCRISPR
‐CasC
RISPR
‐associated
Cry2A
acrystaltoxins2A
asubfam
ilyproduced
byBtFS
full‐sibG
BS
genotyping
bysequencingG
EBVsgenomic‐estim
ated
breeding
valuesG
MOg
enetically
modi-
fied
organism
GS
genomicselection
GWAS
genomew
ideassociationstudy
INRAF
renchNationalInstituteforAgriculturalR
esearch
IPintellectualp
roperty
IPTisopentenyltransferase
ISSR
inter‐SSR
MflorMiscant-
husflo
ridulusMsacMsacchariflo
rusMsinMsinensisM
AS
marker‐a
ssistedselection
MISEDITm
iscanthusgene
editing
forseed‐propagatedtriploidsNACn
oapicalmeristemA
TAF1
2and
cup‐shaped
cotyledon2
‐lik
eQTLq
uantitativ
etraitlocusS
NP
singlenucleotid
epolymorphismS
SRsim
plesequence
repeatsUSD
AU
SDepartm
ento
fAgriculture
a ISO
Alpha‐2
lettercountrycodes
16 | CLIFTON‐BROWN ET AL
High seed production from field crossing trials con-ducted in locations where flowering in both seed andpollen parents is likely to happen synchronously
Scalable and adapted harvesting threshing and seed pro-cessing methods for producing high seed quality
The results of these parallel activities need to becombined to identify the upscaling pathway for eachhybrid if this cannot be achieved the hybrid will likelynot be commercially viable The UK‐led programme withpartners in Italy and Germany shows that seedbased mul-tiplication rates of 12000 are achievable several inter-specific hybrids (Clifton‐Brown et al 2017) Themultiplication rate of M sinensis is higher probably15000ndash10000 Conventional cloning from rhizome islimited to around 120 that is one ha could provide rhi-zomes for around 20 ha of new plantation
Multilocation field testing of wild and novel miscanthushybrids selected by breeding programmes in the Nether-lands and the United Kingdom was performed as part ofthe project Optimizing Miscanthus Biomass Production(OPTIMISC 2012ndash2016) These trials showed that com-mercial yields and biomass qualities (Kiesel et al 2017Van der Weijde et al 2017a Van der Weijde Kieselet al 2017) could be produced in a wide range of climatesand soil conditions from the temperate maritime climate ofwestern Wales to the continental climate of eastern Russiaand the Ukraine (Kalinina et al 2017) Extensive environ-mental measurements of soil and climate combined withgrowth monitoring are being used to understand abioticstresses (Nunn et al 2017 Van der Weijde Huxley et al2017) and develop genotype‐specific scenarios similar tothose reported earlier in Hastings et al (2009) Phenomicsexperiments on drought tolerance have been conducted onwild and improved germplasm (Malinowska Donnison ampRobson 2017 Van der Weijde Huxley et al 2017)Recently produced interspecific hybrids displaying excep-tional yield under drought (~30 greater than control Mxg)in field trials in Poland and Moldova are being furtherstudied in detail in the phenomics and genomics facility atAberystwyth to better understand genendashtrait associationswhich can be fed back into breeding
Intraspecific seeded hybrids of M sinensis produced inthe Netherlands and interspecific M sacchariflorus times M si-nensis hybrids produced by the UK‐led breeding programmehave entered yield testing in 2018 with the recently EU‐funded project ldquoGRowing Advanced industrial Crops on mar-ginal lands for biorEfineries (GRACE)rdquo (httpswwwgrace-bbieu) Substantial variation in biomass quality for sacchari-fication efficiency (glucose release as of dry matter) ashcontent and melting point has already been generated inintraspecific M sinensis hybrids (Van der Weijde Kiesel
et al 2017) across environments (Weijde Dolstra et al2017a) GRACE aims to establish more than 20 hectares ofnew inter‐ and intraspecific seeded hybrids across six Euro-pean countries This project is building the know‐how andagronomy needed to transition from small research plots tocommercial‐scale field sites and linking biomass productiondirectly to industrial applications The biomass produced byhybrids in different locations will be supplied to innovativeindustrial end‐users making a wide range of biobased prod-ucts both for chemicals and for energy In the United Statesmulti‐location yield were initiated in 2018 to evaluate new tri-ploid M times giganteus genotypes developed at Illinois Cur-rently infertile hybrids are favoured in the United Statesbecause this eliminates the risk of invasiveness from naturallydispersed viable seed The precautionary principle is appliedas fertile miscanthus has naturalized in several states (QuinnAllen amp Stewart 2010) North European multilocation fieldtrials in the EMI and OPTIMISC projects have shown thereis minimal risk of invasiveness even in years when fertileflowering hybrids produce viable seed Naturalized standshave not established here due perhaps to low dormancy pooroverwintering and low seedling competitive strength In addi-tion to breeding for nonshattering or sterile seeded hybridsQuinn et al (2010) suggest management strategies which canfurther minimize environmental opportunities to manage therisk of invasiveness
31 | Molecular breeding and biotechnology
In miscanthus new plant breeding techniques (Table 5)have focussed on developing molecular markers forbreeding in Europe the United States South Korea andJapan There are several publications on QTL mappingpopulations for key traits such as flowering (AtienzaRamirez amp Martin 2003) and compositional traits(Atienza Satovic Petersen Dolstra amp Martin 2003) Inthe United States and United Kingdom independent andinterconnected bi‐parental ldquomappingrdquo families have beenstudied (Dong et al 2018 Gifford Chae SwaminathanMoose amp Juvik 2015) alongside panels of diverse germ-plasm accessions for GWAS (Slavov et al 2013) Fur-ther developments calibrating GS with very large panelsof parents and cross progeny are underway (Daveyet al 2017) The recently completed first miscanthus ref-erence genome sequence is expected to improve the effi-ciency of MAS strategies and especially GWAS(httpsphytozomejgidoegovpzportalhtmlinfoalias=Org_Msinensis_er) For example without a referencegenome sequence Clark et al (2014) obtained 21207RADseq SNPs (single nucleotide polymorphisms) on apanel of 767 miscanthus genotypes (mostly M sinensis)but subsequent reanalysis of the RADseq data using the
CLIFTON‐BROWN ET AL | 17
new reference genome resulted in hundreds of thousandsof SNPs being called
Robust and effective in vitro regeneration systems havebeen developed for Miscanthus sinensis M times giganteusand M sacchariflorus (Dalton 2013 Guo et al 2013Hwang Cho et al 2014 Rambaud et al 2013 Ślusarkie-wicz‐Jarzina et al 2017 Wang et al 2011 Zhang et al2012) However there is still significant genotype speci-ficity and these methods need ldquoin‐houserdquo optimization anddevelopment to be used routinely They provide potentialroutes for rapid clonal propagation and also as a basis forgenetic transformation Stable transformation using bothbiolistics (Wang et al 2011) and Agrobacterium tumefa-ciens DNA delivery methods (Hwang Cho et al 2014Hwang Lim et al 2014) has been achieved in M sinen-sis The development of miscanthus transformation andgene editing to generate diplogametes for producing seed‐propagated triploid hybrids are performed as part of theFrench project MISEDIT (miscanthus gene editing forseed‐propagated triploids) There are no reports of genomeediting in any miscanthus species but new breeding inno-vations including genome editing are particularly relevantin this slow‐to‐breed nonfood bioenergy crop (Table 4)
4 | WILLOW
Willow (Salix spp) is a very diverse group of catkin‐bear-ing trees and shrubs Willow belongs to the family Sali-caceae which also includes the Populus genus There areapproximately 350 willow species (Argus 2007) foundmostly in temperate and arctic zones in the northern hemi-sphere A few are adapted to subtropical and tropicalzones The centre of diversity is believed to be in Asiawith over 200 species in China Around 120 species arefound in the former Soviet Union over 100 in NorthAmerica and around 65 species in Europe and one speciesis native to South America (Karp et al 2011) Willows aredioecious thus obligate outcrossers and highly heterozy-gous The haploid chromosome number is 19 (Hanley ampKarp 2014) Around 40 of species are polyploid (Sudaamp Argus 1968) ranging from triploids to the atypicaldodecaploid S maxxaliana with 2n=190 (Zsuffa et al1984)
Although almost exclusively native to the NorthernHemisphere willow has been grown around the globe formany thousands of years to support a wide range of appli-cations (Kuzovkina amp Quigley 2005 Stott 1992) How-ever it has been the focus of domestication for bioenergypurposes for only a relatively short period since the 1970sin North America and Europe For bioenergy breedershave focused their efforts on the shrub willows (subgenusVetix) because of their rapid juvenile growth rates as aresponse to coppicing on a 2‐ to 4‐year cycle that can be
accomplished using farm machinery rather than forestryequipment (Shield Macalpine Hanley amp Karp 2015Smart amp Cameron 2012)
Since shrub willow was not generally recognized as anagricultural crop until very recently there has been littlecommitment to building and maintaining germplasm reposi-tories of willow to support long‐term breeding One excep-tion is the United Kingdom where a large and well‐characterized Salix germplasm collection comprising over1500 accessions is held at Rothamsted Research (Stott1992 Trybush et al 2008) Originally initiated for use inbasketry in 1923 accessions have been added ever sinceIn the United States a germplasm collection of gt350accessions is located at Cornell University to support thebreeding programme there The UK and Cornell collectionshave a relatively small number of accessions in common(around 20) Taken together they represent much of thespecies diversity but only a small fraction of the overallgenetic diversity within the genus There are three activewillow breeding programmes in Europe RothamstedResearch (UK) Salixenergi Europa AB (SEE) and a pro-gramme at the University of Warmia and Mazury in Olsz-tyn (Poland) (abbreviations used in Table 1) There is oneactive US programme based at Cornell University Culti-vars are still being marketed by the European WillowBreeding Programme (EWBP) (UK) which was activelybreeding biomass varieties from 1996 to 2002 Cultivarsare protected by plant breedersrsquo rights (PBRs) in Europeand by plant patents in the United States The sharing ofgenetic resources in the willow community is generallyregulated by material transfer agreements (MTA) and tai-lored licensing agreements although the import of cuttingsinto North America is prohibited except under special quar-antine permit conditions
Efforts to augment breeding germplasm collection fromnature are continuing with phenotypic screening of wildgermplasm performed in field experiments with 177 S pur-purea genotypes in the United States (at sites in Genevaand Portland NY and Morgantown WV) that have beengenotyped using genotyping by sequencing (GBS) (Elshireet al 2011) In addition there are approximately 400accessions of S viminalis in Europe (near Pustnaumls Upp-sala Sweden and Woburn UK (Berlin et al 2014 Hal-lingbaumlck et al 2016) The S viminalis accessions wereinitially genotyped using 38 simple sequence repeats (SSR)markers to assess genetic diversity and screened with~1600 SNPs in genes of potential interest for phenologyand biomass traits Genetics and genomics combined withextensive phenotyping have substantially improved thegenetic basis of biomass‐related traits in willow and arenow being developed in targeted breeding via MAS Thisunderpinning work has been conducted on large specifi-cally developed biparental Salix mapping populations
18 | CLIFTON‐BROWN ET AL
(Hanley amp Karp 2014 Zhou et al 2018) as well asGWAS panels (Hallingbaumlck et al 2016)
Once promising parental combinations are identifiedcrosses are usually performed using fresh pollen frommaterial that has been subject to a phased removal fromcold storage (minus4degC) (Lindegaard amp Barker 1997 Macal-pine Shield Trybush Hayes amp Karp 2008 Mosseler1990) Pollen storage is useful in certain interspecific com-binations where flowering is not naturally synchronizedThis can be overcome by using pollen collection and stor-age protocol which involves extracting pollen using toluene(Kopp Maynard Niella Smart amp Abrahamson 2002)
The main breeding approach to improve willow yieldsrelies on species hybridization to capture hybrid vigour(Fabio et al 2017 Serapiglia Gouker amp Smart 2014) Inthe absence of genotypic models for heterosis breedershave extensively tested general and specific combiningability of parents to produce superior progeny The UKbreeding programmes (EWBP 1996ndash2002 and RothamstedResearch from 2003 on) have performed more than 1500exploratory cross‐pollinations The Cornell programme hassuccessfully completed about 550 crosses since 1998Investment into the characterization of genetic diversitycombined with progeny tests from exploratory crosses hasbeen used to produce hundreds of targeted intraspeciescrosses in the United Kingdom and United States respec-tively (see Table 1) To achieve long‐term gains beyond F1hybrids four intraspecific recurrent selection populationshave been created in the United Kingdom (for S dasycla-dos S viminalis and S miyabeana) and Cornell is pursu-ing recurrent selection of S purpurea Interspecifichybridizations with genotypes selected from the recurrentselection cycles are well advanced in willow with suchcrosses to date totalling 420 in the United Kingdom andover 100 in the United States
While species hybridization is common in Salix it isnot universal Of the crosses attempted about 50 hybri-dize and produce seed (Macalpine Shield amp Karp 2010)As the viability of seed from successful crosses is short (amatter of days at ambient temperatures) proper seed rear-ing and storage protocols are essential (Maroder PregoFacciuto amp Maldonado 2000)
Progeny from crosses are treated in different waysamong the breeding programmes at the seedling stage Inthe United States seedlings are planted into an irrigatedfield where plants are screened for two seasons beforebeing progressed to further field trials In the United King-dom seedlings are planted into trays of compost wherethey remain containerized in an irrigated nursery for theremainder of year one In the United Kingdom seedlingsare subject to two rounds of selection in the nursery yearThe first round takes place in September to select againstsusceptibility to rust infection (Melampsora spp) A second
round of selection in winter assesses tip damage from frostand giant willow aphid infestation In the United Stateswhere the rust pressure is lower screening for Melampsoraspp cannot be performed at the nursery stage Both pro-grammes monitor Melampsora spp pest susceptibilityyield and architecture over multiple years in field trialsSelected material is subject to two rounds of field trials fol-lowed by a final multilocation yield trial to identify vari-eties for commercialization
Promising selections (ie potential cultivars) need to beclonally propagated A rapid in vitro tissue culture propa-gation method has been developed (Palomo‐Riacuteos et al2015) This method can generate about 5000 viable trans-plantable clones from a single plant in just 24 weeks Anin vitro system can also accommodate early selection viamolecular or biochemical markers to increase selectionspeed Conventional breeding systems take 13 years viafour rounds of selection from crossing to selecting a variety(Figure 1) but this has the potential to be reduced to7 years if micropropagation and MAS selection are adopted(Hanley amp Karp 2014 Palomo‐Riacuteos et al 2015)
Willows are currently propagated commercially byplanting winter‐dormant stem cuttings in spring Commer-cial planting systems for willow use mechanical plantersthat cut and insert stem sections from whips into a well‐prepared soil One hectare of stock plants grown in specificmultiplication beds planted at 40000 plants per ha pro-duces planting material for 80 hectares of commercialshort‐rotation coppice willow annually (planted at 15000cuttings per hectare) (Whittaker et al 2016) When com-mercial plantations are established the industry standard isto plant intimate mixtures of ~5 diverse rust (Melampsoraspp)‐resistant varieties (McCracken amp Dawson 1997 VanDen Broek et al 2001)
The foundations for using new plant breeding tech-niques have been established with funding from both thepublic and the private sectors To establish QTL maps 16mapping populations from biparental crosses are understudy in the United Kingdom Nine are under study in theUnited States The average number of individuals in thesefamilies ranges from 150 to 947 (Hanley amp Karp 2014)GS is also being evaluated in S viminalis and preliminaryresults indicate that multiomic approaches combininggenomic and metabolomic data have great potential (Sla-vov amp Davey 2017) For both QTL and GS approachesthe field phenotyping demands are large as several thou-sand individuals need to be phenotyped for a wide rangeof traits These include the following dates of bud burstand growth succession stem height stem density wooddensity and disease resistance The greater the number ofindividuals the more precise the QTL marker maps andGS models are However the logistical and financial chal-lenges of phenotyping large numbers of individuals are
CLIFTON‐BROWN ET AL | 19
considerable because the willow crop is gt5 m tall in thesecond year There is tremendous potential to improve thethroughput of phenotyping using unmanned aerial systemswhich is being tested in the USDA National Institute ofFood and Agriculture (NIFA) Willow SkyCAP project atCornell Further investment in these approaches needs tobe sustained over many years fully realizes the potentialof a marker‐assisted selection programme for willow
To date despite considerable efforts in Europe and theUnited States to establish a routine transformation systemthere has not been a breakthrough in willow but attemptsare ongoing As some form of transformation is typically aprerequisite for genome editing techniques these have notyet been applied to willow
In Europe there are 53 short‐rotation coppice (SRC) bio-mass willow cultivars registered with the Community PlantVariety Office (CPVO) for PBRs of which ~25 are availablecommercially in the United Kingdom There are eightpatented cultivars commercially available in the UnitedStates In Sweden there are nine commercial cultivars regis-tered in Europe and two others which are unregistered(httpssalixenergiseplanting-material) Furthermore thereare about 20 precommercial hybrids in final yield trials inboth the United States and the United Kingdom It has beenestimated that it would take two years to produce the stockrequired to plant 50 ha commercially from the plant stock inthe final yield trials Breeding programmes have alreadydelivered rust‐resistant varieties and increases in yield to themarket The adoption of advanced breeding technologies willlikely lead to a step change in improving traits of interest
5 | POPLAR
Poplar a fast‐growing tree from the northern hemispherewith a small genome size has been adopted for commercialforestry and scientific purposes The genus Populus con-sists of about 29 species classified in six different sectionsPopulus (formerly Leuce) Tacamahaca Aigeiros AbasoTuranga and Leucoides (Eckenwalder 1996) The Populusspecies of most interest for breeding and testing in the Uni-ted States and Europe are P nigra P deltoides P maxi-mowiczii and P trichocarpa (Stanton 2014) Populusclones for biomass production are being developed byintra‐ and interspecies hybridization (DeWoody Trewin ampTaylor 2015 Richardson Isebrands amp Ball 2014 van derSchoot et al 2000) Recurrent selection approaches areused for gradual population improvement and to create eliteclonal lines for commercialization (Berguson McMahon ampRiemenschneider 2017 Neale amp Kremer 2011) Currentlypoplar breeding in the United States occurs in industrialand academic programmes located in the Southeast theMidwest and the Pacific Northwest These use six speciesand five interspecific taxa (Stanton 2014)
The southeastern programme historically focused onrecurrent selection of P deltoides from accessions made inthe lower Mississippi River alluvial plain (Robison Rous-seau amp Zhang 2006) More recently the genetic base hasbeen broadened to produce interspecific hybrids with resis-tance to the fungal infection Septoria musiva which causescankers
In the midwest of the United States populationimprovement efforts are focused on P deltoides selectionsfrom native provenances and hybrid crosses with acces-sions introduced from Europe Interspecific intercontinental(Europe and America) hybrid crosses between P nigra andP deltoides (P times canadensis) are behind many of theleading commercial hybrids which are the most advancedbreeding materials for many applications and regions InMinnesota previous breeding experience and efforts utiliz-ing P maximowiczii and P trichocarpa have been discon-tinued due to Septoria susceptibility and a lack of coldhardiness (Berguson et al 2017) Traits targeted forimprovement include yieldgrowth rate cold hardinessadventitious rooting resistance to Septoria and Melamp-sora leaf rust and stem form The Upper Midwest pro-gramme also carries out wide hybridizations within thesection Populus The P times wettsteinii (P trem-ula times P tremuloides) taxon is bred for gains in growthrate wood quality and resistance to the fungus Entoleucamammata which causes hypoxylon canker (David ampAnderson 2002)
In the Pacific Northwest GreenWood Resources Incleads poplar breeding that emphasizes interspecifichybrid improvement of P times generosa (P deltoides timesP trichocarpa and reciprocal) and P deltoides times P maxi-mowiczii taxa for coastal regions and the P times canadensistaxon for the drier continental regions Intraspecificimprovement of second‐generation breeding populations ofP deltoides P nigra P maximowiczii and P trichocarpaare also involved (Stanton et al 2010) The present focusof GreenWood Resourcesrsquo hybridization is bioenergy feed-stock improvement concentrating on coppice yield wood‐specific gravity and rate of sugar release
Industrial interest in poplar in the United States has his-torically come from the pulp and paper sector althoughveneer and dimensional lumber markets have been pursuedat times Currently the biomass market for liquid trans-portation fuels is being emphasized along with the use oftraditional and improved poplar genotypes for ecosystemservices such as phytoremediation (Tuskan amp Walsh 2001Zalesny et al 2016)
In Europe there are breeding programmes in FranceGermany Italy and Sweden These include the following(a) Alasia Franco Vivai (AFV) programme in northernItaly (b) the French programme led by the poplar Scien-tific Interest Group (GIS Peuplier) and carried out
20 | CLIFTON‐BROWN ET AL
collaboratively between the National Institute for Agricul-tural Research (INRA) the National Research Unit ofScience and Technology for Environment and Agriculture(IRSTEA) and the Forest Cellulose Wood Constructionand Furniture Technology Institute (FCBA) (c) the Ger-man programme at Northwest German Forest Research Sta-tion (NW‐FVA) at Hannoversch Muumlnden and (d) theSwedish programme at the Swedish University of Agricul-tural Sciences and SweTree Technologies AB (Table 1)
AFV leads an Italian poplar breeding programme usingextensive field‐grown germplasm collections of P albaP deltoides P nigra and P trichocarpa While interspeci-fic hybridization uses several taxa the focus is onP times canadensis The breeding programme addresses dis-ease resistance (Marssonina brunnea Melampsora larici‐populina Discosporium populeum and poplar mosaicvirus) growth rate and photoperiod adaptation AFV andGreenWood Resources collaborate in poplar improvementin Europe through the exchange of frozen pollen and seedfor reciprocal breeding projects Plantations in Poland andRomania are currently the focus of the collaboration
The ongoing French GIS Peuplier is developing a long‐term breeding programme based on intraspecific recurrentselection for the four parental species (P deltoides P tri-chocarpa P nigra and P maximowiczii) designed to betterbenefit from hybrid vigour demonstrated by the interspeci-fic crosses P canadensis P deltoides times P trichocarpaand P trichocarpa times P maximowiczii Current selectionpriorities are targeting adaptation to soil and climate condi-tions resistance and tolerance to the most economicallyimportant diseases and pests high volume production underSRC and traditional poplar cultivation regimes as well aswood quality of interest by different markets Currentlygenomic selection is under exploration to increase selectionaccuracy and selection intensity while maintaining geneticdiversity over generations
The German NW‐FVA programme is breeding intersec-tional AigeirosndashTacamahaca hybrids with a focus on resis-tance to Pollaccia elegans Xanthomonas populiDothichiza spp Marssonina brunnea and Melampsoraspp (Stanton 2014) Various cross combinations ofP maximowiczii P trichocarpa P nigra and P deltoideshave led to new cultivars suitable for deployment in vari-etal mixtures of five to ten genotypes of complementarystature high productivity and phenotypic stability (Weis-gerber 1993) The current priority is the selection of culti-vars for high‐yield short‐rotation biomass production Sixhundred P nigra genotypes are maintained in an ex situconservation programme An in situ P nigra conservationeffort involves an inventory of native stands which havebeen molecular fingerprinted for identity and diversity
The Swedish programme is concentrating on locallyadapted genotypes used for short‐rotation forestry (SRF)
because these meet the needs of the current pulping mar-kets Several field trials have shown that commercial poplarclones tested and deployed in Southern and Central Europeare not well adapted to photoperiods and low temperaturesin Sweden and in the Baltics Consequently SwedishUniversity of Agricultural Sciences and SweTree Technolo-gies AB started breeding in Sweden in 1990s to producepoplar clones better adapted to local climates and markets
51 | Molecular breeding technologies
Poplar genetic improvement cannot be rapidly achievedthrough traditional methods alone because of the longbreeding cycles outcrossing breeding systems and highheterozygosity Integrating modern genetic genomic andphenomics techniques with conventional breeding has thepotential to expedite poplar improvement
The genome of poplar has been sequenced (Tuskanet al 2006) It has an estimated genome size of485 plusmn 10 Mbp divided into 19 chromosomes This is smal-ler than other PBCs and makes poplar more amenable togenetic engineering (transgenesis) GS and genome editingPoplar has seen major investment in both the United Statesand Europe being the model system for woody perennialplant genetics and genomics research
52 | Targets for genetic modification
Traits targeted include wood properties (lignin content andcomposition) earlylate flowering male sterility to addressbiosafety regulation issues enhanced yield traits and herbi-cide tolerance These extensive transgenic experiments haveshown differences in recalcitrance to in vitro regenerationand genetic transformation in some of the most importantcommercial hybrid poplars (Alburquerque et al 2016)Further transgene expression stability is being studied Sofar China is the only country known to have commerciallyused transgenic insect‐resistant poplar A precommercialherbicide‐tolerant poplar was trialled for 8 years in the Uni-ted States (Li Meilan Ma Barish amp Strauss 2008) butcould not be released due to stringent environmental riskassessments required for regulatory approval This increasestranslation costs and delays reducing investor confidencefor commercial deployment (Harfouche et al 2011)
The first field trials of transgenic poplar were performedin France in 1987 (Fillatti Sellmer Mccown Haissig ampComai 1987) and in Belgium in 1988 (Deblock 1990)Although there have been a total of 28 research‐scale GMpoplar field trials approved in the European Union underCouncil Directive 90220EEC since October 1991 (inPoland Belgium Finland France Germany Spain Swe-den and in the United Kingdom (Pilate et al 2016) onlyauthorizations in Poland and Belgium are in place today In
CLIFTON‐BROWN ET AL | 21
the United States regulatory notifications and permits fornearly 20000 transgenic poplar trees derived from approxi-mately 600 different constructs have been issued since1995 by the USDAs Animal and Plant Health InspectionService (APHIS) (Strauss et al 2016)
53 | Genome editing CRISPR technologies
Clustered regularly interspaced palindromic repeats(CRISPR) and the CRISPR‐associated (CRISPR‐Cas)nucleases are a groundbreaking genome‐engineering toolthat complements classical plant breeding and transgenicmethods (Moreno‐Mateos et al 2017) Only two publishedstudies in poplar have applied the CRISPRCas9 technol-ogy One is in P tomentosa in which an endogenous phy-toene desaturase gene (PtoPDS) was successfully disruptedsite specifically in the first generation of transgenic plantsresulting in an albino and dwarf phenotype (Fan et al2015) The second was in P tremula times alba in whichhigh CRISPR‐Cas9 mutational efficiency was achieved forthree 4‐coumarateCoA ligase (4Cl) genes 4CL1 4CL2and 4CL5 associated with lignin and flavonoid biosynthe-sis (Zhou et al 2015) Due to its low cost precision andrapidness it is very probable that cultivars or clones pro-duced using CRISPR technology will be ready for market-ing in the near future (Yin et al 2017) Recently aCRISPR with a smaller associated endonuclease has beendiscovered from Prevotella and Francisella 1 (Cpf1) whichmay have advantages over Cas9 In addition there arereports of DNA‐free editing in plants using both CRISPRCpf1 and CRISPR Cas9 for example Ref (Kim et al2017 Mahfouz 2017 Zaidi et al 2017)
It remains unresolved whether plants modified by genomeediting will be regulated as genetically modified organisms(GMOs) by the relevant authorities in different countries(Lozano‐Juste amp Cutler 2014) Regulations to cover thesenew breeding techniques are still evolving but those coun-tries who have published specific guidance (including UnitedStates Argentina and Chile) are indicating that plants pos-sessing simple genome edits will not be regulated as conven-tional transgenesis (Jones 2015b) The first generation ofgenome‐edited crops will likely be phenocopy gene knock-outs that already exist to produce ldquonature identicalrdquo traits thatis traits that could also be derived by conventional breedingDespite this confidence in applying these new powerfulbreeding tools remains limited owing to the uncertain regula-tory environment in many parts of the world (Gao 2018)including the recent ECJ 2018 rulings mentioned earlier
54 | Genomics‐based breeding technologies
Poplar breeding programs are becoming well equipped withuseful genomics tools and resources that are critical to
explore genomewide variability and make use of the varia-tion for enhancing genetic gains Deep transcriptomesequencing resequencing of alternate genomes and GBStechnology for genomewide marker detection using next‐generation sequencing (NGS) are yielding valuable geno-mics tools GWAS with NGS‐based markers facilitatesmarker identification for MAS breeding by design and GS
GWAS approaches have provided a deeper understand-ing of genome function as well as allelic architectures ofcomplex traits (Huang et al 2010) and have been widelyimplemented in poplar for wood characteristics (Porthet al 2013) stomatal patterning carbon gain versus dis-ease resistance (McKown et al 2014) height and phenol-ogy (Evans et al 2014) cell wall chemistry (Mucheroet al 2015) growth and cell walls traits (Fahrenkroget al 2017) bark roughness (Bdeir et al 2017) andheight and diameter growth (Liu et al 2018) Using high‐throughput sequencing and genotyping platforms an enor-mous amount of SNP markers have been used to character-ize the linkage disequilibrium (LD) in poplar (eg Slavovet al 2012 discussed below)
The genetic architecture of photoperiodic traits in peren-nial trees is complex involving many loci However itshows high levels of conservation during evolution (Mau-rya amp Bhalerao 2017) These genomics tools can thereforebe used to address adaptation issues and fine‐tune themovement of elite lines into new environments For exam-ple poor timing of spring bud burst and autumn bud setcan result in frost damage resulting in yield losses (Ilstedt1996) These have been studied in P tremula genotypesalong a latitudinal cline in Sweden (~56ndash66oN) and haverevealed high nucleotide polymorphism in two nonsynony-mous SNPs within and around the phytochrome B2 locus(Ingvarsson Garcia Hall Luquez amp Jansson 2006 Ing-varsson Garcia Luquez Hall amp Jansson 2008) Rese-quencing 94 of these P tremula genotypes for GWASshowed that noncoding variation of a single genomicregion containing the PtFT2 gene described 65 ofobserved genetic variation in bud set along the latitudinalcline (Tan 2018)
Resequencing genomes is currently the most rapid andeffective method detecting genetic differences betweenvariants and for linking loci to complex and importantagronomical and biomass traits thus addressing breedingchallenges associated with long‐lived plants like poplars
To date whole genome resequencing initiatives havebeen launched for several poplar species and genotypes InPopulus LD studies based on genome resequencing sug-gested the feasibility of GWAS in undomesticated popula-tions (Slavov et al 2012) This plant population is beingused to inform breeding for bioenergy development Forexample the detection of reliable phenotypegenotype asso-ciations and molecular signatures of selection requires a
22 | CLIFTON‐BROWN ET AL
detailed knowledge about genomewide patterns of allelefrequency variation LD and recombination suggesting thatGWAS and GS in undomesticated populations may bemore feasible in Populus than previously assumed Slavovet al (2012) have resequenced 16 genomes of P tri-chocarpa and genotyped 120 trees from 10 subpopulationsusing 29213 SNPs (Geraldes et al 2013) The largest everSNP data set of genetic variations in poplar has recentlybeen released providing useful information for breedinghttpswwwbioenergycenterorgbescgwasindexcfmAlso deep sequencing of transcriptomes using RNA‐Seqhas been used for identification of functional genes andmolecular markers that is polymorphism markers andSSRs A multitissue and multiple experimental data set forP trichocarpa RNA‐Seq is publicly available (httpsjgidoegovdoe-jgi-plant-flagship-gene-atlas)
The availability of genomic information of DNA‐con-taining cell organelles (nucleus chloroplast and mitochon-dria) will also allow a holistic approach in poplarmolecular breeding in the near future (Kersten et al2016) Complete Populus genome sequences are availablefor nucleus (P trichocarpa section Tacamahaca) andchloroplasts (seven species and two clones from P trem-ula W52 and P tremula times P alba 717ndash1B4) A compara-tive approach revealed structural and functionalinformation broadening the knowledge base of PopuluscpDNA and stimulating future diagnostic marker develop-ment The availability of whole genome sequences of thesecellular compartments of P tremula holds promise forboosting marker‐assisted poplar breeding Other nucleargenome sequences from additional Populus species arenow available (eg P deltoides (httpsphytozomejgidoegovpz) and will become available in the forthcomingyears (eg P tremula and P tremuloidesmdashPopGenIE(Sjodin Street Sandberg Gustafsson amp Jansson 2009))Recently the characterization of the poplar pan‐genome bygenomewide identification of structural variation in threecrossable poplar species P nigra P deltoides and P tri-chocarpa revealed a deeper understanding of the role ofinter‐ and intraspecific structural variants in poplar pheno-type and may have important implications for breedingparticularly interspecific hybrids (Pinosio et al 2016)
GS has been proposed as an alternative to MAS in cropimprovement (Bernardo amp Yu 2007 Heffner Sorrells ampJannink 2009) GS is particularly well suited for specieswith long generation times for characteristics that displaymoderate‐to‐low heritability for traits that are expensive tomeasure and for selection of traits expressed late in the lifecycle as is the case for most traits of commercial value inforestry (Harfouche et al 2012) Current joint genomesequencing efforts to implement GS in poplar using geno-mic‐estimated breeding values for bioenergy conversiontraits from 49 P trichocarpa families and 20 full‐sib
progeny are taking place at the Oak Ridge National Labo-ratory and GreenWood Resources (Brian Stanton personalcommunication httpscbiornlgov) These data togetherwith the resequenced GWAS population data will be thebasis for developing GS algorithms Genomic breedingtools have been developed for the intraspecific programmetargeting yield resistance to Venturia shoot blightMelampsora leaf rust resistance to Cryptorhynchus lapathistem form wood‐specific gravity and wind firmness (Evanset al 2014 Guerra et al 2016) A newly developedldquobreeding with rare defective allelesrdquo (BRDA) technologyhas been developed to exploit natural variation of P nigraand identify defective variants of genes predicted by priortransgenic research to impact lignin properties Individualtrees carrying naturally defective alleles can then be incor-porated directly into breeding programs thereby bypassingthe need for transgenics (Vanholme et al 2013) Thisnovel breeding technology offers a reverse genetics com-plement to emerging GS for targeted improvement of quan-titative traits (Tsai 2013)
55 | Phenomics‐assisted breeding technology
Phenomics involves the characterization of phenomesmdashthefull set of phenotypes of given individual plants (HouleGovindaraju amp Omholt 2010) Traditional phenotypingtools which inefficiently measure a limited set of pheno-types have become a bottleneck in plant breeding studiesHigh‐throughput plant phenotyping facilities provide accu-rate screening of thousands of plant breeding lines clones orpopulations over time (Fu 2015) are critical for acceleratinggenomics‐based breeding Automated image collection andanalysis phenomics technologies allow accurate and nonde-structive measurements of a diversity of phenotypic traits inlarge breeding populations (Gegas Gay Camargo amp Doo-nan 2014 Goggin Lorence amp Topp 2015 Ludovisi et al2017 Shakoor Lee amp Mockler 2017) One important con-sideration is the identification of relevant and quantifiabletarget traits that are early diagnostic indicators of biomassyield Good progress has been made in elucidating theseunderpinning morpho‐physiological traits that are amenableto remote sensing in Populus (Harfouche Meilan amp Altman2014 Rae Robinson Street amp Taylor 2004) Morerecently Ludovisi et al (2017) developed a novel methodol-ogy for field phenomics of drought stress in a P nigra F2partially inbred population using thermal infrared imagesrecorded from an unmanned aerial vehicle‐based platform
Energy is the current main market for poplar biomass butthe market return provided is not sufficient to support pro-duction expansion even with added demand for environmen-tal and land management ldquoecosystem servicesrdquo such as thetreatment of effluent phytoremediation riparian buffer zonesand agro‐forestry plantings Aviation fuel is a significant
CLIFTON‐BROWN ET AL | 23
target market (Crawford et al 2016) To serve this marketand to reduce current carbon costs of production (Budsberget al 2016) key improvement traits in addition to yield(eg coppice regeneration pestdisease resistance water‐and nutrient‐use efficiencies) will be trace greenhouse gas(GHG) emissions (eg isoprene volatiles) site adaptabilityand biomass conversion efficiency Efforts are underway tohave national environmental protection agenciesrsquo approvalfor poplar hybrids qualifying for renewable energy credits
6 | REFLECTIONS ON THECOMMERCIALIZATIONCHALLENGE
The research and innovation activities reviewed in thispaper aim to advance the genetic improvement of speciesthat can provide feedstocks for bioenergy applicationsshould those markets eventually develop These marketsneed to generate sufficient revenue and adequately dis-tribute it to the actors along the value chain The work onall four crops shares one thing in common long‐termefforts to integrate fundamental knowledge into breedingand crop development along a research and development(RampD) pipeline The development of miscanthus led byAberystwyth University exemplifies the concerted researcheffort that has integrated the RampD activities from eight pro-jects over 14 years with background core research fundingalong an emerging innovation chain (Figure 2) This pro-gramme has produced a first range of conventionally bredseeded interspecies hybrids which are now in upscaling tri-als (Table 3) The application of molecular approaches(Table 5) with further conventional breeding (Table 4)offers the prospect of a second range of improved seededhybrids This example shows that research‐based support ofthe development of new crops or crop types requires along‐term commitment that goes beyond that normallyavailable from project‐based funding (Figure 1) Innovationin this sector requires continuous resourcing of conven-tional breeding operations and capability to minimize timeand investment losses caused by funding discontinuities
This challenge is increased further by the well‐knownmarket failure in the breeding of many agricultural cropspecies The UK Department for Environment Food andRural Affairs (Defra) examined the role of geneticimprovement in relation to nonmarket outcomes such asenvironmental protection and concluded that public invest-ment in breeding was required if profound market failure isto be addressed (Defra 2002) With the exception ofwidely grown hybrid crops such as maize and some high‐value horticultural crops royalties arising from plant breed-ersrsquo rights or other returns to breeders fail to adequatelycompensate for the full cost for research‐based plant breed-ing The result even for major crops such as wheat is sub‐
optimal investment and suboptimal returns for society Thismarket failure is especially acute for perennial crops devel-oped for improved sustainability rather than consumerappeal (Tracy et al 2016) Figure 3 illustrates the underly-ing challenge of capturing value for the breeding effortThe ldquovalley of deathrdquo that results from the low and delayedreturns to investment applies generally to the research‐to‐product innovation pipeline (Beard Ford Koutsky amp Spi-wak 2009) and certainly to most agricultural crop speciesHowever this schematic is particularly relevant to PBCsMost of the value for society from the improved breedingof these crops comes from changes in how agricultural landis used that is it depends on the increased production ofthese crops The value for society includes many ecosys-tems benefits the effects of a return to seminatural peren-nial crop cover that protects soils the increase in soilcarbon storage the protection of vulnerable land or the cul-tivation of polluted soils and the reductions in GHG emis-sions (Lewandowski 2016) By its very nature theproduction of biomass on agricultural land marginal forfood production challenges farm‐level profitability Thecosts of planting material and one‐time nature of cropestablishment are major early‐stage costs and therefore theopportunities for conventional royalty capture by breedersthat are manifold for annual crops are limited for PBCs(Hastings et al 2017) Public investment in developingPBCs for the nonfood biobased sector needs to providemore long‐term support for this critical foundation to a sus-tainable bioeconomy
7 | CONCLUSIONS
This paper provides an overview of research‐based plantbreeding in four leading PBCs For all four PBC generasignificant progress has been made in genetic improvementthrough collaboration between research scientists and thoseoperating ongoing breeding programmes Compared withthe main food crops most PBC breeding programmes dateback only a few decades (Table 1) This breeding efforthas thus co‐evolved with molecular biology and the result-ing ‐omics technologies that can support breeding Thedevelopment of all four PBCs has depended strongly onpublic investment in research and innovation The natureand driver of the investment varied In close associationwith public research organizations or universities all theseprogrammes started with germplasm collection and charac-terization which underpin the selection of parents forexploratory wide crosses for progeny testing (Figure 1Table 4)
Public support for switchgrass in North America wasexplicitly linked to plant breeding with 12 breeding pro-grammes supported in the United States and CanadaSwitchgrass breeding efforts to date using conventional
24 | CLIFTON‐BROWN ET AL
breeding have resulted in over 36 registered cultivars inthe United States (Table 1) with the development of dedi-cated biomass‐type cultivars coming within the past fewyears While ‐omics technologies have been incorporatedinto several of these breeding programmes they have notyet led to commercial deployment in either conventional orhybrid cultivars
Willow genetic improvement was led by the researchcommunity closely linked to plant breeding programmesWillow and poplar have the longest record of public invest-ment in genetic improvement that can be traced back to1920s in the United Kingdom and United States respec-tively Like switchgrass breeding programmes for willoware connected to public research efforts The United King-dom in partnership with the programme based at CornellUniversity remains the European leader in willowimprovement with a long‐term breeding effort closelylinked to supporting biological research at Rothamsted Inwillow F1 hybrids have produced impressive yield gainsover parental germplasm by capturing hybrid vigour Over30 willow clones are commercially available in the UnitedStates and Europe and a further ~90 are under precommer-cial testing (Table 1)
Compared with willow and poplar miscanthus is a rela-tive newcomer with all the current breeding programmesstarting in the 2000s Clonal M times giganteus propagated byrhizomes is expected to be replaced by more readily scal-able seeded hybrids from intra‐ (M sinensis) and inter‐(M sacchariflorus times M sinensis) species crosses with highseed multiplication rates (of gt2000) The first group ofhybrid cultivars is expected to be market‐ready around2022
Of the four genera used as PBCs Populus is the mostadvanced in terms of achievements in biological researchas a result of its use as a model for basic research of treesMuch of this biological research is not directly connectedto plant breeding Nevertheless reflecting the fact thatpoplar is widely grown as a single‐stem tree in SRF thereare about 60 commercially available clones and an addi-tional 80 clones in commercial pipelines (Table 1) Trans-genic poplar hybrids have moved beyond proof of conceptto commercial reality in China
Many PBC programmes have initiated long‐term con-ventional recurrent selection breeding cycles for populationimprovement which is a key process in increasing yieldthrough hybrid vigour As this approach requires manyyears most programmes are experimenting with molecularbreeding methods as these have the potential to accelerateprecision breeding For all four PBCs investments in basicgenetic and genomic resources including the developmentof mapping populations for QTLs and whole genomesequences are available to support long‐term advancesMore recently association genetics with panels of diversegermplasm are being used as training populations for GSmodels (Table 5) These efforts are benefitting from pub-licly available DNA sequences and whole genome assem-blies in crop databases Key to these accelerated breedingtechnologies are developments in novel phenomics tech-nologies to bridge the genotypephenotype gap In poplarnovel remote sensing field phenotyping is now beingdeployed to assist breeders These advances are being com-bined with in vitro and in planta modern molecular breed-ing techniques such as CRISPR (Table 5) CRISPRtechnology for genome editing has been proven in poplar
FIGURE 2 A schematic developmentpathway for miscanthus in the UnitedKingdom related to the investment in RampDprojects at Aberystwyth (top coloured areasfor projects in the three categories basicresearch breeding and commercialupscaling) leading to a projected croppingarea of 350000 ha by 2030 with clonal andsuccessive ranges of improved seed‐basedhybrids Purple represents theBiotechnology and Biological SciencesResearch Council (BBSRC) and brown theDepartment for Environment Food andRural Affairs (Defra) (UK Nationalfunding) blue bars represent EU fundingand green private sector funding (Terravestaand CERES) and GIANT‐LINK andMiscanthus Upscaling Technology (MUST)are publicndashprivate‐initiatives (PPI)
CLIFTON‐BROWN ET AL | 25
This technology is also being applied in switchgrass andmiscanthus (Table 5) but the future of CRISPR in com-mercial breeding for the European market is uncertain inthe light of recent ECJ 2018 rulings
There is integration of research and plant breeding itselfin all four PBCs Therefore estimating the ongoing costs ofmaintaining these breeding programmes is difficult Invest-ment in research also seeks wider benefits associated withtechnological advances in plant science rather than cultivardevelopment per se However in all cases the conventionalbreeding cycle shown in Figure 1 is the basic ldquoenginerdquo withmolecular technologies (‐omics) serving to accelerate thisengine The history of the development shows that the exis-tence of these breeding programmes is essential to gain ben-efits from the biological research Despite this it is thisessential step that is at most risk from reductions in invest-ment A conventional breeding programme typicallyrequires a breeder and several technicians who are supported
over the long term (20ndash30 years Figure 1) at costs of about05 to 10 million Euro per year (as of 2018) The analysisreported here shows that the time needed to perform onecycle of conventional breeding bringing germplasm fromthe wild to a commercial hybrid ranged from 11 years inswitchgrass to 26 years in poplar (Figure 1) In a maturecrop grown on over 100000 ha with effective cultivar pro-tection and a suitable business model this level of revenuecould come from royalties Until such levels are reachedPBCs lie in the innovation valley of death (Figure 3) andneed public support
Applying industrial ldquotechnology readiness levelsrdquo (TRL)originally developed for aerospace (Heacuteder 2017) to ourplant breeding efforts we estimate many promising hybridscultivars are at TRL levels of 3ndash4 In Table 3 experts ineach crop estimate that it would take 3 years from now toupscale planting material from leading cultivars in plot tri-als to 100 ha
FIGURE 3 A schematic relating some of the steps in the innovation chain from relatively basic crop science research through to thedeployment in commercial cropping systems and value chains The shape of the funnel above the expanding development and deploymentrepresents the availability of investment along the development chain from relatively basic research at the top to the upscaled deployment at thebottom Plant breeding links the research effort with the development of cropping systems The constriction represents the constrained fundingfor breeding that links conventional public research investment and the potential returns from commercial development The handover pointsbetween publicly funded work to develop the germplasm resources (often known as prebreeding) the breeding and the subsequent cropdevelopment are shown on the left The constriction point is aggravated by the lack academic rewards for this essential breeding activity Theoutcome is such that this innovation system is constrained by the precarious resourcing of plant breeding The authorsrsquo assessment ofdevelopment status of the four species is shown (poplar having two one for short‐rotation coppice (SRC) poplar and one for the more traditionalshort‐rotation forestry (SRF)) The four new perennial biomass crops (PBCs) are now in the critical phase of depending of plant breedingprogress without the income stream from a large crop production base
26 | CLIFTON‐BROWN ET AL
Taking the UK example mentioned in the introductionplanting rates of ~35000 ha per year from 2022 onwardsare needed to reach over 1 m ha by 2050 Ongoing workin the UK‐funded Miscanthus Upscaling Technology(MUST) project shows that ramping annual hybrid seedproduction from the current level of sufficient seed for10 ha in 2018 to 35000 ha would take about 5 yearsassuming no setbacks If current hybrids of any of the fourPBCs in the upscaling pipeline fail on any step for exam-ple lower than expected multiplication rates or unforeseenagronomic barriers then further selections from ongoingbreeding are needed to replace earlier candidates
In conclusion the breeding foundations have been laidwell for switchgrass miscanthus willow and poplar owingto public funding over the long time periods necessaryImproved cultivars or genotypes are available that could bescaled up over a few years if real sustained market oppor-tunities emerged in response to sustained favourable poli-cies and industrial market pull The potential contributionsof growing and using these PBCs for socioeconomic andenvironmental benefits are clear but how farmers andothers in commercial value chains are rewarded for mass‐scale deployment as is necessary is not obvious at presentTherefore mass‐scale deployment of these lignocellulosecrops needs developments outside the breeding arenas todrive breeding activities more rapidly and extensively
8 | UNCITED REFERENCE
Huang et al (2019)
ACKNOWLEDGEMENTS
UK The UK‐led miscanthus research and breeding wasmainly supported by the Biotechnology and BiologicalSciences Research Council (BBSRC) Department for Envi-ronment Food and Rural Affairs (Defra) the BBSRC CSPstrategic funding grant BBCSP17301 Innovate UKBBSRC ldquoMUSTrdquo BBN0161491 CERES Inc and Ter-ravesta Ltd through the GIANT‐LINK project (LK0863)Genomic selection and genomewide association study activi-ties were supported by BBSRC grant BBK01711X1 theBBSRC strategic programme grant on Energy Grasses ampBio‐refining BBSEW10963A01 The UK‐led willow RampDwork reported here was supported by BBSRC (BBSEC00005199 BBSEC00005201 BBG0162161 BBE0068331 BBG00580X1 and BBSEC000I0410) Defra(NF0424 NF0426) and the Department of Trade and Indus-try (DTI) (BW6005990000) IT The Brain Gain Program(Rientro dei cervelli) of the Italian Ministry of EducationUniversity and Research supports Antoine Harfouche USContributions by Gerald Tuskan to this manuscript were sup-ported by the Center for Bioenergy Innovation a US
Department of Energy Bioenergy Research Center supportedby the Office of Biological and Environmental Research inthe DOE Office of Science under contract number DE‐AC05‐00OR22725 Willow breeding efforts at CornellUniversity have been supported by grants from the USDepartment of Agriculture National Institute of Food andAgriculture Contributions by the University of Illinois weresupported primarily by the DOE Office of Science Office ofBiological and Environmental Research (BER) grant nosDE‐SC0006634 DE‐SC0012379 and DE‐SC0018420 (Cen-ter for Advanced Bioenergy and Bioproducts Innovation)and the Energy Biosciences Institute EU We would like tofurther acknowledge contributions from the EU projectsldquoOPTIMISCrdquo FP7‐289159 on miscanthus and ldquoWATBIOrdquoFP7‐311929 on poplar and miscanthus as well as ldquoGRACErdquoH2020‐EU326 Biobased Industries Joint Technology Ini-tiative (BBI‐JTI) Project ID 745012 on miscanthus
CONFLICT OF INTEREST
The authors declare that progress reported in this paperwhich includes input from industrial partners is not biasedby their business interests
ORCID
John Clifton‐Brown httporcidorg0000-0001-6477-5452Antoine Harfouche httporcidorg0000-0003-3998-0694Lawrence B Smart httporcidorg0000-0002-7812-7736Danny Awty‐Carroll httporcidorg0000-0001-5855-0775VasileBotnari httpsorcidorg0000-0002-0470-0384Lindsay V Clark httporcidorg0000-0002-3881-9252SalvatoreCosentino httporcidorg0000-0001-7076-8777Lin SHuang httporcidorg0000-0003-3079-326XJon P McCalmont httporcidorg0000-0002-5978-9574Paul Robson httporcidorg0000-0003-1841-3594AnatoliiSandu httporcidorg0000-0002-7900-448XDaniloScordia httporcidorg0000-0002-3822-788XRezaShafiei httporcidorg0000-0003-0891-0730Gail Taylor httporcidorg0000-0001-8470-6390IrisLewandowski httporcidorg0000-0002-0388-4521
REFERENCES
Alburquerque N Baldacci‐Cresp F Baucher M Casacuberta JM Collonnier C ElJaziri M hellip Burgos L (2016) New trans-formation technologies for trees In C Vettori F Gallardo H
CLIFTON‐BROWN ET AL | 27
Haumlggman V Kazana F Migliacci G Pilateamp M Fladung(Eds) Biosafety of forest transgenic trees (pp 31ndash66) DordrechtNetherlands Springer
Argus G W (2007) Salix (Salicaceae) distribution maps and a syn-opsis of their classification in North America north of MexicoHarvard Papers in Botany 12 335ndash368 httpsdoiorg1031001043-4534(2007)12[335SSDMAA]20CO2
Atienza S G Ramirez M C amp Martin A (2003) Mapping‐QTLscontrolling flowering date in Miscanthus sinensis Anderss CerealResearch Communications 31 265ndash271
Atienza S G Satovic Z Petersen K K Dolstra O amp Martin A(2003) Identification of QTLs influencing agronomic traits inMiscanthus sinensis Anderss I Total height flag‐leaf height andstem diameter Theoretical and Applied Genetics 107 123ndash129
Atienza S G Satovic Z Petersen K K Dolstra O amp Martin A(2003) Influencing combustion quality in Miscanthus sinensisAnderss Identification of QTLs for calcium phosphorus and sul-phur content Plant Breeding 122 141ndash145
Bdeir R Muchero W Yordanov Y Tuskan G A Busov V ampGailing O (2017) Quantitative trait locus mapping of Populusbark features and stem diameter BMC Plant Biology 17 224httpsdoiorg101186s12870-017-1166-4
Beard T R Ford G S Koutsky T M amp Spiwak L J (2009) AValley of Death in the innovation sequence An economic investi-gation Research Evaluation 18 343ndash356 httpsdoiorg103152095820209X481057
Berguson W McMahon B amp Riemenschneider D (2017) Addi-tive and non‐additive genetic variances for tree growth in severalhybrid poplar populations and implications regarding breedingstrategy Silvae Genetica 66(1) 33ndash39 httpsdoiorg101515sg-2017-0005
Berlin S Trybush S O Fogelqvist J Gyllenstrand N Halling-baumlck H R Aringhman I hellip Hanley S J (2014) Genetic diversitypopulation structure and phenotypic variation in European Salixviminalis L (Salicaceae) Tree Genetics amp Genomes 10 1595ndash1610 httpsdoiorg101007s11295-014-0782-5
Bernardo R amp Yu J (2007) Prospects for genomewide selectionfor quantitative traits in maize Crop Science 47 1082ndash1090httpsdoiorg102135cropsci2006110690
Biswal A K Atmodjo M A Li M Baxter H L Yoo C GPu Y hellip Mohnen D (2018) Sugar release and growth of bio-fuel crops are improved by downregulation of pectin biosynthesisNature Biotechnology 36 249ndash257
Bmu (2009) National biomass action plan for Germany (p 17) Bun-desministerium fuumlr Umwelt Naturschutz und ReaktorsicherheitBundesministerium fuumlr Ernaumlhrung (BMU) amp Landwirtschaft undVerbraucherschutz (BMELV) 11055 Berlin | Germany Retrievedfrom httpwwwbmeldeSharedDocsDownloadsENPublicationsBiomassActionPlanpdf__blob=publicationFile
Brummer E C (1999) Capturing heterosis in forage crop cultivardevelopment Crop Science 39 943ndash954 httpsdoiorg102135cropsci19990011183X003900040001x
Budsberg E Crawford J T Morgan H Chin W S Bura R ampGustafson R (2016) Hydrocarbon bio‐jet fuel from bioconver-sion of poplar biomass Life cycle assessment Biotechnology forBiofuels 9 170 httpsdoiorg101186s13068-016-0582-2
Burris K P Dlugosz E M Collins A G Stewart C N amp Lena-ghan S C (2016) Development of a rapid low‐cost protoplasttransfection system for switchgrass (Panicum virgatum L) Plant
Cell Reports 35 693ndash704 httpsdoiorg101007s00299-015-1913-7
Casler M (2012) Switchgrass breeding genetics and genomics InA Monti (Ed) Switchgrass (pp 29ndash54) London UK Springer
Casler M Mitchell R amp Vogel K (2012) Switchgrass In CKole C P Joshi amp D R Shonnard (Eds) Handbook of bioen-ergy crop plants Vol 2 New York NY Taylor amp Francis
Casler M D amp Ramstein G P (2018) Breeding for biomass yieldin switchgrass using surrogate measures of yield BioEnergyResearch 11 6ndash12
Casler M D Sosa S Hoffman L Mayton H Ernst C Adler PR hellip Bonos S A (2017) Biomass Yield of Switchgrass Culti-vars under High‐ versus Low‐Input Conditions Crop Science 57821ndash832 httpsdoiorg102135cropsci2016080698
Casler M D amp Vogel K P (2014) Selection for biomass yield inupland lowland and hybrid switchgrass Crop Science 54 626ndash636 httpsdoiorg102135cropsci2013040239
Casler M D Vogel K P amp Harrison M (2015) Switchgrassgermplasm resources Crop Science 55 2463ndash2478 httpsdoiorg102135cropsci2015020076
Casler M D Vogel K P Lee D K Mitchell R B Adler P RSulc R M hellip Moore K J (2018) 30 Years of progress towardincreasing biomass yield of switchgrass and big bluestem CropScience 58 1242
Clark L V Brummer J E Głowacka K Hall M C Heo K PengJ hellip Sacks E J (2014) A footprint of past climate change on thediversity and population structure of Miscanthus sinensis Annals ofBotany 114 97ndash107 httpsdoiorg101093aobmcu084
Clark L V Dzyubenko E Dzyubenko N Bagmet L SabitovA Chebukin P hellip Sacks E J (2016) Ecological characteristicsand in situ genetic associations for yield‐component traits of wildMiscanthus from eastern Russia Annals of Botany 118 941ndash955
Clark L V Jin X Petersen K K Anzoua K G Bagmet LChebukin P hellip Sacks E J(2018) Population structure of Mis-canthus sacchariflorus reveals two major polyploidization eventstetraploid‐mediated unidirectional introgression from diploid Msinensis and diversity centred around the Yellow Sea Annals ofBotany 20 1ndash18 httpsdoiorg101093aobmcy1161
Clark L V Stewart J R Nishiwaki A Toma Y Kjeldsen J BJoslashrgensen U hellip Sacks E J (2015) Genetic structure ofMiscanthus sinensis and Miscanthus sacchariflorus in Japan indi-cates a gradient of bidirectional but asymmetric introgressionJournal of Experimental Botany 66 4213ndash4225
Clifton‐Brown J Hastings A Mos M McCalmont J P Ashman CAwty‐Carroll DhellipCracroft‐EleyW (2017) Progress in upscalingMis-canthus biomass production for the European bio‐economy with seed‐based hybridsGlobal Change Biology Bioenergy 9 6ndash17
Clifton‐Brown J Schwarz K U amp Hastings A (2015) History ofthe development of Miscanthus as a bioenergy crop From smallbeginnings to potential realisation Biology and Environment‐Pro-ceedings of the Royal Irish Academy 115B 45ndash57
Clifton‐Brown J Senior H Purdy S Horsnell R Lankamp BMuumlennekhoff A-K hellip Bentley A R (2018) Investigating thepotential of novel non‐woven fabrics to increase pollination effi-ciency in plant breeding PloS One (Accepted)
Crawford J T Shan CW Budsberg E Morgan H Bura R ampGus-tafson R (2016) Hydrocarbon bio‐jet fuel from bioconversion ofpoplar biomass Techno‐economic assessment Biotechnology forBiofuels 9 141 httpsdoiorg101186s13068-016-0545-7
28 | CLIFTON‐BROWN ET AL
Dalton S (2013) Biotechnology of Miscanthus In S M Jain amp SD Gupta (Eds) Biotechnology of neglected and underutilizedcrops (pp 243ndash294) Dordrecht Netherlands Springer
Davey C L Robson P Hawkins S Farrar K Clifton‐Brown J CDonnison I S amp Slavov G T (2017) Genetic relationshipsbetween spring emergence canopy phenology and biomass yieldincrease the accuracy of genomic prediction in Miscanthus Journalof Experimental Botany 68 5093ndash5102 httpsdoiorg101093jxberx339
David X amp Anderson X (2002) Aspen and larch genetics cooper-ative annual report 13 St Paul MN Department of ForestResources University of Minnesota
Deblock M (1990) Factors influencing the tissue‐culture and theAgrobacterium‐Tumefaciens‐mediated transformation of hybridaspen and poplar clones Plant Physiology 93 1110ndash1116httpsdoiorg101104pp9331110
Defra (2002) The role of future public research investment in thegenetic improvement of UK‐grown crops (p 220) LondonRetrieved from sciencesearchdefragovukDefaultaspxMenu=MenuampModule=MoreampLocation=NoneampCompleted=220ampProjectID=10412
Dewoody J Trewin H amp Taylor G (2015) Genetic and morpho-logical differentiation in Populus nigra L Isolation by coloniza-tion or isolation by adaptation Molecular Ecology 24 2641ndash2655
Dong H Liu S Clark L V Sharma S Gifford J M Juvik JA hellip Sacks E J (2018) Genetic mapping of biomass yield inthree interconnected Miscanthus populations Global Change Biol-ogy Bioenergy 10 165ndash185
Dukes (2017) Digest of UK Energy Statistics (DUKES) (p 264) Lon-don UK Department for Business Energy amp Industrial StrategyRetrieved from wwwgovukgovernmentcollectionsdigest-of-uk-energy-statistics-dukes
Eckenwalder J (1996) Systematics and evolution of Populus In RStettler H Bradshaw J Heilman amp T Hinckley (Eds) Biologyof Populus and its implications for management and conservation(pp 7‐32) Ottawa ON National Research Council of Canada
Elshire R J Glaubitz J C Sun Q Poland J A Kawamoto KBuckler E S amp Mitchell S E (2011) A robust simple geno-typing‐by‐sequencing (GBS) approach for high diversity speciesPloS One 6 e19379
Evans H (2016) Bioenergy crops in the UK Case studies of suc-cessful whole farm integration (p 23) Loughborough UKEnergy Technologies Institute Retrieved from httpswwweticouklibrarybioenergy-crops-in-the-uk-case-studies-on-successful-whole-farm-integration-evidence-pack
Evans H (2017) Increasing UK biomass production through moreproductive use of land (p 7) Loughborough UK Energy Tech-nologies Institute Retrieved from httpswwweticouklibraryan-eti-perspective-increasing-uk-biomass-production-through-more-productive-use-of-land
Evans J Sanciangco M D Lau K H Crisovan E Barry KDaum C hellip Buell C R (2018) Extensive genetic diversity ispresent within North American switchgrass germplasm The PlantGenome 11 1ndash16 httpsdoiorg103835plantgenome2017060055
Evans L M Slavov G T Rodgers‐Melnick E Martin J RanjanP Muchero W hellip DiFazio S P (2014) Population genomicsof Populus trichocarpa identifies signatures of selection and
adaptive trait associations Nature Genetics 46 1089ndash1096httpsdoiorg101038ng3075
Fabio E S Volk T A Miller R O Serapiglia M J Gauch HG Van Rees K C J hellip Smart L B (2017) Genotypetimes envi-ronment interaction analysis of North American shrub willowyield trials confirms superior performance of triploid hybrids Glo-bal Change Biology Bioenergy 9 445ndash459 httpsdoiorg101111gcbb12344
Fahrenkrog A M Neves L G Resende M F R Vazquez A Ide los Campos G Dervinis C hellip Kirst M (2017) Genome‐wide association study reveals putative regulators of bioenergytraits in Populus deltoides New Phytologist 213 799ndash811
Fan D Liu T T Li C F Jiao B Li S Hou Y S amp Luo KM (2015) Efficient CRISPRCas9‐mediated targeted mutagenesisin Populus in the first generation Scientific Reports 5 12217
Feng X P Lourgant K Castric V Saumitou-Laprade P ZhengB S Jiang D M amp Brancourt-Hulmel M (2014) The discov-ery of natural accessions related to Miscanthus x giganteus usingchloroplast DNA Crop Science 54 1645ndash1655
Fillatti J J Sellmer J Mccown B Haissig B amp Comai L(1987) Agrobacterium-mediated transformation and regenerationof Populus Molecular amp General Genetics 206 192ndash199
Fu Y B (2015) Understanding crop genetic diversity under modernplant breeding Theoretical and Applied Genetics 128 2131ndash2142 httpsdoiorg101007s00122-015-2585-y
Fu C Mielenz J R Xiao X Ge Y Hamilton C Y RodriguezM hellip Wang Z‐Y (2011) Genetic manipulation of ligninreduces recalcitrance and improves ethanol production fromswitchgrass Proceedings of the National Academy of Sciences ofthe United States of America 108 3803ndash3808 httpsdoiorg101073pnas1100310108
Gao C (2018) The future of CRISPR technologies in agricultureNature Reviews Molecular Cell Biology 19(5) 275ndash276
Gegas V Gay A Camargo A amp Doonan J (2014) Challengesof Crop Phenomics in the Post-genomic Era In J Hancock (Ed)Phenomics (pp 142ndash171) Boca Raton FL CRC Press
Geraldes A DiFazio S P Slavov G T Ranjan P Muchero WHannemann J hellip Tuskan G A (2013) A 34K SNP genotypingarray for Populus trichocarpa Design application to the study ofnatural populations and transferability to other Populus speciesMolecular Ecology Resources 13 306ndash323
Gifford J M Chae W B Swaminathan K Moose S P amp JuvikJ A (2015) Mapping the genome of Miscanthus sinensis forQTL associated with biomass productivity Global Change Biol-ogy Bioenergy 7 797ndash810
Goggin F L Lorence A amp Topp C N (2015) Applying high‐throughput phenotyping to plant‐insect interactions Picturingmore resistant crops Current Opinion in Insect Science 9 69ndash76httpsdoiorg101016jcois201503002
Grabowski P P Evans J Daum C Deshpande S Barry K WKennedy M hellip Casler M D (2017) Genome‐wide associationswith flowering time in switchgrass using exome‐capture sequenc-ing data New Phytologist 213 154ndash169 httpsdoiorg101111nph14101
Greef J M amp Deuter M (1993) Syntaxonomy of Miscanthus xgiganteus GREEF et DEU Angewandte Botanik 67 87ndash90
Guerra F P Richards J H Fiehn O Famula R Stanton B JShuren R hellip Neale D B (2016) Analysis of the genetic varia-tion in growth ecophysiology and chemical and metabolomic
CLIFTON‐BROWN ET AL | 29
composition of wood of Populus trichocarpa provenances TreeGenetics amp Genomes 12 6 httpsdoiorg101007s11295-015-0965-8
Guo H P Shao R X Hong C T Hu H K Zheng B S ampZhang Q X (2013) Rapid in vitro propagation of bioenergy cropMiscanthus sacchariflorus Applied Mechanics and Materials 260181ndash186
Hallingbaumlck H R Fogelqvist J Powers S J Turrion‐Gomez JRossiter R Amey J hellip Roumlnnberg‐Waumlstljung A‐C (2016)Association mapping in Salix viminalis L (Salicaceae)ndashIdentifica-tion of candidate genes associated with growth and phenologyGlobal Change Biology Bioenergy 8 670ndash685
Hanley S J amp Karp A (2014) Genetic strategies for dissectingcomplex traits in biomass willows (Salix spp) Tree Physiology34 1167ndash1180 httpsdoiorg101093treephystpt089
Harfouche A Meilan R amp Altman A (2011) Tree genetic engi-neering and applications to sustainable forestry and biomass pro-duction Trends in Biotechnology 29 9ndash17 httpsdoiorg101016jtibtech201009003
Harfouche A Meilan R amp Altman A (2014) Molecular and phys-iological responses to abiotic stress in forest trees and their rele-vance to tree improvement Tree Physiology 34 1181ndash1198httpsdoiorg101093treephystpu012
Harfouche A Meilan R Kirst M Morgante M Boerjan WSabatti M amp Mugnozza G S (2012) Accelerating the domesti-cation of forest trees in a changing world Trends in PlantScience 17 64ndash72 httpsdoiorg101016jtplants201111005
Hastings A Clifton‐Brown J Wattenbach M Mitchell C PStampfl P amp Smith P (2009) Future energy potential of Mis-canthus in Europe Global Change Biology Bioenergy 1 180ndash196
Hastings A Mos M Yesufu J AMccalmont J Schwarz KShafei RhellipClifton-Brown J (2017) Economic and environmen-tal assessment of seed and rhizome propagated Miscanthus in theUK Frontiers in Plant Science 8 16 httpsdoiorg103389fpls201701058
Heacuteder M (2017) From NASA to EU The evolution of the TRLscale in Public Sector Innovation The Innovation Journal 22 1ndash23 httpsdoiorg103389fpls201701058
Heffner E L Sorrells M E amp Jannink J L (2009) Genomicselection for crop improvement Crop Science 49 1ndash12 httpsdoiorg102135cropsci2008080512
Hodkinson T R Klaas M Jones M B Prickett R amp Barth S(2015) Miscanthus A case study for the utilization of naturalgenetic variation Plant Genetic Resources‐Characterization andUtilization 13 219ndash237 httpsdoiorg101017S147926211400094X
Hodkinson T R amp Renvoize S (2001) Nomenclature of Miscant-hus xgiganteus (Poaceae) Kew Bulletin 56 759ndash760 httpsdoiorg1023074117709
Houle D Govindaraju D R amp Omholt S (2010) Phenomics Thenext challenge Nature Reviews Genetics 11 855ndash866 httpsdoiorg101038nrg2897
Huang X H Wei X H Sang T Zhao Q Feng Q Zhao Yhellip Han B (2010) Genome‐wide association studies of 14 agro-nomic traits in rice landraces Nature Genetics 42 961ndashU976
Hwang O‐J Cho M‐A Han Y‐J Kim Y‐M Lim S‐H KimD‐S hellip Kim J‐I (2014) Agrobacterium‐mediated genetic trans-formation of Miscanthus sinensis Plant Cell Tissue and OrganCulture (PCTOC) 117 51ndash63
Hwang O‐J Lim S‐H Han Y‐J Shin A‐Y Kim D‐S ampKim J‐I (2014) Phenotypic characterization of transgenic Mis-canthus sinensis plants overexpressing Arabidopsis phytochromeB International Journal of Photoenergy 2014501016
Ilstedt B (1996) Genetics and performance of Belgian poplar clonestested in Sweden International Journal of Forest Genetics 3183ndash195
Ingvarsson P K Garcia M V Hall D Luquez V amp Jansson S(2006) Clinal variation in phyB2 a candidate gene for day‐length‐induced growth cessation and bud set across a latitudinalgradient in European aspen (Populus tremula) Genetics 1721845ndash1853
Ingvarsson P K Garcia M V Luquez V Hall D amp Jansson S(2008) Nucleotide polymorphism and phenotypic associationswithin and around the phytochrome B2 locus in European aspen(Populus tremula Salicaceae) Genetics 178 2217ndash2226 httpsdoiorg101534genetics107082354
Jannink J‐L Lorenz A J amp Iwata H (2010) Genomic selectionin plant breeding From theory to practice Briefings in FunctionalGenomics 9 166ndash177 httpsdoiorg101093bfgpelq001
Jiang J Guan Y Mccormick S Juvik J Lubberstedt T amp FeiS‐Z (2017) Gametophytic self‐incompatibility is operative inMiscanthus sinensis (Poaceae) and is affected by pistil age CropScience 57 1948ndash1956
Jones H D (2015a) Future of breeding by genome editing is in thehands of regulators GM Crops amp Food 6 223ndash232
Jones H D (2015b) Regulatory uncertainty over genome editingNature Plants 1 14011
Kalinina O Nunn C Sanderson R Hastings A F S van der Weijde TOumlzguumlven MhellipClifton‐Brown J C (2017) ExtendingMiscanthus cul-tivation with novel germplasm at six contrasting sites Frontiers in PlantScience 8563 httpsdoiorg103389fpls201700563
Karp A Hanley S J Trybush S O Macalpine W Pei M ampShield I (2011) Genetic improvement of willow for bioenergyand biofuels free access Journal of Integrative Plant Biology 53151ndash165 httpsdoiorg101111j1744-7909201001015x
Kersten B Faivre Rampant P Mader M Le Paslier M‐C Bou-non R Berard A hellip Fladung M (2016) Genome sequences ofPopulus tremula chloroplast and mitochondrion Implications forholistic poplar breeding PloS One 11 e0147209 httpsdoiorg101371journalpone0147209
Kiesel A Nunn C Iqbal Y Van der Weijde T Wagner MOumlzguumlven M hellip Lewandowski I (2017) Site‐specific manage-ment of Miscanthus genotypes for combustion and anaerobicdigestion A comparison of energy yields Frontiers in PlantScience 8 347 httpsdoiorg103389fpls201700347
Kim H Kim S T Ryu J Kang B C Kim J S amp Kim S G(2017) CRISPRCpf1‐mediated DNA‐free plant genome editingNature Communications 8 14406 httpsdoiorg101038ncomms14406
King Z R Bray A L Lafayette P R amp Parrott W A (2014)Biolistic transformation of elite genotypes of switchgrass (Pan-icum virgatum L) Plant Cell Reports 33 313ndash322 httpsdoiorg101007s00299-013-1531-1
Kopp R F Maynard C A De Niella P R Smart L B amp Abra-hamson L P (2002) Collection and storage of pollen from Salix(Salicaceae) American Journal of Botany 89 248ndash252
Krzyżak J Pogrzeba M Rusinowski S Clifton‐Brown J McCal-mont J P Kiesel A hellip Mos M (2017) Heavy metal uptake
30 | CLIFTON‐BROWN ET AL
by novel miscanthus seed‐based hybrids cultivated in heavy metalcontaminated soil Civil and Environmental Engineering Reports26 121ndash132 httpsdoiorg101515ceer-2017-0040
Kuzovkina Y A amp Quigley M F (2005) Willows beyond wet-lands Uses of Salix L species for environmental projects WaterAir and Soil Pollution 162 183ndash204 httpsdoiorg101007s11270-005-6272-5
Lazarus W Headlee W L amp Zalesny R S (2015) Impacts of sup-plyshed‐level differences in productivity and land costs on the eco-nomics of hybrid poplar production in Minnesota USA BioEnergyResearch 8 231ndash248 httpsdoiorg101007s12155-014-9520-y
Lewandowski I (2015) Securing a sustainable biomass supply in agrowing bioeconomy Global Food Security 6 34ndash42 httpsdoiorg101016jgfs201510001
Lewandowski I (2016) The role of perennial biomass crops in agrowing bioeconomy In S Barth D Murphy‐Bokern O Kalin-ina G Taylor amp M Jones (Eds) Perennial biomass crops for aresource‐constrained world (pp 3ndash13) Cham SwitzerlandSpringer Switzerland AG
Li J L Meilan R Ma C Barish M amp Strauss S H (2008)Stability of herbicide resistance over 8 years of coppice in field‐grown genetically engineered poplars Western Journal of AppliedForestry 23 89ndash93
Li R Y amp Qu R D (2011) High throughput Agrobacterium‐medi-ated switchgrass transformation Biomass amp Bioenergy 35 1046ndash1054 httpsdoiorg101016jbiombioe201011025
Lindegaard K N amp Barker J H A (1997) Breeding willows forbiomass Aspects of Applied Biology 49 155ndash162
Linde‐Laursen I B (1993) Cytogenetic analysis of MiscanthusGiganteus an interspecific hybrid Hereditas 119 297ndash300httpsdoiorg101111j1601-5223199300297x
Liu Y Merrick P Zhang Z Z Ji C H Yang B amp Fei S Z(2018) Targeted mutagenesis in tetraploid switchgrass (Panicumvirgatum L) using CRISPRCas9 Plant Biotechnology Journal16 381ndash393
Liu L L Wu Y Q Wang Y W amp Samuels T (2012) A high‐density simple sequence repeat‐based genetic linkage map ofswitchgrass G3 (Bethesda Md) 2 357ndash370
Lovett A Suumlnnenberg G amp Dockerty T (2014) The availabilityof land for perennial energy crops in Great Britain GlobalChange Biology Bioenergy 6 99ndash107 httpsdoiorg101111gcbb12147
Lozano‐Juste J amp Cutler S R (2014) Plant genome engineering infull bloom Trends in Plant Science 19 284ndash287 httpsdoiorg101016jtplants201402014
Lu F Lipka A E Glaubitz J Elshire R Cherney J H CaslerM D hellip Costich D E (2013) Switchgrass Genomic DiversityPloidy and Evolution Novel Insights from a Network‐Based SNPDiscovery Protocol Plos Genetics 9 e1003215
Ludovisi R Tauro F Salvati R Khoury S Mugnozza G S ampHarfouche A (2017) UAV‐based thermal imaging for high‐throughput field phenotyping of black poplar response to droughtFrontiers in Plant Science 81681
Macalpine W Shield I amp Karp A (2010) Seed to near market vari-ety the BEGIN willow breeding pipeline 2003ndash2010 and beyond InBioten conference proceedings Birmingham pp 21ndash23
Macalpine W J Shield I F Trybush S O Hayes C M ampKarp A (2008) Overcoming barriers to crossing in willow (Salixspp) breeding Aspects of Applied Biology 90 173ndash180
Mahfouz M M (2017) The efficient tool CRISPR‐Cpf1 NaturePlants 3 17028
Malinowska M Donnison I S amp Robson P R (2017) Phenomicsanalysis of drought responses in Miscanthus collected from differ-ent geographical locations Global Change Biology Bioenergy 978ndash91
Maroder H L Prego I A Facciuto G R amp Maldonado S B(2000) Storage behaviour of Salix alba and Salix matsudanaseeds Annals of Botany 86 1017ndash1021 httpsdoiorg101006anbo20001265
Martinez‐Reyna J amp Vogel K (2002) Incompatibility systems inswitchgrass Crop Science 42 1800ndash1805 httpsdoiorg102135cropsci20021800
Maurya J P amp Bhalerao R P (2017) Photoperiod‐and tempera-ture‐mediated control of growth cessation and dormancy in treesA molecular perspective Annals of Botany 120 351ndash360httpsdoiorg101093aobmcx061
McCalmont J P Hastings A Mcnamara N P Richter G MRobson P Donnison I S amp Clifton‐Brown J (2017) Environ-mental costs and benefits of growing Miscanthus for bioenergy inthe UK Global Change Biology Bioenergy 9 489ndash507
McCracken A R amp Dawson W M (1997) Using mixtures of wil-low clones as a means of controlling rust disease Aspects ofApplied Biology 49 97ndash103
McKown A D Klaacutepště J Guy R D Geraldes A Porth I Hanne-mann JhellipDouglas C J (2014) Genome‐wide association implicatesnumerous genes underlying ecological trait variation in natural popula-tions ofPopulus trichocarpaNew Phytologist 203 535ndash553
Merrick P amp Fei S Z (2015) Plant regeneration and genetic transforma-tion in switchgrass ndash A review Journal of Integrative Agriculture 14483ndash493 httpsdoiorg101016S2095-3119(14)60921-7
Moreno‐Mateos M A Fernandez J P Rouet R Lane M AVejnar C E Mis E hellip Giraldez A J (2017) CRISPR‐Cpf1mediates efficient homology‐directed repair and temperature‐con-trolled genome editing Nature Communications 8 2024 httpsdoiorg101038s41467-017-01836-2
Mosseler A (1990) Hybrid performance and species crossabilityrelationships in willows (Salix) Canadian Journal of Botany 682329ndash2338
Muchero W Guo J DiFazio S P Chen J‐G Ranjan P Slavov GT hellip Tuskan G A (2015) High‐resolution genetic mapping ofallelic variants associated with cell wall chemistry in Populus BMCGenomics 16 24 httpsdoiorg101186s12864-015-1215-z
Neale D B amp Kremer A (2011) Forest tree genomics Growingresources and applications Nature Reviews Genetics 12 111ndash122 httpsdoiorg101038nrg2931
Nunn C Hastings A F S J Kalinina O Oumlzguumlven M SchuumlleH Tarakanov I G hellip Clifton‐Brown J C (2017) Environmen-tal influences on the growing season duration and ripening ofdiverse Miscanthus germplasm grown in six countries Frontiersin Plant Science 8 907 httpsdoiorg103389fpls201700907
Ogawa Y Honda M Kondo Y amp Hara‐Nishimura I (2016) Anefficient Agrobacterium‐mediated transformation method forswitchgrass genotypes using Type I callus Plant Biotechnology33 19ndash26
Ogawa Y Shirakawa M Koumoto Y Honda M Asami Y KondoY ampHara‐Nishimura I (2014) A simple and reliable multi‐gene trans-formation method for switchgrass Plant Cell Reports 33 1161ndash1172httpsdoiorg101007s00299-014-1605-8
CLIFTON‐BROWN ET AL | 31
Okada M Lanzatella C Saha M C Bouton J Wu R L amp Tobias CM (2010) Complete switchgrass genetic maps reveal subgenomecollinearity preferential pairing and multilocus interactions Genetics185 745ndash760 httpsdoiorg101534genetics110113910
Palomo‐Riacuteos E Macalpine W Shield I Amey J Karaoğlu C WestJ hellip Jones H D (2015) Efficient method for rapid multiplication ofclean and healthy willow clones via in vitro propagation with broadgenotype applicability Canadian Journal of Forest Research 451662ndash1667 httpsdoiorg101139cjfr-2015-0055
Pilate G Allona I Boerjan W Deacutejardin A Fladung M Gal-lardo F hellip Halpin C (2016) Lessons from 25 years of GM treefield trials in Europe and prospects for the future In C Vettori FGallardo H Haumlggman V Kazana F Migliacci G Pilateamp MFladung (Eds) Biosafety of forest transgenic trees (pp 67ndash100)Dordrecht Netherlands Springer Switzerland AG
Pinosio S Giacomello S Faivre-Rampant P Taylor G Jorge VLe Paslier M C amp Marroni F (2016) Characterization of thepoplar pan-genome by genome-wide identification of structuralvariation Molecular Biology and Evolution 33 2706ndash2719
Porth I Klaacutepště J Skyba O Friedmann M C Hannemann JEhlting J hellip Douglas C J (2013) Network analysis reveals therelationship among wood properties gene expression levels andgenotypes of natural Populus trichocarpa accessions New Phytol-ogist 200 727ndash742
Pugesgaard S Schelde K Larsen S U Laeligrke P E amp JoslashrgensenU (2015) Comparing annual and perennial crops for bioenergyproductionndashinfluence on nitrate leaching and energy balance Glo-bal Change Biology Bioenergy 7 1136ndash1149 httpsdoiorg101111gcbb12215
Quinn L D Allen D J amp Stewart J R (2010) Invasivenesspotential of Miscanthus sinensis Implications for bioenergy pro-duction in the United States Global Change Biology Bioenergy2 310ndash320 httpsdoiorg101111j1757-1707201001062x
Rae A M Robinson K M Street N R amp Taylor G (2004)Morphological and physiological traits influencing biomass pro-ductivity in short‐rotation coppice poplar Canadian Journal ofForest Research‐Revue Canadienne De Recherche Forestiere 341488ndash1498 httpsdoiorg101139x04-033
Rambaud C Arnoult S Bluteau A Mansard M C Blassiau Camp Brancourt‐Hulmel M (2013) Shoot organogenesis in threeMiscanthus species and evaluation for genetic uniformity usingAFLP analysis Plant Cell Tissue and Organ Culture (PCTOC)113 437ndash448 httpsdoiorg101007s11240-012-0284-9
Ramstein G P Evans J Kaeppler S M Mitchell R B Vogel K PBuell C R amp Casler M D (2016) Accuracy of genomic prediction inswitchgrass (Panicum virgatum L) improved by accounting for linkagedisequilibriumG3 (Bethesda Md) 6 1049ndash1062
Rayburn A L Crawford J Rayburn C M amp Juvik J A (2009)Genome size of three Miscanthus species Plant Molecular Biol-ogy Reporter 27 184 httpsdoiorg101007s11105-008-0070-3
Richardson J Isebrands J amp Ball J (2014) Ecology and physiol-ogy of poplars and willows In J Isebrands amp J Richardson(Eds) Poplars and willows Trees for society and the environ-ment (pp 92‐123) Boston MA and Rome CABI and FAO
Robison T L Rousseau R J amp Zhang J (2006) Biomass produc-tivity improvement for eastern cottonwood Biomass amp Bioenergy30 735ndash739 httpsdoiorg101016jbiombioe200601012
Robson P R Farrar K Gay A P Jensen E F Clifton‐Brown JC amp Donnison I S (2013) Variation in canopy duration in the
perennial biofuel crop Miscanthus reveals complex associationswith yield Journal of Experimental Botany 64 2373ndash2383
Robson P Jensen E Hawkins S White S R Kenobi K Clif-ton‐Brown J hellip Farrar K (2013) Accelerating the domestica-tion of a bioenergy crop Identifying and modelling morphologicaltargets for sustainable yield increase in Miscanthus Journal ofExperimental Botany 64 4143ndash4155
Sanderson M A Adler P R Boateng A A Casler M D ampSarath G (2006) Switchgrass as a biofuels feedstock in theUSA Canadian Journal of Plant Science 86 1315ndash1325httpsdoiorg104141P06-136
Scarlat N Dallemand J F Monforti‐Ferrario F amp Nita V(2015) The role of biomass and bioenergy in a future bioecon-omy Policies and facts Environmental Development 15 3ndash34httpsdoiorg101016jenvdev201503006
Serapiglia M J Gouker F E amp Smart L B (2014) Early selec-tion of novel triploid hybrids of shrub willow with improved bio-mass yield relative to diploids BMC Plant Biology 14 74httpsdoiorg1011861471-2229-14-74
Serba D Wu L Daverdin G Bahri B A Wang X Kilian Ahellip Devos K M (2013) Linkage maps of lowland and upland tet-raploid switchgrass ecotypes BioEnergy Research 6 953ndash965httpsdoiorg101007s12155-013-9315-6
Shakoor N Lee S amp Mockler T C (2017) High throughput phe-notyping to accelerate crop breeding and monitoring of diseases inthe field Current Opinion in Plant Biology 38 184ndash192 httpsdoiorg101016jpbi201705006
Shield I Macalpine W Hanley S amp Karp A (2015) Breedingwillow for short rotation coppice energy cropping In Industrialcrops Breeding for BioEnergy and Bioproducts (pp 67ndash80) NewYork NY Springer
Sjodin A Street N R Sandberg G Gustafsson P amp Jansson S (2009)The Populus Genome Integrative Explorer (PopGenIE) A new resourcefor exploring the Populus genomeNewPhytologist 182 1013ndash1025
Slavov G amp Davey C (2017) Integrated genomic prediction inbioenergy crops In Abstracts from the International Conferenceon Developing biomass crops for future climates pp 34 24ndash27September 2017 Oriel College Oxford wwwwatbioeu
Slavov G Davey C Bosch M Robson P Donnison I ampMackay I (2018a) Genomic index selection provides a pragmaticframework for setting and refining multi‐objective breeding targetsin Miscanthus Annals of Botany (in press)
Slavov G T DiFazio S P Martin J Schackwitz W MucheroW Rodgers‐Melnick E hellip Tuskan G A (2012) Genome rese-quencing reveals multiscale geographic structure and extensivelinkage disequilibrium in the forest tree Populus trichocarpa NewPhytologist 196 713ndash725
Slavov G Davey C Robson P Donnison I amp Mackay I(2018b) Domestication of Miscanthus through genomic indexselection In Plant and Animal Genome Conference 13 January2018 San Diego CA Retrieved from httpspagconfexcompagxxvimeetingappcgiPaper30622
Slavov G T Nipper R Robson P Farrar K Allison G GBosch M hellip Jensen E (2013) Genome‐wide association studiesand prediction of 17 traits related to phenology biomass and cellwall composition in the energy grass Miscanthus sinensis NewPhytologist 201 1227ndash1239
Slavov G Robson P Jensen E Hodgson E Farrar K AllisonG hellip Donnison I (2014) Contrasting geographic patterns of
32 | CLIFTON‐BROWN ET AL
genetic variation for molecular markers vs phenotypic traits in theenergy grass Miscanthus sinensis Global Change Biology Bioen-ergy 5 562ndash571
Ślusarkiewicz‐Jarzina A Ponitka A Cerazy‐Waliszewska J Woj-ciechowicz M K Sobańska K Jeżowski S amp Pniewski T(2017) Effective and simple in vitro regeneration system of Mis-canthus sinensis Mtimes giganteus and M sacchariflorus for plant-ing and biotechnology purposes Biomass and Bioenergy 107219ndash226 httpsdoiorg101016jbiombioe201710012
Smart L amp Cameron K (2012) Shrub willow In C Kole S Joshiamp D Shonnard (Eds) Handbook of bioenergy crop plants (pp687ndash708) Boca Raton FL Taylor and Francis Group
Stanton B J (2014) The domestication and conservation of Populusand Salix genetic resources (Chapter 4) In Isebrands J ampRichardson J (Eds) Poplars and willows Trees for society andthe environment (pp 124ndash200) Rome Italy CAB InternationalFood and Agricultural Organization of the United Nations
Stanton B J Neale D B amp Li S (2010) Populus breeding Fromthe classical to the genomic approach In S Jansson R Bha-lerao amp A T Groover (Eds) Genetics and genomics of popu-lus (pp 309ndash348) New York NY Springer
Stewart J R Toma Y Fernandez F G Nishiwaki A Yamada T ampBollero G (2009) The ecology and agronomy ofMiscanthus sinensisa species important to bioenergy crop development in its native range inJapan A reviewGlobal Change Biology Bioenergy 1 126ndash153
Stott K G (1992) Willows in the service of man Proceedings of the RoyalSociety of Edinburgh Section B Biological Sciences 98 169ndash182
Strauss S H Ma C Ault K amp Klocko A L (2016) Lessonsfrom two decades of field trials with genetically modified trees inthe USA Biology and regulatory compliance In C Vettori FGallardo H Haumlggman V Kazana F Migliacci G Pilate amp MFladung (Eds) Biosafety of forest transgenic trees (pp 101ndash124)Dordrecht Netherlands Springer
Suda Y amp Argus G W (1968) Chromosome numbers of some NorthAmerican Salix Brittonia 20 191ndash197 httpsdoiorg1023072805440
Tan B (2018) Genomic selection and genome‐wide association studies todissect quantitative traits in forest trees Umearing Sweden University
Tornqvist C Taylor M Jiang Y Evans J Buell C KaepplerS amp Casler M (2018) Quantitative trait locus mapping for flow-ering time in a lowland x upland switchgrass pseudo‐F2 popula-tion The Plant Genome 11 170093
Toacuteth G Hermann T Da Silva M amp Montanarella L (2016)Heavy metals in agricultural soils of the European Union withimplications for food safety Environment International 88 299ndash309 httpsdoiorg101016jenvint201512017
Tracy W Dawson J Moore V amp Fisch J (2016) Intellectual propertyrights and public plant breeding Recommendations and proceedings ofa conference on best practices for intellectual property protection of pub-lically developed plant germplasm (p 70) University of Wisconsin‐Madison Retrieved from httpshostcalswisceduagronomywp-contentuploadssites1620162005Proceedings-IPR-Finalpdf
Trybush S Jahodovaacute Š Macalpine W amp Karp A (2008) A geneticstudy of a Salix germplasm resource reveals new insights into rela-tionships among subgenera sections and species BioEnergyResearch 1 67ndash79 httpsdoiorg101007s12155-008-9007-9
Tsai C J (2013) Next‐generation sequencing for next‐generationbreeding and more New Phytologist 198 635ndash637 httpsdoiorg101111nph12245
Tuskan G A Difazio S Jansson S Bohlmann J Grigoriev I HellstenU hellip Rokhsar D (2006) The genome of black cottonwood Populustrichocarpa (Torr ampGray) Science 313 1596ndash1604
Tuskan G A amp Walsh M E (2001) Short‐rotation woody cropsystems atmospheric carbon dioxide and carbon management AUS case study The Forestry Chronicle 77 259ndash264 httpsdoiorg105558tfc77259-2
Van Den Broek R Treffers D‐J Meeusen M Van Wijk ANieuwlaar E amp Turkenburg W (2001) Green energy or organicfood A life‐cycle assessment comparing two uses of set‐asideland Journal of Industrial Ecology 5 65ndash87 httpsdoiorg101162108819801760049477
Van Der Schoot J Pospiskova M Vosman B amp Smulders M J M(2000) Development and characterization of microsatellite markers inblack poplar (Populus nigra L) Theoretical and Applied Genetics 101317ndash322 httpsdoiorg101007s001220051485
Van Der Weijde T Dolstra O Visser R G amp Trindade L M(2017a) Stability of cell wall composition and saccharificationefficiency in Miscanthus across diverse environments Frontiers inPlant Science 7 2004
Van der Weijde T Huxley L M Hawkins S Sembiring E HFarrar K Dolstra O hellip Trindade L M (2017) Impact ofdrought stress on growth and quality of miscanthus for biofuelproduction Global Change Biology Bioenergy 9 770ndash782
Van der Weijde T Kamei C L A Severing E I Torres A FGomez L D Dolstra O hellip Trindade L M (2017) Geneticcomplexity of miscanthus cell wall composition and biomass qual-ity for biofuels BMC Genomics 18 406
Van der Weijde T Kiesel A Iqbal Y Muylle H Dolstra O Visser RG FhellipTrindade LM (2017) Evaluation ofMiscanthus sinensis bio-mass quality as feedstock for conversion into different bioenergy prod-uctsGlobal Change Biology Bioenergy 9 176ndash190
Vanholme B Cesarino I Goeminne G Kim H Marroni F VanAcker R hellip Boerjan W (2013) Breeding with rare defectivealleles (BRDA) A natural Populus nigra HCT mutant with modi-fied lignin as a case study New Phytologist 198 765ndash776
Vogel K P Mitchell R B Casler M D amp Sarath G (2014)Registration of Liberty Switchgrass Journal of Plant Registra-tions 8 242ndash247 httpsdoiorg103198jpr2013120076crc
Wang X Yamada T Kong F‐J Abe Y Hoshino Y Sato Hhellip Yamada T (2011) Establishment of an efficient in vitro cul-ture and particle bombardment‐mediated transformation systems inMiscanthus sinensis Anderss a potential bioenergy crop GlobalChange Biology Bioenergy 3 322ndash332 httpsdoiorg101111j1757-1707201101090x
Watrud L S Lee E H Fairbrother A Burdick C Reichman JR Bollman M hellip Van de Water P K (2004) Evidence forlandscape‐level pollen‐mediated gene flow from genetically modi-fied creeping bentgrass with CP4 EPSPS as a marker Proceedingsof the National Academy of Sciences of the United States of Amer-ica 101 14533ndash14538
Weisgerber H (1993) Poplar breeding for the purpose of biomassproduction in short‐rotation periods in Germany ndash Problems andfirst findings Forestry Chronicle 69 727ndash729 httpsdoiorg105558tfc69727-6
Wheeler N C Steiner K C Schlarbaum S E amp Neale D B(2015) The evolution of forest genetics and tree improvementresearch in the United States Journal of Forestry 113 500ndash510httpsdoiorg105849jof14-120
CLIFTON‐BROWN ET AL | 33
Whittaker C Macalpine W J Yates N E amp Shield I (2016)Dry matter losses and methane emissions during wood chip stor-age The impact on full life cycle greenhouse gas savings of shortrotation coppice willow for heat BioEnergy Research 9 820ndash835 httpsdoiorg101007s12155-016-9728-0
Xi Q (2000) Investigation on the distribution and potential of giant grassesin China PhD thesis Goettingen Germany Cuvillier Verlag
Xi Q amp Jezowkski S (2004) Plant resources of Triarrhena andMiscanthus species in China and its meaning for Europe PlantBreeding and Seed Science 49 63ndash77
Xue S Kalinina O amp Lewandowski I (2015) Present and future optionsfor the improvement of Miscanthus propagation techniques Renewableand Sustainable Energy Reviews 49 1233ndash1246
Yin K Q Gao C X amp Qiu J L (2017) Progress and prospectsin plant genome editing Nature Plants 3 httpsdoiorg101038nplants2017107
YookM (2016)Genetic diversity and transcriptome analysis for salt toler-ance inMiscanthus PhD thesis Seoul National University Korea
Zaidi S S E A Mahfouz M M amp Mansoor S (2017) CRISPR‐Cpf1 A new tool for plant genome editing Trends in PlantScience 22 550ndash553 httpsdoiorg101016jtplants201705001
Zalesny R S Stanturf J A Gardiner E S Perdue J H YoungT M Coyle D R hellip Hass A (2016) Ecosystem services ofwoody crop production systems BioEnergy Research 9 465ndash491 httpsdoiorg101007s12155-016-9737-z
Zhang Q X Sun Y Hu H K Chen B Hong C T Guo H Phellip Zheng B S (2012) Micropropagation and plant regenerationfrom embryogenic callus of Miscanthus sinensis Vitro Cellular ampDevelopmental Biology‐Plant 48 50ndash57 httpsdoiorg101007s11627-011-9387-y
Zhang Y Zalapa J E Jakubowski A R Price D L AcharyaA Wei Y hellip Casler M D (2011) Post‐glacial evolution of
Panicum virgatum Centers of diversity and gene pools revealedby SSR markers and cpDNA sequences Genetica 139 933ndash948httpsdoiorg101007s10709-011-9597-6
Zhao H Wang B He J Yang J Pan L Sun D amp Peng J(2013) Genetic diversity and population structure of Miscanthussinensis germplasm in China PloS One 8 e75672
Zhou X H Jacobs T B Xue L J Harding S A amp Tsai C J(2015) Exploiting SNPs for biallelic CRISPR mutations in theoutcrossing woody perennial Populus reveals 4‐coumarate CoAligase specificity and redundancy New Phytologist 208 298ndash301
Zhou R Macaya‐Sanz D Rodgers‐Melnick E Carlson C HGouker F E Evans L M hellip DiFazio S P (2018) Characteri-zation of a large sex determination region in Salix purpurea L(Salicaceae) Molecular Genetics and Genomics 1ndash16 httpsdoiorg101007s00438-018-1473-y
Zsuffa L Mosseler A amp Raj Y (1984) Prospects for interspecifichybridization in willow for biomass production In K L Perttu(Ed) Ecology and management of forest biomass production sys-tems Report 15 (pp 261ndash281) Uppsala Sweden SwedishUniversity of Agricultural Sciences
How to cite this article Clifton‐Brown J HarfoucheA Casler MD et al Breeding progress andpreparedness for mass‐scale deployment of perenniallignocellulosic biomass crops switchgrassmiscanthus willow and poplar GCB Bioenergy201800000ndash000 httpsdoiorg101111gcbb12566
34 | CLIFTON‐BROWN ET AL
Graphical AbstractThe contents of this page will be used as part of the graphical abstract of html only
It will not be published as part of main article
Plant breeding links the research effort with commercial mass upscaling The authorsrsquo assessment of development status ofthe four species is shown (poplar having two one for short rotation coppice (SRC) poplar and one for the more traditionalshort rotation forestry (SRF)) Mass scale deployment needs developments outside the breeding arenas to drive breedingactivities more rapidly and extensively
TABLE
1(Contin
ued)
Species
Switc
hgrass
Miscanthu
sW
illow
Pop
lar
Current
commercial
varietieson
the
market
US
Nocommercial
hybrids
CA2
EU1(M
timesgfrom
different
origins)
+selected
Msinforthatching
inDK
US
3(but
2aregenetically
identical)
UK25
US
8EUDElt10
FR44
IT10ndash1
5SE
~1
4US
8ndash12
Southeast8ndash
14Upp
erMidwest10
PacificNorthwest
Precom
mercial
cultivars
expected
tobe
onthemarketin
3years
US
36registered
cultivars
(halfare
rand
omseed
increasesfrom
natural
prairiesandhalfarebred
varieties)
Mostarepu
blic
releasesfew
are
protected
patented
orlicensed
NL8
seeded
hybridsvanDinterSemo
MTA
UK4
seeded
hybridsCERES(Land
OLakes)andTerravestaLtdMTA
US
Non
eFR
Non
e
EU53
registered
with
CPV
OforPB
RUK20
registered
with
CPV
OforPB
RUS
18clon
esfrom
Cornellin
multisite
trials
CAun
quantified
UAlberta
andQuebec
FR8ndash
12SR
Fclon
eslicence‐based
IT5SR
Cand6SR
FAFV
MTA
orlicence
SE14
STTlicence‐based
andMTA
US
~19So
utheast5ndash7Upp
erMidwest
from
UMD
NRRIMTA6ndash12
North
Central
from
GWRMTAPacific
Northwest8clon
esGWRMTA24
inmultilocationyieldtrialsGWR
Com
mercial
yield
(tDM
haminus1year
minus1 )
US
3ndash18
EU8ndash
12CN20
ndash30
US
10ndash25
EU7ndash20
UK8ndash
14US
8ndash14
EU5ndash20
(SEandBaltic
Cou
ntries8ndash
12)
US
10ndash2
2(10ndash
12North
Central12
ndash16
Southeast15ndash2
2PacificNortheast)
Harvestrotatio
nand
commercial
stand
lifespan
Ann
ualfor10
ndash12years
Ann
ualfor10ndash2
5years(M
sac
hasbeen
used
for~3
0yearsin
China)
2‐to
4‐year
cyclefor22
ndash30years
SRC3‐
to7‐year
cyclefor20
years
SRF
10‐to12‐yearcycleforgt50
years
Adaptiverang
eOpen‐po
llinatedandsynthetic
cultivars
arelim
itedin
adaptatio
nby
temperature
andprecipitatio
n(~8breeding
zonesin
USandCA)
Standard
Mtimesgiswidelyadaptedin
EU
butislim
itedin
theUnitedStates
byinsufficient
winterh
ardiness
forN
orthern
Midwestand
heatintolerancein
the
southNovelMsactimesMsinhyb
rids
andMsintimesMsinhybrids
selected
incontinentalD
Ehave
show
nawide
adaptiv
erang
ein
EU(K
alininaetal
2017
)Ongoing
trialson
heavymetal
contam
inated
soils
indicatetoleranceby
exclusion(K
rzyżak
etal20
17)
Different
hybridsareneeded
fordifferent
zones
Besthy
bridsshow
someG
timesE(U
S)
andsomehy
bridslow
GtimesE(Fabio
etal20
17)
Different
hybridsareneeded
ford
ifferent
clim
aticzonesHyb
rids
thatareadapted
forg
rowingseason
sof
~6mon
thsand
relativ
elyshortd
aysin
Southern
Europ
earemaladaptedto
shortg
rowing
season
sof
~4mon
thsandrelativ
ely
long
days
inNorthernEU
The
mostb
roadly
adaptedvarietiescome
from
thePcana
densistaxo
n
Note
AFV
AlasiaFranco
VivaiC
ornell
CornellUniversity
CPV
OC
ommunity
PlantV
ariety
OfficeFC
BAF
orestCelluloseW
ood
ConstructionandFu
rnitu
reTechnologyInstitu
teG
timesEg
enotype‐by
‐environmentinterac-
tion
GWRG
reenWoodResourcesINRAF
renchNationalInstituteforA
griculturalR
esearch
IRST
EAN
ationalR
esearchUnito
fScience
andTechnologyforE
nvironmentand
AgricultureM
sacMsacchariflo
rusandM
timesg
(Mtimes
giganteus)M
sin
MiscanthussinensisM
BIMendelB
iotechnology
IncM
bpm
egabase
pairM
SStateM
ississippi
StateUniversity
MTAm
aterialtransferagreem
entNICS
NationalInstituteof
CropScience
NW‐
FVANorthwestGerman
ForestResearchInstitu
tePB
Rplantbreeders
rightRDARural
DevelopmentAdm
inistration
REAP
ResourceEfficient
Agriculture
Productio
nSL
USw
edishUniversity
ofAgriculturalSciences
SNUS
eoul
NationalU
niSR
Cshort‐ro
tatio
ncoppice
SRF
short‐rotationforestryS
TTS
weT
reeTech
tDM
haminus1year
minus1 tons
ofdrymatterperhectareperyearU
AlbertaUniversity
ofAlbertaU
MD
NRRIUniversity
ofMinnesotaDuluthsNaturalResources
ResearchInstitu
teU
MNU
niversity
ofMinnesotaU
WU
niversity
ofWashington
UWMU
niversity
ofWarmiaandMazury
a ISO
Alpha‐2
lettercountrycodes
b EU
isused
forEurope
CLIFTON‐BROWN ET AL | 5
2011) Genotype‐by‐environment interactions (G times E) arestrong and must be considered in breeding (Casler 2012Casler Mitchell amp Vogel 2012) Adaptation to environ-ment is regulated principally by responses to day‐lengthand temperature There are also strong genotype times environ-ment interactions between the drier western regions and thewetter eastern regions (Casler et al 2017)
The growing regions of North America are divided intofour adaptation zones for switchgrass each roughly corre-sponding to two official hardiness zones The lowland eco-types are generally late flowering high yielding andadapted to warmer climates but have lower drought andcold resistance than upland ecotypes (Casler 2012 Casleret al 2012)
In 2015 the US Department of Agriculture (USDA)National Plant Germplasm System GRIN (httpswwwars-gringovnpgs) had 181 switchgrass accessions of whichonly 96 were available for distribution due to limitationsassociated with seed multiplication (Casler Vogel amp Har-rison 2015) There are well over 2000 additional uncata-logued accessions (Table 1) held by various universities butthe USDA access to these is also constrained by the effortneeded in seed multiplication Switchgrass is a model herba-ceous species for conducting scientific research on biomass(Sanderson Adler Boateng Casler amp Sarath 2006) but lit-tle funding is available for the critical prebreeding work thatis necessary to link this biological research to commercialbreeding More than a million genotypes from ~2000 acces-sions (seed accessions contain many genotypes) have beenphenotypically screened in spaced plant nurseries and tenthousand of the most useful have been genotyped with differ-ent technologies depending on the technology available at
the time when these were performed From these character-ized genotypes parents are selected for exploratory pairwisecrosses to produce synthetic populations within ecotypesSwitchgrass like many grasses is outcrossing due to astrong genetically controlled self‐incompatibly (akin to theS‐Z‐locus system of other grasses (Martinez‐Reyna ampVogel 2002)) Thus the normal breeding approaches usedare F1 wide crosses and recurrent selection cycles within syn-thetic populations
The scale of these programmes varies from small‐scaleconventional breeding based solely on phenotypic selec-tion (eg REAP Canada Montreal Quebec) to large pro-grammes incorporating modern molecular breedingmethods (eg USDA‐ARS Madison Wisconsin) Earlyagronomic research and biomass production efforts werefocused on the seed‐based multiplication of promising wildaccessions from natural prairies Cultivars Alamo Kanlowand Cave‐in‐Rock were popular due to high yield andmoderate‐to‐wide adaptation Conventional breedingapproaches focussed on biomass production traits and haveled to the development of five cultivars particularly suitedto biomass production Cimarron EG1101 EG1102EG2101 and Liberty The first four of these represent thelowland ecotype and were developed either in Oklahomaor Georgia Liberty is a derivative of lowland times uplandhybrids developed in Nebraska following selection for lateflowering the high yield of the lowland ecotype and coldtolerance of the upland ecotype (Vogel et al 2014) Thesefive cultivars were all approximately 25ndash30 years in themaking counting from the initiation of these breeding pro-grammes Many more biomass‐type cultivars are expectedwithin the next few years as these and other breeding
TABLE 2 Generalized improvement targets for perennial biomass crops (PBCs)
Net energy yield per hectare
Increased yield
Reduced moisture content at harvest
Physical and chemical composition for different end‐use applications
Increased lignin content and decreased corrosive elements for thermal conversion
Reduced recalcitrance through decreased lignin content andor modified lignin monomer composition to reduce pretreatment requirementsfor next‐generation biofuels by saccharification and fermentation
Plant morphological differences which influence biomass harvest transport and storage (eg stem thickness)
Propagation costs
Improved cloning systems (trees and grasses)
Seed systems (grasses)
Optimizing agronomy for each new cultivar
Resilience through enhanced
Abiotic stress toleranceresistance (eg drought salinity and high and low temperature )
Biotic stress resistance (eg insects fungal bacterial and viral diseases)
Site adaptability especially to those of marginalcontaminated agricultural land
6 | CLIFTON‐BROWN ET AL
TABLE3
Preparedness
formassupscaling
currentmarketvalueandresearch
investmentin
four
perennialbiom
asscrops(PBCs)
Species
Switc
hgrass
Miscanthu
sW
illow
Pop
lar
Current
commercial
plantin
gcostsperha
USa20
0USD
bin
the
establishm
entyear
DE3
375
Eurob
(XueK
alininaamp
Lew
ando
wski20
15)
UK2
153
GBPb
(Evans201
6)redu
cedto
116
9GBPwith
Mxg
rhizom
ewhere
thefarm
erdo
estheland
preparation(w
wwterravestacom
)US
173
0ndash222
5USD
UK150
0ndash173
9GBPplus
land
preparation(Evans20
16)
US
197
6USD
(=80
0USD
acre)
IT110
0Euro
FRSE100
0ndash200
0Euro
US
863USD
ha(LazarusHeadleeamp
Zalesny
20
15)
Current
marketvalue
ofthebiom
asspert
DM
US
80ndash1
00USD
UK~8
0GBP(bales)(Terravesta
person
alcommun
ication)
US
94USD
(chipp
ed)
UK49
41GBP(chipp
ed)(Evans20
16)
US
55ndash7
0USD
IT10
0Euro
US
80ndash90USD
deliv
ered
basedon
40milesof
haulage
Scienceforg
enetic
improv
ementprojects
inthelast10
years
USgt50
projectsfund
edby
USDOEand
USD
ANIFA
CNgt
30projectsfund
edby
CN‐N
SFCandMBI
EUc 6projects(O
PTIM
ISCO
PTIM
A
WATBIO
SUNLIBBF
IBRAandGRACE)
FRB
FFSK
2projectsfund
edby
IPETandPM
BC
US
EBICABBImultip
leDOEFeedstock
Genom
icsandUSD
AAFR
Iprojects
UKB
SBECB
EGIN
(200
3ndash20
10)
US
USDOEJG
IGenom
eSequ
encing
(200
9ndash20
122
015ndash
2018
)USD
ANortheastSu
nGrant
(200
9ndash20
12)
USD
ANIFANEWBio
Con
sortium
(201
2ndash20
17)USD
ANIFAWillow
SkyC
AP(201
8ndash20
21)
EUF
P7(EnergyPo
plarN
ovelTreeWATBIO
Tree4Fu
ture)
SEC
limate‐adaptedpo
plarsandNanow
ood
SLU
andST
TUS
Bioenergy
Science(O
RNL)USD
Afeedstocks
geno
micsprog
rammeBRDIa
ndBRC‐CBI
Major
projects
supp
ortin
gcrossing
andselectioncycles
inthelast10
years
USandCA12
projects
NLRUEmiscanthu
sPP
P20
15ndash2
019
50ndash
100K
Euroyear
UKGIA
NT‐LIN
KPP
I20
11ndash2
016
13M
GBPyear
US
EBICABBI~0
5M
USD
year
UKBEGIN
2000
ndash201
0US
USD
ANortheastSu
nGrant
(200
9ndash20
12)USD
ANIFA
NEWBio
Con
sortium
(201
2ndash20
17)USD
ANIFA
Willow
SkyC
AP(201
8ndash20
21)
FR1
4region
alandnatio
nalp
rojects10
0ndash15
0k
Euroyear
SES
TTo
nebreeding
effortas
aninternalproject
in20
10with
aresulting
prog
enytrial
US
DOESu
nGrant
feedstockdevelopm
ent
partnershipprog
ramme(including
Willow
)
Current
annu
alinvestmentin
projects
fortranslationinto
commercial
hybrids
USgt20
MUSD
year
UKMUST
201
6ndash20
1905M
GBPyear
US
CABBIUSD
AAFR
I20
17ndash2
02205M
USD
year
US
USD
ANIFA
NEWBio
~04
MUSD
year
FRScienceforim
prov
ement~2
0kEuroyear
US
USD
Aun
derAFR
ICo‐ordinatedAg
Prod
ucer
projectsAnd
severaltranslational
geno
micsprojectswith
outpu
blic
fund
ing
Upscalin
gtim
e(years)
toproducesufficient
propagules
toplantgt10
0ha
US
Using
seed
2ndash3years
UKRhizomefor1ndash200
0hectares
canbe
ready
in6mon
ths
UKNL2
0ha
ofseed
hybridsplantedin
2018
In
2019
sufficientseedfor5
0ndash10
0ha
isexpected
UK3yearsusingconv
entio
nalcutting
sfaster
usingmicroprop
agation
FRIT3yearsby
vegetativ
eprop
agation
SEgt3yearsby
vegetativ
eprop
agation(cuttin
gs)
and3yearsby
microprop
agation
Note
AFR
IAgriculture
andFo
odResearchInitiativein
theUnitedStatesBEGIN
BiomassforEnergyGenetic
Improvem
entNetwork
BFF
BiomassFo
rtheFu
tureBRC‐CBI
Bioenergy
ResearchCentre‐Centrefor
Bioenergy
Innovatio
nBRDIBiomassresearch
developm
entinitiative
BSB
ECBBSR
CSu
stainableBioenergy
Centre
CABBICenterforadvanced
bioenergyandbioproductsinnovatio
nin
theUnitedStatesCN‐N
SFC
Natural
ScienceFo
undatio
nof
ChinaDOEDepartm
entof
Energyin
theUnitedStatesEBIEnergyBiosciences
Institu
teFIBRAFibrecropsas
sustainablesource
ofbiobased
materialforindustrial
productsin
Europeand
ChinaFP
7SeventhFram
eworkProgrammein
theEUGIA
NT‐LIN
KGenetic
improvem
entof
miscanthusas
asustainablefeedstockforbioenergyin
theUnitedKingdom
GRACEGRow
ingAdvancedindustrial
Crops
onmarginallandsforbiorEfineriesIPETKorea
Institu
teof
Planning
andEvaluationforTechnologyin
FoodA
gricultureFo
restry
andFisheriesJG
IJointGenom
eInstitu
teMBIMendelBiotechnology
IncMUST
Miscant-
husUpS
calin
gTechnology
NEWBioNortheast
WoodyW
arm‐seasonBiomassConsortiumNIFANationalInstitu
teof
Food
andAgriculture
intheUnitedStatesNovelTree
Novel
tree
breeding
strategiesOPT
IMAOpti-
mizationof
perennialgrassesforbiom
assproductio
nin
theMediterraneanareaOPT
IMISCOptim
izingbioenergyproductio
nfrom
Miscanthus
ORNLOak
Ridge
NationalLabPM
BCPlantMolecular
BreedingCenterof
theNextGenerationBiogreenResearchCenters
oftheRepublic
ofKoreaPP
Ipublicndashp
rivate
investmentPP
Ppublicndashp
rivate
partnership
RUEradiationuseefficiencySk
yCAP
Coordinated
AgriculturalProjectSL
U
SwedishUniversity
ofAgriculturalSciencesST
TSw
eTreeTech
SUNLIBBSu
stainableLiquidBiofuelsfrom
BiomassBiorefiningTree4Fu
tureDesigning
Trees
forthefutureUSD
AUSDepartm
entof
Agriculture
WATBIO
Developmentof
improved
perennialnonfoodbiom
assandbioproduct
cropsforwater
stressed
environm
ents
a ISO
Alpha‐2
lettercountrycodes
b Local
currencies
areused
asat
2018Euro
GBP(G
reat
Britain
Pound)USD
(USDollar)c EU
isused
forEurope
CLIFTON‐BROWN ET AL | 7
TABLE4
Prebreedingresearch
andthestatus
ofconventio
nalbreeding
infour
leadingperennialbiom
asscrops(PBCs)
Breeding
techno
logy
step
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sWillow
Poplar
1Collected
wild
accessions
or
secondarysources
availablefor
breeding
Provideabroadbase
of
useful
traits
Respect
forCBD
and
Nagoyaprotocol
on
collections
after2014
Not
allindigenous
genetic
resourcesareaccessible
under
CBD
forpolitical
reasons
USa~2
000
(181
inofficial
GRIN
gene
bank
ofwhich
96
wereavailable
others
in
variousunofficial
collections)
CNChangsha
~1000
andNanjin
g
2000from
China
DK~1
20from
JP
FRworking
collectionof
~100
(mainlyM
sin)
JP~1
500
from
JPS
KandCN
NLworking
collectionof
~300
Msin
SK~7
00from
SKC
NJP
andRU
UK~1
500
accessions
(from
~500
sites)
ofwhich
1000arein
field
nurseriesin
2018
US
14in
GRIN
global
US
Illin
ois~1
500
collections
made
inAsiawith
25
currently
available
inUnitedStates
forbreeding
UK1500with
about20
incommon
with
the
UnitedStates
US
350largelyunique
FR3370Populus
delto
ides
(650)Pnigra(2000)
Ptrichocarpa(600)and
Pmaximow
iczii(120)
IT30000P
delto
idsPnigra
Pmaximow
icziiand
Ptrichocarpa
SE13000
accessionsmainly
Ptrichocarpa
byST
TandSL
U
US
GWR150Ptrichocarpaand
300Pmaximow
icziiaccessions
forPacificNorthwestprogram
535Pdelto
ides
and~2
00
Pnigraaccessions
for
Southeastprogrammeand77
Psimoniifrom
Mongolia
UMD
NRRI550+
clonal
accessions
(primarily
Ptimescanadnesisa
long
with
Pdelto
ides
andPnigra
parents)
forUpper
Midwest
program
2Wild
accessions
which
have
undergone
phenotypic
screeningin
field
trials
Selectionof
parental
lines
with
useful
traits
Strong
partnerships
torun
multilocationfieldtrials
tophenotypeconsistently
theaccessionsgenotypes
indifferentenvironm
ents
anddatabases
Costof
runningmultilocation
US
gt500000genotypes
CN~1
250
inChangsha
~1700
in
HunanJiangsuS
handong
Hainan
NL~250genotypes
JP~1
200
SK~4
00
UK‐le
d~1
000
genotypesin
a
replicated
trial
US‐led
~1200
genotypesin
multi‐
locatio
nreplicated
trialsin
Asiaand
North
America
MsinSK
CNJP
CA
(Ontario)US(Illinoisand
Colorado)MsacSK
(Kangw
on)
CN
(Zhejiang)JP
(Sapporo)CA
(Onatario)U
S(Illinois)DK
(Aarhus)
UKgt400
US
180
FR2720
IT10000
SE~1
50
US
GWR~1
500
wild
Ptrichocarpaphenotypes
and
OPseedlots(total
of70)from
thePmaximow
icziispecies
complex
(suaveolens
cathayana
koreana
ussuriensismaximow
iczii)and
2000ndash3000Pdelto
ides
collections
(based
on104s‐
generatio
nfamilies)
UMD
NRRIscreened
seedlin
gsfrom
aPdelto
ides
(OPor
CP)
breeding
programme(1996ndash2016)
gt7200OPseedlin
gsof
PnigraunderMinnesota
clim
atic
conditions(improve
parental
populatio
n)
3Wild
germ
plasm
genotyping
Amanaged
living
collectionof
clonal
types
US
~20000genotypes
CNChangsha
~1000Nanjin
g37
FR44
bycpDNA
(Fenget
al
UK~4
00
US
225
(Con
tinues)
8 | CLIFTON‐BROWN ET AL
TABLE
4(Contin
ued)
Breeding
techno
logy
step
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sWillow
Poplar
Construct
phylogenetic
treesanddissim
ilarity
indices
The
type
ofmolecular
analysis
mdashAFL
PcpDNAR
ADseq
whole
genomesequencing
2014)
JP~1
200
NL250genotypesby
RADseq
UK~1
000
byRADseq
SK~3
00
US
Illin
oisRADseqon
allwild
accessions
FR2310
SE150
US
1500
4Exploratory
crossing
and
progenytests
Discovergood
parental
combinatio
nsmdashgeneral
combining
ability
Geographicseparatio
n
speciesor
phylogenetic
trees
Costly
long‐te
rmmultilocation
trialsforprogeny
US
gt10000
CN~1
0
FR~2
00
JP~3
0
NL3600
SK~2
0
UK~4
000
(~1500Msin)
US
500ndash1000
UK gt700since2003
gt800EWBP(1996ndash
2002)
US
800
FRCloned13
times13
factorial
matingdesign
with
3species[10
Pdelto
ides8Ptrichocarpa
8
Pnigra)
US
GWR1000exploratory
crossingsof
Pfrem
ontii
Psimoniiforim
proved
droughttolerance
UMD
NRRIcrossednativ
e
Pdelto
ides
with
non‐nativ
e
Ptrichocarpaand
Pmaximow
icziiMajority
of
progenypopulatio
nssuffered
from
clim
atic
andor
pathogenic
susceptib
ility
issues
5Wideintraspecies
hybridization
Discovergood
parental
combinatio
nsmdashgeneral
combining
ability
1to
4above
inform
atics
Flow
eringsynchronizationand
low
seed
set
US
~200
CN~1
0
NL1800
SK~1
0
UK~5
00
UK276
US
400
IT500
SE120
US
50ndash100
6With
inspecies
recurrentselection
Concentratio
nof
positiv
e
traits
Identificationof
theright
heterotic
groups
insteps
1ndash5
Difficultto
introducenew
germ
plasm
with
outdilutio
n
ofbesttraits
US
gt40
populatio
ns
undergoing
recurrentselection
NL2populatio
ns
UK6populatio
ns
US
~12populatio
nsto
beestablished
by2019
UK4
US
150
FRPdelto
ides
(gt30
FSfamilies
ndash900clones)Pnigra(gt40
FS
families
ndash1600clones)and
Ptrichocarpa(15FS
Families
minus350clones)
IT80
SEca50
parent
breeding
populatio
nswith
in
Ptrichocarpa
US
Goalsis~1
00parent
breeding
populatio
nswith
inPtrichocarpa
Pdelto
ides
PnigraandPmaximow
iczii
7Interspecies
hybrid
breeding
Com
bining
complem
entary
traitsto
producegood
morphotypes
and
heterosiseffects
Allthesteps1ndash6
Inearlystage
improvem
ents
areunpredictable
Awide
base
iscostly
tomanage
None
CNChangsha
3Nanjin
g
120MsintimesMflo
r
JP~5
with
Saccharum
spontaneum
SK5
UK420
US
250
FRPcanadensis(1800)
Pdelto
ides
timesPtrichocarpa
(800)and
PtrichocarpatimesPmaximow
iczii
(1200)
(Con
tinues)
CLIFTON‐BROWN ET AL | 9
TABLE
4(Contin
ued)
Breeding
techno
logy
step
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sWillow
Poplar
IT~1
20
US
525genotypesin
eastside
hybrid
programme
205westside
hybrid
programmeand~1
200
in
SoutheastPdelto
ides
program
8Chrom
osom
e
doublin
g
Arouteto
triploid
seeded
typesdoublin
gadiploid
parent
ofknow
n
breeding
valuefrom
recurrentselection
Needs
1ndash7to
help
identify
therightparental
lines
Doubled
plantsarenotorious
forrevertingto
diploid
US
~50plantstakenfrom
tetraploid
tooctaploid
which
arenow
infieldtrials
CN~1
0Msactriploids
DKSeveralattemptsin
mid
90s
FR5genotypesof
Msin
UK2Msin1Mflora
nd1Msac
US
~30
UKAttempted
butnot
routinelyused
US
Not
partof
theregular
breeding
becausethere
existnaturalpolyploids
NA
9DoubleHaploids
Arouteto
producing
homogeneous
progeny
andalso
forthe
introductio
nof
transgenes
orforgenome
editing
Needs
1ndash7to
help
identify
therightparental
lines
Fully
homozygousplantsare
weakandeasily
die
andmay
notflow
ersynchronously
US
Noneyetattempteddueto
poor
vigour
andviability
of
haploids
CN2genotypesof
Msin
andMflor
UKDihaploidsof
MsinMflor
and
Msacand2transgenic
Msinvia
anther
cultu
re
US
6createdandused
toidentify
miss‐identifying
paralogueloci
(causedby
recent
genome
duplicationin
miscanthus)Usedin
creatin
gthereferencegenomefor
MsinVeryweakplantsand
difficultto
retain
thelin
es
NA
NA
10Embryo
rescue
Anattractiv
etechniquefor
recovering
plantsfrom
sexual
crosses
The
majority
ofem
bryos
cannot
survivein
vivo
or
becomelong‐timedorm
ant
None
UKone3times
hybrid
UKEmbryosrescue
protocol
proved
robustat
8days
post‐pollin
ation
FRAnem
bryo
rescue
protocol
proved
robustandim
proved
hybridizationsuccessfor
Pdelto
ides
timesPtrichocarpa
crosses
11Po
llenstorage
Flexibility
tocross
interestingparents
with
outneed
of
flow
ering
synchronization
None
Nosuccessto
daten
otoriously
difficultwith
grassesincluding
sugarcane
Freshpollencommonly
used
incrossing
Pollenstoragein
some
interspecificcrosses
FRBothstored
(cryobank)
and
freshindividual
pollen
Note
AFL
Pam
plifiedfragmentlength
polymorphismCBDconventio
non
biological
diversity
CP
controlledpollinatio
nCpD
NAchloroplastDNAEWBP
Europeanwillow
breeding
programme
FSfull‐sib
GtimesE
genotype‐by‐environm
entGRIN
germ
plasm
resourcesinform
ationnetwork
GWRGreenWoodResourcesMflorMiscanthusflo
ridulusMsacMsacchariflo
rus
MsinMsinensisNAnotapplicableNRRINatural
Resources
ResearchInstitu
teOP
open
pollinatio
nRADseq
restrictionsite‐associatedDNA
sequencingSL
USw
edishUniversity
ofAgriculturalSciencesST
TSw
eTreeTech
UMDUniversity
ofMinnesota
Duluth
a ISO
Alpha‐2
lettercountrycodes
10 | CLIFTON‐BROWN ET AL
programmes mature The average rate of gain for biomassyield in long‐term switchgrass breeding programmes hasbeen 1ndash4 per year depending on ecotype populationand location of the breeding programme (Casler amp Vogel2014 Casler et al 2018) The hybrid derivative Libertyhas a biomass yield 43 higher than the better of its twoparents (Casler amp Vogel 2014 Vogel et al 2014) Thedevelopment of cold‐tolerant and late‐flowering lowland‐ecotype populations for the northern United States hasincreased biomass yields by 27 (Casler et al 2018)
Currently more than 20 recurrent selection populationsare being managed in the United States to select parentsfor improved yield yield resilience and compositional qual-ity of the biomass For the agronomic development andupscaling high seed multiplication rates need to be com-bined with lower seed dormancy to reduce both crop estab-lishment costs and risks Expresso is the first cultivar withsignificantly reduced seed dormancy which is the first steptowards development of domesticated populations (Casleret al 2015) Most phenotypic traits of interest to breeders
require a minimum of 2 years to be fully expressed whichresults in a breeding cycle that is at least two years Morecomplicated breeding programmes or traits that requiremore time to evaluate can extend the breeding cycle to 4ndash8 years per generation for example progeny testing forbiomass yield or field‐based selection for cold toleranceBreeding for a range of traits with such long cycles callsfor the development of molecular methods to reduce time-scales and improve breeding efficiency
Two association panels of switchgrass have been pheno-typically and genotypically characterized to identify quanti-tative trait loci (QTLs) that control important biomasstraits The northern panel consists of 60 populationsapproximately 65 from the upland ecotype The southernpanel consists of 48 populations approximately 65 fromthe lowland ecotype Numerous QTLs have been identifiedwithin the northern panel to date (Grabowski et al 2017)Both panels are the subject of additional studies focused onbiomass quality flowering and phenology and cold toler-ance Additionally numerous linkage maps have been
FIGURE 1 Cumulative minimum years needed for the conventional breeding cycle through the steps from wild germplasm to thecommercial hybrids in switchgrass miscanthus willow and poplar Information links between the steps are indicated by dotted arrows andhighlight the importance of long‐term informatics to maximize breeding gain
CLIFTON‐BROWN ET AL | 11
created by the pairwise crossing of individuals with diver-gent characteristics often to generate four‐way crosses thatare analysed as pseudo‐F2 crosses (Liu Wu Wang ampSamuels 2012 Okada et al 2010 Serba et al 2013Tornqvist et al 2018) Individual markers and QTLs iden-tified can be used to design marker‐assisted selection(MAS) programmes to accelerate breeding and increase itsefficiency Genomic prediction and selection (GS) holdseven more promise with the potential to double or triplethe rate of gain for biomass yield and other highly complexquantitative traits of switchgrass (Casler amp Ramstein 2018Ramstein et al 2016) The genome of switchgrass hasrecently been made public through the Joint Genome Insti-tute (httpsphytozomejgidoegov)
Transgenic approaches have been heavily relied upon togenerate unique genetic variants principally for traitsrelated to biomass quality (Merrick amp Fei 2015) Switch-grass is highly transformable using either Agrobacterium‐mediated transformation or biolistics bombardment butregeneration of plants is the bottleneck to these systemsTraditionally plants from the cultivar Alamo were the onlyregenerable genotypes but recent efforts have begun toidentify more genotypes from different populations that arecapable of both transformation and subsequent regeneration(King Bray Lafayette amp Parrott 2014 Li amp Qu 2011Ogawa et al 2014 Ogawa Honda Kondo amp Hara‐Nishi-mura 2016) Cell wall recalcitrance and improved sugarrelease are the most common targets for modification (Bis-wal et al 2018 Fu et al 2011) Transgenic approacheshave the potential to provide traits that cannot be bredusing natural genetic variability However they will stillrequire about 10ndash15 years and will cost $70ndash100 millionfor cultivar development and deployment (Harfouche Mei-lan amp Altman 2011) In addition there is commercialuncertainty due to the significant costs and unpredictabletimescales and outcomes of the regulatory approval processin the countries targeted for seed sales As seen in maizeone advantage of transgenic approaches is that they caneasily be incorporated into F1 hybrid cultivars (Casler2012 Casler et al 2012) but this does not decrease thetime required for cultivar development due to field evalua-tion and seed multiplication requirements
The potential impacts of unintentional gene flow andestablishment of non‐native transgene sequences in nativeprairie species via cross‐pollination are also major issues forthe environmental risk assessment These limit further thecommercialization of varieties made using these technolo-gies Although there is active research into switchgrass steril-ity mechanisms to curb unintended pollen‐mediated genetransfer it is likely that the first transgenic cultivars proposedfor release in the United States will be met with considerableopposition due to the potential for pollen flow to remainingwild prairie sites which account for lt1 of the original
prairie land area and are highly protected by various govern-mental and nongovernmental organizations (Casler et al2015) Evidence for landscape‐level pollen‐mediated geneflow from genetically modified Agrostis seed multiplicationfields (over a mountain range) to pollinate wild relatives(Watrud et al 2004) confirms the challenge of using trans-genic approaches Looking ahead genome editing technolo-gies hold considerable promise for creating targeted changesin phenotype (Burris Dlugosz Collins Stewart amp Lena-ghan 2016 Liu et al 2018) and at least in some jurisdic-tions it is likely that cultivars resulting from gene editingwill not need the same regulatory approval as GMOs (Jones2015a) However in July 2018 the European Court of Justice(ECJ) ruled that cultivars carrying mutations resulting fromgene editing should be regulated in the same way as GMOsThe ECJ ruled that such cultivars be distinguished from thosearising from untargeted mutation breeding which isexempted from regulation under Directive 200118EC
3 | MISCANTHUS
Miscanthus is indigenous to eastern Asia and Oceania whereit is traditionally used for forage thatching and papermaking(Xi 2000 Xi amp Jezowkski 2004) In the 1960s the highbiomass potential of a Japanese genotype introduced to Eur-ope by Danish nurseryman Aksel Olsen in 1935 was firstrecognized in Denmark (Linde‐Laursen 1993) Later thisaccession was characterized described and named asldquoM times giganteusrdquo (Greef amp Deuter 1993 Hodkinson ampRenvoize 2001) commonly abbreviated as Mxg It is a natu-rally occurring interspecies triploid hybrid between tetraploidM sacchariflorus (2n = 4x) and diploid M sinensis(2n = 2x) Despite its favourable agronomic characteristicsand ability to produce high yields in a wide range of environ-ments in Europe (Kalinina et al 2017) the risks of relianceon it as a single clone have been recognized Miscanthuslike switchgrass is outcrossing due to self‐incompatibility(Jiang et al 2017) Thus seeded hybrids are an option forcommercial breeding Miscanthus can also be vegetativelypropagated by rhizome or in vitro culture which allows thedevelopment of clones The breeding approaches are usuallybased on F1 crosses and recurrent selection cycles within thesynthetic populations There are several breeding pro-grammes that target improvement of miscanthus traitsincluding stress resilience targeted regional adaptation agro-nomic ldquoscalabilityrdquo through cheaper propagation fasterestablishment lower moisture and ash contents and greaterusable yield (Clifton‐Brown et al 2017)
Germplasm collections specifically to support breedingfor biomass started in the late 1980s and early 1990s inDenmark Germany and the United Kingdom (Clifton‐Brown Schwarz amp Hastings 2015) These collectionshave continued with successive expeditions from European
12 | CLIFTON‐BROWN ET AL
and US teams assembling diverse collections from a widegeographic range in eastern Asia including from ChinaJapan South Korea Russia and Taiwan (Hodkinson KlaasJones Prickett amp Barth 2015 Stewart et al 2009) Threekey miscanthus species for biomass production are M si-nensis M floridulus and M sacchariflorus M sinensis iswidely distributed throughout eastern Asia with an adap-tive range from the subtropics to southern Russia (Zhaoet al 2013) This species has small rhizomes and producesmany tightly packed shoots forming a ldquotuftrdquo M floridulushas a more southerly adaptive range with a rather similarmorphology to M sinensis but grows taller with thickerstems and is evergreen and less cold‐tolerant than the othermiscanthus species M sacchariflorus is the most northern‐adapted species ranging to 50 degN in eastern Russia (Clarket al 2016) Populations of diploid and tetraploid M sac-chariflorus are found in China (Xi 2000) and South Korea(Yook 2016) and eastern Russia but only tetraploids havebeen found in Japan (Clark Jin amp Petersen 2018)
Germplasm has been assembled from multiple collec-tions over the last century though some early collectionsare poorly documented This historical germplasm has beenused to initiate breeding programmes largely based on phe-notypic and genotypic characterization As many of theaccessions from these collections are ldquoorigin unknownrdquocrucial environmental envelope data are not available UK‐led expeditions started in 2006 and continued until 2011with European and Asian partners and have built up a com-prehensive collection of 1500 accessions from 500 sitesacross Eastern Asia including China Japan South Koreaand Taiwan These collections were guided using spatialclimatic data to identify variation in abiotic stress toleranceAccessions from these recent collections were planted fol-lowing quarantine in multilocation nursery trials at severallocations in Europe to examine trait expression in differentenvironments Based on the resulting phenotypic andmolecular marker data several studies (a) characterized pat-terns of population genetic structure (Slavov et al 20132014) (b) evaluated the statistical power of genomewideassociation studies (GWASs) and identified preliminarymarkerndashtrait associations (Slavov et al 2013 2014) and(c) assessed the potential of genomic prediction (Daveyet al 2017 Slavov et al 2014 2018b) Genomic indexselection in particular offers the possibility of exploringscenarios for different locations or industrial markets (Sla-vov et al 2018a 2018b)
Separately US‐led expeditions also collected about1500 accessions between 2010 and 2014 (Clark et al2014 2016 2018 2015) A comprehensive genetic analy-sis of the population structure has been produced by RAD-seq for M sinensis (Clark et al 2015 Van der WeijdeKamei et al 2017) and M sacchariflorus (Clark et al2018) Multilocation replicated field trials have also been
conducted on these materials in North America and inAsia GWAS has been conducted for both M sinensis anda subset of M sacchariflorus accessions (Clark et al2016) To date about 75 of these recent US‐led collec-tions are in nursery trials outside the United States Due tolengthy US quarantine procedures these are not yet avail-able for breeding in the United States However molecularanalyses have allowed us to identify and prioritize sets ofgenotypes that best encompass the genetic variation in eachspecies
While mostM sinensis accessions flower in northern Eur-ope very few M sacchariflorus accessions flower even inheated glasshouses For this reason the European pro-grammes in the United Kingdom the Netherlands and Francehave performed mainly M sinensis (intraspecies) hybridiza-tions (Table 4) Selected progeny become the parents of latergenerations (recurrent selection as in switchgrass) Seed setsof up to 400 seed per panicle occur inM sinensis In Aberyst-wyth and Illinois significant efforts to induce synchronousflowering in M sacchariflorus and M sinensis have beenmade because interspecies hybrids have proven higher yieldperformance and wide adaptability (Kalinina et al 2017) Ininterspecies pairwise crosses in glasshouses breathable bagsandor large crossing tubes or chambers in which two or morewhole plants fit are used for pollination control Seed sets arelower in bags than in the open air because bags restrict pollenmovement while increasing temperatures and reducinghumidity (Clifton‐Brown Senior amp Purdy 2018) About30 of attempted crosses produced 10 to 60 seeds per baggedpanicle The seed (thousand seed mass ranges from 05 to09 g) is threshed from the inflorescences and sown into mod-ular trays to produce plug plants which are then planted infield nurseries to identify key parental combinations
A breeding programme of this scale must serve theneeds of different environments accepting the commonpurpose is to optimize the interception of solar radiationAn ideal hybrid for a given environment combines adapta-tion to date of emergence with optimization of traits suchas height number of stems per plant flowering and senes-cence time to optimize solar interception to produce a highbiomass yield with low moisture content at harvest (Rob-son Farrar et al 2013 Robson Jensen et al 2013) By20132014 conventional breeding in Europe had producedintra‐ and interspecific fertile seeded hybrids When acohort (typically about 5) of outstanding crosses have beenidentified it is important to work on related upscaling mat-ters in parallel These are as follows
Assessment of the yield and critical traits in selectedhybrids using a network of field trials
Efficient cloning of the seed parents While in vitro andmacro‐cloning techniques are used some genotypes areamenable to neither technique
CLIFTON‐BROWN ET AL | 13
TABLE5
Status
ofmod
ernplantbreeding
techniqu
esin
four
leadingperennialbiom
asscrop
s(PBCs)
Breeding
techno
logy
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sW
illow
Pop
lar
MAS
Use
ofmarker
sequ
encesthat
correlatewith
atrait
allowingearly
prog
enyselectionor
rejectionlocatin
gkn
owngenesfor
useful
traits(suchas
height)from
other
speciesin
your
crop
Breeding
prog
ramme
relevant
biparental
crosses(Table
4)to
create
ldquomapping
popu
latio
nsrdquoand
QTLidentification
andestim
ationto
identifyrobu
stmarkers
alternatively
acomprehensive
panelof
unrelated
geno
typesfor
GWAS
Ineffectivewhere
traits
areaffected
bymany
geneswith
small
effects
US
Literally
dozens
ofstud
iesto
identify
SNPs
andmarkers
ofinterestNoattempts
atMASas
yetlargely
dueto
popu
latio
nspecificity
CNS
SRISS
Rmarkers
forM
sinMsacand
Mflor
FR2
Msin
mapping
popu
latio
ns(BFF
project)
NL2
Msin
mapping
popu
latio
ns(A
tienza
Ram
irezetal
2003
AtienzaSatovic
PetersenD
olstraamp
Martin
200
3Van
Der
Weijdeetal20
17a
Van
derW
eijde
Hux
ley
etal20
17
Van
derW
eijdeKam
ei
etal20
17V
ander
WeijdeKieseletal
2017
)SK
2GWASpanels(1
collectionof
Msin1
diploidbiparentalF 1
popu
latio
n)in
the
pipelin
eUK1
2families
tofind
QTLin
common
indifferentfam
ilies
ofthe
sameanddifferent
speciesandhy
brids
US
2largeGWAS
panels4
diploidbi‐
parentalF 1
popu
latio
ns
1F 2
popu
latio
n(add
ition
alin
pipelin
e)
UK16
families
for
QTLdiscov
ery
2crossesMASscreened
1po
tentialvariety
selected
US
9families
geno
typedby
GBS(8
areF 1o
neisF 2)a
numberof
prom
ising
QTLdevelopm
entof
markers
inprog
ress
FR(INRA)tested
inlargeFS
families
and
infactorialmating
design
low
efficiency
dueto
family
specificity
US
49families
GS
Metho
dto
accelerate
breeding
throug
hredu
cing
the
resourcesforcross
attemptsby
Aldquotraining
popu
latio
nrdquoof
individu
alsfrom
stages
abov
ethat
have
been
both
Risks
ofpo
orpredictio
nProg
eny
testingneedsto
becontinueddu
ring
the
timethetraining
setis
US
One
prog
rammeso
farmdash
USD
Ain
WisconsinThree
cycles
ofgeno
mic
selectioncompleted
CN~1
000
geno
types
NLprelim
inary
mod
elsforMsin
developed
UK~1
000
geno
types
Verysuitable
application
not
attempted
yet
Maybe
challeng
ingfor
interspecies
hybrids
FR(INRA)120
0geno
type
ofPop
ulus
nigraarebeingused
todevelop
intraspecificGS
(Con
tinues)
14 | CLIFTON‐BROWN ET AL
TABLE
5(Contin
ued)
Breeding
techno
logy
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sW
illow
Pop
lar
predictin
gthe
performance
ofprog
enyof
crosses
(and
inresearch
topredictbestparents
tousein
biparental
crosses)
geno
typedand
phenotyp
edto
developamod
elthattakesgeno
typic
datafrom
aldquocandidate
popu
latio
nrdquoof
untested
individu
als
andprod
uces
GEBVs(Jannink
Lorenzamp
Iwata
2010
)
beingph
enotyp
edand
geno
typed
Inthis
time
next‐generation
germ
plasm
andreadyto
beginthe
second
roun
dof
training
and
recalib
ratio
nof
geno
mic
predictio
nmod
els
US
Prelim
inary
mod
elsforMsac
and
Msin
developed
Work
isun
derw
ayto
deal
moreeffectivelywith
polyploidgenetics
calib
ratio
nsUS
125
0geno
types
arebeingused
todevelopGS
calib
ratio
ns
Traditio
nal
transgenesis
Efficient
introd
uctio
nof
ldquoforeign
rdquotraits
(possiblyfrom
other
genu
sspecies)
into
anelite
plant(ega
prov
enparent
ora
hybrid)that
needsa
simpletraitto
beim
prov
edValidate
cand
idategenesfrom
QTLstud
ies
MASand
know
ledg
eof
the
biolog
yof
thetrait
andsource
ofgenesto
confer
the
relevant
changesto
phenotyp
eWorking
transformation
protocol
IPissuescostof
regu
latory
approv
al
GMO
labelling
marketin
gissuestricky
tousetransgenes
inou
t‐breedersbecause
ofcomplexity
oftransformingandgene
flow
risks
US
Manyprog
ramsin
UnitedStates
are
creatin
gtransgenic
plantsusingAlamoas
asource
oftransformable
geno
typesManytraits
ofinterestNothing
commercial
yet
CNC
hang
shaBtg
ene
transformed
into
Msac
in20
04M
sin
transformed
with
MINAC2gene
and
Msactransform
edwith
Cry2A
ageneb
othwith
amarkerg
eneby
Agrob
acterium
JP1
Msintransgenic
forlow
temperature
tolerancewith
increased
expression
offructans
(unp
ublished)
NLImprov
ingprotocol
forM
sin
transformation
SK2
Msin
with
Arabido
psisand
Brachypod
ium
PhytochromeBgenes
(Hwang
Cho
etal
2014
Hwang
Lim
etal20
14)
UK2
Msin
and
1Mflorg
enotyp
es
UKRou
tine
transformationno
tyet
possibleResearchto
overcomerecalcitrance
ongo
ing
Currently
trying
differentspecies
andcond
ition
sPo
plar
transformationused
atpresent
US
Frequently
attempted
with
very
limitedsuccess
US
600transgenic
lines
have
been
characterized
mostly
performed
inthe
aspenhy
brids
reprod
uctiv
esterility
drou
ghttolerance
with
Kno
ckdo
wns
(Con
tinues)
CLIFTON‐BROWN ET AL | 15
TABLE
5(Contin
ued)
Breeding
techno
logy
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sW
illow
Pop
lar
variou
slytransformed
with
four
cellwall
genesiptand
uidA
genesusingtwo
selectionsystem
sby
biolisticsand
Agrob
acterium
Transform
edplantsare
beinganalysed
US
Prelim
inarywork
will
betakenforw
ardin
2018
ndash202
2in
CABBI
Genom
eediting
CRISPR
Refinem
entofexisting
traits
inusefulparents
orprom
isinghybrids
bygeneratingtargeted
mutations
ingenes
know
ntocontrolthe
traitof
interestF
irst
adouble‐stranded
breakismadeinthe
DNAwhich
isrepairedby
natural
DNArepair
machineryL
eads
tofram
eshiftSN
Por
can
usealdquorepair
templaterdquo
orcanbe
used
toinserta
transgene
intoaldquosafe
harbourlocusrdquoIt
couldbe
used
todeleterepressors(or
transcriptionfactors)
Identification
mapping
and
sequ
encing
oftarget
genes(from
DNA
sequ
ence)
avoiding
screening
outof
unintend
ededits
Manyregu
latory
authorities
have
not
decidedwhether
CRISPR
andother
geno
meediting
techno
logies
are
GMOsor
notIf
GMOsseecomment
onstage10
abov
eIf
noteditedcrop
swill
beregu
latedas
conv
entio
nalvarieties
CATechn
olog
yisstill
toonewand
switchg
rass
geno
meis
very
complexOther
labo
ratories
are
interestedbu
tno
tyet
mov
ingon
this
US
One
prog
rammeso
farmdash
USD
Ain
Albany
FRInitiated
in20
16(M
ISEDIT
project)
NLInitiatingin
2018
US
Initiatingin
2018
in
CABBI
UKNot
started
Not
possible
until
transformation
achieved
US
CRISPR
‐Cas9and
Cpf1have
been
successfully
developedin
Pop
ulus
andarehigh
lyefficientExamples
forlig
nin
biosyn
thesis
Note
BFF
BiomassFo
rtheFu
tureBtBacillus
thuringiensisCABBICenterforadvanced
bioenergyandbioproductsinnovatio
nCpf1
CRISPR
from
Prevotella
andFrancisella
1CRISPR
clusteredregularlyinterspaced
palin
drom
icrepeatsCRISPR
‐CasC
RISPR
‐associated
Cry2A
acrystaltoxins2A
asubfam
ilyproduced
byBtFS
full‐sibG
BS
genotyping
bysequencingG
EBVsgenomic‐estim
ated
breeding
valuesG
MOg
enetically
modi-
fied
organism
GS
genomicselection
GWAS
genomew
ideassociationstudy
INRAF
renchNationalInstituteforAgriculturalR
esearch
IPintellectualp
roperty
IPTisopentenyltransferase
ISSR
inter‐SSR
MflorMiscant-
husflo
ridulusMsacMsacchariflo
rusMsinMsinensisM
AS
marker‐a
ssistedselection
MISEDITm
iscanthusgene
editing
forseed‐propagatedtriploidsNACn
oapicalmeristemA
TAF1
2and
cup‐shaped
cotyledon2
‐lik
eQTLq
uantitativ
etraitlocusS
NP
singlenucleotid
epolymorphismS
SRsim
plesequence
repeatsUSD
AU
SDepartm
ento
fAgriculture
a ISO
Alpha‐2
lettercountrycodes
16 | CLIFTON‐BROWN ET AL
High seed production from field crossing trials con-ducted in locations where flowering in both seed andpollen parents is likely to happen synchronously
Scalable and adapted harvesting threshing and seed pro-cessing methods for producing high seed quality
The results of these parallel activities need to becombined to identify the upscaling pathway for eachhybrid if this cannot be achieved the hybrid will likelynot be commercially viable The UK‐led programme withpartners in Italy and Germany shows that seedbased mul-tiplication rates of 12000 are achievable several inter-specific hybrids (Clifton‐Brown et al 2017) Themultiplication rate of M sinensis is higher probably15000ndash10000 Conventional cloning from rhizome islimited to around 120 that is one ha could provide rhi-zomes for around 20 ha of new plantation
Multilocation field testing of wild and novel miscanthushybrids selected by breeding programmes in the Nether-lands and the United Kingdom was performed as part ofthe project Optimizing Miscanthus Biomass Production(OPTIMISC 2012ndash2016) These trials showed that com-mercial yields and biomass qualities (Kiesel et al 2017Van der Weijde et al 2017a Van der Weijde Kieselet al 2017) could be produced in a wide range of climatesand soil conditions from the temperate maritime climate ofwestern Wales to the continental climate of eastern Russiaand the Ukraine (Kalinina et al 2017) Extensive environ-mental measurements of soil and climate combined withgrowth monitoring are being used to understand abioticstresses (Nunn et al 2017 Van der Weijde Huxley et al2017) and develop genotype‐specific scenarios similar tothose reported earlier in Hastings et al (2009) Phenomicsexperiments on drought tolerance have been conducted onwild and improved germplasm (Malinowska Donnison ampRobson 2017 Van der Weijde Huxley et al 2017)Recently produced interspecific hybrids displaying excep-tional yield under drought (~30 greater than control Mxg)in field trials in Poland and Moldova are being furtherstudied in detail in the phenomics and genomics facility atAberystwyth to better understand genendashtrait associationswhich can be fed back into breeding
Intraspecific seeded hybrids of M sinensis produced inthe Netherlands and interspecific M sacchariflorus times M si-nensis hybrids produced by the UK‐led breeding programmehave entered yield testing in 2018 with the recently EU‐funded project ldquoGRowing Advanced industrial Crops on mar-ginal lands for biorEfineries (GRACE)rdquo (httpswwwgrace-bbieu) Substantial variation in biomass quality for sacchari-fication efficiency (glucose release as of dry matter) ashcontent and melting point has already been generated inintraspecific M sinensis hybrids (Van der Weijde Kiesel
et al 2017) across environments (Weijde Dolstra et al2017a) GRACE aims to establish more than 20 hectares ofnew inter‐ and intraspecific seeded hybrids across six Euro-pean countries This project is building the know‐how andagronomy needed to transition from small research plots tocommercial‐scale field sites and linking biomass productiondirectly to industrial applications The biomass produced byhybrids in different locations will be supplied to innovativeindustrial end‐users making a wide range of biobased prod-ucts both for chemicals and for energy In the United Statesmulti‐location yield were initiated in 2018 to evaluate new tri-ploid M times giganteus genotypes developed at Illinois Cur-rently infertile hybrids are favoured in the United Statesbecause this eliminates the risk of invasiveness from naturallydispersed viable seed The precautionary principle is appliedas fertile miscanthus has naturalized in several states (QuinnAllen amp Stewart 2010) North European multilocation fieldtrials in the EMI and OPTIMISC projects have shown thereis minimal risk of invasiveness even in years when fertileflowering hybrids produce viable seed Naturalized standshave not established here due perhaps to low dormancy pooroverwintering and low seedling competitive strength In addi-tion to breeding for nonshattering or sterile seeded hybridsQuinn et al (2010) suggest management strategies which canfurther minimize environmental opportunities to manage therisk of invasiveness
31 | Molecular breeding and biotechnology
In miscanthus new plant breeding techniques (Table 5)have focussed on developing molecular markers forbreeding in Europe the United States South Korea andJapan There are several publications on QTL mappingpopulations for key traits such as flowering (AtienzaRamirez amp Martin 2003) and compositional traits(Atienza Satovic Petersen Dolstra amp Martin 2003) Inthe United States and United Kingdom independent andinterconnected bi‐parental ldquomappingrdquo families have beenstudied (Dong et al 2018 Gifford Chae SwaminathanMoose amp Juvik 2015) alongside panels of diverse germ-plasm accessions for GWAS (Slavov et al 2013) Fur-ther developments calibrating GS with very large panelsof parents and cross progeny are underway (Daveyet al 2017) The recently completed first miscanthus ref-erence genome sequence is expected to improve the effi-ciency of MAS strategies and especially GWAS(httpsphytozomejgidoegovpzportalhtmlinfoalias=Org_Msinensis_er) For example without a referencegenome sequence Clark et al (2014) obtained 21207RADseq SNPs (single nucleotide polymorphisms) on apanel of 767 miscanthus genotypes (mostly M sinensis)but subsequent reanalysis of the RADseq data using the
CLIFTON‐BROWN ET AL | 17
new reference genome resulted in hundreds of thousandsof SNPs being called
Robust and effective in vitro regeneration systems havebeen developed for Miscanthus sinensis M times giganteusand M sacchariflorus (Dalton 2013 Guo et al 2013Hwang Cho et al 2014 Rambaud et al 2013 Ślusarkie-wicz‐Jarzina et al 2017 Wang et al 2011 Zhang et al2012) However there is still significant genotype speci-ficity and these methods need ldquoin‐houserdquo optimization anddevelopment to be used routinely They provide potentialroutes for rapid clonal propagation and also as a basis forgenetic transformation Stable transformation using bothbiolistics (Wang et al 2011) and Agrobacterium tumefa-ciens DNA delivery methods (Hwang Cho et al 2014Hwang Lim et al 2014) has been achieved in M sinen-sis The development of miscanthus transformation andgene editing to generate diplogametes for producing seed‐propagated triploid hybrids are performed as part of theFrench project MISEDIT (miscanthus gene editing forseed‐propagated triploids) There are no reports of genomeediting in any miscanthus species but new breeding inno-vations including genome editing are particularly relevantin this slow‐to‐breed nonfood bioenergy crop (Table 4)
4 | WILLOW
Willow (Salix spp) is a very diverse group of catkin‐bear-ing trees and shrubs Willow belongs to the family Sali-caceae which also includes the Populus genus There areapproximately 350 willow species (Argus 2007) foundmostly in temperate and arctic zones in the northern hemi-sphere A few are adapted to subtropical and tropicalzones The centre of diversity is believed to be in Asiawith over 200 species in China Around 120 species arefound in the former Soviet Union over 100 in NorthAmerica and around 65 species in Europe and one speciesis native to South America (Karp et al 2011) Willows aredioecious thus obligate outcrossers and highly heterozy-gous The haploid chromosome number is 19 (Hanley ampKarp 2014) Around 40 of species are polyploid (Sudaamp Argus 1968) ranging from triploids to the atypicaldodecaploid S maxxaliana with 2n=190 (Zsuffa et al1984)
Although almost exclusively native to the NorthernHemisphere willow has been grown around the globe formany thousands of years to support a wide range of appli-cations (Kuzovkina amp Quigley 2005 Stott 1992) How-ever it has been the focus of domestication for bioenergypurposes for only a relatively short period since the 1970sin North America and Europe For bioenergy breedershave focused their efforts on the shrub willows (subgenusVetix) because of their rapid juvenile growth rates as aresponse to coppicing on a 2‐ to 4‐year cycle that can be
accomplished using farm machinery rather than forestryequipment (Shield Macalpine Hanley amp Karp 2015Smart amp Cameron 2012)
Since shrub willow was not generally recognized as anagricultural crop until very recently there has been littlecommitment to building and maintaining germplasm reposi-tories of willow to support long‐term breeding One excep-tion is the United Kingdom where a large and well‐characterized Salix germplasm collection comprising over1500 accessions is held at Rothamsted Research (Stott1992 Trybush et al 2008) Originally initiated for use inbasketry in 1923 accessions have been added ever sinceIn the United States a germplasm collection of gt350accessions is located at Cornell University to support thebreeding programme there The UK and Cornell collectionshave a relatively small number of accessions in common(around 20) Taken together they represent much of thespecies diversity but only a small fraction of the overallgenetic diversity within the genus There are three activewillow breeding programmes in Europe RothamstedResearch (UK) Salixenergi Europa AB (SEE) and a pro-gramme at the University of Warmia and Mazury in Olsz-tyn (Poland) (abbreviations used in Table 1) There is oneactive US programme based at Cornell University Culti-vars are still being marketed by the European WillowBreeding Programme (EWBP) (UK) which was activelybreeding biomass varieties from 1996 to 2002 Cultivarsare protected by plant breedersrsquo rights (PBRs) in Europeand by plant patents in the United States The sharing ofgenetic resources in the willow community is generallyregulated by material transfer agreements (MTA) and tai-lored licensing agreements although the import of cuttingsinto North America is prohibited except under special quar-antine permit conditions
Efforts to augment breeding germplasm collection fromnature are continuing with phenotypic screening of wildgermplasm performed in field experiments with 177 S pur-purea genotypes in the United States (at sites in Genevaand Portland NY and Morgantown WV) that have beengenotyped using genotyping by sequencing (GBS) (Elshireet al 2011) In addition there are approximately 400accessions of S viminalis in Europe (near Pustnaumls Upp-sala Sweden and Woburn UK (Berlin et al 2014 Hal-lingbaumlck et al 2016) The S viminalis accessions wereinitially genotyped using 38 simple sequence repeats (SSR)markers to assess genetic diversity and screened with~1600 SNPs in genes of potential interest for phenologyand biomass traits Genetics and genomics combined withextensive phenotyping have substantially improved thegenetic basis of biomass‐related traits in willow and arenow being developed in targeted breeding via MAS Thisunderpinning work has been conducted on large specifi-cally developed biparental Salix mapping populations
18 | CLIFTON‐BROWN ET AL
(Hanley amp Karp 2014 Zhou et al 2018) as well asGWAS panels (Hallingbaumlck et al 2016)
Once promising parental combinations are identifiedcrosses are usually performed using fresh pollen frommaterial that has been subject to a phased removal fromcold storage (minus4degC) (Lindegaard amp Barker 1997 Macal-pine Shield Trybush Hayes amp Karp 2008 Mosseler1990) Pollen storage is useful in certain interspecific com-binations where flowering is not naturally synchronizedThis can be overcome by using pollen collection and stor-age protocol which involves extracting pollen using toluene(Kopp Maynard Niella Smart amp Abrahamson 2002)
The main breeding approach to improve willow yieldsrelies on species hybridization to capture hybrid vigour(Fabio et al 2017 Serapiglia Gouker amp Smart 2014) Inthe absence of genotypic models for heterosis breedershave extensively tested general and specific combiningability of parents to produce superior progeny The UKbreeding programmes (EWBP 1996ndash2002 and RothamstedResearch from 2003 on) have performed more than 1500exploratory cross‐pollinations The Cornell programme hassuccessfully completed about 550 crosses since 1998Investment into the characterization of genetic diversitycombined with progeny tests from exploratory crosses hasbeen used to produce hundreds of targeted intraspeciescrosses in the United Kingdom and United States respec-tively (see Table 1) To achieve long‐term gains beyond F1hybrids four intraspecific recurrent selection populationshave been created in the United Kingdom (for S dasycla-dos S viminalis and S miyabeana) and Cornell is pursu-ing recurrent selection of S purpurea Interspecifichybridizations with genotypes selected from the recurrentselection cycles are well advanced in willow with suchcrosses to date totalling 420 in the United Kingdom andover 100 in the United States
While species hybridization is common in Salix it isnot universal Of the crosses attempted about 50 hybri-dize and produce seed (Macalpine Shield amp Karp 2010)As the viability of seed from successful crosses is short (amatter of days at ambient temperatures) proper seed rear-ing and storage protocols are essential (Maroder PregoFacciuto amp Maldonado 2000)
Progeny from crosses are treated in different waysamong the breeding programmes at the seedling stage Inthe United States seedlings are planted into an irrigatedfield where plants are screened for two seasons beforebeing progressed to further field trials In the United King-dom seedlings are planted into trays of compost wherethey remain containerized in an irrigated nursery for theremainder of year one In the United Kingdom seedlingsare subject to two rounds of selection in the nursery yearThe first round takes place in September to select againstsusceptibility to rust infection (Melampsora spp) A second
round of selection in winter assesses tip damage from frostand giant willow aphid infestation In the United Stateswhere the rust pressure is lower screening for Melampsoraspp cannot be performed at the nursery stage Both pro-grammes monitor Melampsora spp pest susceptibilityyield and architecture over multiple years in field trialsSelected material is subject to two rounds of field trials fol-lowed by a final multilocation yield trial to identify vari-eties for commercialization
Promising selections (ie potential cultivars) need to beclonally propagated A rapid in vitro tissue culture propa-gation method has been developed (Palomo‐Riacuteos et al2015) This method can generate about 5000 viable trans-plantable clones from a single plant in just 24 weeks Anin vitro system can also accommodate early selection viamolecular or biochemical markers to increase selectionspeed Conventional breeding systems take 13 years viafour rounds of selection from crossing to selecting a variety(Figure 1) but this has the potential to be reduced to7 years if micropropagation and MAS selection are adopted(Hanley amp Karp 2014 Palomo‐Riacuteos et al 2015)
Willows are currently propagated commercially byplanting winter‐dormant stem cuttings in spring Commer-cial planting systems for willow use mechanical plantersthat cut and insert stem sections from whips into a well‐prepared soil One hectare of stock plants grown in specificmultiplication beds planted at 40000 plants per ha pro-duces planting material for 80 hectares of commercialshort‐rotation coppice willow annually (planted at 15000cuttings per hectare) (Whittaker et al 2016) When com-mercial plantations are established the industry standard isto plant intimate mixtures of ~5 diverse rust (Melampsoraspp)‐resistant varieties (McCracken amp Dawson 1997 VanDen Broek et al 2001)
The foundations for using new plant breeding tech-niques have been established with funding from both thepublic and the private sectors To establish QTL maps 16mapping populations from biparental crosses are understudy in the United Kingdom Nine are under study in theUnited States The average number of individuals in thesefamilies ranges from 150 to 947 (Hanley amp Karp 2014)GS is also being evaluated in S viminalis and preliminaryresults indicate that multiomic approaches combininggenomic and metabolomic data have great potential (Sla-vov amp Davey 2017) For both QTL and GS approachesthe field phenotyping demands are large as several thou-sand individuals need to be phenotyped for a wide rangeof traits These include the following dates of bud burstand growth succession stem height stem density wooddensity and disease resistance The greater the number ofindividuals the more precise the QTL marker maps andGS models are However the logistical and financial chal-lenges of phenotyping large numbers of individuals are
CLIFTON‐BROWN ET AL | 19
considerable because the willow crop is gt5 m tall in thesecond year There is tremendous potential to improve thethroughput of phenotyping using unmanned aerial systemswhich is being tested in the USDA National Institute ofFood and Agriculture (NIFA) Willow SkyCAP project atCornell Further investment in these approaches needs tobe sustained over many years fully realizes the potentialof a marker‐assisted selection programme for willow
To date despite considerable efforts in Europe and theUnited States to establish a routine transformation systemthere has not been a breakthrough in willow but attemptsare ongoing As some form of transformation is typically aprerequisite for genome editing techniques these have notyet been applied to willow
In Europe there are 53 short‐rotation coppice (SRC) bio-mass willow cultivars registered with the Community PlantVariety Office (CPVO) for PBRs of which ~25 are availablecommercially in the United Kingdom There are eightpatented cultivars commercially available in the UnitedStates In Sweden there are nine commercial cultivars regis-tered in Europe and two others which are unregistered(httpssalixenergiseplanting-material) Furthermore thereare about 20 precommercial hybrids in final yield trials inboth the United States and the United Kingdom It has beenestimated that it would take two years to produce the stockrequired to plant 50 ha commercially from the plant stock inthe final yield trials Breeding programmes have alreadydelivered rust‐resistant varieties and increases in yield to themarket The adoption of advanced breeding technologies willlikely lead to a step change in improving traits of interest
5 | POPLAR
Poplar a fast‐growing tree from the northern hemispherewith a small genome size has been adopted for commercialforestry and scientific purposes The genus Populus con-sists of about 29 species classified in six different sectionsPopulus (formerly Leuce) Tacamahaca Aigeiros AbasoTuranga and Leucoides (Eckenwalder 1996) The Populusspecies of most interest for breeding and testing in the Uni-ted States and Europe are P nigra P deltoides P maxi-mowiczii and P trichocarpa (Stanton 2014) Populusclones for biomass production are being developed byintra‐ and interspecies hybridization (DeWoody Trewin ampTaylor 2015 Richardson Isebrands amp Ball 2014 van derSchoot et al 2000) Recurrent selection approaches areused for gradual population improvement and to create eliteclonal lines for commercialization (Berguson McMahon ampRiemenschneider 2017 Neale amp Kremer 2011) Currentlypoplar breeding in the United States occurs in industrialand academic programmes located in the Southeast theMidwest and the Pacific Northwest These use six speciesand five interspecific taxa (Stanton 2014)
The southeastern programme historically focused onrecurrent selection of P deltoides from accessions made inthe lower Mississippi River alluvial plain (Robison Rous-seau amp Zhang 2006) More recently the genetic base hasbeen broadened to produce interspecific hybrids with resis-tance to the fungal infection Septoria musiva which causescankers
In the midwest of the United States populationimprovement efforts are focused on P deltoides selectionsfrom native provenances and hybrid crosses with acces-sions introduced from Europe Interspecific intercontinental(Europe and America) hybrid crosses between P nigra andP deltoides (P times canadensis) are behind many of theleading commercial hybrids which are the most advancedbreeding materials for many applications and regions InMinnesota previous breeding experience and efforts utiliz-ing P maximowiczii and P trichocarpa have been discon-tinued due to Septoria susceptibility and a lack of coldhardiness (Berguson et al 2017) Traits targeted forimprovement include yieldgrowth rate cold hardinessadventitious rooting resistance to Septoria and Melamp-sora leaf rust and stem form The Upper Midwest pro-gramme also carries out wide hybridizations within thesection Populus The P times wettsteinii (P trem-ula times P tremuloides) taxon is bred for gains in growthrate wood quality and resistance to the fungus Entoleucamammata which causes hypoxylon canker (David ampAnderson 2002)
In the Pacific Northwest GreenWood Resources Incleads poplar breeding that emphasizes interspecifichybrid improvement of P times generosa (P deltoides timesP trichocarpa and reciprocal) and P deltoides times P maxi-mowiczii taxa for coastal regions and the P times canadensistaxon for the drier continental regions Intraspecificimprovement of second‐generation breeding populations ofP deltoides P nigra P maximowiczii and P trichocarpaare also involved (Stanton et al 2010) The present focusof GreenWood Resourcesrsquo hybridization is bioenergy feed-stock improvement concentrating on coppice yield wood‐specific gravity and rate of sugar release
Industrial interest in poplar in the United States has his-torically come from the pulp and paper sector althoughveneer and dimensional lumber markets have been pursuedat times Currently the biomass market for liquid trans-portation fuels is being emphasized along with the use oftraditional and improved poplar genotypes for ecosystemservices such as phytoremediation (Tuskan amp Walsh 2001Zalesny et al 2016)
In Europe there are breeding programmes in FranceGermany Italy and Sweden These include the following(a) Alasia Franco Vivai (AFV) programme in northernItaly (b) the French programme led by the poplar Scien-tific Interest Group (GIS Peuplier) and carried out
20 | CLIFTON‐BROWN ET AL
collaboratively between the National Institute for Agricul-tural Research (INRA) the National Research Unit ofScience and Technology for Environment and Agriculture(IRSTEA) and the Forest Cellulose Wood Constructionand Furniture Technology Institute (FCBA) (c) the Ger-man programme at Northwest German Forest Research Sta-tion (NW‐FVA) at Hannoversch Muumlnden and (d) theSwedish programme at the Swedish University of Agricul-tural Sciences and SweTree Technologies AB (Table 1)
AFV leads an Italian poplar breeding programme usingextensive field‐grown germplasm collections of P albaP deltoides P nigra and P trichocarpa While interspeci-fic hybridization uses several taxa the focus is onP times canadensis The breeding programme addresses dis-ease resistance (Marssonina brunnea Melampsora larici‐populina Discosporium populeum and poplar mosaicvirus) growth rate and photoperiod adaptation AFV andGreenWood Resources collaborate in poplar improvementin Europe through the exchange of frozen pollen and seedfor reciprocal breeding projects Plantations in Poland andRomania are currently the focus of the collaboration
The ongoing French GIS Peuplier is developing a long‐term breeding programme based on intraspecific recurrentselection for the four parental species (P deltoides P tri-chocarpa P nigra and P maximowiczii) designed to betterbenefit from hybrid vigour demonstrated by the interspeci-fic crosses P canadensis P deltoides times P trichocarpaand P trichocarpa times P maximowiczii Current selectionpriorities are targeting adaptation to soil and climate condi-tions resistance and tolerance to the most economicallyimportant diseases and pests high volume production underSRC and traditional poplar cultivation regimes as well aswood quality of interest by different markets Currentlygenomic selection is under exploration to increase selectionaccuracy and selection intensity while maintaining geneticdiversity over generations
The German NW‐FVA programme is breeding intersec-tional AigeirosndashTacamahaca hybrids with a focus on resis-tance to Pollaccia elegans Xanthomonas populiDothichiza spp Marssonina brunnea and Melampsoraspp (Stanton 2014) Various cross combinations ofP maximowiczii P trichocarpa P nigra and P deltoideshave led to new cultivars suitable for deployment in vari-etal mixtures of five to ten genotypes of complementarystature high productivity and phenotypic stability (Weis-gerber 1993) The current priority is the selection of culti-vars for high‐yield short‐rotation biomass production Sixhundred P nigra genotypes are maintained in an ex situconservation programme An in situ P nigra conservationeffort involves an inventory of native stands which havebeen molecular fingerprinted for identity and diversity
The Swedish programme is concentrating on locallyadapted genotypes used for short‐rotation forestry (SRF)
because these meet the needs of the current pulping mar-kets Several field trials have shown that commercial poplarclones tested and deployed in Southern and Central Europeare not well adapted to photoperiods and low temperaturesin Sweden and in the Baltics Consequently SwedishUniversity of Agricultural Sciences and SweTree Technolo-gies AB started breeding in Sweden in 1990s to producepoplar clones better adapted to local climates and markets
51 | Molecular breeding technologies
Poplar genetic improvement cannot be rapidly achievedthrough traditional methods alone because of the longbreeding cycles outcrossing breeding systems and highheterozygosity Integrating modern genetic genomic andphenomics techniques with conventional breeding has thepotential to expedite poplar improvement
The genome of poplar has been sequenced (Tuskanet al 2006) It has an estimated genome size of485 plusmn 10 Mbp divided into 19 chromosomes This is smal-ler than other PBCs and makes poplar more amenable togenetic engineering (transgenesis) GS and genome editingPoplar has seen major investment in both the United Statesand Europe being the model system for woody perennialplant genetics and genomics research
52 | Targets for genetic modification
Traits targeted include wood properties (lignin content andcomposition) earlylate flowering male sterility to addressbiosafety regulation issues enhanced yield traits and herbi-cide tolerance These extensive transgenic experiments haveshown differences in recalcitrance to in vitro regenerationand genetic transformation in some of the most importantcommercial hybrid poplars (Alburquerque et al 2016)Further transgene expression stability is being studied Sofar China is the only country known to have commerciallyused transgenic insect‐resistant poplar A precommercialherbicide‐tolerant poplar was trialled for 8 years in the Uni-ted States (Li Meilan Ma Barish amp Strauss 2008) butcould not be released due to stringent environmental riskassessments required for regulatory approval This increasestranslation costs and delays reducing investor confidencefor commercial deployment (Harfouche et al 2011)
The first field trials of transgenic poplar were performedin France in 1987 (Fillatti Sellmer Mccown Haissig ampComai 1987) and in Belgium in 1988 (Deblock 1990)Although there have been a total of 28 research‐scale GMpoplar field trials approved in the European Union underCouncil Directive 90220EEC since October 1991 (inPoland Belgium Finland France Germany Spain Swe-den and in the United Kingdom (Pilate et al 2016) onlyauthorizations in Poland and Belgium are in place today In
CLIFTON‐BROWN ET AL | 21
the United States regulatory notifications and permits fornearly 20000 transgenic poplar trees derived from approxi-mately 600 different constructs have been issued since1995 by the USDAs Animal and Plant Health InspectionService (APHIS) (Strauss et al 2016)
53 | Genome editing CRISPR technologies
Clustered regularly interspaced palindromic repeats(CRISPR) and the CRISPR‐associated (CRISPR‐Cas)nucleases are a groundbreaking genome‐engineering toolthat complements classical plant breeding and transgenicmethods (Moreno‐Mateos et al 2017) Only two publishedstudies in poplar have applied the CRISPRCas9 technol-ogy One is in P tomentosa in which an endogenous phy-toene desaturase gene (PtoPDS) was successfully disruptedsite specifically in the first generation of transgenic plantsresulting in an albino and dwarf phenotype (Fan et al2015) The second was in P tremula times alba in whichhigh CRISPR‐Cas9 mutational efficiency was achieved forthree 4‐coumarateCoA ligase (4Cl) genes 4CL1 4CL2and 4CL5 associated with lignin and flavonoid biosynthe-sis (Zhou et al 2015) Due to its low cost precision andrapidness it is very probable that cultivars or clones pro-duced using CRISPR technology will be ready for market-ing in the near future (Yin et al 2017) Recently aCRISPR with a smaller associated endonuclease has beendiscovered from Prevotella and Francisella 1 (Cpf1) whichmay have advantages over Cas9 In addition there arereports of DNA‐free editing in plants using both CRISPRCpf1 and CRISPR Cas9 for example Ref (Kim et al2017 Mahfouz 2017 Zaidi et al 2017)
It remains unresolved whether plants modified by genomeediting will be regulated as genetically modified organisms(GMOs) by the relevant authorities in different countries(Lozano‐Juste amp Cutler 2014) Regulations to cover thesenew breeding techniques are still evolving but those coun-tries who have published specific guidance (including UnitedStates Argentina and Chile) are indicating that plants pos-sessing simple genome edits will not be regulated as conven-tional transgenesis (Jones 2015b) The first generation ofgenome‐edited crops will likely be phenocopy gene knock-outs that already exist to produce ldquonature identicalrdquo traits thatis traits that could also be derived by conventional breedingDespite this confidence in applying these new powerfulbreeding tools remains limited owing to the uncertain regula-tory environment in many parts of the world (Gao 2018)including the recent ECJ 2018 rulings mentioned earlier
54 | Genomics‐based breeding technologies
Poplar breeding programs are becoming well equipped withuseful genomics tools and resources that are critical to
explore genomewide variability and make use of the varia-tion for enhancing genetic gains Deep transcriptomesequencing resequencing of alternate genomes and GBStechnology for genomewide marker detection using next‐generation sequencing (NGS) are yielding valuable geno-mics tools GWAS with NGS‐based markers facilitatesmarker identification for MAS breeding by design and GS
GWAS approaches have provided a deeper understand-ing of genome function as well as allelic architectures ofcomplex traits (Huang et al 2010) and have been widelyimplemented in poplar for wood characteristics (Porthet al 2013) stomatal patterning carbon gain versus dis-ease resistance (McKown et al 2014) height and phenol-ogy (Evans et al 2014) cell wall chemistry (Mucheroet al 2015) growth and cell walls traits (Fahrenkroget al 2017) bark roughness (Bdeir et al 2017) andheight and diameter growth (Liu et al 2018) Using high‐throughput sequencing and genotyping platforms an enor-mous amount of SNP markers have been used to character-ize the linkage disequilibrium (LD) in poplar (eg Slavovet al 2012 discussed below)
The genetic architecture of photoperiodic traits in peren-nial trees is complex involving many loci However itshows high levels of conservation during evolution (Mau-rya amp Bhalerao 2017) These genomics tools can thereforebe used to address adaptation issues and fine‐tune themovement of elite lines into new environments For exam-ple poor timing of spring bud burst and autumn bud setcan result in frost damage resulting in yield losses (Ilstedt1996) These have been studied in P tremula genotypesalong a latitudinal cline in Sweden (~56ndash66oN) and haverevealed high nucleotide polymorphism in two nonsynony-mous SNPs within and around the phytochrome B2 locus(Ingvarsson Garcia Hall Luquez amp Jansson 2006 Ing-varsson Garcia Luquez Hall amp Jansson 2008) Rese-quencing 94 of these P tremula genotypes for GWASshowed that noncoding variation of a single genomicregion containing the PtFT2 gene described 65 ofobserved genetic variation in bud set along the latitudinalcline (Tan 2018)
Resequencing genomes is currently the most rapid andeffective method detecting genetic differences betweenvariants and for linking loci to complex and importantagronomical and biomass traits thus addressing breedingchallenges associated with long‐lived plants like poplars
To date whole genome resequencing initiatives havebeen launched for several poplar species and genotypes InPopulus LD studies based on genome resequencing sug-gested the feasibility of GWAS in undomesticated popula-tions (Slavov et al 2012) This plant population is beingused to inform breeding for bioenergy development Forexample the detection of reliable phenotypegenotype asso-ciations and molecular signatures of selection requires a
22 | CLIFTON‐BROWN ET AL
detailed knowledge about genomewide patterns of allelefrequency variation LD and recombination suggesting thatGWAS and GS in undomesticated populations may bemore feasible in Populus than previously assumed Slavovet al (2012) have resequenced 16 genomes of P tri-chocarpa and genotyped 120 trees from 10 subpopulationsusing 29213 SNPs (Geraldes et al 2013) The largest everSNP data set of genetic variations in poplar has recentlybeen released providing useful information for breedinghttpswwwbioenergycenterorgbescgwasindexcfmAlso deep sequencing of transcriptomes using RNA‐Seqhas been used for identification of functional genes andmolecular markers that is polymorphism markers andSSRs A multitissue and multiple experimental data set forP trichocarpa RNA‐Seq is publicly available (httpsjgidoegovdoe-jgi-plant-flagship-gene-atlas)
The availability of genomic information of DNA‐con-taining cell organelles (nucleus chloroplast and mitochon-dria) will also allow a holistic approach in poplarmolecular breeding in the near future (Kersten et al2016) Complete Populus genome sequences are availablefor nucleus (P trichocarpa section Tacamahaca) andchloroplasts (seven species and two clones from P trem-ula W52 and P tremula times P alba 717ndash1B4) A compara-tive approach revealed structural and functionalinformation broadening the knowledge base of PopuluscpDNA and stimulating future diagnostic marker develop-ment The availability of whole genome sequences of thesecellular compartments of P tremula holds promise forboosting marker‐assisted poplar breeding Other nucleargenome sequences from additional Populus species arenow available (eg P deltoides (httpsphytozomejgidoegovpz) and will become available in the forthcomingyears (eg P tremula and P tremuloidesmdashPopGenIE(Sjodin Street Sandberg Gustafsson amp Jansson 2009))Recently the characterization of the poplar pan‐genome bygenomewide identification of structural variation in threecrossable poplar species P nigra P deltoides and P tri-chocarpa revealed a deeper understanding of the role ofinter‐ and intraspecific structural variants in poplar pheno-type and may have important implications for breedingparticularly interspecific hybrids (Pinosio et al 2016)
GS has been proposed as an alternative to MAS in cropimprovement (Bernardo amp Yu 2007 Heffner Sorrells ampJannink 2009) GS is particularly well suited for specieswith long generation times for characteristics that displaymoderate‐to‐low heritability for traits that are expensive tomeasure and for selection of traits expressed late in the lifecycle as is the case for most traits of commercial value inforestry (Harfouche et al 2012) Current joint genomesequencing efforts to implement GS in poplar using geno-mic‐estimated breeding values for bioenergy conversiontraits from 49 P trichocarpa families and 20 full‐sib
progeny are taking place at the Oak Ridge National Labo-ratory and GreenWood Resources (Brian Stanton personalcommunication httpscbiornlgov) These data togetherwith the resequenced GWAS population data will be thebasis for developing GS algorithms Genomic breedingtools have been developed for the intraspecific programmetargeting yield resistance to Venturia shoot blightMelampsora leaf rust resistance to Cryptorhynchus lapathistem form wood‐specific gravity and wind firmness (Evanset al 2014 Guerra et al 2016) A newly developedldquobreeding with rare defective allelesrdquo (BRDA) technologyhas been developed to exploit natural variation of P nigraand identify defective variants of genes predicted by priortransgenic research to impact lignin properties Individualtrees carrying naturally defective alleles can then be incor-porated directly into breeding programs thereby bypassingthe need for transgenics (Vanholme et al 2013) Thisnovel breeding technology offers a reverse genetics com-plement to emerging GS for targeted improvement of quan-titative traits (Tsai 2013)
55 | Phenomics‐assisted breeding technology
Phenomics involves the characterization of phenomesmdashthefull set of phenotypes of given individual plants (HouleGovindaraju amp Omholt 2010) Traditional phenotypingtools which inefficiently measure a limited set of pheno-types have become a bottleneck in plant breeding studiesHigh‐throughput plant phenotyping facilities provide accu-rate screening of thousands of plant breeding lines clones orpopulations over time (Fu 2015) are critical for acceleratinggenomics‐based breeding Automated image collection andanalysis phenomics technologies allow accurate and nonde-structive measurements of a diversity of phenotypic traits inlarge breeding populations (Gegas Gay Camargo amp Doo-nan 2014 Goggin Lorence amp Topp 2015 Ludovisi et al2017 Shakoor Lee amp Mockler 2017) One important con-sideration is the identification of relevant and quantifiabletarget traits that are early diagnostic indicators of biomassyield Good progress has been made in elucidating theseunderpinning morpho‐physiological traits that are amenableto remote sensing in Populus (Harfouche Meilan amp Altman2014 Rae Robinson Street amp Taylor 2004) Morerecently Ludovisi et al (2017) developed a novel methodol-ogy for field phenomics of drought stress in a P nigra F2partially inbred population using thermal infrared imagesrecorded from an unmanned aerial vehicle‐based platform
Energy is the current main market for poplar biomass butthe market return provided is not sufficient to support pro-duction expansion even with added demand for environmen-tal and land management ldquoecosystem servicesrdquo such as thetreatment of effluent phytoremediation riparian buffer zonesand agro‐forestry plantings Aviation fuel is a significant
CLIFTON‐BROWN ET AL | 23
target market (Crawford et al 2016) To serve this marketand to reduce current carbon costs of production (Budsberget al 2016) key improvement traits in addition to yield(eg coppice regeneration pestdisease resistance water‐and nutrient‐use efficiencies) will be trace greenhouse gas(GHG) emissions (eg isoprene volatiles) site adaptabilityand biomass conversion efficiency Efforts are underway tohave national environmental protection agenciesrsquo approvalfor poplar hybrids qualifying for renewable energy credits
6 | REFLECTIONS ON THECOMMERCIALIZATIONCHALLENGE
The research and innovation activities reviewed in thispaper aim to advance the genetic improvement of speciesthat can provide feedstocks for bioenergy applicationsshould those markets eventually develop These marketsneed to generate sufficient revenue and adequately dis-tribute it to the actors along the value chain The work onall four crops shares one thing in common long‐termefforts to integrate fundamental knowledge into breedingand crop development along a research and development(RampD) pipeline The development of miscanthus led byAberystwyth University exemplifies the concerted researcheffort that has integrated the RampD activities from eight pro-jects over 14 years with background core research fundingalong an emerging innovation chain (Figure 2) This pro-gramme has produced a first range of conventionally bredseeded interspecies hybrids which are now in upscaling tri-als (Table 3) The application of molecular approaches(Table 5) with further conventional breeding (Table 4)offers the prospect of a second range of improved seededhybrids This example shows that research‐based support ofthe development of new crops or crop types requires along‐term commitment that goes beyond that normallyavailable from project‐based funding (Figure 1) Innovationin this sector requires continuous resourcing of conven-tional breeding operations and capability to minimize timeand investment losses caused by funding discontinuities
This challenge is increased further by the well‐knownmarket failure in the breeding of many agricultural cropspecies The UK Department for Environment Food andRural Affairs (Defra) examined the role of geneticimprovement in relation to nonmarket outcomes such asenvironmental protection and concluded that public invest-ment in breeding was required if profound market failure isto be addressed (Defra 2002) With the exception ofwidely grown hybrid crops such as maize and some high‐value horticultural crops royalties arising from plant breed-ersrsquo rights or other returns to breeders fail to adequatelycompensate for the full cost for research‐based plant breed-ing The result even for major crops such as wheat is sub‐
optimal investment and suboptimal returns for society Thismarket failure is especially acute for perennial crops devel-oped for improved sustainability rather than consumerappeal (Tracy et al 2016) Figure 3 illustrates the underly-ing challenge of capturing value for the breeding effortThe ldquovalley of deathrdquo that results from the low and delayedreturns to investment applies generally to the research‐to‐product innovation pipeline (Beard Ford Koutsky amp Spi-wak 2009) and certainly to most agricultural crop speciesHowever this schematic is particularly relevant to PBCsMost of the value for society from the improved breedingof these crops comes from changes in how agricultural landis used that is it depends on the increased production ofthese crops The value for society includes many ecosys-tems benefits the effects of a return to seminatural peren-nial crop cover that protects soils the increase in soilcarbon storage the protection of vulnerable land or the cul-tivation of polluted soils and the reductions in GHG emis-sions (Lewandowski 2016) By its very nature theproduction of biomass on agricultural land marginal forfood production challenges farm‐level profitability Thecosts of planting material and one‐time nature of cropestablishment are major early‐stage costs and therefore theopportunities for conventional royalty capture by breedersthat are manifold for annual crops are limited for PBCs(Hastings et al 2017) Public investment in developingPBCs for the nonfood biobased sector needs to providemore long‐term support for this critical foundation to a sus-tainable bioeconomy
7 | CONCLUSIONS
This paper provides an overview of research‐based plantbreeding in four leading PBCs For all four PBC generasignificant progress has been made in genetic improvementthrough collaboration between research scientists and thoseoperating ongoing breeding programmes Compared withthe main food crops most PBC breeding programmes dateback only a few decades (Table 1) This breeding efforthas thus co‐evolved with molecular biology and the result-ing ‐omics technologies that can support breeding Thedevelopment of all four PBCs has depended strongly onpublic investment in research and innovation The natureand driver of the investment varied In close associationwith public research organizations or universities all theseprogrammes started with germplasm collection and charac-terization which underpin the selection of parents forexploratory wide crosses for progeny testing (Figure 1Table 4)
Public support for switchgrass in North America wasexplicitly linked to plant breeding with 12 breeding pro-grammes supported in the United States and CanadaSwitchgrass breeding efforts to date using conventional
24 | CLIFTON‐BROWN ET AL
breeding have resulted in over 36 registered cultivars inthe United States (Table 1) with the development of dedi-cated biomass‐type cultivars coming within the past fewyears While ‐omics technologies have been incorporatedinto several of these breeding programmes they have notyet led to commercial deployment in either conventional orhybrid cultivars
Willow genetic improvement was led by the researchcommunity closely linked to plant breeding programmesWillow and poplar have the longest record of public invest-ment in genetic improvement that can be traced back to1920s in the United Kingdom and United States respec-tively Like switchgrass breeding programmes for willoware connected to public research efforts The United King-dom in partnership with the programme based at CornellUniversity remains the European leader in willowimprovement with a long‐term breeding effort closelylinked to supporting biological research at Rothamsted Inwillow F1 hybrids have produced impressive yield gainsover parental germplasm by capturing hybrid vigour Over30 willow clones are commercially available in the UnitedStates and Europe and a further ~90 are under precommer-cial testing (Table 1)
Compared with willow and poplar miscanthus is a rela-tive newcomer with all the current breeding programmesstarting in the 2000s Clonal M times giganteus propagated byrhizomes is expected to be replaced by more readily scal-able seeded hybrids from intra‐ (M sinensis) and inter‐(M sacchariflorus times M sinensis) species crosses with highseed multiplication rates (of gt2000) The first group ofhybrid cultivars is expected to be market‐ready around2022
Of the four genera used as PBCs Populus is the mostadvanced in terms of achievements in biological researchas a result of its use as a model for basic research of treesMuch of this biological research is not directly connectedto plant breeding Nevertheless reflecting the fact thatpoplar is widely grown as a single‐stem tree in SRF thereare about 60 commercially available clones and an addi-tional 80 clones in commercial pipelines (Table 1) Trans-genic poplar hybrids have moved beyond proof of conceptto commercial reality in China
Many PBC programmes have initiated long‐term con-ventional recurrent selection breeding cycles for populationimprovement which is a key process in increasing yieldthrough hybrid vigour As this approach requires manyyears most programmes are experimenting with molecularbreeding methods as these have the potential to accelerateprecision breeding For all four PBCs investments in basicgenetic and genomic resources including the developmentof mapping populations for QTLs and whole genomesequences are available to support long‐term advancesMore recently association genetics with panels of diversegermplasm are being used as training populations for GSmodels (Table 5) These efforts are benefitting from pub-licly available DNA sequences and whole genome assem-blies in crop databases Key to these accelerated breedingtechnologies are developments in novel phenomics tech-nologies to bridge the genotypephenotype gap In poplarnovel remote sensing field phenotyping is now beingdeployed to assist breeders These advances are being com-bined with in vitro and in planta modern molecular breed-ing techniques such as CRISPR (Table 5) CRISPRtechnology for genome editing has been proven in poplar
FIGURE 2 A schematic developmentpathway for miscanthus in the UnitedKingdom related to the investment in RampDprojects at Aberystwyth (top coloured areasfor projects in the three categories basicresearch breeding and commercialupscaling) leading to a projected croppingarea of 350000 ha by 2030 with clonal andsuccessive ranges of improved seed‐basedhybrids Purple represents theBiotechnology and Biological SciencesResearch Council (BBSRC) and brown theDepartment for Environment Food andRural Affairs (Defra) (UK Nationalfunding) blue bars represent EU fundingand green private sector funding (Terravestaand CERES) and GIANT‐LINK andMiscanthus Upscaling Technology (MUST)are publicndashprivate‐initiatives (PPI)
CLIFTON‐BROWN ET AL | 25
This technology is also being applied in switchgrass andmiscanthus (Table 5) but the future of CRISPR in com-mercial breeding for the European market is uncertain inthe light of recent ECJ 2018 rulings
There is integration of research and plant breeding itselfin all four PBCs Therefore estimating the ongoing costs ofmaintaining these breeding programmes is difficult Invest-ment in research also seeks wider benefits associated withtechnological advances in plant science rather than cultivardevelopment per se However in all cases the conventionalbreeding cycle shown in Figure 1 is the basic ldquoenginerdquo withmolecular technologies (‐omics) serving to accelerate thisengine The history of the development shows that the exis-tence of these breeding programmes is essential to gain ben-efits from the biological research Despite this it is thisessential step that is at most risk from reductions in invest-ment A conventional breeding programme typicallyrequires a breeder and several technicians who are supported
over the long term (20ndash30 years Figure 1) at costs of about05 to 10 million Euro per year (as of 2018) The analysisreported here shows that the time needed to perform onecycle of conventional breeding bringing germplasm fromthe wild to a commercial hybrid ranged from 11 years inswitchgrass to 26 years in poplar (Figure 1) In a maturecrop grown on over 100000 ha with effective cultivar pro-tection and a suitable business model this level of revenuecould come from royalties Until such levels are reachedPBCs lie in the innovation valley of death (Figure 3) andneed public support
Applying industrial ldquotechnology readiness levelsrdquo (TRL)originally developed for aerospace (Heacuteder 2017) to ourplant breeding efforts we estimate many promising hybridscultivars are at TRL levels of 3ndash4 In Table 3 experts ineach crop estimate that it would take 3 years from now toupscale planting material from leading cultivars in plot tri-als to 100 ha
FIGURE 3 A schematic relating some of the steps in the innovation chain from relatively basic crop science research through to thedeployment in commercial cropping systems and value chains The shape of the funnel above the expanding development and deploymentrepresents the availability of investment along the development chain from relatively basic research at the top to the upscaled deployment at thebottom Plant breeding links the research effort with the development of cropping systems The constriction represents the constrained fundingfor breeding that links conventional public research investment and the potential returns from commercial development The handover pointsbetween publicly funded work to develop the germplasm resources (often known as prebreeding) the breeding and the subsequent cropdevelopment are shown on the left The constriction point is aggravated by the lack academic rewards for this essential breeding activity Theoutcome is such that this innovation system is constrained by the precarious resourcing of plant breeding The authorsrsquo assessment ofdevelopment status of the four species is shown (poplar having two one for short‐rotation coppice (SRC) poplar and one for the more traditionalshort‐rotation forestry (SRF)) The four new perennial biomass crops (PBCs) are now in the critical phase of depending of plant breedingprogress without the income stream from a large crop production base
26 | CLIFTON‐BROWN ET AL
Taking the UK example mentioned in the introductionplanting rates of ~35000 ha per year from 2022 onwardsare needed to reach over 1 m ha by 2050 Ongoing workin the UK‐funded Miscanthus Upscaling Technology(MUST) project shows that ramping annual hybrid seedproduction from the current level of sufficient seed for10 ha in 2018 to 35000 ha would take about 5 yearsassuming no setbacks If current hybrids of any of the fourPBCs in the upscaling pipeline fail on any step for exam-ple lower than expected multiplication rates or unforeseenagronomic barriers then further selections from ongoingbreeding are needed to replace earlier candidates
In conclusion the breeding foundations have been laidwell for switchgrass miscanthus willow and poplar owingto public funding over the long time periods necessaryImproved cultivars or genotypes are available that could bescaled up over a few years if real sustained market oppor-tunities emerged in response to sustained favourable poli-cies and industrial market pull The potential contributionsof growing and using these PBCs for socioeconomic andenvironmental benefits are clear but how farmers andothers in commercial value chains are rewarded for mass‐scale deployment as is necessary is not obvious at presentTherefore mass‐scale deployment of these lignocellulosecrops needs developments outside the breeding arenas todrive breeding activities more rapidly and extensively
8 | UNCITED REFERENCE
Huang et al (2019)
ACKNOWLEDGEMENTS
UK The UK‐led miscanthus research and breeding wasmainly supported by the Biotechnology and BiologicalSciences Research Council (BBSRC) Department for Envi-ronment Food and Rural Affairs (Defra) the BBSRC CSPstrategic funding grant BBCSP17301 Innovate UKBBSRC ldquoMUSTrdquo BBN0161491 CERES Inc and Ter-ravesta Ltd through the GIANT‐LINK project (LK0863)Genomic selection and genomewide association study activi-ties were supported by BBSRC grant BBK01711X1 theBBSRC strategic programme grant on Energy Grasses ampBio‐refining BBSEW10963A01 The UK‐led willow RampDwork reported here was supported by BBSRC (BBSEC00005199 BBSEC00005201 BBG0162161 BBE0068331 BBG00580X1 and BBSEC000I0410) Defra(NF0424 NF0426) and the Department of Trade and Indus-try (DTI) (BW6005990000) IT The Brain Gain Program(Rientro dei cervelli) of the Italian Ministry of EducationUniversity and Research supports Antoine Harfouche USContributions by Gerald Tuskan to this manuscript were sup-ported by the Center for Bioenergy Innovation a US
Department of Energy Bioenergy Research Center supportedby the Office of Biological and Environmental Research inthe DOE Office of Science under contract number DE‐AC05‐00OR22725 Willow breeding efforts at CornellUniversity have been supported by grants from the USDepartment of Agriculture National Institute of Food andAgriculture Contributions by the University of Illinois weresupported primarily by the DOE Office of Science Office ofBiological and Environmental Research (BER) grant nosDE‐SC0006634 DE‐SC0012379 and DE‐SC0018420 (Cen-ter for Advanced Bioenergy and Bioproducts Innovation)and the Energy Biosciences Institute EU We would like tofurther acknowledge contributions from the EU projectsldquoOPTIMISCrdquo FP7‐289159 on miscanthus and ldquoWATBIOrdquoFP7‐311929 on poplar and miscanthus as well as ldquoGRACErdquoH2020‐EU326 Biobased Industries Joint Technology Ini-tiative (BBI‐JTI) Project ID 745012 on miscanthus
CONFLICT OF INTEREST
The authors declare that progress reported in this paperwhich includes input from industrial partners is not biasedby their business interests
ORCID
John Clifton‐Brown httporcidorg0000-0001-6477-5452Antoine Harfouche httporcidorg0000-0003-3998-0694Lawrence B Smart httporcidorg0000-0002-7812-7736Danny Awty‐Carroll httporcidorg0000-0001-5855-0775VasileBotnari httpsorcidorg0000-0002-0470-0384Lindsay V Clark httporcidorg0000-0002-3881-9252SalvatoreCosentino httporcidorg0000-0001-7076-8777Lin SHuang httporcidorg0000-0003-3079-326XJon P McCalmont httporcidorg0000-0002-5978-9574Paul Robson httporcidorg0000-0003-1841-3594AnatoliiSandu httporcidorg0000-0002-7900-448XDaniloScordia httporcidorg0000-0002-3822-788XRezaShafiei httporcidorg0000-0003-0891-0730Gail Taylor httporcidorg0000-0001-8470-6390IrisLewandowski httporcidorg0000-0002-0388-4521
REFERENCES
Alburquerque N Baldacci‐Cresp F Baucher M Casacuberta JM Collonnier C ElJaziri M hellip Burgos L (2016) New trans-formation technologies for trees In C Vettori F Gallardo H
CLIFTON‐BROWN ET AL | 27
Haumlggman V Kazana F Migliacci G Pilateamp M Fladung(Eds) Biosafety of forest transgenic trees (pp 31ndash66) DordrechtNetherlands Springer
Argus G W (2007) Salix (Salicaceae) distribution maps and a syn-opsis of their classification in North America north of MexicoHarvard Papers in Botany 12 335ndash368 httpsdoiorg1031001043-4534(2007)12[335SSDMAA]20CO2
Atienza S G Ramirez M C amp Martin A (2003) Mapping‐QTLscontrolling flowering date in Miscanthus sinensis Anderss CerealResearch Communications 31 265ndash271
Atienza S G Satovic Z Petersen K K Dolstra O amp Martin A(2003) Identification of QTLs influencing agronomic traits inMiscanthus sinensis Anderss I Total height flag‐leaf height andstem diameter Theoretical and Applied Genetics 107 123ndash129
Atienza S G Satovic Z Petersen K K Dolstra O amp Martin A(2003) Influencing combustion quality in Miscanthus sinensisAnderss Identification of QTLs for calcium phosphorus and sul-phur content Plant Breeding 122 141ndash145
Bdeir R Muchero W Yordanov Y Tuskan G A Busov V ampGailing O (2017) Quantitative trait locus mapping of Populusbark features and stem diameter BMC Plant Biology 17 224httpsdoiorg101186s12870-017-1166-4
Beard T R Ford G S Koutsky T M amp Spiwak L J (2009) AValley of Death in the innovation sequence An economic investi-gation Research Evaluation 18 343ndash356 httpsdoiorg103152095820209X481057
Berguson W McMahon B amp Riemenschneider D (2017) Addi-tive and non‐additive genetic variances for tree growth in severalhybrid poplar populations and implications regarding breedingstrategy Silvae Genetica 66(1) 33ndash39 httpsdoiorg101515sg-2017-0005
Berlin S Trybush S O Fogelqvist J Gyllenstrand N Halling-baumlck H R Aringhman I hellip Hanley S J (2014) Genetic diversitypopulation structure and phenotypic variation in European Salixviminalis L (Salicaceae) Tree Genetics amp Genomes 10 1595ndash1610 httpsdoiorg101007s11295-014-0782-5
Bernardo R amp Yu J (2007) Prospects for genomewide selectionfor quantitative traits in maize Crop Science 47 1082ndash1090httpsdoiorg102135cropsci2006110690
Biswal A K Atmodjo M A Li M Baxter H L Yoo C GPu Y hellip Mohnen D (2018) Sugar release and growth of bio-fuel crops are improved by downregulation of pectin biosynthesisNature Biotechnology 36 249ndash257
Bmu (2009) National biomass action plan for Germany (p 17) Bun-desministerium fuumlr Umwelt Naturschutz und ReaktorsicherheitBundesministerium fuumlr Ernaumlhrung (BMU) amp Landwirtschaft undVerbraucherschutz (BMELV) 11055 Berlin | Germany Retrievedfrom httpwwwbmeldeSharedDocsDownloadsENPublicationsBiomassActionPlanpdf__blob=publicationFile
Brummer E C (1999) Capturing heterosis in forage crop cultivardevelopment Crop Science 39 943ndash954 httpsdoiorg102135cropsci19990011183X003900040001x
Budsberg E Crawford J T Morgan H Chin W S Bura R ampGustafson R (2016) Hydrocarbon bio‐jet fuel from bioconver-sion of poplar biomass Life cycle assessment Biotechnology forBiofuels 9 170 httpsdoiorg101186s13068-016-0582-2
Burris K P Dlugosz E M Collins A G Stewart C N amp Lena-ghan S C (2016) Development of a rapid low‐cost protoplasttransfection system for switchgrass (Panicum virgatum L) Plant
Cell Reports 35 693ndash704 httpsdoiorg101007s00299-015-1913-7
Casler M (2012) Switchgrass breeding genetics and genomics InA Monti (Ed) Switchgrass (pp 29ndash54) London UK Springer
Casler M Mitchell R amp Vogel K (2012) Switchgrass In CKole C P Joshi amp D R Shonnard (Eds) Handbook of bioen-ergy crop plants Vol 2 New York NY Taylor amp Francis
Casler M D amp Ramstein G P (2018) Breeding for biomass yieldin switchgrass using surrogate measures of yield BioEnergyResearch 11 6ndash12
Casler M D Sosa S Hoffman L Mayton H Ernst C Adler PR hellip Bonos S A (2017) Biomass Yield of Switchgrass Culti-vars under High‐ versus Low‐Input Conditions Crop Science 57821ndash832 httpsdoiorg102135cropsci2016080698
Casler M D amp Vogel K P (2014) Selection for biomass yield inupland lowland and hybrid switchgrass Crop Science 54 626ndash636 httpsdoiorg102135cropsci2013040239
Casler M D Vogel K P amp Harrison M (2015) Switchgrassgermplasm resources Crop Science 55 2463ndash2478 httpsdoiorg102135cropsci2015020076
Casler M D Vogel K P Lee D K Mitchell R B Adler P RSulc R M hellip Moore K J (2018) 30 Years of progress towardincreasing biomass yield of switchgrass and big bluestem CropScience 58 1242
Clark L V Brummer J E Głowacka K Hall M C Heo K PengJ hellip Sacks E J (2014) A footprint of past climate change on thediversity and population structure of Miscanthus sinensis Annals ofBotany 114 97ndash107 httpsdoiorg101093aobmcu084
Clark L V Dzyubenko E Dzyubenko N Bagmet L SabitovA Chebukin P hellip Sacks E J (2016) Ecological characteristicsand in situ genetic associations for yield‐component traits of wildMiscanthus from eastern Russia Annals of Botany 118 941ndash955
Clark L V Jin X Petersen K K Anzoua K G Bagmet LChebukin P hellip Sacks E J(2018) Population structure of Mis-canthus sacchariflorus reveals two major polyploidization eventstetraploid‐mediated unidirectional introgression from diploid Msinensis and diversity centred around the Yellow Sea Annals ofBotany 20 1ndash18 httpsdoiorg101093aobmcy1161
Clark L V Stewart J R Nishiwaki A Toma Y Kjeldsen J BJoslashrgensen U hellip Sacks E J (2015) Genetic structure ofMiscanthus sinensis and Miscanthus sacchariflorus in Japan indi-cates a gradient of bidirectional but asymmetric introgressionJournal of Experimental Botany 66 4213ndash4225
Clifton‐Brown J Hastings A Mos M McCalmont J P Ashman CAwty‐Carroll DhellipCracroft‐EleyW (2017) Progress in upscalingMis-canthus biomass production for the European bio‐economy with seed‐based hybridsGlobal Change Biology Bioenergy 9 6ndash17
Clifton‐Brown J Schwarz K U amp Hastings A (2015) History ofthe development of Miscanthus as a bioenergy crop From smallbeginnings to potential realisation Biology and Environment‐Pro-ceedings of the Royal Irish Academy 115B 45ndash57
Clifton‐Brown J Senior H Purdy S Horsnell R Lankamp BMuumlennekhoff A-K hellip Bentley A R (2018) Investigating thepotential of novel non‐woven fabrics to increase pollination effi-ciency in plant breeding PloS One (Accepted)
Crawford J T Shan CW Budsberg E Morgan H Bura R ampGus-tafson R (2016) Hydrocarbon bio‐jet fuel from bioconversion ofpoplar biomass Techno‐economic assessment Biotechnology forBiofuels 9 141 httpsdoiorg101186s13068-016-0545-7
28 | CLIFTON‐BROWN ET AL
Dalton S (2013) Biotechnology of Miscanthus In S M Jain amp SD Gupta (Eds) Biotechnology of neglected and underutilizedcrops (pp 243ndash294) Dordrecht Netherlands Springer
Davey C L Robson P Hawkins S Farrar K Clifton‐Brown J CDonnison I S amp Slavov G T (2017) Genetic relationshipsbetween spring emergence canopy phenology and biomass yieldincrease the accuracy of genomic prediction in Miscanthus Journalof Experimental Botany 68 5093ndash5102 httpsdoiorg101093jxberx339
David X amp Anderson X (2002) Aspen and larch genetics cooper-ative annual report 13 St Paul MN Department of ForestResources University of Minnesota
Deblock M (1990) Factors influencing the tissue‐culture and theAgrobacterium‐Tumefaciens‐mediated transformation of hybridaspen and poplar clones Plant Physiology 93 1110ndash1116httpsdoiorg101104pp9331110
Defra (2002) The role of future public research investment in thegenetic improvement of UK‐grown crops (p 220) LondonRetrieved from sciencesearchdefragovukDefaultaspxMenu=MenuampModule=MoreampLocation=NoneampCompleted=220ampProjectID=10412
Dewoody J Trewin H amp Taylor G (2015) Genetic and morpho-logical differentiation in Populus nigra L Isolation by coloniza-tion or isolation by adaptation Molecular Ecology 24 2641ndash2655
Dong H Liu S Clark L V Sharma S Gifford J M Juvik JA hellip Sacks E J (2018) Genetic mapping of biomass yield inthree interconnected Miscanthus populations Global Change Biol-ogy Bioenergy 10 165ndash185
Dukes (2017) Digest of UK Energy Statistics (DUKES) (p 264) Lon-don UK Department for Business Energy amp Industrial StrategyRetrieved from wwwgovukgovernmentcollectionsdigest-of-uk-energy-statistics-dukes
Eckenwalder J (1996) Systematics and evolution of Populus In RStettler H Bradshaw J Heilman amp T Hinckley (Eds) Biologyof Populus and its implications for management and conservation(pp 7‐32) Ottawa ON National Research Council of Canada
Elshire R J Glaubitz J C Sun Q Poland J A Kawamoto KBuckler E S amp Mitchell S E (2011) A robust simple geno-typing‐by‐sequencing (GBS) approach for high diversity speciesPloS One 6 e19379
Evans H (2016) Bioenergy crops in the UK Case studies of suc-cessful whole farm integration (p 23) Loughborough UKEnergy Technologies Institute Retrieved from httpswwweticouklibrarybioenergy-crops-in-the-uk-case-studies-on-successful-whole-farm-integration-evidence-pack
Evans H (2017) Increasing UK biomass production through moreproductive use of land (p 7) Loughborough UK Energy Tech-nologies Institute Retrieved from httpswwweticouklibraryan-eti-perspective-increasing-uk-biomass-production-through-more-productive-use-of-land
Evans J Sanciangco M D Lau K H Crisovan E Barry KDaum C hellip Buell C R (2018) Extensive genetic diversity ispresent within North American switchgrass germplasm The PlantGenome 11 1ndash16 httpsdoiorg103835plantgenome2017060055
Evans L M Slavov G T Rodgers‐Melnick E Martin J RanjanP Muchero W hellip DiFazio S P (2014) Population genomicsof Populus trichocarpa identifies signatures of selection and
adaptive trait associations Nature Genetics 46 1089ndash1096httpsdoiorg101038ng3075
Fabio E S Volk T A Miller R O Serapiglia M J Gauch HG Van Rees K C J hellip Smart L B (2017) Genotypetimes envi-ronment interaction analysis of North American shrub willowyield trials confirms superior performance of triploid hybrids Glo-bal Change Biology Bioenergy 9 445ndash459 httpsdoiorg101111gcbb12344
Fahrenkrog A M Neves L G Resende M F R Vazquez A Ide los Campos G Dervinis C hellip Kirst M (2017) Genome‐wide association study reveals putative regulators of bioenergytraits in Populus deltoides New Phytologist 213 799ndash811
Fan D Liu T T Li C F Jiao B Li S Hou Y S amp Luo KM (2015) Efficient CRISPRCas9‐mediated targeted mutagenesisin Populus in the first generation Scientific Reports 5 12217
Feng X P Lourgant K Castric V Saumitou-Laprade P ZhengB S Jiang D M amp Brancourt-Hulmel M (2014) The discov-ery of natural accessions related to Miscanthus x giganteus usingchloroplast DNA Crop Science 54 1645ndash1655
Fillatti J J Sellmer J Mccown B Haissig B amp Comai L(1987) Agrobacterium-mediated transformation and regenerationof Populus Molecular amp General Genetics 206 192ndash199
Fu Y B (2015) Understanding crop genetic diversity under modernplant breeding Theoretical and Applied Genetics 128 2131ndash2142 httpsdoiorg101007s00122-015-2585-y
Fu C Mielenz J R Xiao X Ge Y Hamilton C Y RodriguezM hellip Wang Z‐Y (2011) Genetic manipulation of ligninreduces recalcitrance and improves ethanol production fromswitchgrass Proceedings of the National Academy of Sciences ofthe United States of America 108 3803ndash3808 httpsdoiorg101073pnas1100310108
Gao C (2018) The future of CRISPR technologies in agricultureNature Reviews Molecular Cell Biology 19(5) 275ndash276
Gegas V Gay A Camargo A amp Doonan J (2014) Challengesof Crop Phenomics in the Post-genomic Era In J Hancock (Ed)Phenomics (pp 142ndash171) Boca Raton FL CRC Press
Geraldes A DiFazio S P Slavov G T Ranjan P Muchero WHannemann J hellip Tuskan G A (2013) A 34K SNP genotypingarray for Populus trichocarpa Design application to the study ofnatural populations and transferability to other Populus speciesMolecular Ecology Resources 13 306ndash323
Gifford J M Chae W B Swaminathan K Moose S P amp JuvikJ A (2015) Mapping the genome of Miscanthus sinensis forQTL associated with biomass productivity Global Change Biol-ogy Bioenergy 7 797ndash810
Goggin F L Lorence A amp Topp C N (2015) Applying high‐throughput phenotyping to plant‐insect interactions Picturingmore resistant crops Current Opinion in Insect Science 9 69ndash76httpsdoiorg101016jcois201503002
Grabowski P P Evans J Daum C Deshpande S Barry K WKennedy M hellip Casler M D (2017) Genome‐wide associationswith flowering time in switchgrass using exome‐capture sequenc-ing data New Phytologist 213 154ndash169 httpsdoiorg101111nph14101
Greef J M amp Deuter M (1993) Syntaxonomy of Miscanthus xgiganteus GREEF et DEU Angewandte Botanik 67 87ndash90
Guerra F P Richards J H Fiehn O Famula R Stanton B JShuren R hellip Neale D B (2016) Analysis of the genetic varia-tion in growth ecophysiology and chemical and metabolomic
CLIFTON‐BROWN ET AL | 29
composition of wood of Populus trichocarpa provenances TreeGenetics amp Genomes 12 6 httpsdoiorg101007s11295-015-0965-8
Guo H P Shao R X Hong C T Hu H K Zheng B S ampZhang Q X (2013) Rapid in vitro propagation of bioenergy cropMiscanthus sacchariflorus Applied Mechanics and Materials 260181ndash186
Hallingbaumlck H R Fogelqvist J Powers S J Turrion‐Gomez JRossiter R Amey J hellip Roumlnnberg‐Waumlstljung A‐C (2016)Association mapping in Salix viminalis L (Salicaceae)ndashIdentifica-tion of candidate genes associated with growth and phenologyGlobal Change Biology Bioenergy 8 670ndash685
Hanley S J amp Karp A (2014) Genetic strategies for dissectingcomplex traits in biomass willows (Salix spp) Tree Physiology34 1167ndash1180 httpsdoiorg101093treephystpt089
Harfouche A Meilan R amp Altman A (2011) Tree genetic engi-neering and applications to sustainable forestry and biomass pro-duction Trends in Biotechnology 29 9ndash17 httpsdoiorg101016jtibtech201009003
Harfouche A Meilan R amp Altman A (2014) Molecular and phys-iological responses to abiotic stress in forest trees and their rele-vance to tree improvement Tree Physiology 34 1181ndash1198httpsdoiorg101093treephystpu012
Harfouche A Meilan R Kirst M Morgante M Boerjan WSabatti M amp Mugnozza G S (2012) Accelerating the domesti-cation of forest trees in a changing world Trends in PlantScience 17 64ndash72 httpsdoiorg101016jtplants201111005
Hastings A Clifton‐Brown J Wattenbach M Mitchell C PStampfl P amp Smith P (2009) Future energy potential of Mis-canthus in Europe Global Change Biology Bioenergy 1 180ndash196
Hastings A Mos M Yesufu J AMccalmont J Schwarz KShafei RhellipClifton-Brown J (2017) Economic and environmen-tal assessment of seed and rhizome propagated Miscanthus in theUK Frontiers in Plant Science 8 16 httpsdoiorg103389fpls201701058
Heacuteder M (2017) From NASA to EU The evolution of the TRLscale in Public Sector Innovation The Innovation Journal 22 1ndash23 httpsdoiorg103389fpls201701058
Heffner E L Sorrells M E amp Jannink J L (2009) Genomicselection for crop improvement Crop Science 49 1ndash12 httpsdoiorg102135cropsci2008080512
Hodkinson T R Klaas M Jones M B Prickett R amp Barth S(2015) Miscanthus A case study for the utilization of naturalgenetic variation Plant Genetic Resources‐Characterization andUtilization 13 219ndash237 httpsdoiorg101017S147926211400094X
Hodkinson T R amp Renvoize S (2001) Nomenclature of Miscant-hus xgiganteus (Poaceae) Kew Bulletin 56 759ndash760 httpsdoiorg1023074117709
Houle D Govindaraju D R amp Omholt S (2010) Phenomics Thenext challenge Nature Reviews Genetics 11 855ndash866 httpsdoiorg101038nrg2897
Huang X H Wei X H Sang T Zhao Q Feng Q Zhao Yhellip Han B (2010) Genome‐wide association studies of 14 agro-nomic traits in rice landraces Nature Genetics 42 961ndashU976
Hwang O‐J Cho M‐A Han Y‐J Kim Y‐M Lim S‐H KimD‐S hellip Kim J‐I (2014) Agrobacterium‐mediated genetic trans-formation of Miscanthus sinensis Plant Cell Tissue and OrganCulture (PCTOC) 117 51ndash63
Hwang O‐J Lim S‐H Han Y‐J Shin A‐Y Kim D‐S ampKim J‐I (2014) Phenotypic characterization of transgenic Mis-canthus sinensis plants overexpressing Arabidopsis phytochromeB International Journal of Photoenergy 2014501016
Ilstedt B (1996) Genetics and performance of Belgian poplar clonestested in Sweden International Journal of Forest Genetics 3183ndash195
Ingvarsson P K Garcia M V Hall D Luquez V amp Jansson S(2006) Clinal variation in phyB2 a candidate gene for day‐length‐induced growth cessation and bud set across a latitudinalgradient in European aspen (Populus tremula) Genetics 1721845ndash1853
Ingvarsson P K Garcia M V Luquez V Hall D amp Jansson S(2008) Nucleotide polymorphism and phenotypic associationswithin and around the phytochrome B2 locus in European aspen(Populus tremula Salicaceae) Genetics 178 2217ndash2226 httpsdoiorg101534genetics107082354
Jannink J‐L Lorenz A J amp Iwata H (2010) Genomic selectionin plant breeding From theory to practice Briefings in FunctionalGenomics 9 166ndash177 httpsdoiorg101093bfgpelq001
Jiang J Guan Y Mccormick S Juvik J Lubberstedt T amp FeiS‐Z (2017) Gametophytic self‐incompatibility is operative inMiscanthus sinensis (Poaceae) and is affected by pistil age CropScience 57 1948ndash1956
Jones H D (2015a) Future of breeding by genome editing is in thehands of regulators GM Crops amp Food 6 223ndash232
Jones H D (2015b) Regulatory uncertainty over genome editingNature Plants 1 14011
Kalinina O Nunn C Sanderson R Hastings A F S van der Weijde TOumlzguumlven MhellipClifton‐Brown J C (2017) ExtendingMiscanthus cul-tivation with novel germplasm at six contrasting sites Frontiers in PlantScience 8563 httpsdoiorg103389fpls201700563
Karp A Hanley S J Trybush S O Macalpine W Pei M ampShield I (2011) Genetic improvement of willow for bioenergyand biofuels free access Journal of Integrative Plant Biology 53151ndash165 httpsdoiorg101111j1744-7909201001015x
Kersten B Faivre Rampant P Mader M Le Paslier M‐C Bou-non R Berard A hellip Fladung M (2016) Genome sequences ofPopulus tremula chloroplast and mitochondrion Implications forholistic poplar breeding PloS One 11 e0147209 httpsdoiorg101371journalpone0147209
Kiesel A Nunn C Iqbal Y Van der Weijde T Wagner MOumlzguumlven M hellip Lewandowski I (2017) Site‐specific manage-ment of Miscanthus genotypes for combustion and anaerobicdigestion A comparison of energy yields Frontiers in PlantScience 8 347 httpsdoiorg103389fpls201700347
Kim H Kim S T Ryu J Kang B C Kim J S amp Kim S G(2017) CRISPRCpf1‐mediated DNA‐free plant genome editingNature Communications 8 14406 httpsdoiorg101038ncomms14406
King Z R Bray A L Lafayette P R amp Parrott W A (2014)Biolistic transformation of elite genotypes of switchgrass (Pan-icum virgatum L) Plant Cell Reports 33 313ndash322 httpsdoiorg101007s00299-013-1531-1
Kopp R F Maynard C A De Niella P R Smart L B amp Abra-hamson L P (2002) Collection and storage of pollen from Salix(Salicaceae) American Journal of Botany 89 248ndash252
Krzyżak J Pogrzeba M Rusinowski S Clifton‐Brown J McCal-mont J P Kiesel A hellip Mos M (2017) Heavy metal uptake
30 | CLIFTON‐BROWN ET AL
by novel miscanthus seed‐based hybrids cultivated in heavy metalcontaminated soil Civil and Environmental Engineering Reports26 121ndash132 httpsdoiorg101515ceer-2017-0040
Kuzovkina Y A amp Quigley M F (2005) Willows beyond wet-lands Uses of Salix L species for environmental projects WaterAir and Soil Pollution 162 183ndash204 httpsdoiorg101007s11270-005-6272-5
Lazarus W Headlee W L amp Zalesny R S (2015) Impacts of sup-plyshed‐level differences in productivity and land costs on the eco-nomics of hybrid poplar production in Minnesota USA BioEnergyResearch 8 231ndash248 httpsdoiorg101007s12155-014-9520-y
Lewandowski I (2015) Securing a sustainable biomass supply in agrowing bioeconomy Global Food Security 6 34ndash42 httpsdoiorg101016jgfs201510001
Lewandowski I (2016) The role of perennial biomass crops in agrowing bioeconomy In S Barth D Murphy‐Bokern O Kalin-ina G Taylor amp M Jones (Eds) Perennial biomass crops for aresource‐constrained world (pp 3ndash13) Cham SwitzerlandSpringer Switzerland AG
Li J L Meilan R Ma C Barish M amp Strauss S H (2008)Stability of herbicide resistance over 8 years of coppice in field‐grown genetically engineered poplars Western Journal of AppliedForestry 23 89ndash93
Li R Y amp Qu R D (2011) High throughput Agrobacterium‐medi-ated switchgrass transformation Biomass amp Bioenergy 35 1046ndash1054 httpsdoiorg101016jbiombioe201011025
Lindegaard K N amp Barker J H A (1997) Breeding willows forbiomass Aspects of Applied Biology 49 155ndash162
Linde‐Laursen I B (1993) Cytogenetic analysis of MiscanthusGiganteus an interspecific hybrid Hereditas 119 297ndash300httpsdoiorg101111j1601-5223199300297x
Liu Y Merrick P Zhang Z Z Ji C H Yang B amp Fei S Z(2018) Targeted mutagenesis in tetraploid switchgrass (Panicumvirgatum L) using CRISPRCas9 Plant Biotechnology Journal16 381ndash393
Liu L L Wu Y Q Wang Y W amp Samuels T (2012) A high‐density simple sequence repeat‐based genetic linkage map ofswitchgrass G3 (Bethesda Md) 2 357ndash370
Lovett A Suumlnnenberg G amp Dockerty T (2014) The availabilityof land for perennial energy crops in Great Britain GlobalChange Biology Bioenergy 6 99ndash107 httpsdoiorg101111gcbb12147
Lozano‐Juste J amp Cutler S R (2014) Plant genome engineering infull bloom Trends in Plant Science 19 284ndash287 httpsdoiorg101016jtplants201402014
Lu F Lipka A E Glaubitz J Elshire R Cherney J H CaslerM D hellip Costich D E (2013) Switchgrass Genomic DiversityPloidy and Evolution Novel Insights from a Network‐Based SNPDiscovery Protocol Plos Genetics 9 e1003215
Ludovisi R Tauro F Salvati R Khoury S Mugnozza G S ampHarfouche A (2017) UAV‐based thermal imaging for high‐throughput field phenotyping of black poplar response to droughtFrontiers in Plant Science 81681
Macalpine W Shield I amp Karp A (2010) Seed to near market vari-ety the BEGIN willow breeding pipeline 2003ndash2010 and beyond InBioten conference proceedings Birmingham pp 21ndash23
Macalpine W J Shield I F Trybush S O Hayes C M ampKarp A (2008) Overcoming barriers to crossing in willow (Salixspp) breeding Aspects of Applied Biology 90 173ndash180
Mahfouz M M (2017) The efficient tool CRISPR‐Cpf1 NaturePlants 3 17028
Malinowska M Donnison I S amp Robson P R (2017) Phenomicsanalysis of drought responses in Miscanthus collected from differ-ent geographical locations Global Change Biology Bioenergy 978ndash91
Maroder H L Prego I A Facciuto G R amp Maldonado S B(2000) Storage behaviour of Salix alba and Salix matsudanaseeds Annals of Botany 86 1017ndash1021 httpsdoiorg101006anbo20001265
Martinez‐Reyna J amp Vogel K (2002) Incompatibility systems inswitchgrass Crop Science 42 1800ndash1805 httpsdoiorg102135cropsci20021800
Maurya J P amp Bhalerao R P (2017) Photoperiod‐and tempera-ture‐mediated control of growth cessation and dormancy in treesA molecular perspective Annals of Botany 120 351ndash360httpsdoiorg101093aobmcx061
McCalmont J P Hastings A Mcnamara N P Richter G MRobson P Donnison I S amp Clifton‐Brown J (2017) Environ-mental costs and benefits of growing Miscanthus for bioenergy inthe UK Global Change Biology Bioenergy 9 489ndash507
McCracken A R amp Dawson W M (1997) Using mixtures of wil-low clones as a means of controlling rust disease Aspects ofApplied Biology 49 97ndash103
McKown A D Klaacutepště J Guy R D Geraldes A Porth I Hanne-mann JhellipDouglas C J (2014) Genome‐wide association implicatesnumerous genes underlying ecological trait variation in natural popula-tions ofPopulus trichocarpaNew Phytologist 203 535ndash553
Merrick P amp Fei S Z (2015) Plant regeneration and genetic transforma-tion in switchgrass ndash A review Journal of Integrative Agriculture 14483ndash493 httpsdoiorg101016S2095-3119(14)60921-7
Moreno‐Mateos M A Fernandez J P Rouet R Lane M AVejnar C E Mis E hellip Giraldez A J (2017) CRISPR‐Cpf1mediates efficient homology‐directed repair and temperature‐con-trolled genome editing Nature Communications 8 2024 httpsdoiorg101038s41467-017-01836-2
Mosseler A (1990) Hybrid performance and species crossabilityrelationships in willows (Salix) Canadian Journal of Botany 682329ndash2338
Muchero W Guo J DiFazio S P Chen J‐G Ranjan P Slavov GT hellip Tuskan G A (2015) High‐resolution genetic mapping ofallelic variants associated with cell wall chemistry in Populus BMCGenomics 16 24 httpsdoiorg101186s12864-015-1215-z
Neale D B amp Kremer A (2011) Forest tree genomics Growingresources and applications Nature Reviews Genetics 12 111ndash122 httpsdoiorg101038nrg2931
Nunn C Hastings A F S J Kalinina O Oumlzguumlven M SchuumlleH Tarakanov I G hellip Clifton‐Brown J C (2017) Environmen-tal influences on the growing season duration and ripening ofdiverse Miscanthus germplasm grown in six countries Frontiersin Plant Science 8 907 httpsdoiorg103389fpls201700907
Ogawa Y Honda M Kondo Y amp Hara‐Nishimura I (2016) Anefficient Agrobacterium‐mediated transformation method forswitchgrass genotypes using Type I callus Plant Biotechnology33 19ndash26
Ogawa Y Shirakawa M Koumoto Y Honda M Asami Y KondoY ampHara‐Nishimura I (2014) A simple and reliable multi‐gene trans-formation method for switchgrass Plant Cell Reports 33 1161ndash1172httpsdoiorg101007s00299-014-1605-8
CLIFTON‐BROWN ET AL | 31
Okada M Lanzatella C Saha M C Bouton J Wu R L amp Tobias CM (2010) Complete switchgrass genetic maps reveal subgenomecollinearity preferential pairing and multilocus interactions Genetics185 745ndash760 httpsdoiorg101534genetics110113910
Palomo‐Riacuteos E Macalpine W Shield I Amey J Karaoğlu C WestJ hellip Jones H D (2015) Efficient method for rapid multiplication ofclean and healthy willow clones via in vitro propagation with broadgenotype applicability Canadian Journal of Forest Research 451662ndash1667 httpsdoiorg101139cjfr-2015-0055
Pilate G Allona I Boerjan W Deacutejardin A Fladung M Gal-lardo F hellip Halpin C (2016) Lessons from 25 years of GM treefield trials in Europe and prospects for the future In C Vettori FGallardo H Haumlggman V Kazana F Migliacci G Pilateamp MFladung (Eds) Biosafety of forest transgenic trees (pp 67ndash100)Dordrecht Netherlands Springer Switzerland AG
Pinosio S Giacomello S Faivre-Rampant P Taylor G Jorge VLe Paslier M C amp Marroni F (2016) Characterization of thepoplar pan-genome by genome-wide identification of structuralvariation Molecular Biology and Evolution 33 2706ndash2719
Porth I Klaacutepště J Skyba O Friedmann M C Hannemann JEhlting J hellip Douglas C J (2013) Network analysis reveals therelationship among wood properties gene expression levels andgenotypes of natural Populus trichocarpa accessions New Phytol-ogist 200 727ndash742
Pugesgaard S Schelde K Larsen S U Laeligrke P E amp JoslashrgensenU (2015) Comparing annual and perennial crops for bioenergyproductionndashinfluence on nitrate leaching and energy balance Glo-bal Change Biology Bioenergy 7 1136ndash1149 httpsdoiorg101111gcbb12215
Quinn L D Allen D J amp Stewart J R (2010) Invasivenesspotential of Miscanthus sinensis Implications for bioenergy pro-duction in the United States Global Change Biology Bioenergy2 310ndash320 httpsdoiorg101111j1757-1707201001062x
Rae A M Robinson K M Street N R amp Taylor G (2004)Morphological and physiological traits influencing biomass pro-ductivity in short‐rotation coppice poplar Canadian Journal ofForest Research‐Revue Canadienne De Recherche Forestiere 341488ndash1498 httpsdoiorg101139x04-033
Rambaud C Arnoult S Bluteau A Mansard M C Blassiau Camp Brancourt‐Hulmel M (2013) Shoot organogenesis in threeMiscanthus species and evaluation for genetic uniformity usingAFLP analysis Plant Cell Tissue and Organ Culture (PCTOC)113 437ndash448 httpsdoiorg101007s11240-012-0284-9
Ramstein G P Evans J Kaeppler S M Mitchell R B Vogel K PBuell C R amp Casler M D (2016) Accuracy of genomic prediction inswitchgrass (Panicum virgatum L) improved by accounting for linkagedisequilibriumG3 (Bethesda Md) 6 1049ndash1062
Rayburn A L Crawford J Rayburn C M amp Juvik J A (2009)Genome size of three Miscanthus species Plant Molecular Biol-ogy Reporter 27 184 httpsdoiorg101007s11105-008-0070-3
Richardson J Isebrands J amp Ball J (2014) Ecology and physiol-ogy of poplars and willows In J Isebrands amp J Richardson(Eds) Poplars and willows Trees for society and the environ-ment (pp 92‐123) Boston MA and Rome CABI and FAO
Robison T L Rousseau R J amp Zhang J (2006) Biomass produc-tivity improvement for eastern cottonwood Biomass amp Bioenergy30 735ndash739 httpsdoiorg101016jbiombioe200601012
Robson P R Farrar K Gay A P Jensen E F Clifton‐Brown JC amp Donnison I S (2013) Variation in canopy duration in the
perennial biofuel crop Miscanthus reveals complex associationswith yield Journal of Experimental Botany 64 2373ndash2383
Robson P Jensen E Hawkins S White S R Kenobi K Clif-ton‐Brown J hellip Farrar K (2013) Accelerating the domestica-tion of a bioenergy crop Identifying and modelling morphologicaltargets for sustainable yield increase in Miscanthus Journal ofExperimental Botany 64 4143ndash4155
Sanderson M A Adler P R Boateng A A Casler M D ampSarath G (2006) Switchgrass as a biofuels feedstock in theUSA Canadian Journal of Plant Science 86 1315ndash1325httpsdoiorg104141P06-136
Scarlat N Dallemand J F Monforti‐Ferrario F amp Nita V(2015) The role of biomass and bioenergy in a future bioecon-omy Policies and facts Environmental Development 15 3ndash34httpsdoiorg101016jenvdev201503006
Serapiglia M J Gouker F E amp Smart L B (2014) Early selec-tion of novel triploid hybrids of shrub willow with improved bio-mass yield relative to diploids BMC Plant Biology 14 74httpsdoiorg1011861471-2229-14-74
Serba D Wu L Daverdin G Bahri B A Wang X Kilian Ahellip Devos K M (2013) Linkage maps of lowland and upland tet-raploid switchgrass ecotypes BioEnergy Research 6 953ndash965httpsdoiorg101007s12155-013-9315-6
Shakoor N Lee S amp Mockler T C (2017) High throughput phe-notyping to accelerate crop breeding and monitoring of diseases inthe field Current Opinion in Plant Biology 38 184ndash192 httpsdoiorg101016jpbi201705006
Shield I Macalpine W Hanley S amp Karp A (2015) Breedingwillow for short rotation coppice energy cropping In Industrialcrops Breeding for BioEnergy and Bioproducts (pp 67ndash80) NewYork NY Springer
Sjodin A Street N R Sandberg G Gustafsson P amp Jansson S (2009)The Populus Genome Integrative Explorer (PopGenIE) A new resourcefor exploring the Populus genomeNewPhytologist 182 1013ndash1025
Slavov G amp Davey C (2017) Integrated genomic prediction inbioenergy crops In Abstracts from the International Conferenceon Developing biomass crops for future climates pp 34 24ndash27September 2017 Oriel College Oxford wwwwatbioeu
Slavov G Davey C Bosch M Robson P Donnison I ampMackay I (2018a) Genomic index selection provides a pragmaticframework for setting and refining multi‐objective breeding targetsin Miscanthus Annals of Botany (in press)
Slavov G T DiFazio S P Martin J Schackwitz W MucheroW Rodgers‐Melnick E hellip Tuskan G A (2012) Genome rese-quencing reveals multiscale geographic structure and extensivelinkage disequilibrium in the forest tree Populus trichocarpa NewPhytologist 196 713ndash725
Slavov G Davey C Robson P Donnison I amp Mackay I(2018b) Domestication of Miscanthus through genomic indexselection In Plant and Animal Genome Conference 13 January2018 San Diego CA Retrieved from httpspagconfexcompagxxvimeetingappcgiPaper30622
Slavov G T Nipper R Robson P Farrar K Allison G GBosch M hellip Jensen E (2013) Genome‐wide association studiesand prediction of 17 traits related to phenology biomass and cellwall composition in the energy grass Miscanthus sinensis NewPhytologist 201 1227ndash1239
Slavov G Robson P Jensen E Hodgson E Farrar K AllisonG hellip Donnison I (2014) Contrasting geographic patterns of
32 | CLIFTON‐BROWN ET AL
genetic variation for molecular markers vs phenotypic traits in theenergy grass Miscanthus sinensis Global Change Biology Bioen-ergy 5 562ndash571
Ślusarkiewicz‐Jarzina A Ponitka A Cerazy‐Waliszewska J Woj-ciechowicz M K Sobańska K Jeżowski S amp Pniewski T(2017) Effective and simple in vitro regeneration system of Mis-canthus sinensis Mtimes giganteus and M sacchariflorus for plant-ing and biotechnology purposes Biomass and Bioenergy 107219ndash226 httpsdoiorg101016jbiombioe201710012
Smart L amp Cameron K (2012) Shrub willow In C Kole S Joshiamp D Shonnard (Eds) Handbook of bioenergy crop plants (pp687ndash708) Boca Raton FL Taylor and Francis Group
Stanton B J (2014) The domestication and conservation of Populusand Salix genetic resources (Chapter 4) In Isebrands J ampRichardson J (Eds) Poplars and willows Trees for society andthe environment (pp 124ndash200) Rome Italy CAB InternationalFood and Agricultural Organization of the United Nations
Stanton B J Neale D B amp Li S (2010) Populus breeding Fromthe classical to the genomic approach In S Jansson R Bha-lerao amp A T Groover (Eds) Genetics and genomics of popu-lus (pp 309ndash348) New York NY Springer
Stewart J R Toma Y Fernandez F G Nishiwaki A Yamada T ampBollero G (2009) The ecology and agronomy ofMiscanthus sinensisa species important to bioenergy crop development in its native range inJapan A reviewGlobal Change Biology Bioenergy 1 126ndash153
Stott K G (1992) Willows in the service of man Proceedings of the RoyalSociety of Edinburgh Section B Biological Sciences 98 169ndash182
Strauss S H Ma C Ault K amp Klocko A L (2016) Lessonsfrom two decades of field trials with genetically modified trees inthe USA Biology and regulatory compliance In C Vettori FGallardo H Haumlggman V Kazana F Migliacci G Pilate amp MFladung (Eds) Biosafety of forest transgenic trees (pp 101ndash124)Dordrecht Netherlands Springer
Suda Y amp Argus G W (1968) Chromosome numbers of some NorthAmerican Salix Brittonia 20 191ndash197 httpsdoiorg1023072805440
Tan B (2018) Genomic selection and genome‐wide association studies todissect quantitative traits in forest trees Umearing Sweden University
Tornqvist C Taylor M Jiang Y Evans J Buell C KaepplerS amp Casler M (2018) Quantitative trait locus mapping for flow-ering time in a lowland x upland switchgrass pseudo‐F2 popula-tion The Plant Genome 11 170093
Toacuteth G Hermann T Da Silva M amp Montanarella L (2016)Heavy metals in agricultural soils of the European Union withimplications for food safety Environment International 88 299ndash309 httpsdoiorg101016jenvint201512017
Tracy W Dawson J Moore V amp Fisch J (2016) Intellectual propertyrights and public plant breeding Recommendations and proceedings ofa conference on best practices for intellectual property protection of pub-lically developed plant germplasm (p 70) University of Wisconsin‐Madison Retrieved from httpshostcalswisceduagronomywp-contentuploadssites1620162005Proceedings-IPR-Finalpdf
Trybush S Jahodovaacute Š Macalpine W amp Karp A (2008) A geneticstudy of a Salix germplasm resource reveals new insights into rela-tionships among subgenera sections and species BioEnergyResearch 1 67ndash79 httpsdoiorg101007s12155-008-9007-9
Tsai C J (2013) Next‐generation sequencing for next‐generationbreeding and more New Phytologist 198 635ndash637 httpsdoiorg101111nph12245
Tuskan G A Difazio S Jansson S Bohlmann J Grigoriev I HellstenU hellip Rokhsar D (2006) The genome of black cottonwood Populustrichocarpa (Torr ampGray) Science 313 1596ndash1604
Tuskan G A amp Walsh M E (2001) Short‐rotation woody cropsystems atmospheric carbon dioxide and carbon management AUS case study The Forestry Chronicle 77 259ndash264 httpsdoiorg105558tfc77259-2
Van Den Broek R Treffers D‐J Meeusen M Van Wijk ANieuwlaar E amp Turkenburg W (2001) Green energy or organicfood A life‐cycle assessment comparing two uses of set‐asideland Journal of Industrial Ecology 5 65ndash87 httpsdoiorg101162108819801760049477
Van Der Schoot J Pospiskova M Vosman B amp Smulders M J M(2000) Development and characterization of microsatellite markers inblack poplar (Populus nigra L) Theoretical and Applied Genetics 101317ndash322 httpsdoiorg101007s001220051485
Van Der Weijde T Dolstra O Visser R G amp Trindade L M(2017a) Stability of cell wall composition and saccharificationefficiency in Miscanthus across diverse environments Frontiers inPlant Science 7 2004
Van der Weijde T Huxley L M Hawkins S Sembiring E HFarrar K Dolstra O hellip Trindade L M (2017) Impact ofdrought stress on growth and quality of miscanthus for biofuelproduction Global Change Biology Bioenergy 9 770ndash782
Van der Weijde T Kamei C L A Severing E I Torres A FGomez L D Dolstra O hellip Trindade L M (2017) Geneticcomplexity of miscanthus cell wall composition and biomass qual-ity for biofuels BMC Genomics 18 406
Van der Weijde T Kiesel A Iqbal Y Muylle H Dolstra O Visser RG FhellipTrindade LM (2017) Evaluation ofMiscanthus sinensis bio-mass quality as feedstock for conversion into different bioenergy prod-uctsGlobal Change Biology Bioenergy 9 176ndash190
Vanholme B Cesarino I Goeminne G Kim H Marroni F VanAcker R hellip Boerjan W (2013) Breeding with rare defectivealleles (BRDA) A natural Populus nigra HCT mutant with modi-fied lignin as a case study New Phytologist 198 765ndash776
Vogel K P Mitchell R B Casler M D amp Sarath G (2014)Registration of Liberty Switchgrass Journal of Plant Registra-tions 8 242ndash247 httpsdoiorg103198jpr2013120076crc
Wang X Yamada T Kong F‐J Abe Y Hoshino Y Sato Hhellip Yamada T (2011) Establishment of an efficient in vitro cul-ture and particle bombardment‐mediated transformation systems inMiscanthus sinensis Anderss a potential bioenergy crop GlobalChange Biology Bioenergy 3 322ndash332 httpsdoiorg101111j1757-1707201101090x
Watrud L S Lee E H Fairbrother A Burdick C Reichman JR Bollman M hellip Van de Water P K (2004) Evidence forlandscape‐level pollen‐mediated gene flow from genetically modi-fied creeping bentgrass with CP4 EPSPS as a marker Proceedingsof the National Academy of Sciences of the United States of Amer-ica 101 14533ndash14538
Weisgerber H (1993) Poplar breeding for the purpose of biomassproduction in short‐rotation periods in Germany ndash Problems andfirst findings Forestry Chronicle 69 727ndash729 httpsdoiorg105558tfc69727-6
Wheeler N C Steiner K C Schlarbaum S E amp Neale D B(2015) The evolution of forest genetics and tree improvementresearch in the United States Journal of Forestry 113 500ndash510httpsdoiorg105849jof14-120
CLIFTON‐BROWN ET AL | 33
Whittaker C Macalpine W J Yates N E amp Shield I (2016)Dry matter losses and methane emissions during wood chip stor-age The impact on full life cycle greenhouse gas savings of shortrotation coppice willow for heat BioEnergy Research 9 820ndash835 httpsdoiorg101007s12155-016-9728-0
Xi Q (2000) Investigation on the distribution and potential of giant grassesin China PhD thesis Goettingen Germany Cuvillier Verlag
Xi Q amp Jezowkski S (2004) Plant resources of Triarrhena andMiscanthus species in China and its meaning for Europe PlantBreeding and Seed Science 49 63ndash77
Xue S Kalinina O amp Lewandowski I (2015) Present and future optionsfor the improvement of Miscanthus propagation techniques Renewableand Sustainable Energy Reviews 49 1233ndash1246
Yin K Q Gao C X amp Qiu J L (2017) Progress and prospectsin plant genome editing Nature Plants 3 httpsdoiorg101038nplants2017107
YookM (2016)Genetic diversity and transcriptome analysis for salt toler-ance inMiscanthus PhD thesis Seoul National University Korea
Zaidi S S E A Mahfouz M M amp Mansoor S (2017) CRISPR‐Cpf1 A new tool for plant genome editing Trends in PlantScience 22 550ndash553 httpsdoiorg101016jtplants201705001
Zalesny R S Stanturf J A Gardiner E S Perdue J H YoungT M Coyle D R hellip Hass A (2016) Ecosystem services ofwoody crop production systems BioEnergy Research 9 465ndash491 httpsdoiorg101007s12155-016-9737-z
Zhang Q X Sun Y Hu H K Chen B Hong C T Guo H Phellip Zheng B S (2012) Micropropagation and plant regenerationfrom embryogenic callus of Miscanthus sinensis Vitro Cellular ampDevelopmental Biology‐Plant 48 50ndash57 httpsdoiorg101007s11627-011-9387-y
Zhang Y Zalapa J E Jakubowski A R Price D L AcharyaA Wei Y hellip Casler M D (2011) Post‐glacial evolution of
Panicum virgatum Centers of diversity and gene pools revealedby SSR markers and cpDNA sequences Genetica 139 933ndash948httpsdoiorg101007s10709-011-9597-6
Zhao H Wang B He J Yang J Pan L Sun D amp Peng J(2013) Genetic diversity and population structure of Miscanthussinensis germplasm in China PloS One 8 e75672
Zhou X H Jacobs T B Xue L J Harding S A amp Tsai C J(2015) Exploiting SNPs for biallelic CRISPR mutations in theoutcrossing woody perennial Populus reveals 4‐coumarate CoAligase specificity and redundancy New Phytologist 208 298ndash301
Zhou R Macaya‐Sanz D Rodgers‐Melnick E Carlson C HGouker F E Evans L M hellip DiFazio S P (2018) Characteri-zation of a large sex determination region in Salix purpurea L(Salicaceae) Molecular Genetics and Genomics 1ndash16 httpsdoiorg101007s00438-018-1473-y
Zsuffa L Mosseler A amp Raj Y (1984) Prospects for interspecifichybridization in willow for biomass production In K L Perttu(Ed) Ecology and management of forest biomass production sys-tems Report 15 (pp 261ndash281) Uppsala Sweden SwedishUniversity of Agricultural Sciences
How to cite this article Clifton‐Brown J HarfoucheA Casler MD et al Breeding progress andpreparedness for mass‐scale deployment of perenniallignocellulosic biomass crops switchgrassmiscanthus willow and poplar GCB Bioenergy201800000ndash000 httpsdoiorg101111gcbb12566
34 | CLIFTON‐BROWN ET AL
Graphical AbstractThe contents of this page will be used as part of the graphical abstract of html only
It will not be published as part of main article
Plant breeding links the research effort with commercial mass upscaling The authorsrsquo assessment of development status ofthe four species is shown (poplar having two one for short rotation coppice (SRC) poplar and one for the more traditionalshort rotation forestry (SRF)) Mass scale deployment needs developments outside the breeding arenas to drive breedingactivities more rapidly and extensively
2011) Genotype‐by‐environment interactions (G times E) arestrong and must be considered in breeding (Casler 2012Casler Mitchell amp Vogel 2012) Adaptation to environ-ment is regulated principally by responses to day‐lengthand temperature There are also strong genotype times environ-ment interactions between the drier western regions and thewetter eastern regions (Casler et al 2017)
The growing regions of North America are divided intofour adaptation zones for switchgrass each roughly corre-sponding to two official hardiness zones The lowland eco-types are generally late flowering high yielding andadapted to warmer climates but have lower drought andcold resistance than upland ecotypes (Casler 2012 Casleret al 2012)
In 2015 the US Department of Agriculture (USDA)National Plant Germplasm System GRIN (httpswwwars-gringovnpgs) had 181 switchgrass accessions of whichonly 96 were available for distribution due to limitationsassociated with seed multiplication (Casler Vogel amp Har-rison 2015) There are well over 2000 additional uncata-logued accessions (Table 1) held by various universities butthe USDA access to these is also constrained by the effortneeded in seed multiplication Switchgrass is a model herba-ceous species for conducting scientific research on biomass(Sanderson Adler Boateng Casler amp Sarath 2006) but lit-tle funding is available for the critical prebreeding work thatis necessary to link this biological research to commercialbreeding More than a million genotypes from ~2000 acces-sions (seed accessions contain many genotypes) have beenphenotypically screened in spaced plant nurseries and tenthousand of the most useful have been genotyped with differ-ent technologies depending on the technology available at
the time when these were performed From these character-ized genotypes parents are selected for exploratory pairwisecrosses to produce synthetic populations within ecotypesSwitchgrass like many grasses is outcrossing due to astrong genetically controlled self‐incompatibly (akin to theS‐Z‐locus system of other grasses (Martinez‐Reyna ampVogel 2002)) Thus the normal breeding approaches usedare F1 wide crosses and recurrent selection cycles within syn-thetic populations
The scale of these programmes varies from small‐scaleconventional breeding based solely on phenotypic selec-tion (eg REAP Canada Montreal Quebec) to large pro-grammes incorporating modern molecular breedingmethods (eg USDA‐ARS Madison Wisconsin) Earlyagronomic research and biomass production efforts werefocused on the seed‐based multiplication of promising wildaccessions from natural prairies Cultivars Alamo Kanlowand Cave‐in‐Rock were popular due to high yield andmoderate‐to‐wide adaptation Conventional breedingapproaches focussed on biomass production traits and haveled to the development of five cultivars particularly suitedto biomass production Cimarron EG1101 EG1102EG2101 and Liberty The first four of these represent thelowland ecotype and were developed either in Oklahomaor Georgia Liberty is a derivative of lowland times uplandhybrids developed in Nebraska following selection for lateflowering the high yield of the lowland ecotype and coldtolerance of the upland ecotype (Vogel et al 2014) Thesefive cultivars were all approximately 25ndash30 years in themaking counting from the initiation of these breeding pro-grammes Many more biomass‐type cultivars are expectedwithin the next few years as these and other breeding
TABLE 2 Generalized improvement targets for perennial biomass crops (PBCs)
Net energy yield per hectare
Increased yield
Reduced moisture content at harvest
Physical and chemical composition for different end‐use applications
Increased lignin content and decreased corrosive elements for thermal conversion
Reduced recalcitrance through decreased lignin content andor modified lignin monomer composition to reduce pretreatment requirementsfor next‐generation biofuels by saccharification and fermentation
Plant morphological differences which influence biomass harvest transport and storage (eg stem thickness)
Propagation costs
Improved cloning systems (trees and grasses)
Seed systems (grasses)
Optimizing agronomy for each new cultivar
Resilience through enhanced
Abiotic stress toleranceresistance (eg drought salinity and high and low temperature )
Biotic stress resistance (eg insects fungal bacterial and viral diseases)
Site adaptability especially to those of marginalcontaminated agricultural land
6 | CLIFTON‐BROWN ET AL
TABLE3
Preparedness
formassupscaling
currentmarketvalueandresearch
investmentin
four
perennialbiom
asscrops(PBCs)
Species
Switc
hgrass
Miscanthu
sW
illow
Pop
lar
Current
commercial
plantin
gcostsperha
USa20
0USD
bin
the
establishm
entyear
DE3
375
Eurob
(XueK
alininaamp
Lew
ando
wski20
15)
UK2
153
GBPb
(Evans201
6)redu
cedto
116
9GBPwith
Mxg
rhizom
ewhere
thefarm
erdo
estheland
preparation(w
wwterravestacom
)US
173
0ndash222
5USD
UK150
0ndash173
9GBPplus
land
preparation(Evans20
16)
US
197
6USD
(=80
0USD
acre)
IT110
0Euro
FRSE100
0ndash200
0Euro
US
863USD
ha(LazarusHeadleeamp
Zalesny
20
15)
Current
marketvalue
ofthebiom
asspert
DM
US
80ndash1
00USD
UK~8
0GBP(bales)(Terravesta
person
alcommun
ication)
US
94USD
(chipp
ed)
UK49
41GBP(chipp
ed)(Evans20
16)
US
55ndash7
0USD
IT10
0Euro
US
80ndash90USD
deliv
ered
basedon
40milesof
haulage
Scienceforg
enetic
improv
ementprojects
inthelast10
years
USgt50
projectsfund
edby
USDOEand
USD
ANIFA
CNgt
30projectsfund
edby
CN‐N
SFCandMBI
EUc 6projects(O
PTIM
ISCO
PTIM
A
WATBIO
SUNLIBBF
IBRAandGRACE)
FRB
FFSK
2projectsfund
edby
IPETandPM
BC
US
EBICABBImultip
leDOEFeedstock
Genom
icsandUSD
AAFR
Iprojects
UKB
SBECB
EGIN
(200
3ndash20
10)
US
USDOEJG
IGenom
eSequ
encing
(200
9ndash20
122
015ndash
2018
)USD
ANortheastSu
nGrant
(200
9ndash20
12)
USD
ANIFANEWBio
Con
sortium
(201
2ndash20
17)USD
ANIFAWillow
SkyC
AP(201
8ndash20
21)
EUF
P7(EnergyPo
plarN
ovelTreeWATBIO
Tree4Fu
ture)
SEC
limate‐adaptedpo
plarsandNanow
ood
SLU
andST
TUS
Bioenergy
Science(O
RNL)USD
Afeedstocks
geno
micsprog
rammeBRDIa
ndBRC‐CBI
Major
projects
supp
ortin
gcrossing
andselectioncycles
inthelast10
years
USandCA12
projects
NLRUEmiscanthu
sPP
P20
15ndash2
019
50ndash
100K
Euroyear
UKGIA
NT‐LIN
KPP
I20
11ndash2
016
13M
GBPyear
US
EBICABBI~0
5M
USD
year
UKBEGIN
2000
ndash201
0US
USD
ANortheastSu
nGrant
(200
9ndash20
12)USD
ANIFA
NEWBio
Con
sortium
(201
2ndash20
17)USD
ANIFA
Willow
SkyC
AP(201
8ndash20
21)
FR1
4region
alandnatio
nalp
rojects10
0ndash15
0k
Euroyear
SES
TTo
nebreeding
effortas
aninternalproject
in20
10with
aresulting
prog
enytrial
US
DOESu
nGrant
feedstockdevelopm
ent
partnershipprog
ramme(including
Willow
)
Current
annu
alinvestmentin
projects
fortranslationinto
commercial
hybrids
USgt20
MUSD
year
UKMUST
201
6ndash20
1905M
GBPyear
US
CABBIUSD
AAFR
I20
17ndash2
02205M
USD
year
US
USD
ANIFA
NEWBio
~04
MUSD
year
FRScienceforim
prov
ement~2
0kEuroyear
US
USD
Aun
derAFR
ICo‐ordinatedAg
Prod
ucer
projectsAnd
severaltranslational
geno
micsprojectswith
outpu
blic
fund
ing
Upscalin
gtim
e(years)
toproducesufficient
propagules
toplantgt10
0ha
US
Using
seed
2ndash3years
UKRhizomefor1ndash200
0hectares
canbe
ready
in6mon
ths
UKNL2
0ha
ofseed
hybridsplantedin
2018
In
2019
sufficientseedfor5
0ndash10
0ha
isexpected
UK3yearsusingconv
entio
nalcutting
sfaster
usingmicroprop
agation
FRIT3yearsby
vegetativ
eprop
agation
SEgt3yearsby
vegetativ
eprop
agation(cuttin
gs)
and3yearsby
microprop
agation
Note
AFR
IAgriculture
andFo
odResearchInitiativein
theUnitedStatesBEGIN
BiomassforEnergyGenetic
Improvem
entNetwork
BFF
BiomassFo
rtheFu
tureBRC‐CBI
Bioenergy
ResearchCentre‐Centrefor
Bioenergy
Innovatio
nBRDIBiomassresearch
developm
entinitiative
BSB
ECBBSR
CSu
stainableBioenergy
Centre
CABBICenterforadvanced
bioenergyandbioproductsinnovatio
nin
theUnitedStatesCN‐N
SFC
Natural
ScienceFo
undatio
nof
ChinaDOEDepartm
entof
Energyin
theUnitedStatesEBIEnergyBiosciences
Institu
teFIBRAFibrecropsas
sustainablesource
ofbiobased
materialforindustrial
productsin
Europeand
ChinaFP
7SeventhFram
eworkProgrammein
theEUGIA
NT‐LIN
KGenetic
improvem
entof
miscanthusas
asustainablefeedstockforbioenergyin
theUnitedKingdom
GRACEGRow
ingAdvancedindustrial
Crops
onmarginallandsforbiorEfineriesIPETKorea
Institu
teof
Planning
andEvaluationforTechnologyin
FoodA
gricultureFo
restry
andFisheriesJG
IJointGenom
eInstitu
teMBIMendelBiotechnology
IncMUST
Miscant-
husUpS
calin
gTechnology
NEWBioNortheast
WoodyW
arm‐seasonBiomassConsortiumNIFANationalInstitu
teof
Food
andAgriculture
intheUnitedStatesNovelTree
Novel
tree
breeding
strategiesOPT
IMAOpti-
mizationof
perennialgrassesforbiom
assproductio
nin
theMediterraneanareaOPT
IMISCOptim
izingbioenergyproductio
nfrom
Miscanthus
ORNLOak
Ridge
NationalLabPM
BCPlantMolecular
BreedingCenterof
theNextGenerationBiogreenResearchCenters
oftheRepublic
ofKoreaPP
Ipublicndashp
rivate
investmentPP
Ppublicndashp
rivate
partnership
RUEradiationuseefficiencySk
yCAP
Coordinated
AgriculturalProjectSL
U
SwedishUniversity
ofAgriculturalSciencesST
TSw
eTreeTech
SUNLIBBSu
stainableLiquidBiofuelsfrom
BiomassBiorefiningTree4Fu
tureDesigning
Trees
forthefutureUSD
AUSDepartm
entof
Agriculture
WATBIO
Developmentof
improved
perennialnonfoodbiom
assandbioproduct
cropsforwater
stressed
environm
ents
a ISO
Alpha‐2
lettercountrycodes
b Local
currencies
areused
asat
2018Euro
GBP(G
reat
Britain
Pound)USD
(USDollar)c EU
isused
forEurope
CLIFTON‐BROWN ET AL | 7
TABLE4
Prebreedingresearch
andthestatus
ofconventio
nalbreeding
infour
leadingperennialbiom
asscrops(PBCs)
Breeding
techno
logy
step
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sWillow
Poplar
1Collected
wild
accessions
or
secondarysources
availablefor
breeding
Provideabroadbase
of
useful
traits
Respect
forCBD
and
Nagoyaprotocol
on
collections
after2014
Not
allindigenous
genetic
resourcesareaccessible
under
CBD
forpolitical
reasons
USa~2
000
(181
inofficial
GRIN
gene
bank
ofwhich
96
wereavailable
others
in
variousunofficial
collections)
CNChangsha
~1000
andNanjin
g
2000from
China
DK~1
20from
JP
FRworking
collectionof
~100
(mainlyM
sin)
JP~1
500
from
JPS
KandCN
NLworking
collectionof
~300
Msin
SK~7
00from
SKC
NJP
andRU
UK~1
500
accessions
(from
~500
sites)
ofwhich
1000arein
field
nurseriesin
2018
US
14in
GRIN
global
US
Illin
ois~1
500
collections
made
inAsiawith
25
currently
available
inUnitedStates
forbreeding
UK1500with
about20
incommon
with
the
UnitedStates
US
350largelyunique
FR3370Populus
delto
ides
(650)Pnigra(2000)
Ptrichocarpa(600)and
Pmaximow
iczii(120)
IT30000P
delto
idsPnigra
Pmaximow
icziiand
Ptrichocarpa
SE13000
accessionsmainly
Ptrichocarpa
byST
TandSL
U
US
GWR150Ptrichocarpaand
300Pmaximow
icziiaccessions
forPacificNorthwestprogram
535Pdelto
ides
and~2
00
Pnigraaccessions
for
Southeastprogrammeand77
Psimoniifrom
Mongolia
UMD
NRRI550+
clonal
accessions
(primarily
Ptimescanadnesisa
long
with
Pdelto
ides
andPnigra
parents)
forUpper
Midwest
program
2Wild
accessions
which
have
undergone
phenotypic
screeningin
field
trials
Selectionof
parental
lines
with
useful
traits
Strong
partnerships
torun
multilocationfieldtrials
tophenotypeconsistently
theaccessionsgenotypes
indifferentenvironm
ents
anddatabases
Costof
runningmultilocation
US
gt500000genotypes
CN~1
250
inChangsha
~1700
in
HunanJiangsuS
handong
Hainan
NL~250genotypes
JP~1
200
SK~4
00
UK‐le
d~1
000
genotypesin
a
replicated
trial
US‐led
~1200
genotypesin
multi‐
locatio
nreplicated
trialsin
Asiaand
North
America
MsinSK
CNJP
CA
(Ontario)US(Illinoisand
Colorado)MsacSK
(Kangw
on)
CN
(Zhejiang)JP
(Sapporo)CA
(Onatario)U
S(Illinois)DK
(Aarhus)
UKgt400
US
180
FR2720
IT10000
SE~1
50
US
GWR~1
500
wild
Ptrichocarpaphenotypes
and
OPseedlots(total
of70)from
thePmaximow
icziispecies
complex
(suaveolens
cathayana
koreana
ussuriensismaximow
iczii)and
2000ndash3000Pdelto
ides
collections
(based
on104s‐
generatio
nfamilies)
UMD
NRRIscreened
seedlin
gsfrom
aPdelto
ides
(OPor
CP)
breeding
programme(1996ndash2016)
gt7200OPseedlin
gsof
PnigraunderMinnesota
clim
atic
conditions(improve
parental
populatio
n)
3Wild
germ
plasm
genotyping
Amanaged
living
collectionof
clonal
types
US
~20000genotypes
CNChangsha
~1000Nanjin
g37
FR44
bycpDNA
(Fenget
al
UK~4
00
US
225
(Con
tinues)
8 | CLIFTON‐BROWN ET AL
TABLE
4(Contin
ued)
Breeding
techno
logy
step
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sWillow
Poplar
Construct
phylogenetic
treesanddissim
ilarity
indices
The
type
ofmolecular
analysis
mdashAFL
PcpDNAR
ADseq
whole
genomesequencing
2014)
JP~1
200
NL250genotypesby
RADseq
UK~1
000
byRADseq
SK~3
00
US
Illin
oisRADseqon
allwild
accessions
FR2310
SE150
US
1500
4Exploratory
crossing
and
progenytests
Discovergood
parental
combinatio
nsmdashgeneral
combining
ability
Geographicseparatio
n
speciesor
phylogenetic
trees
Costly
long‐te
rmmultilocation
trialsforprogeny
US
gt10000
CN~1
0
FR~2
00
JP~3
0
NL3600
SK~2
0
UK~4
000
(~1500Msin)
US
500ndash1000
UK gt700since2003
gt800EWBP(1996ndash
2002)
US
800
FRCloned13
times13
factorial
matingdesign
with
3species[10
Pdelto
ides8Ptrichocarpa
8
Pnigra)
US
GWR1000exploratory
crossingsof
Pfrem
ontii
Psimoniiforim
proved
droughttolerance
UMD
NRRIcrossednativ
e
Pdelto
ides
with
non‐nativ
e
Ptrichocarpaand
Pmaximow
icziiMajority
of
progenypopulatio
nssuffered
from
clim
atic
andor
pathogenic
susceptib
ility
issues
5Wideintraspecies
hybridization
Discovergood
parental
combinatio
nsmdashgeneral
combining
ability
1to
4above
inform
atics
Flow
eringsynchronizationand
low
seed
set
US
~200
CN~1
0
NL1800
SK~1
0
UK~5
00
UK276
US
400
IT500
SE120
US
50ndash100
6With
inspecies
recurrentselection
Concentratio
nof
positiv
e
traits
Identificationof
theright
heterotic
groups
insteps
1ndash5
Difficultto
introducenew
germ
plasm
with
outdilutio
n
ofbesttraits
US
gt40
populatio
ns
undergoing
recurrentselection
NL2populatio
ns
UK6populatio
ns
US
~12populatio
nsto
beestablished
by2019
UK4
US
150
FRPdelto
ides
(gt30
FSfamilies
ndash900clones)Pnigra(gt40
FS
families
ndash1600clones)and
Ptrichocarpa(15FS
Families
minus350clones)
IT80
SEca50
parent
breeding
populatio
nswith
in
Ptrichocarpa
US
Goalsis~1
00parent
breeding
populatio
nswith
inPtrichocarpa
Pdelto
ides
PnigraandPmaximow
iczii
7Interspecies
hybrid
breeding
Com
bining
complem
entary
traitsto
producegood
morphotypes
and
heterosiseffects
Allthesteps1ndash6
Inearlystage
improvem
ents
areunpredictable
Awide
base
iscostly
tomanage
None
CNChangsha
3Nanjin
g
120MsintimesMflo
r
JP~5
with
Saccharum
spontaneum
SK5
UK420
US
250
FRPcanadensis(1800)
Pdelto
ides
timesPtrichocarpa
(800)and
PtrichocarpatimesPmaximow
iczii
(1200)
(Con
tinues)
CLIFTON‐BROWN ET AL | 9
TABLE
4(Contin
ued)
Breeding
techno
logy
step
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sWillow
Poplar
IT~1
20
US
525genotypesin
eastside
hybrid
programme
205westside
hybrid
programmeand~1
200
in
SoutheastPdelto
ides
program
8Chrom
osom
e
doublin
g
Arouteto
triploid
seeded
typesdoublin
gadiploid
parent
ofknow
n
breeding
valuefrom
recurrentselection
Needs
1ndash7to
help
identify
therightparental
lines
Doubled
plantsarenotorious
forrevertingto
diploid
US
~50plantstakenfrom
tetraploid
tooctaploid
which
arenow
infieldtrials
CN~1
0Msactriploids
DKSeveralattemptsin
mid
90s
FR5genotypesof
Msin
UK2Msin1Mflora
nd1Msac
US
~30
UKAttempted
butnot
routinelyused
US
Not
partof
theregular
breeding
becausethere
existnaturalpolyploids
NA
9DoubleHaploids
Arouteto
producing
homogeneous
progeny
andalso
forthe
introductio
nof
transgenes
orforgenome
editing
Needs
1ndash7to
help
identify
therightparental
lines
Fully
homozygousplantsare
weakandeasily
die
andmay
notflow
ersynchronously
US
Noneyetattempteddueto
poor
vigour
andviability
of
haploids
CN2genotypesof
Msin
andMflor
UKDihaploidsof
MsinMflor
and
Msacand2transgenic
Msinvia
anther
cultu
re
US
6createdandused
toidentify
miss‐identifying
paralogueloci
(causedby
recent
genome
duplicationin
miscanthus)Usedin
creatin
gthereferencegenomefor
MsinVeryweakplantsand
difficultto
retain
thelin
es
NA
NA
10Embryo
rescue
Anattractiv
etechniquefor
recovering
plantsfrom
sexual
crosses
The
majority
ofem
bryos
cannot
survivein
vivo
or
becomelong‐timedorm
ant
None
UKone3times
hybrid
UKEmbryosrescue
protocol
proved
robustat
8days
post‐pollin
ation
FRAnem
bryo
rescue
protocol
proved
robustandim
proved
hybridizationsuccessfor
Pdelto
ides
timesPtrichocarpa
crosses
11Po
llenstorage
Flexibility
tocross
interestingparents
with
outneed
of
flow
ering
synchronization
None
Nosuccessto
daten
otoriously
difficultwith
grassesincluding
sugarcane
Freshpollencommonly
used
incrossing
Pollenstoragein
some
interspecificcrosses
FRBothstored
(cryobank)
and
freshindividual
pollen
Note
AFL
Pam
plifiedfragmentlength
polymorphismCBDconventio
non
biological
diversity
CP
controlledpollinatio
nCpD
NAchloroplastDNAEWBP
Europeanwillow
breeding
programme
FSfull‐sib
GtimesE
genotype‐by‐environm
entGRIN
germ
plasm
resourcesinform
ationnetwork
GWRGreenWoodResourcesMflorMiscanthusflo
ridulusMsacMsacchariflo
rus
MsinMsinensisNAnotapplicableNRRINatural
Resources
ResearchInstitu
teOP
open
pollinatio
nRADseq
restrictionsite‐associatedDNA
sequencingSL
USw
edishUniversity
ofAgriculturalSciencesST
TSw
eTreeTech
UMDUniversity
ofMinnesota
Duluth
a ISO
Alpha‐2
lettercountrycodes
10 | CLIFTON‐BROWN ET AL
programmes mature The average rate of gain for biomassyield in long‐term switchgrass breeding programmes hasbeen 1ndash4 per year depending on ecotype populationand location of the breeding programme (Casler amp Vogel2014 Casler et al 2018) The hybrid derivative Libertyhas a biomass yield 43 higher than the better of its twoparents (Casler amp Vogel 2014 Vogel et al 2014) Thedevelopment of cold‐tolerant and late‐flowering lowland‐ecotype populations for the northern United States hasincreased biomass yields by 27 (Casler et al 2018)
Currently more than 20 recurrent selection populationsare being managed in the United States to select parentsfor improved yield yield resilience and compositional qual-ity of the biomass For the agronomic development andupscaling high seed multiplication rates need to be com-bined with lower seed dormancy to reduce both crop estab-lishment costs and risks Expresso is the first cultivar withsignificantly reduced seed dormancy which is the first steptowards development of domesticated populations (Casleret al 2015) Most phenotypic traits of interest to breeders
require a minimum of 2 years to be fully expressed whichresults in a breeding cycle that is at least two years Morecomplicated breeding programmes or traits that requiremore time to evaluate can extend the breeding cycle to 4ndash8 years per generation for example progeny testing forbiomass yield or field‐based selection for cold toleranceBreeding for a range of traits with such long cycles callsfor the development of molecular methods to reduce time-scales and improve breeding efficiency
Two association panels of switchgrass have been pheno-typically and genotypically characterized to identify quanti-tative trait loci (QTLs) that control important biomasstraits The northern panel consists of 60 populationsapproximately 65 from the upland ecotype The southernpanel consists of 48 populations approximately 65 fromthe lowland ecotype Numerous QTLs have been identifiedwithin the northern panel to date (Grabowski et al 2017)Both panels are the subject of additional studies focused onbiomass quality flowering and phenology and cold toler-ance Additionally numerous linkage maps have been
FIGURE 1 Cumulative minimum years needed for the conventional breeding cycle through the steps from wild germplasm to thecommercial hybrids in switchgrass miscanthus willow and poplar Information links between the steps are indicated by dotted arrows andhighlight the importance of long‐term informatics to maximize breeding gain
CLIFTON‐BROWN ET AL | 11
created by the pairwise crossing of individuals with diver-gent characteristics often to generate four‐way crosses thatare analysed as pseudo‐F2 crosses (Liu Wu Wang ampSamuels 2012 Okada et al 2010 Serba et al 2013Tornqvist et al 2018) Individual markers and QTLs iden-tified can be used to design marker‐assisted selection(MAS) programmes to accelerate breeding and increase itsefficiency Genomic prediction and selection (GS) holdseven more promise with the potential to double or triplethe rate of gain for biomass yield and other highly complexquantitative traits of switchgrass (Casler amp Ramstein 2018Ramstein et al 2016) The genome of switchgrass hasrecently been made public through the Joint Genome Insti-tute (httpsphytozomejgidoegov)
Transgenic approaches have been heavily relied upon togenerate unique genetic variants principally for traitsrelated to biomass quality (Merrick amp Fei 2015) Switch-grass is highly transformable using either Agrobacterium‐mediated transformation or biolistics bombardment butregeneration of plants is the bottleneck to these systemsTraditionally plants from the cultivar Alamo were the onlyregenerable genotypes but recent efforts have begun toidentify more genotypes from different populations that arecapable of both transformation and subsequent regeneration(King Bray Lafayette amp Parrott 2014 Li amp Qu 2011Ogawa et al 2014 Ogawa Honda Kondo amp Hara‐Nishi-mura 2016) Cell wall recalcitrance and improved sugarrelease are the most common targets for modification (Bis-wal et al 2018 Fu et al 2011) Transgenic approacheshave the potential to provide traits that cannot be bredusing natural genetic variability However they will stillrequire about 10ndash15 years and will cost $70ndash100 millionfor cultivar development and deployment (Harfouche Mei-lan amp Altman 2011) In addition there is commercialuncertainty due to the significant costs and unpredictabletimescales and outcomes of the regulatory approval processin the countries targeted for seed sales As seen in maizeone advantage of transgenic approaches is that they caneasily be incorporated into F1 hybrid cultivars (Casler2012 Casler et al 2012) but this does not decrease thetime required for cultivar development due to field evalua-tion and seed multiplication requirements
The potential impacts of unintentional gene flow andestablishment of non‐native transgene sequences in nativeprairie species via cross‐pollination are also major issues forthe environmental risk assessment These limit further thecommercialization of varieties made using these technolo-gies Although there is active research into switchgrass steril-ity mechanisms to curb unintended pollen‐mediated genetransfer it is likely that the first transgenic cultivars proposedfor release in the United States will be met with considerableopposition due to the potential for pollen flow to remainingwild prairie sites which account for lt1 of the original
prairie land area and are highly protected by various govern-mental and nongovernmental organizations (Casler et al2015) Evidence for landscape‐level pollen‐mediated geneflow from genetically modified Agrostis seed multiplicationfields (over a mountain range) to pollinate wild relatives(Watrud et al 2004) confirms the challenge of using trans-genic approaches Looking ahead genome editing technolo-gies hold considerable promise for creating targeted changesin phenotype (Burris Dlugosz Collins Stewart amp Lena-ghan 2016 Liu et al 2018) and at least in some jurisdic-tions it is likely that cultivars resulting from gene editingwill not need the same regulatory approval as GMOs (Jones2015a) However in July 2018 the European Court of Justice(ECJ) ruled that cultivars carrying mutations resulting fromgene editing should be regulated in the same way as GMOsThe ECJ ruled that such cultivars be distinguished from thosearising from untargeted mutation breeding which isexempted from regulation under Directive 200118EC
3 | MISCANTHUS
Miscanthus is indigenous to eastern Asia and Oceania whereit is traditionally used for forage thatching and papermaking(Xi 2000 Xi amp Jezowkski 2004) In the 1960s the highbiomass potential of a Japanese genotype introduced to Eur-ope by Danish nurseryman Aksel Olsen in 1935 was firstrecognized in Denmark (Linde‐Laursen 1993) Later thisaccession was characterized described and named asldquoM times giganteusrdquo (Greef amp Deuter 1993 Hodkinson ampRenvoize 2001) commonly abbreviated as Mxg It is a natu-rally occurring interspecies triploid hybrid between tetraploidM sacchariflorus (2n = 4x) and diploid M sinensis(2n = 2x) Despite its favourable agronomic characteristicsand ability to produce high yields in a wide range of environ-ments in Europe (Kalinina et al 2017) the risks of relianceon it as a single clone have been recognized Miscanthuslike switchgrass is outcrossing due to self‐incompatibility(Jiang et al 2017) Thus seeded hybrids are an option forcommercial breeding Miscanthus can also be vegetativelypropagated by rhizome or in vitro culture which allows thedevelopment of clones The breeding approaches are usuallybased on F1 crosses and recurrent selection cycles within thesynthetic populations There are several breeding pro-grammes that target improvement of miscanthus traitsincluding stress resilience targeted regional adaptation agro-nomic ldquoscalabilityrdquo through cheaper propagation fasterestablishment lower moisture and ash contents and greaterusable yield (Clifton‐Brown et al 2017)
Germplasm collections specifically to support breedingfor biomass started in the late 1980s and early 1990s inDenmark Germany and the United Kingdom (Clifton‐Brown Schwarz amp Hastings 2015) These collectionshave continued with successive expeditions from European
12 | CLIFTON‐BROWN ET AL
and US teams assembling diverse collections from a widegeographic range in eastern Asia including from ChinaJapan South Korea Russia and Taiwan (Hodkinson KlaasJones Prickett amp Barth 2015 Stewart et al 2009) Threekey miscanthus species for biomass production are M si-nensis M floridulus and M sacchariflorus M sinensis iswidely distributed throughout eastern Asia with an adap-tive range from the subtropics to southern Russia (Zhaoet al 2013) This species has small rhizomes and producesmany tightly packed shoots forming a ldquotuftrdquo M floridulushas a more southerly adaptive range with a rather similarmorphology to M sinensis but grows taller with thickerstems and is evergreen and less cold‐tolerant than the othermiscanthus species M sacchariflorus is the most northern‐adapted species ranging to 50 degN in eastern Russia (Clarket al 2016) Populations of diploid and tetraploid M sac-chariflorus are found in China (Xi 2000) and South Korea(Yook 2016) and eastern Russia but only tetraploids havebeen found in Japan (Clark Jin amp Petersen 2018)
Germplasm has been assembled from multiple collec-tions over the last century though some early collectionsare poorly documented This historical germplasm has beenused to initiate breeding programmes largely based on phe-notypic and genotypic characterization As many of theaccessions from these collections are ldquoorigin unknownrdquocrucial environmental envelope data are not available UK‐led expeditions started in 2006 and continued until 2011with European and Asian partners and have built up a com-prehensive collection of 1500 accessions from 500 sitesacross Eastern Asia including China Japan South Koreaand Taiwan These collections were guided using spatialclimatic data to identify variation in abiotic stress toleranceAccessions from these recent collections were planted fol-lowing quarantine in multilocation nursery trials at severallocations in Europe to examine trait expression in differentenvironments Based on the resulting phenotypic andmolecular marker data several studies (a) characterized pat-terns of population genetic structure (Slavov et al 20132014) (b) evaluated the statistical power of genomewideassociation studies (GWASs) and identified preliminarymarkerndashtrait associations (Slavov et al 2013 2014) and(c) assessed the potential of genomic prediction (Daveyet al 2017 Slavov et al 2014 2018b) Genomic indexselection in particular offers the possibility of exploringscenarios for different locations or industrial markets (Sla-vov et al 2018a 2018b)
Separately US‐led expeditions also collected about1500 accessions between 2010 and 2014 (Clark et al2014 2016 2018 2015) A comprehensive genetic analy-sis of the population structure has been produced by RAD-seq for M sinensis (Clark et al 2015 Van der WeijdeKamei et al 2017) and M sacchariflorus (Clark et al2018) Multilocation replicated field trials have also been
conducted on these materials in North America and inAsia GWAS has been conducted for both M sinensis anda subset of M sacchariflorus accessions (Clark et al2016) To date about 75 of these recent US‐led collec-tions are in nursery trials outside the United States Due tolengthy US quarantine procedures these are not yet avail-able for breeding in the United States However molecularanalyses have allowed us to identify and prioritize sets ofgenotypes that best encompass the genetic variation in eachspecies
While mostM sinensis accessions flower in northern Eur-ope very few M sacchariflorus accessions flower even inheated glasshouses For this reason the European pro-grammes in the United Kingdom the Netherlands and Francehave performed mainly M sinensis (intraspecies) hybridiza-tions (Table 4) Selected progeny become the parents of latergenerations (recurrent selection as in switchgrass) Seed setsof up to 400 seed per panicle occur inM sinensis In Aberyst-wyth and Illinois significant efforts to induce synchronousflowering in M sacchariflorus and M sinensis have beenmade because interspecies hybrids have proven higher yieldperformance and wide adaptability (Kalinina et al 2017) Ininterspecies pairwise crosses in glasshouses breathable bagsandor large crossing tubes or chambers in which two or morewhole plants fit are used for pollination control Seed sets arelower in bags than in the open air because bags restrict pollenmovement while increasing temperatures and reducinghumidity (Clifton‐Brown Senior amp Purdy 2018) About30 of attempted crosses produced 10 to 60 seeds per baggedpanicle The seed (thousand seed mass ranges from 05 to09 g) is threshed from the inflorescences and sown into mod-ular trays to produce plug plants which are then planted infield nurseries to identify key parental combinations
A breeding programme of this scale must serve theneeds of different environments accepting the commonpurpose is to optimize the interception of solar radiationAn ideal hybrid for a given environment combines adapta-tion to date of emergence with optimization of traits suchas height number of stems per plant flowering and senes-cence time to optimize solar interception to produce a highbiomass yield with low moisture content at harvest (Rob-son Farrar et al 2013 Robson Jensen et al 2013) By20132014 conventional breeding in Europe had producedintra‐ and interspecific fertile seeded hybrids When acohort (typically about 5) of outstanding crosses have beenidentified it is important to work on related upscaling mat-ters in parallel These are as follows
Assessment of the yield and critical traits in selectedhybrids using a network of field trials
Efficient cloning of the seed parents While in vitro andmacro‐cloning techniques are used some genotypes areamenable to neither technique
CLIFTON‐BROWN ET AL | 13
TABLE5
Status
ofmod
ernplantbreeding
techniqu
esin
four
leadingperennialbiom
asscrop
s(PBCs)
Breeding
techno
logy
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sW
illow
Pop
lar
MAS
Use
ofmarker
sequ
encesthat
correlatewith
atrait
allowingearly
prog
enyselectionor
rejectionlocatin
gkn
owngenesfor
useful
traits(suchas
height)from
other
speciesin
your
crop
Breeding
prog
ramme
relevant
biparental
crosses(Table
4)to
create
ldquomapping
popu
latio
nsrdquoand
QTLidentification
andestim
ationto
identifyrobu
stmarkers
alternatively
acomprehensive
panelof
unrelated
geno
typesfor
GWAS
Ineffectivewhere
traits
areaffected
bymany
geneswith
small
effects
US
Literally
dozens
ofstud
iesto
identify
SNPs
andmarkers
ofinterestNoattempts
atMASas
yetlargely
dueto
popu
latio
nspecificity
CNS
SRISS
Rmarkers
forM
sinMsacand
Mflor
FR2
Msin
mapping
popu
latio
ns(BFF
project)
NL2
Msin
mapping
popu
latio
ns(A
tienza
Ram
irezetal
2003
AtienzaSatovic
PetersenD
olstraamp
Martin
200
3Van
Der
Weijdeetal20
17a
Van
derW
eijde
Hux
ley
etal20
17
Van
derW
eijdeKam
ei
etal20
17V
ander
WeijdeKieseletal
2017
)SK
2GWASpanels(1
collectionof
Msin1
diploidbiparentalF 1
popu
latio
n)in
the
pipelin
eUK1
2families
tofind
QTLin
common
indifferentfam
ilies
ofthe
sameanddifferent
speciesandhy
brids
US
2largeGWAS
panels4
diploidbi‐
parentalF 1
popu
latio
ns
1F 2
popu
latio
n(add
ition
alin
pipelin
e)
UK16
families
for
QTLdiscov
ery
2crossesMASscreened
1po
tentialvariety
selected
US
9families
geno
typedby
GBS(8
areF 1o
neisF 2)a
numberof
prom
ising
QTLdevelopm
entof
markers
inprog
ress
FR(INRA)tested
inlargeFS
families
and
infactorialmating
design
low
efficiency
dueto
family
specificity
US
49families
GS
Metho
dto
accelerate
breeding
throug
hredu
cing
the
resourcesforcross
attemptsby
Aldquotraining
popu
latio
nrdquoof
individu
alsfrom
stages
abov
ethat
have
been
both
Risks
ofpo
orpredictio
nProg
eny
testingneedsto
becontinueddu
ring
the
timethetraining
setis
US
One
prog
rammeso
farmdash
USD
Ain
WisconsinThree
cycles
ofgeno
mic
selectioncompleted
CN~1
000
geno
types
NLprelim
inary
mod
elsforMsin
developed
UK~1
000
geno
types
Verysuitable
application
not
attempted
yet
Maybe
challeng
ingfor
interspecies
hybrids
FR(INRA)120
0geno
type
ofPop
ulus
nigraarebeingused
todevelop
intraspecificGS
(Con
tinues)
14 | CLIFTON‐BROWN ET AL
TABLE
5(Contin
ued)
Breeding
techno
logy
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sW
illow
Pop
lar
predictin
gthe
performance
ofprog
enyof
crosses
(and
inresearch
topredictbestparents
tousein
biparental
crosses)
geno
typedand
phenotyp
edto
developamod
elthattakesgeno
typic
datafrom
aldquocandidate
popu
latio
nrdquoof
untested
individu
als
andprod
uces
GEBVs(Jannink
Lorenzamp
Iwata
2010
)
beingph
enotyp
edand
geno
typed
Inthis
time
next‐generation
germ
plasm
andreadyto
beginthe
second
roun
dof
training
and
recalib
ratio
nof
geno
mic
predictio
nmod
els
US
Prelim
inary
mod
elsforMsac
and
Msin
developed
Work
isun
derw
ayto
deal
moreeffectivelywith
polyploidgenetics
calib
ratio
nsUS
125
0geno
types
arebeingused
todevelopGS
calib
ratio
ns
Traditio
nal
transgenesis
Efficient
introd
uctio
nof
ldquoforeign
rdquotraits
(possiblyfrom
other
genu
sspecies)
into
anelite
plant(ega
prov
enparent
ora
hybrid)that
needsa
simpletraitto
beim
prov
edValidate
cand
idategenesfrom
QTLstud
ies
MASand
know
ledg
eof
the
biolog
yof
thetrait
andsource
ofgenesto
confer
the
relevant
changesto
phenotyp
eWorking
transformation
protocol
IPissuescostof
regu
latory
approv
al
GMO
labelling
marketin
gissuestricky
tousetransgenes
inou
t‐breedersbecause
ofcomplexity
oftransformingandgene
flow
risks
US
Manyprog
ramsin
UnitedStates
are
creatin
gtransgenic
plantsusingAlamoas
asource
oftransformable
geno
typesManytraits
ofinterestNothing
commercial
yet
CNC
hang
shaBtg
ene
transformed
into
Msac
in20
04M
sin
transformed
with
MINAC2gene
and
Msactransform
edwith
Cry2A
ageneb
othwith
amarkerg
eneby
Agrob
acterium
JP1
Msintransgenic
forlow
temperature
tolerancewith
increased
expression
offructans
(unp
ublished)
NLImprov
ingprotocol
forM
sin
transformation
SK2
Msin
with
Arabido
psisand
Brachypod
ium
PhytochromeBgenes
(Hwang
Cho
etal
2014
Hwang
Lim
etal20
14)
UK2
Msin
and
1Mflorg
enotyp
es
UKRou
tine
transformationno
tyet
possibleResearchto
overcomerecalcitrance
ongo
ing
Currently
trying
differentspecies
andcond
ition
sPo
plar
transformationused
atpresent
US
Frequently
attempted
with
very
limitedsuccess
US
600transgenic
lines
have
been
characterized
mostly
performed
inthe
aspenhy
brids
reprod
uctiv
esterility
drou
ghttolerance
with
Kno
ckdo
wns
(Con
tinues)
CLIFTON‐BROWN ET AL | 15
TABLE
5(Contin
ued)
Breeding
techno
logy
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sW
illow
Pop
lar
variou
slytransformed
with
four
cellwall
genesiptand
uidA
genesusingtwo
selectionsystem
sby
biolisticsand
Agrob
acterium
Transform
edplantsare
beinganalysed
US
Prelim
inarywork
will
betakenforw
ardin
2018
ndash202
2in
CABBI
Genom
eediting
CRISPR
Refinem
entofexisting
traits
inusefulparents
orprom
isinghybrids
bygeneratingtargeted
mutations
ingenes
know
ntocontrolthe
traitof
interestF
irst
adouble‐stranded
breakismadeinthe
DNAwhich
isrepairedby
natural
DNArepair
machineryL
eads
tofram
eshiftSN
Por
can
usealdquorepair
templaterdquo
orcanbe
used
toinserta
transgene
intoaldquosafe
harbourlocusrdquoIt
couldbe
used
todeleterepressors(or
transcriptionfactors)
Identification
mapping
and
sequ
encing
oftarget
genes(from
DNA
sequ
ence)
avoiding
screening
outof
unintend
ededits
Manyregu
latory
authorities
have
not
decidedwhether
CRISPR
andother
geno
meediting
techno
logies
are
GMOsor
notIf
GMOsseecomment
onstage10
abov
eIf
noteditedcrop
swill
beregu
latedas
conv
entio
nalvarieties
CATechn
olog
yisstill
toonewand
switchg
rass
geno
meis
very
complexOther
labo
ratories
are
interestedbu
tno
tyet
mov
ingon
this
US
One
prog
rammeso
farmdash
USD
Ain
Albany
FRInitiated
in20
16(M
ISEDIT
project)
NLInitiatingin
2018
US
Initiatingin
2018
in
CABBI
UKNot
started
Not
possible
until
transformation
achieved
US
CRISPR
‐Cas9and
Cpf1have
been
successfully
developedin
Pop
ulus
andarehigh
lyefficientExamples
forlig
nin
biosyn
thesis
Note
BFF
BiomassFo
rtheFu
tureBtBacillus
thuringiensisCABBICenterforadvanced
bioenergyandbioproductsinnovatio
nCpf1
CRISPR
from
Prevotella
andFrancisella
1CRISPR
clusteredregularlyinterspaced
palin
drom
icrepeatsCRISPR
‐CasC
RISPR
‐associated
Cry2A
acrystaltoxins2A
asubfam
ilyproduced
byBtFS
full‐sibG
BS
genotyping
bysequencingG
EBVsgenomic‐estim
ated
breeding
valuesG
MOg
enetically
modi-
fied
organism
GS
genomicselection
GWAS
genomew
ideassociationstudy
INRAF
renchNationalInstituteforAgriculturalR
esearch
IPintellectualp
roperty
IPTisopentenyltransferase
ISSR
inter‐SSR
MflorMiscant-
husflo
ridulusMsacMsacchariflo
rusMsinMsinensisM
AS
marker‐a
ssistedselection
MISEDITm
iscanthusgene
editing
forseed‐propagatedtriploidsNACn
oapicalmeristemA
TAF1
2and
cup‐shaped
cotyledon2
‐lik
eQTLq
uantitativ
etraitlocusS
NP
singlenucleotid
epolymorphismS
SRsim
plesequence
repeatsUSD
AU
SDepartm
ento
fAgriculture
a ISO
Alpha‐2
lettercountrycodes
16 | CLIFTON‐BROWN ET AL
High seed production from field crossing trials con-ducted in locations where flowering in both seed andpollen parents is likely to happen synchronously
Scalable and adapted harvesting threshing and seed pro-cessing methods for producing high seed quality
The results of these parallel activities need to becombined to identify the upscaling pathway for eachhybrid if this cannot be achieved the hybrid will likelynot be commercially viable The UK‐led programme withpartners in Italy and Germany shows that seedbased mul-tiplication rates of 12000 are achievable several inter-specific hybrids (Clifton‐Brown et al 2017) Themultiplication rate of M sinensis is higher probably15000ndash10000 Conventional cloning from rhizome islimited to around 120 that is one ha could provide rhi-zomes for around 20 ha of new plantation
Multilocation field testing of wild and novel miscanthushybrids selected by breeding programmes in the Nether-lands and the United Kingdom was performed as part ofthe project Optimizing Miscanthus Biomass Production(OPTIMISC 2012ndash2016) These trials showed that com-mercial yields and biomass qualities (Kiesel et al 2017Van der Weijde et al 2017a Van der Weijde Kieselet al 2017) could be produced in a wide range of climatesand soil conditions from the temperate maritime climate ofwestern Wales to the continental climate of eastern Russiaand the Ukraine (Kalinina et al 2017) Extensive environ-mental measurements of soil and climate combined withgrowth monitoring are being used to understand abioticstresses (Nunn et al 2017 Van der Weijde Huxley et al2017) and develop genotype‐specific scenarios similar tothose reported earlier in Hastings et al (2009) Phenomicsexperiments on drought tolerance have been conducted onwild and improved germplasm (Malinowska Donnison ampRobson 2017 Van der Weijde Huxley et al 2017)Recently produced interspecific hybrids displaying excep-tional yield under drought (~30 greater than control Mxg)in field trials in Poland and Moldova are being furtherstudied in detail in the phenomics and genomics facility atAberystwyth to better understand genendashtrait associationswhich can be fed back into breeding
Intraspecific seeded hybrids of M sinensis produced inthe Netherlands and interspecific M sacchariflorus times M si-nensis hybrids produced by the UK‐led breeding programmehave entered yield testing in 2018 with the recently EU‐funded project ldquoGRowing Advanced industrial Crops on mar-ginal lands for biorEfineries (GRACE)rdquo (httpswwwgrace-bbieu) Substantial variation in biomass quality for sacchari-fication efficiency (glucose release as of dry matter) ashcontent and melting point has already been generated inintraspecific M sinensis hybrids (Van der Weijde Kiesel
et al 2017) across environments (Weijde Dolstra et al2017a) GRACE aims to establish more than 20 hectares ofnew inter‐ and intraspecific seeded hybrids across six Euro-pean countries This project is building the know‐how andagronomy needed to transition from small research plots tocommercial‐scale field sites and linking biomass productiondirectly to industrial applications The biomass produced byhybrids in different locations will be supplied to innovativeindustrial end‐users making a wide range of biobased prod-ucts both for chemicals and for energy In the United Statesmulti‐location yield were initiated in 2018 to evaluate new tri-ploid M times giganteus genotypes developed at Illinois Cur-rently infertile hybrids are favoured in the United Statesbecause this eliminates the risk of invasiveness from naturallydispersed viable seed The precautionary principle is appliedas fertile miscanthus has naturalized in several states (QuinnAllen amp Stewart 2010) North European multilocation fieldtrials in the EMI and OPTIMISC projects have shown thereis minimal risk of invasiveness even in years when fertileflowering hybrids produce viable seed Naturalized standshave not established here due perhaps to low dormancy pooroverwintering and low seedling competitive strength In addi-tion to breeding for nonshattering or sterile seeded hybridsQuinn et al (2010) suggest management strategies which canfurther minimize environmental opportunities to manage therisk of invasiveness
31 | Molecular breeding and biotechnology
In miscanthus new plant breeding techniques (Table 5)have focussed on developing molecular markers forbreeding in Europe the United States South Korea andJapan There are several publications on QTL mappingpopulations for key traits such as flowering (AtienzaRamirez amp Martin 2003) and compositional traits(Atienza Satovic Petersen Dolstra amp Martin 2003) Inthe United States and United Kingdom independent andinterconnected bi‐parental ldquomappingrdquo families have beenstudied (Dong et al 2018 Gifford Chae SwaminathanMoose amp Juvik 2015) alongside panels of diverse germ-plasm accessions for GWAS (Slavov et al 2013) Fur-ther developments calibrating GS with very large panelsof parents and cross progeny are underway (Daveyet al 2017) The recently completed first miscanthus ref-erence genome sequence is expected to improve the effi-ciency of MAS strategies and especially GWAS(httpsphytozomejgidoegovpzportalhtmlinfoalias=Org_Msinensis_er) For example without a referencegenome sequence Clark et al (2014) obtained 21207RADseq SNPs (single nucleotide polymorphisms) on apanel of 767 miscanthus genotypes (mostly M sinensis)but subsequent reanalysis of the RADseq data using the
CLIFTON‐BROWN ET AL | 17
new reference genome resulted in hundreds of thousandsof SNPs being called
Robust and effective in vitro regeneration systems havebeen developed for Miscanthus sinensis M times giganteusand M sacchariflorus (Dalton 2013 Guo et al 2013Hwang Cho et al 2014 Rambaud et al 2013 Ślusarkie-wicz‐Jarzina et al 2017 Wang et al 2011 Zhang et al2012) However there is still significant genotype speci-ficity and these methods need ldquoin‐houserdquo optimization anddevelopment to be used routinely They provide potentialroutes for rapid clonal propagation and also as a basis forgenetic transformation Stable transformation using bothbiolistics (Wang et al 2011) and Agrobacterium tumefa-ciens DNA delivery methods (Hwang Cho et al 2014Hwang Lim et al 2014) has been achieved in M sinen-sis The development of miscanthus transformation andgene editing to generate diplogametes for producing seed‐propagated triploid hybrids are performed as part of theFrench project MISEDIT (miscanthus gene editing forseed‐propagated triploids) There are no reports of genomeediting in any miscanthus species but new breeding inno-vations including genome editing are particularly relevantin this slow‐to‐breed nonfood bioenergy crop (Table 4)
4 | WILLOW
Willow (Salix spp) is a very diverse group of catkin‐bear-ing trees and shrubs Willow belongs to the family Sali-caceae which also includes the Populus genus There areapproximately 350 willow species (Argus 2007) foundmostly in temperate and arctic zones in the northern hemi-sphere A few are adapted to subtropical and tropicalzones The centre of diversity is believed to be in Asiawith over 200 species in China Around 120 species arefound in the former Soviet Union over 100 in NorthAmerica and around 65 species in Europe and one speciesis native to South America (Karp et al 2011) Willows aredioecious thus obligate outcrossers and highly heterozy-gous The haploid chromosome number is 19 (Hanley ampKarp 2014) Around 40 of species are polyploid (Sudaamp Argus 1968) ranging from triploids to the atypicaldodecaploid S maxxaliana with 2n=190 (Zsuffa et al1984)
Although almost exclusively native to the NorthernHemisphere willow has been grown around the globe formany thousands of years to support a wide range of appli-cations (Kuzovkina amp Quigley 2005 Stott 1992) How-ever it has been the focus of domestication for bioenergypurposes for only a relatively short period since the 1970sin North America and Europe For bioenergy breedershave focused their efforts on the shrub willows (subgenusVetix) because of their rapid juvenile growth rates as aresponse to coppicing on a 2‐ to 4‐year cycle that can be
accomplished using farm machinery rather than forestryequipment (Shield Macalpine Hanley amp Karp 2015Smart amp Cameron 2012)
Since shrub willow was not generally recognized as anagricultural crop until very recently there has been littlecommitment to building and maintaining germplasm reposi-tories of willow to support long‐term breeding One excep-tion is the United Kingdom where a large and well‐characterized Salix germplasm collection comprising over1500 accessions is held at Rothamsted Research (Stott1992 Trybush et al 2008) Originally initiated for use inbasketry in 1923 accessions have been added ever sinceIn the United States a germplasm collection of gt350accessions is located at Cornell University to support thebreeding programme there The UK and Cornell collectionshave a relatively small number of accessions in common(around 20) Taken together they represent much of thespecies diversity but only a small fraction of the overallgenetic diversity within the genus There are three activewillow breeding programmes in Europe RothamstedResearch (UK) Salixenergi Europa AB (SEE) and a pro-gramme at the University of Warmia and Mazury in Olsz-tyn (Poland) (abbreviations used in Table 1) There is oneactive US programme based at Cornell University Culti-vars are still being marketed by the European WillowBreeding Programme (EWBP) (UK) which was activelybreeding biomass varieties from 1996 to 2002 Cultivarsare protected by plant breedersrsquo rights (PBRs) in Europeand by plant patents in the United States The sharing ofgenetic resources in the willow community is generallyregulated by material transfer agreements (MTA) and tai-lored licensing agreements although the import of cuttingsinto North America is prohibited except under special quar-antine permit conditions
Efforts to augment breeding germplasm collection fromnature are continuing with phenotypic screening of wildgermplasm performed in field experiments with 177 S pur-purea genotypes in the United States (at sites in Genevaand Portland NY and Morgantown WV) that have beengenotyped using genotyping by sequencing (GBS) (Elshireet al 2011) In addition there are approximately 400accessions of S viminalis in Europe (near Pustnaumls Upp-sala Sweden and Woburn UK (Berlin et al 2014 Hal-lingbaumlck et al 2016) The S viminalis accessions wereinitially genotyped using 38 simple sequence repeats (SSR)markers to assess genetic diversity and screened with~1600 SNPs in genes of potential interest for phenologyand biomass traits Genetics and genomics combined withextensive phenotyping have substantially improved thegenetic basis of biomass‐related traits in willow and arenow being developed in targeted breeding via MAS Thisunderpinning work has been conducted on large specifi-cally developed biparental Salix mapping populations
18 | CLIFTON‐BROWN ET AL
(Hanley amp Karp 2014 Zhou et al 2018) as well asGWAS panels (Hallingbaumlck et al 2016)
Once promising parental combinations are identifiedcrosses are usually performed using fresh pollen frommaterial that has been subject to a phased removal fromcold storage (minus4degC) (Lindegaard amp Barker 1997 Macal-pine Shield Trybush Hayes amp Karp 2008 Mosseler1990) Pollen storage is useful in certain interspecific com-binations where flowering is not naturally synchronizedThis can be overcome by using pollen collection and stor-age protocol which involves extracting pollen using toluene(Kopp Maynard Niella Smart amp Abrahamson 2002)
The main breeding approach to improve willow yieldsrelies on species hybridization to capture hybrid vigour(Fabio et al 2017 Serapiglia Gouker amp Smart 2014) Inthe absence of genotypic models for heterosis breedershave extensively tested general and specific combiningability of parents to produce superior progeny The UKbreeding programmes (EWBP 1996ndash2002 and RothamstedResearch from 2003 on) have performed more than 1500exploratory cross‐pollinations The Cornell programme hassuccessfully completed about 550 crosses since 1998Investment into the characterization of genetic diversitycombined with progeny tests from exploratory crosses hasbeen used to produce hundreds of targeted intraspeciescrosses in the United Kingdom and United States respec-tively (see Table 1) To achieve long‐term gains beyond F1hybrids four intraspecific recurrent selection populationshave been created in the United Kingdom (for S dasycla-dos S viminalis and S miyabeana) and Cornell is pursu-ing recurrent selection of S purpurea Interspecifichybridizations with genotypes selected from the recurrentselection cycles are well advanced in willow with suchcrosses to date totalling 420 in the United Kingdom andover 100 in the United States
While species hybridization is common in Salix it isnot universal Of the crosses attempted about 50 hybri-dize and produce seed (Macalpine Shield amp Karp 2010)As the viability of seed from successful crosses is short (amatter of days at ambient temperatures) proper seed rear-ing and storage protocols are essential (Maroder PregoFacciuto amp Maldonado 2000)
Progeny from crosses are treated in different waysamong the breeding programmes at the seedling stage Inthe United States seedlings are planted into an irrigatedfield where plants are screened for two seasons beforebeing progressed to further field trials In the United King-dom seedlings are planted into trays of compost wherethey remain containerized in an irrigated nursery for theremainder of year one In the United Kingdom seedlingsare subject to two rounds of selection in the nursery yearThe first round takes place in September to select againstsusceptibility to rust infection (Melampsora spp) A second
round of selection in winter assesses tip damage from frostand giant willow aphid infestation In the United Stateswhere the rust pressure is lower screening for Melampsoraspp cannot be performed at the nursery stage Both pro-grammes monitor Melampsora spp pest susceptibilityyield and architecture over multiple years in field trialsSelected material is subject to two rounds of field trials fol-lowed by a final multilocation yield trial to identify vari-eties for commercialization
Promising selections (ie potential cultivars) need to beclonally propagated A rapid in vitro tissue culture propa-gation method has been developed (Palomo‐Riacuteos et al2015) This method can generate about 5000 viable trans-plantable clones from a single plant in just 24 weeks Anin vitro system can also accommodate early selection viamolecular or biochemical markers to increase selectionspeed Conventional breeding systems take 13 years viafour rounds of selection from crossing to selecting a variety(Figure 1) but this has the potential to be reduced to7 years if micropropagation and MAS selection are adopted(Hanley amp Karp 2014 Palomo‐Riacuteos et al 2015)
Willows are currently propagated commercially byplanting winter‐dormant stem cuttings in spring Commer-cial planting systems for willow use mechanical plantersthat cut and insert stem sections from whips into a well‐prepared soil One hectare of stock plants grown in specificmultiplication beds planted at 40000 plants per ha pro-duces planting material for 80 hectares of commercialshort‐rotation coppice willow annually (planted at 15000cuttings per hectare) (Whittaker et al 2016) When com-mercial plantations are established the industry standard isto plant intimate mixtures of ~5 diverse rust (Melampsoraspp)‐resistant varieties (McCracken amp Dawson 1997 VanDen Broek et al 2001)
The foundations for using new plant breeding tech-niques have been established with funding from both thepublic and the private sectors To establish QTL maps 16mapping populations from biparental crosses are understudy in the United Kingdom Nine are under study in theUnited States The average number of individuals in thesefamilies ranges from 150 to 947 (Hanley amp Karp 2014)GS is also being evaluated in S viminalis and preliminaryresults indicate that multiomic approaches combininggenomic and metabolomic data have great potential (Sla-vov amp Davey 2017) For both QTL and GS approachesthe field phenotyping demands are large as several thou-sand individuals need to be phenotyped for a wide rangeof traits These include the following dates of bud burstand growth succession stem height stem density wooddensity and disease resistance The greater the number ofindividuals the more precise the QTL marker maps andGS models are However the logistical and financial chal-lenges of phenotyping large numbers of individuals are
CLIFTON‐BROWN ET AL | 19
considerable because the willow crop is gt5 m tall in thesecond year There is tremendous potential to improve thethroughput of phenotyping using unmanned aerial systemswhich is being tested in the USDA National Institute ofFood and Agriculture (NIFA) Willow SkyCAP project atCornell Further investment in these approaches needs tobe sustained over many years fully realizes the potentialof a marker‐assisted selection programme for willow
To date despite considerable efforts in Europe and theUnited States to establish a routine transformation systemthere has not been a breakthrough in willow but attemptsare ongoing As some form of transformation is typically aprerequisite for genome editing techniques these have notyet been applied to willow
In Europe there are 53 short‐rotation coppice (SRC) bio-mass willow cultivars registered with the Community PlantVariety Office (CPVO) for PBRs of which ~25 are availablecommercially in the United Kingdom There are eightpatented cultivars commercially available in the UnitedStates In Sweden there are nine commercial cultivars regis-tered in Europe and two others which are unregistered(httpssalixenergiseplanting-material) Furthermore thereare about 20 precommercial hybrids in final yield trials inboth the United States and the United Kingdom It has beenestimated that it would take two years to produce the stockrequired to plant 50 ha commercially from the plant stock inthe final yield trials Breeding programmes have alreadydelivered rust‐resistant varieties and increases in yield to themarket The adoption of advanced breeding technologies willlikely lead to a step change in improving traits of interest
5 | POPLAR
Poplar a fast‐growing tree from the northern hemispherewith a small genome size has been adopted for commercialforestry and scientific purposes The genus Populus con-sists of about 29 species classified in six different sectionsPopulus (formerly Leuce) Tacamahaca Aigeiros AbasoTuranga and Leucoides (Eckenwalder 1996) The Populusspecies of most interest for breeding and testing in the Uni-ted States and Europe are P nigra P deltoides P maxi-mowiczii and P trichocarpa (Stanton 2014) Populusclones for biomass production are being developed byintra‐ and interspecies hybridization (DeWoody Trewin ampTaylor 2015 Richardson Isebrands amp Ball 2014 van derSchoot et al 2000) Recurrent selection approaches areused for gradual population improvement and to create eliteclonal lines for commercialization (Berguson McMahon ampRiemenschneider 2017 Neale amp Kremer 2011) Currentlypoplar breeding in the United States occurs in industrialand academic programmes located in the Southeast theMidwest and the Pacific Northwest These use six speciesand five interspecific taxa (Stanton 2014)
The southeastern programme historically focused onrecurrent selection of P deltoides from accessions made inthe lower Mississippi River alluvial plain (Robison Rous-seau amp Zhang 2006) More recently the genetic base hasbeen broadened to produce interspecific hybrids with resis-tance to the fungal infection Septoria musiva which causescankers
In the midwest of the United States populationimprovement efforts are focused on P deltoides selectionsfrom native provenances and hybrid crosses with acces-sions introduced from Europe Interspecific intercontinental(Europe and America) hybrid crosses between P nigra andP deltoides (P times canadensis) are behind many of theleading commercial hybrids which are the most advancedbreeding materials for many applications and regions InMinnesota previous breeding experience and efforts utiliz-ing P maximowiczii and P trichocarpa have been discon-tinued due to Septoria susceptibility and a lack of coldhardiness (Berguson et al 2017) Traits targeted forimprovement include yieldgrowth rate cold hardinessadventitious rooting resistance to Septoria and Melamp-sora leaf rust and stem form The Upper Midwest pro-gramme also carries out wide hybridizations within thesection Populus The P times wettsteinii (P trem-ula times P tremuloides) taxon is bred for gains in growthrate wood quality and resistance to the fungus Entoleucamammata which causes hypoxylon canker (David ampAnderson 2002)
In the Pacific Northwest GreenWood Resources Incleads poplar breeding that emphasizes interspecifichybrid improvement of P times generosa (P deltoides timesP trichocarpa and reciprocal) and P deltoides times P maxi-mowiczii taxa for coastal regions and the P times canadensistaxon for the drier continental regions Intraspecificimprovement of second‐generation breeding populations ofP deltoides P nigra P maximowiczii and P trichocarpaare also involved (Stanton et al 2010) The present focusof GreenWood Resourcesrsquo hybridization is bioenergy feed-stock improvement concentrating on coppice yield wood‐specific gravity and rate of sugar release
Industrial interest in poplar in the United States has his-torically come from the pulp and paper sector althoughveneer and dimensional lumber markets have been pursuedat times Currently the biomass market for liquid trans-portation fuels is being emphasized along with the use oftraditional and improved poplar genotypes for ecosystemservices such as phytoremediation (Tuskan amp Walsh 2001Zalesny et al 2016)
In Europe there are breeding programmes in FranceGermany Italy and Sweden These include the following(a) Alasia Franco Vivai (AFV) programme in northernItaly (b) the French programme led by the poplar Scien-tific Interest Group (GIS Peuplier) and carried out
20 | CLIFTON‐BROWN ET AL
collaboratively between the National Institute for Agricul-tural Research (INRA) the National Research Unit ofScience and Technology for Environment and Agriculture(IRSTEA) and the Forest Cellulose Wood Constructionand Furniture Technology Institute (FCBA) (c) the Ger-man programme at Northwest German Forest Research Sta-tion (NW‐FVA) at Hannoversch Muumlnden and (d) theSwedish programme at the Swedish University of Agricul-tural Sciences and SweTree Technologies AB (Table 1)
AFV leads an Italian poplar breeding programme usingextensive field‐grown germplasm collections of P albaP deltoides P nigra and P trichocarpa While interspeci-fic hybridization uses several taxa the focus is onP times canadensis The breeding programme addresses dis-ease resistance (Marssonina brunnea Melampsora larici‐populina Discosporium populeum and poplar mosaicvirus) growth rate and photoperiod adaptation AFV andGreenWood Resources collaborate in poplar improvementin Europe through the exchange of frozen pollen and seedfor reciprocal breeding projects Plantations in Poland andRomania are currently the focus of the collaboration
The ongoing French GIS Peuplier is developing a long‐term breeding programme based on intraspecific recurrentselection for the four parental species (P deltoides P tri-chocarpa P nigra and P maximowiczii) designed to betterbenefit from hybrid vigour demonstrated by the interspeci-fic crosses P canadensis P deltoides times P trichocarpaand P trichocarpa times P maximowiczii Current selectionpriorities are targeting adaptation to soil and climate condi-tions resistance and tolerance to the most economicallyimportant diseases and pests high volume production underSRC and traditional poplar cultivation regimes as well aswood quality of interest by different markets Currentlygenomic selection is under exploration to increase selectionaccuracy and selection intensity while maintaining geneticdiversity over generations
The German NW‐FVA programme is breeding intersec-tional AigeirosndashTacamahaca hybrids with a focus on resis-tance to Pollaccia elegans Xanthomonas populiDothichiza spp Marssonina brunnea and Melampsoraspp (Stanton 2014) Various cross combinations ofP maximowiczii P trichocarpa P nigra and P deltoideshave led to new cultivars suitable for deployment in vari-etal mixtures of five to ten genotypes of complementarystature high productivity and phenotypic stability (Weis-gerber 1993) The current priority is the selection of culti-vars for high‐yield short‐rotation biomass production Sixhundred P nigra genotypes are maintained in an ex situconservation programme An in situ P nigra conservationeffort involves an inventory of native stands which havebeen molecular fingerprinted for identity and diversity
The Swedish programme is concentrating on locallyadapted genotypes used for short‐rotation forestry (SRF)
because these meet the needs of the current pulping mar-kets Several field trials have shown that commercial poplarclones tested and deployed in Southern and Central Europeare not well adapted to photoperiods and low temperaturesin Sweden and in the Baltics Consequently SwedishUniversity of Agricultural Sciences and SweTree Technolo-gies AB started breeding in Sweden in 1990s to producepoplar clones better adapted to local climates and markets
51 | Molecular breeding technologies
Poplar genetic improvement cannot be rapidly achievedthrough traditional methods alone because of the longbreeding cycles outcrossing breeding systems and highheterozygosity Integrating modern genetic genomic andphenomics techniques with conventional breeding has thepotential to expedite poplar improvement
The genome of poplar has been sequenced (Tuskanet al 2006) It has an estimated genome size of485 plusmn 10 Mbp divided into 19 chromosomes This is smal-ler than other PBCs and makes poplar more amenable togenetic engineering (transgenesis) GS and genome editingPoplar has seen major investment in both the United Statesand Europe being the model system for woody perennialplant genetics and genomics research
52 | Targets for genetic modification
Traits targeted include wood properties (lignin content andcomposition) earlylate flowering male sterility to addressbiosafety regulation issues enhanced yield traits and herbi-cide tolerance These extensive transgenic experiments haveshown differences in recalcitrance to in vitro regenerationand genetic transformation in some of the most importantcommercial hybrid poplars (Alburquerque et al 2016)Further transgene expression stability is being studied Sofar China is the only country known to have commerciallyused transgenic insect‐resistant poplar A precommercialherbicide‐tolerant poplar was trialled for 8 years in the Uni-ted States (Li Meilan Ma Barish amp Strauss 2008) butcould not be released due to stringent environmental riskassessments required for regulatory approval This increasestranslation costs and delays reducing investor confidencefor commercial deployment (Harfouche et al 2011)
The first field trials of transgenic poplar were performedin France in 1987 (Fillatti Sellmer Mccown Haissig ampComai 1987) and in Belgium in 1988 (Deblock 1990)Although there have been a total of 28 research‐scale GMpoplar field trials approved in the European Union underCouncil Directive 90220EEC since October 1991 (inPoland Belgium Finland France Germany Spain Swe-den and in the United Kingdom (Pilate et al 2016) onlyauthorizations in Poland and Belgium are in place today In
CLIFTON‐BROWN ET AL | 21
the United States regulatory notifications and permits fornearly 20000 transgenic poplar trees derived from approxi-mately 600 different constructs have been issued since1995 by the USDAs Animal and Plant Health InspectionService (APHIS) (Strauss et al 2016)
53 | Genome editing CRISPR technologies
Clustered regularly interspaced palindromic repeats(CRISPR) and the CRISPR‐associated (CRISPR‐Cas)nucleases are a groundbreaking genome‐engineering toolthat complements classical plant breeding and transgenicmethods (Moreno‐Mateos et al 2017) Only two publishedstudies in poplar have applied the CRISPRCas9 technol-ogy One is in P tomentosa in which an endogenous phy-toene desaturase gene (PtoPDS) was successfully disruptedsite specifically in the first generation of transgenic plantsresulting in an albino and dwarf phenotype (Fan et al2015) The second was in P tremula times alba in whichhigh CRISPR‐Cas9 mutational efficiency was achieved forthree 4‐coumarateCoA ligase (4Cl) genes 4CL1 4CL2and 4CL5 associated with lignin and flavonoid biosynthe-sis (Zhou et al 2015) Due to its low cost precision andrapidness it is very probable that cultivars or clones pro-duced using CRISPR technology will be ready for market-ing in the near future (Yin et al 2017) Recently aCRISPR with a smaller associated endonuclease has beendiscovered from Prevotella and Francisella 1 (Cpf1) whichmay have advantages over Cas9 In addition there arereports of DNA‐free editing in plants using both CRISPRCpf1 and CRISPR Cas9 for example Ref (Kim et al2017 Mahfouz 2017 Zaidi et al 2017)
It remains unresolved whether plants modified by genomeediting will be regulated as genetically modified organisms(GMOs) by the relevant authorities in different countries(Lozano‐Juste amp Cutler 2014) Regulations to cover thesenew breeding techniques are still evolving but those coun-tries who have published specific guidance (including UnitedStates Argentina and Chile) are indicating that plants pos-sessing simple genome edits will not be regulated as conven-tional transgenesis (Jones 2015b) The first generation ofgenome‐edited crops will likely be phenocopy gene knock-outs that already exist to produce ldquonature identicalrdquo traits thatis traits that could also be derived by conventional breedingDespite this confidence in applying these new powerfulbreeding tools remains limited owing to the uncertain regula-tory environment in many parts of the world (Gao 2018)including the recent ECJ 2018 rulings mentioned earlier
54 | Genomics‐based breeding technologies
Poplar breeding programs are becoming well equipped withuseful genomics tools and resources that are critical to
explore genomewide variability and make use of the varia-tion for enhancing genetic gains Deep transcriptomesequencing resequencing of alternate genomes and GBStechnology for genomewide marker detection using next‐generation sequencing (NGS) are yielding valuable geno-mics tools GWAS with NGS‐based markers facilitatesmarker identification for MAS breeding by design and GS
GWAS approaches have provided a deeper understand-ing of genome function as well as allelic architectures ofcomplex traits (Huang et al 2010) and have been widelyimplemented in poplar for wood characteristics (Porthet al 2013) stomatal patterning carbon gain versus dis-ease resistance (McKown et al 2014) height and phenol-ogy (Evans et al 2014) cell wall chemistry (Mucheroet al 2015) growth and cell walls traits (Fahrenkroget al 2017) bark roughness (Bdeir et al 2017) andheight and diameter growth (Liu et al 2018) Using high‐throughput sequencing and genotyping platforms an enor-mous amount of SNP markers have been used to character-ize the linkage disequilibrium (LD) in poplar (eg Slavovet al 2012 discussed below)
The genetic architecture of photoperiodic traits in peren-nial trees is complex involving many loci However itshows high levels of conservation during evolution (Mau-rya amp Bhalerao 2017) These genomics tools can thereforebe used to address adaptation issues and fine‐tune themovement of elite lines into new environments For exam-ple poor timing of spring bud burst and autumn bud setcan result in frost damage resulting in yield losses (Ilstedt1996) These have been studied in P tremula genotypesalong a latitudinal cline in Sweden (~56ndash66oN) and haverevealed high nucleotide polymorphism in two nonsynony-mous SNPs within and around the phytochrome B2 locus(Ingvarsson Garcia Hall Luquez amp Jansson 2006 Ing-varsson Garcia Luquez Hall amp Jansson 2008) Rese-quencing 94 of these P tremula genotypes for GWASshowed that noncoding variation of a single genomicregion containing the PtFT2 gene described 65 ofobserved genetic variation in bud set along the latitudinalcline (Tan 2018)
Resequencing genomes is currently the most rapid andeffective method detecting genetic differences betweenvariants and for linking loci to complex and importantagronomical and biomass traits thus addressing breedingchallenges associated with long‐lived plants like poplars
To date whole genome resequencing initiatives havebeen launched for several poplar species and genotypes InPopulus LD studies based on genome resequencing sug-gested the feasibility of GWAS in undomesticated popula-tions (Slavov et al 2012) This plant population is beingused to inform breeding for bioenergy development Forexample the detection of reliable phenotypegenotype asso-ciations and molecular signatures of selection requires a
22 | CLIFTON‐BROWN ET AL
detailed knowledge about genomewide patterns of allelefrequency variation LD and recombination suggesting thatGWAS and GS in undomesticated populations may bemore feasible in Populus than previously assumed Slavovet al (2012) have resequenced 16 genomes of P tri-chocarpa and genotyped 120 trees from 10 subpopulationsusing 29213 SNPs (Geraldes et al 2013) The largest everSNP data set of genetic variations in poplar has recentlybeen released providing useful information for breedinghttpswwwbioenergycenterorgbescgwasindexcfmAlso deep sequencing of transcriptomes using RNA‐Seqhas been used for identification of functional genes andmolecular markers that is polymorphism markers andSSRs A multitissue and multiple experimental data set forP trichocarpa RNA‐Seq is publicly available (httpsjgidoegovdoe-jgi-plant-flagship-gene-atlas)
The availability of genomic information of DNA‐con-taining cell organelles (nucleus chloroplast and mitochon-dria) will also allow a holistic approach in poplarmolecular breeding in the near future (Kersten et al2016) Complete Populus genome sequences are availablefor nucleus (P trichocarpa section Tacamahaca) andchloroplasts (seven species and two clones from P trem-ula W52 and P tremula times P alba 717ndash1B4) A compara-tive approach revealed structural and functionalinformation broadening the knowledge base of PopuluscpDNA and stimulating future diagnostic marker develop-ment The availability of whole genome sequences of thesecellular compartments of P tremula holds promise forboosting marker‐assisted poplar breeding Other nucleargenome sequences from additional Populus species arenow available (eg P deltoides (httpsphytozomejgidoegovpz) and will become available in the forthcomingyears (eg P tremula and P tremuloidesmdashPopGenIE(Sjodin Street Sandberg Gustafsson amp Jansson 2009))Recently the characterization of the poplar pan‐genome bygenomewide identification of structural variation in threecrossable poplar species P nigra P deltoides and P tri-chocarpa revealed a deeper understanding of the role ofinter‐ and intraspecific structural variants in poplar pheno-type and may have important implications for breedingparticularly interspecific hybrids (Pinosio et al 2016)
GS has been proposed as an alternative to MAS in cropimprovement (Bernardo amp Yu 2007 Heffner Sorrells ampJannink 2009) GS is particularly well suited for specieswith long generation times for characteristics that displaymoderate‐to‐low heritability for traits that are expensive tomeasure and for selection of traits expressed late in the lifecycle as is the case for most traits of commercial value inforestry (Harfouche et al 2012) Current joint genomesequencing efforts to implement GS in poplar using geno-mic‐estimated breeding values for bioenergy conversiontraits from 49 P trichocarpa families and 20 full‐sib
progeny are taking place at the Oak Ridge National Labo-ratory and GreenWood Resources (Brian Stanton personalcommunication httpscbiornlgov) These data togetherwith the resequenced GWAS population data will be thebasis for developing GS algorithms Genomic breedingtools have been developed for the intraspecific programmetargeting yield resistance to Venturia shoot blightMelampsora leaf rust resistance to Cryptorhynchus lapathistem form wood‐specific gravity and wind firmness (Evanset al 2014 Guerra et al 2016) A newly developedldquobreeding with rare defective allelesrdquo (BRDA) technologyhas been developed to exploit natural variation of P nigraand identify defective variants of genes predicted by priortransgenic research to impact lignin properties Individualtrees carrying naturally defective alleles can then be incor-porated directly into breeding programs thereby bypassingthe need for transgenics (Vanholme et al 2013) Thisnovel breeding technology offers a reverse genetics com-plement to emerging GS for targeted improvement of quan-titative traits (Tsai 2013)
55 | Phenomics‐assisted breeding technology
Phenomics involves the characterization of phenomesmdashthefull set of phenotypes of given individual plants (HouleGovindaraju amp Omholt 2010) Traditional phenotypingtools which inefficiently measure a limited set of pheno-types have become a bottleneck in plant breeding studiesHigh‐throughput plant phenotyping facilities provide accu-rate screening of thousands of plant breeding lines clones orpopulations over time (Fu 2015) are critical for acceleratinggenomics‐based breeding Automated image collection andanalysis phenomics technologies allow accurate and nonde-structive measurements of a diversity of phenotypic traits inlarge breeding populations (Gegas Gay Camargo amp Doo-nan 2014 Goggin Lorence amp Topp 2015 Ludovisi et al2017 Shakoor Lee amp Mockler 2017) One important con-sideration is the identification of relevant and quantifiabletarget traits that are early diagnostic indicators of biomassyield Good progress has been made in elucidating theseunderpinning morpho‐physiological traits that are amenableto remote sensing in Populus (Harfouche Meilan amp Altman2014 Rae Robinson Street amp Taylor 2004) Morerecently Ludovisi et al (2017) developed a novel methodol-ogy for field phenomics of drought stress in a P nigra F2partially inbred population using thermal infrared imagesrecorded from an unmanned aerial vehicle‐based platform
Energy is the current main market for poplar biomass butthe market return provided is not sufficient to support pro-duction expansion even with added demand for environmen-tal and land management ldquoecosystem servicesrdquo such as thetreatment of effluent phytoremediation riparian buffer zonesand agro‐forestry plantings Aviation fuel is a significant
CLIFTON‐BROWN ET AL | 23
target market (Crawford et al 2016) To serve this marketand to reduce current carbon costs of production (Budsberget al 2016) key improvement traits in addition to yield(eg coppice regeneration pestdisease resistance water‐and nutrient‐use efficiencies) will be trace greenhouse gas(GHG) emissions (eg isoprene volatiles) site adaptabilityand biomass conversion efficiency Efforts are underway tohave national environmental protection agenciesrsquo approvalfor poplar hybrids qualifying for renewable energy credits
6 | REFLECTIONS ON THECOMMERCIALIZATIONCHALLENGE
The research and innovation activities reviewed in thispaper aim to advance the genetic improvement of speciesthat can provide feedstocks for bioenergy applicationsshould those markets eventually develop These marketsneed to generate sufficient revenue and adequately dis-tribute it to the actors along the value chain The work onall four crops shares one thing in common long‐termefforts to integrate fundamental knowledge into breedingand crop development along a research and development(RampD) pipeline The development of miscanthus led byAberystwyth University exemplifies the concerted researcheffort that has integrated the RampD activities from eight pro-jects over 14 years with background core research fundingalong an emerging innovation chain (Figure 2) This pro-gramme has produced a first range of conventionally bredseeded interspecies hybrids which are now in upscaling tri-als (Table 3) The application of molecular approaches(Table 5) with further conventional breeding (Table 4)offers the prospect of a second range of improved seededhybrids This example shows that research‐based support ofthe development of new crops or crop types requires along‐term commitment that goes beyond that normallyavailable from project‐based funding (Figure 1) Innovationin this sector requires continuous resourcing of conven-tional breeding operations and capability to minimize timeand investment losses caused by funding discontinuities
This challenge is increased further by the well‐knownmarket failure in the breeding of many agricultural cropspecies The UK Department for Environment Food andRural Affairs (Defra) examined the role of geneticimprovement in relation to nonmarket outcomes such asenvironmental protection and concluded that public invest-ment in breeding was required if profound market failure isto be addressed (Defra 2002) With the exception ofwidely grown hybrid crops such as maize and some high‐value horticultural crops royalties arising from plant breed-ersrsquo rights or other returns to breeders fail to adequatelycompensate for the full cost for research‐based plant breed-ing The result even for major crops such as wheat is sub‐
optimal investment and suboptimal returns for society Thismarket failure is especially acute for perennial crops devel-oped for improved sustainability rather than consumerappeal (Tracy et al 2016) Figure 3 illustrates the underly-ing challenge of capturing value for the breeding effortThe ldquovalley of deathrdquo that results from the low and delayedreturns to investment applies generally to the research‐to‐product innovation pipeline (Beard Ford Koutsky amp Spi-wak 2009) and certainly to most agricultural crop speciesHowever this schematic is particularly relevant to PBCsMost of the value for society from the improved breedingof these crops comes from changes in how agricultural landis used that is it depends on the increased production ofthese crops The value for society includes many ecosys-tems benefits the effects of a return to seminatural peren-nial crop cover that protects soils the increase in soilcarbon storage the protection of vulnerable land or the cul-tivation of polluted soils and the reductions in GHG emis-sions (Lewandowski 2016) By its very nature theproduction of biomass on agricultural land marginal forfood production challenges farm‐level profitability Thecosts of planting material and one‐time nature of cropestablishment are major early‐stage costs and therefore theopportunities for conventional royalty capture by breedersthat are manifold for annual crops are limited for PBCs(Hastings et al 2017) Public investment in developingPBCs for the nonfood biobased sector needs to providemore long‐term support for this critical foundation to a sus-tainable bioeconomy
7 | CONCLUSIONS
This paper provides an overview of research‐based plantbreeding in four leading PBCs For all four PBC generasignificant progress has been made in genetic improvementthrough collaboration between research scientists and thoseoperating ongoing breeding programmes Compared withthe main food crops most PBC breeding programmes dateback only a few decades (Table 1) This breeding efforthas thus co‐evolved with molecular biology and the result-ing ‐omics technologies that can support breeding Thedevelopment of all four PBCs has depended strongly onpublic investment in research and innovation The natureand driver of the investment varied In close associationwith public research organizations or universities all theseprogrammes started with germplasm collection and charac-terization which underpin the selection of parents forexploratory wide crosses for progeny testing (Figure 1Table 4)
Public support for switchgrass in North America wasexplicitly linked to plant breeding with 12 breeding pro-grammes supported in the United States and CanadaSwitchgrass breeding efforts to date using conventional
24 | CLIFTON‐BROWN ET AL
breeding have resulted in over 36 registered cultivars inthe United States (Table 1) with the development of dedi-cated biomass‐type cultivars coming within the past fewyears While ‐omics technologies have been incorporatedinto several of these breeding programmes they have notyet led to commercial deployment in either conventional orhybrid cultivars
Willow genetic improvement was led by the researchcommunity closely linked to plant breeding programmesWillow and poplar have the longest record of public invest-ment in genetic improvement that can be traced back to1920s in the United Kingdom and United States respec-tively Like switchgrass breeding programmes for willoware connected to public research efforts The United King-dom in partnership with the programme based at CornellUniversity remains the European leader in willowimprovement with a long‐term breeding effort closelylinked to supporting biological research at Rothamsted Inwillow F1 hybrids have produced impressive yield gainsover parental germplasm by capturing hybrid vigour Over30 willow clones are commercially available in the UnitedStates and Europe and a further ~90 are under precommer-cial testing (Table 1)
Compared with willow and poplar miscanthus is a rela-tive newcomer with all the current breeding programmesstarting in the 2000s Clonal M times giganteus propagated byrhizomes is expected to be replaced by more readily scal-able seeded hybrids from intra‐ (M sinensis) and inter‐(M sacchariflorus times M sinensis) species crosses with highseed multiplication rates (of gt2000) The first group ofhybrid cultivars is expected to be market‐ready around2022
Of the four genera used as PBCs Populus is the mostadvanced in terms of achievements in biological researchas a result of its use as a model for basic research of treesMuch of this biological research is not directly connectedto plant breeding Nevertheless reflecting the fact thatpoplar is widely grown as a single‐stem tree in SRF thereare about 60 commercially available clones and an addi-tional 80 clones in commercial pipelines (Table 1) Trans-genic poplar hybrids have moved beyond proof of conceptto commercial reality in China
Many PBC programmes have initiated long‐term con-ventional recurrent selection breeding cycles for populationimprovement which is a key process in increasing yieldthrough hybrid vigour As this approach requires manyyears most programmes are experimenting with molecularbreeding methods as these have the potential to accelerateprecision breeding For all four PBCs investments in basicgenetic and genomic resources including the developmentof mapping populations for QTLs and whole genomesequences are available to support long‐term advancesMore recently association genetics with panels of diversegermplasm are being used as training populations for GSmodels (Table 5) These efforts are benefitting from pub-licly available DNA sequences and whole genome assem-blies in crop databases Key to these accelerated breedingtechnologies are developments in novel phenomics tech-nologies to bridge the genotypephenotype gap In poplarnovel remote sensing field phenotyping is now beingdeployed to assist breeders These advances are being com-bined with in vitro and in planta modern molecular breed-ing techniques such as CRISPR (Table 5) CRISPRtechnology for genome editing has been proven in poplar
FIGURE 2 A schematic developmentpathway for miscanthus in the UnitedKingdom related to the investment in RampDprojects at Aberystwyth (top coloured areasfor projects in the three categories basicresearch breeding and commercialupscaling) leading to a projected croppingarea of 350000 ha by 2030 with clonal andsuccessive ranges of improved seed‐basedhybrids Purple represents theBiotechnology and Biological SciencesResearch Council (BBSRC) and brown theDepartment for Environment Food andRural Affairs (Defra) (UK Nationalfunding) blue bars represent EU fundingand green private sector funding (Terravestaand CERES) and GIANT‐LINK andMiscanthus Upscaling Technology (MUST)are publicndashprivate‐initiatives (PPI)
CLIFTON‐BROWN ET AL | 25
This technology is also being applied in switchgrass andmiscanthus (Table 5) but the future of CRISPR in com-mercial breeding for the European market is uncertain inthe light of recent ECJ 2018 rulings
There is integration of research and plant breeding itselfin all four PBCs Therefore estimating the ongoing costs ofmaintaining these breeding programmes is difficult Invest-ment in research also seeks wider benefits associated withtechnological advances in plant science rather than cultivardevelopment per se However in all cases the conventionalbreeding cycle shown in Figure 1 is the basic ldquoenginerdquo withmolecular technologies (‐omics) serving to accelerate thisengine The history of the development shows that the exis-tence of these breeding programmes is essential to gain ben-efits from the biological research Despite this it is thisessential step that is at most risk from reductions in invest-ment A conventional breeding programme typicallyrequires a breeder and several technicians who are supported
over the long term (20ndash30 years Figure 1) at costs of about05 to 10 million Euro per year (as of 2018) The analysisreported here shows that the time needed to perform onecycle of conventional breeding bringing germplasm fromthe wild to a commercial hybrid ranged from 11 years inswitchgrass to 26 years in poplar (Figure 1) In a maturecrop grown on over 100000 ha with effective cultivar pro-tection and a suitable business model this level of revenuecould come from royalties Until such levels are reachedPBCs lie in the innovation valley of death (Figure 3) andneed public support
Applying industrial ldquotechnology readiness levelsrdquo (TRL)originally developed for aerospace (Heacuteder 2017) to ourplant breeding efforts we estimate many promising hybridscultivars are at TRL levels of 3ndash4 In Table 3 experts ineach crop estimate that it would take 3 years from now toupscale planting material from leading cultivars in plot tri-als to 100 ha
FIGURE 3 A schematic relating some of the steps in the innovation chain from relatively basic crop science research through to thedeployment in commercial cropping systems and value chains The shape of the funnel above the expanding development and deploymentrepresents the availability of investment along the development chain from relatively basic research at the top to the upscaled deployment at thebottom Plant breeding links the research effort with the development of cropping systems The constriction represents the constrained fundingfor breeding that links conventional public research investment and the potential returns from commercial development The handover pointsbetween publicly funded work to develop the germplasm resources (often known as prebreeding) the breeding and the subsequent cropdevelopment are shown on the left The constriction point is aggravated by the lack academic rewards for this essential breeding activity Theoutcome is such that this innovation system is constrained by the precarious resourcing of plant breeding The authorsrsquo assessment ofdevelopment status of the four species is shown (poplar having two one for short‐rotation coppice (SRC) poplar and one for the more traditionalshort‐rotation forestry (SRF)) The four new perennial biomass crops (PBCs) are now in the critical phase of depending of plant breedingprogress without the income stream from a large crop production base
26 | CLIFTON‐BROWN ET AL
Taking the UK example mentioned in the introductionplanting rates of ~35000 ha per year from 2022 onwardsare needed to reach over 1 m ha by 2050 Ongoing workin the UK‐funded Miscanthus Upscaling Technology(MUST) project shows that ramping annual hybrid seedproduction from the current level of sufficient seed for10 ha in 2018 to 35000 ha would take about 5 yearsassuming no setbacks If current hybrids of any of the fourPBCs in the upscaling pipeline fail on any step for exam-ple lower than expected multiplication rates or unforeseenagronomic barriers then further selections from ongoingbreeding are needed to replace earlier candidates
In conclusion the breeding foundations have been laidwell for switchgrass miscanthus willow and poplar owingto public funding over the long time periods necessaryImproved cultivars or genotypes are available that could bescaled up over a few years if real sustained market oppor-tunities emerged in response to sustained favourable poli-cies and industrial market pull The potential contributionsof growing and using these PBCs for socioeconomic andenvironmental benefits are clear but how farmers andothers in commercial value chains are rewarded for mass‐scale deployment as is necessary is not obvious at presentTherefore mass‐scale deployment of these lignocellulosecrops needs developments outside the breeding arenas todrive breeding activities more rapidly and extensively
8 | UNCITED REFERENCE
Huang et al (2019)
ACKNOWLEDGEMENTS
UK The UK‐led miscanthus research and breeding wasmainly supported by the Biotechnology and BiologicalSciences Research Council (BBSRC) Department for Envi-ronment Food and Rural Affairs (Defra) the BBSRC CSPstrategic funding grant BBCSP17301 Innovate UKBBSRC ldquoMUSTrdquo BBN0161491 CERES Inc and Ter-ravesta Ltd through the GIANT‐LINK project (LK0863)Genomic selection and genomewide association study activi-ties were supported by BBSRC grant BBK01711X1 theBBSRC strategic programme grant on Energy Grasses ampBio‐refining BBSEW10963A01 The UK‐led willow RampDwork reported here was supported by BBSRC (BBSEC00005199 BBSEC00005201 BBG0162161 BBE0068331 BBG00580X1 and BBSEC000I0410) Defra(NF0424 NF0426) and the Department of Trade and Indus-try (DTI) (BW6005990000) IT The Brain Gain Program(Rientro dei cervelli) of the Italian Ministry of EducationUniversity and Research supports Antoine Harfouche USContributions by Gerald Tuskan to this manuscript were sup-ported by the Center for Bioenergy Innovation a US
Department of Energy Bioenergy Research Center supportedby the Office of Biological and Environmental Research inthe DOE Office of Science under contract number DE‐AC05‐00OR22725 Willow breeding efforts at CornellUniversity have been supported by grants from the USDepartment of Agriculture National Institute of Food andAgriculture Contributions by the University of Illinois weresupported primarily by the DOE Office of Science Office ofBiological and Environmental Research (BER) grant nosDE‐SC0006634 DE‐SC0012379 and DE‐SC0018420 (Cen-ter for Advanced Bioenergy and Bioproducts Innovation)and the Energy Biosciences Institute EU We would like tofurther acknowledge contributions from the EU projectsldquoOPTIMISCrdquo FP7‐289159 on miscanthus and ldquoWATBIOrdquoFP7‐311929 on poplar and miscanthus as well as ldquoGRACErdquoH2020‐EU326 Biobased Industries Joint Technology Ini-tiative (BBI‐JTI) Project ID 745012 on miscanthus
CONFLICT OF INTEREST
The authors declare that progress reported in this paperwhich includes input from industrial partners is not biasedby their business interests
ORCID
John Clifton‐Brown httporcidorg0000-0001-6477-5452Antoine Harfouche httporcidorg0000-0003-3998-0694Lawrence B Smart httporcidorg0000-0002-7812-7736Danny Awty‐Carroll httporcidorg0000-0001-5855-0775VasileBotnari httpsorcidorg0000-0002-0470-0384Lindsay V Clark httporcidorg0000-0002-3881-9252SalvatoreCosentino httporcidorg0000-0001-7076-8777Lin SHuang httporcidorg0000-0003-3079-326XJon P McCalmont httporcidorg0000-0002-5978-9574Paul Robson httporcidorg0000-0003-1841-3594AnatoliiSandu httporcidorg0000-0002-7900-448XDaniloScordia httporcidorg0000-0002-3822-788XRezaShafiei httporcidorg0000-0003-0891-0730Gail Taylor httporcidorg0000-0001-8470-6390IrisLewandowski httporcidorg0000-0002-0388-4521
REFERENCES
Alburquerque N Baldacci‐Cresp F Baucher M Casacuberta JM Collonnier C ElJaziri M hellip Burgos L (2016) New trans-formation technologies for trees In C Vettori F Gallardo H
CLIFTON‐BROWN ET AL | 27
Haumlggman V Kazana F Migliacci G Pilateamp M Fladung(Eds) Biosafety of forest transgenic trees (pp 31ndash66) DordrechtNetherlands Springer
Argus G W (2007) Salix (Salicaceae) distribution maps and a syn-opsis of their classification in North America north of MexicoHarvard Papers in Botany 12 335ndash368 httpsdoiorg1031001043-4534(2007)12[335SSDMAA]20CO2
Atienza S G Ramirez M C amp Martin A (2003) Mapping‐QTLscontrolling flowering date in Miscanthus sinensis Anderss CerealResearch Communications 31 265ndash271
Atienza S G Satovic Z Petersen K K Dolstra O amp Martin A(2003) Identification of QTLs influencing agronomic traits inMiscanthus sinensis Anderss I Total height flag‐leaf height andstem diameter Theoretical and Applied Genetics 107 123ndash129
Atienza S G Satovic Z Petersen K K Dolstra O amp Martin A(2003) Influencing combustion quality in Miscanthus sinensisAnderss Identification of QTLs for calcium phosphorus and sul-phur content Plant Breeding 122 141ndash145
Bdeir R Muchero W Yordanov Y Tuskan G A Busov V ampGailing O (2017) Quantitative trait locus mapping of Populusbark features and stem diameter BMC Plant Biology 17 224httpsdoiorg101186s12870-017-1166-4
Beard T R Ford G S Koutsky T M amp Spiwak L J (2009) AValley of Death in the innovation sequence An economic investi-gation Research Evaluation 18 343ndash356 httpsdoiorg103152095820209X481057
Berguson W McMahon B amp Riemenschneider D (2017) Addi-tive and non‐additive genetic variances for tree growth in severalhybrid poplar populations and implications regarding breedingstrategy Silvae Genetica 66(1) 33ndash39 httpsdoiorg101515sg-2017-0005
Berlin S Trybush S O Fogelqvist J Gyllenstrand N Halling-baumlck H R Aringhman I hellip Hanley S J (2014) Genetic diversitypopulation structure and phenotypic variation in European Salixviminalis L (Salicaceae) Tree Genetics amp Genomes 10 1595ndash1610 httpsdoiorg101007s11295-014-0782-5
Bernardo R amp Yu J (2007) Prospects for genomewide selectionfor quantitative traits in maize Crop Science 47 1082ndash1090httpsdoiorg102135cropsci2006110690
Biswal A K Atmodjo M A Li M Baxter H L Yoo C GPu Y hellip Mohnen D (2018) Sugar release and growth of bio-fuel crops are improved by downregulation of pectin biosynthesisNature Biotechnology 36 249ndash257
Bmu (2009) National biomass action plan for Germany (p 17) Bun-desministerium fuumlr Umwelt Naturschutz und ReaktorsicherheitBundesministerium fuumlr Ernaumlhrung (BMU) amp Landwirtschaft undVerbraucherschutz (BMELV) 11055 Berlin | Germany Retrievedfrom httpwwwbmeldeSharedDocsDownloadsENPublicationsBiomassActionPlanpdf__blob=publicationFile
Brummer E C (1999) Capturing heterosis in forage crop cultivardevelopment Crop Science 39 943ndash954 httpsdoiorg102135cropsci19990011183X003900040001x
Budsberg E Crawford J T Morgan H Chin W S Bura R ampGustafson R (2016) Hydrocarbon bio‐jet fuel from bioconver-sion of poplar biomass Life cycle assessment Biotechnology forBiofuels 9 170 httpsdoiorg101186s13068-016-0582-2
Burris K P Dlugosz E M Collins A G Stewart C N amp Lena-ghan S C (2016) Development of a rapid low‐cost protoplasttransfection system for switchgrass (Panicum virgatum L) Plant
Cell Reports 35 693ndash704 httpsdoiorg101007s00299-015-1913-7
Casler M (2012) Switchgrass breeding genetics and genomics InA Monti (Ed) Switchgrass (pp 29ndash54) London UK Springer
Casler M Mitchell R amp Vogel K (2012) Switchgrass In CKole C P Joshi amp D R Shonnard (Eds) Handbook of bioen-ergy crop plants Vol 2 New York NY Taylor amp Francis
Casler M D amp Ramstein G P (2018) Breeding for biomass yieldin switchgrass using surrogate measures of yield BioEnergyResearch 11 6ndash12
Casler M D Sosa S Hoffman L Mayton H Ernst C Adler PR hellip Bonos S A (2017) Biomass Yield of Switchgrass Culti-vars under High‐ versus Low‐Input Conditions Crop Science 57821ndash832 httpsdoiorg102135cropsci2016080698
Casler M D amp Vogel K P (2014) Selection for biomass yield inupland lowland and hybrid switchgrass Crop Science 54 626ndash636 httpsdoiorg102135cropsci2013040239
Casler M D Vogel K P amp Harrison M (2015) Switchgrassgermplasm resources Crop Science 55 2463ndash2478 httpsdoiorg102135cropsci2015020076
Casler M D Vogel K P Lee D K Mitchell R B Adler P RSulc R M hellip Moore K J (2018) 30 Years of progress towardincreasing biomass yield of switchgrass and big bluestem CropScience 58 1242
Clark L V Brummer J E Głowacka K Hall M C Heo K PengJ hellip Sacks E J (2014) A footprint of past climate change on thediversity and population structure of Miscanthus sinensis Annals ofBotany 114 97ndash107 httpsdoiorg101093aobmcu084
Clark L V Dzyubenko E Dzyubenko N Bagmet L SabitovA Chebukin P hellip Sacks E J (2016) Ecological characteristicsand in situ genetic associations for yield‐component traits of wildMiscanthus from eastern Russia Annals of Botany 118 941ndash955
Clark L V Jin X Petersen K K Anzoua K G Bagmet LChebukin P hellip Sacks E J(2018) Population structure of Mis-canthus sacchariflorus reveals two major polyploidization eventstetraploid‐mediated unidirectional introgression from diploid Msinensis and diversity centred around the Yellow Sea Annals ofBotany 20 1ndash18 httpsdoiorg101093aobmcy1161
Clark L V Stewart J R Nishiwaki A Toma Y Kjeldsen J BJoslashrgensen U hellip Sacks E J (2015) Genetic structure ofMiscanthus sinensis and Miscanthus sacchariflorus in Japan indi-cates a gradient of bidirectional but asymmetric introgressionJournal of Experimental Botany 66 4213ndash4225
Clifton‐Brown J Hastings A Mos M McCalmont J P Ashman CAwty‐Carroll DhellipCracroft‐EleyW (2017) Progress in upscalingMis-canthus biomass production for the European bio‐economy with seed‐based hybridsGlobal Change Biology Bioenergy 9 6ndash17
Clifton‐Brown J Schwarz K U amp Hastings A (2015) History ofthe development of Miscanthus as a bioenergy crop From smallbeginnings to potential realisation Biology and Environment‐Pro-ceedings of the Royal Irish Academy 115B 45ndash57
Clifton‐Brown J Senior H Purdy S Horsnell R Lankamp BMuumlennekhoff A-K hellip Bentley A R (2018) Investigating thepotential of novel non‐woven fabrics to increase pollination effi-ciency in plant breeding PloS One (Accepted)
Crawford J T Shan CW Budsberg E Morgan H Bura R ampGus-tafson R (2016) Hydrocarbon bio‐jet fuel from bioconversion ofpoplar biomass Techno‐economic assessment Biotechnology forBiofuels 9 141 httpsdoiorg101186s13068-016-0545-7
28 | CLIFTON‐BROWN ET AL
Dalton S (2013) Biotechnology of Miscanthus In S M Jain amp SD Gupta (Eds) Biotechnology of neglected and underutilizedcrops (pp 243ndash294) Dordrecht Netherlands Springer
Davey C L Robson P Hawkins S Farrar K Clifton‐Brown J CDonnison I S amp Slavov G T (2017) Genetic relationshipsbetween spring emergence canopy phenology and biomass yieldincrease the accuracy of genomic prediction in Miscanthus Journalof Experimental Botany 68 5093ndash5102 httpsdoiorg101093jxberx339
David X amp Anderson X (2002) Aspen and larch genetics cooper-ative annual report 13 St Paul MN Department of ForestResources University of Minnesota
Deblock M (1990) Factors influencing the tissue‐culture and theAgrobacterium‐Tumefaciens‐mediated transformation of hybridaspen and poplar clones Plant Physiology 93 1110ndash1116httpsdoiorg101104pp9331110
Defra (2002) The role of future public research investment in thegenetic improvement of UK‐grown crops (p 220) LondonRetrieved from sciencesearchdefragovukDefaultaspxMenu=MenuampModule=MoreampLocation=NoneampCompleted=220ampProjectID=10412
Dewoody J Trewin H amp Taylor G (2015) Genetic and morpho-logical differentiation in Populus nigra L Isolation by coloniza-tion or isolation by adaptation Molecular Ecology 24 2641ndash2655
Dong H Liu S Clark L V Sharma S Gifford J M Juvik JA hellip Sacks E J (2018) Genetic mapping of biomass yield inthree interconnected Miscanthus populations Global Change Biol-ogy Bioenergy 10 165ndash185
Dukes (2017) Digest of UK Energy Statistics (DUKES) (p 264) Lon-don UK Department for Business Energy amp Industrial StrategyRetrieved from wwwgovukgovernmentcollectionsdigest-of-uk-energy-statistics-dukes
Eckenwalder J (1996) Systematics and evolution of Populus In RStettler H Bradshaw J Heilman amp T Hinckley (Eds) Biologyof Populus and its implications for management and conservation(pp 7‐32) Ottawa ON National Research Council of Canada
Elshire R J Glaubitz J C Sun Q Poland J A Kawamoto KBuckler E S amp Mitchell S E (2011) A robust simple geno-typing‐by‐sequencing (GBS) approach for high diversity speciesPloS One 6 e19379
Evans H (2016) Bioenergy crops in the UK Case studies of suc-cessful whole farm integration (p 23) Loughborough UKEnergy Technologies Institute Retrieved from httpswwweticouklibrarybioenergy-crops-in-the-uk-case-studies-on-successful-whole-farm-integration-evidence-pack
Evans H (2017) Increasing UK biomass production through moreproductive use of land (p 7) Loughborough UK Energy Tech-nologies Institute Retrieved from httpswwweticouklibraryan-eti-perspective-increasing-uk-biomass-production-through-more-productive-use-of-land
Evans J Sanciangco M D Lau K H Crisovan E Barry KDaum C hellip Buell C R (2018) Extensive genetic diversity ispresent within North American switchgrass germplasm The PlantGenome 11 1ndash16 httpsdoiorg103835plantgenome2017060055
Evans L M Slavov G T Rodgers‐Melnick E Martin J RanjanP Muchero W hellip DiFazio S P (2014) Population genomicsof Populus trichocarpa identifies signatures of selection and
adaptive trait associations Nature Genetics 46 1089ndash1096httpsdoiorg101038ng3075
Fabio E S Volk T A Miller R O Serapiglia M J Gauch HG Van Rees K C J hellip Smart L B (2017) Genotypetimes envi-ronment interaction analysis of North American shrub willowyield trials confirms superior performance of triploid hybrids Glo-bal Change Biology Bioenergy 9 445ndash459 httpsdoiorg101111gcbb12344
Fahrenkrog A M Neves L G Resende M F R Vazquez A Ide los Campos G Dervinis C hellip Kirst M (2017) Genome‐wide association study reveals putative regulators of bioenergytraits in Populus deltoides New Phytologist 213 799ndash811
Fan D Liu T T Li C F Jiao B Li S Hou Y S amp Luo KM (2015) Efficient CRISPRCas9‐mediated targeted mutagenesisin Populus in the first generation Scientific Reports 5 12217
Feng X P Lourgant K Castric V Saumitou-Laprade P ZhengB S Jiang D M amp Brancourt-Hulmel M (2014) The discov-ery of natural accessions related to Miscanthus x giganteus usingchloroplast DNA Crop Science 54 1645ndash1655
Fillatti J J Sellmer J Mccown B Haissig B amp Comai L(1987) Agrobacterium-mediated transformation and regenerationof Populus Molecular amp General Genetics 206 192ndash199
Fu Y B (2015) Understanding crop genetic diversity under modernplant breeding Theoretical and Applied Genetics 128 2131ndash2142 httpsdoiorg101007s00122-015-2585-y
Fu C Mielenz J R Xiao X Ge Y Hamilton C Y RodriguezM hellip Wang Z‐Y (2011) Genetic manipulation of ligninreduces recalcitrance and improves ethanol production fromswitchgrass Proceedings of the National Academy of Sciences ofthe United States of America 108 3803ndash3808 httpsdoiorg101073pnas1100310108
Gao C (2018) The future of CRISPR technologies in agricultureNature Reviews Molecular Cell Biology 19(5) 275ndash276
Gegas V Gay A Camargo A amp Doonan J (2014) Challengesof Crop Phenomics in the Post-genomic Era In J Hancock (Ed)Phenomics (pp 142ndash171) Boca Raton FL CRC Press
Geraldes A DiFazio S P Slavov G T Ranjan P Muchero WHannemann J hellip Tuskan G A (2013) A 34K SNP genotypingarray for Populus trichocarpa Design application to the study ofnatural populations and transferability to other Populus speciesMolecular Ecology Resources 13 306ndash323
Gifford J M Chae W B Swaminathan K Moose S P amp JuvikJ A (2015) Mapping the genome of Miscanthus sinensis forQTL associated with biomass productivity Global Change Biol-ogy Bioenergy 7 797ndash810
Goggin F L Lorence A amp Topp C N (2015) Applying high‐throughput phenotyping to plant‐insect interactions Picturingmore resistant crops Current Opinion in Insect Science 9 69ndash76httpsdoiorg101016jcois201503002
Grabowski P P Evans J Daum C Deshpande S Barry K WKennedy M hellip Casler M D (2017) Genome‐wide associationswith flowering time in switchgrass using exome‐capture sequenc-ing data New Phytologist 213 154ndash169 httpsdoiorg101111nph14101
Greef J M amp Deuter M (1993) Syntaxonomy of Miscanthus xgiganteus GREEF et DEU Angewandte Botanik 67 87ndash90
Guerra F P Richards J H Fiehn O Famula R Stanton B JShuren R hellip Neale D B (2016) Analysis of the genetic varia-tion in growth ecophysiology and chemical and metabolomic
CLIFTON‐BROWN ET AL | 29
composition of wood of Populus trichocarpa provenances TreeGenetics amp Genomes 12 6 httpsdoiorg101007s11295-015-0965-8
Guo H P Shao R X Hong C T Hu H K Zheng B S ampZhang Q X (2013) Rapid in vitro propagation of bioenergy cropMiscanthus sacchariflorus Applied Mechanics and Materials 260181ndash186
Hallingbaumlck H R Fogelqvist J Powers S J Turrion‐Gomez JRossiter R Amey J hellip Roumlnnberg‐Waumlstljung A‐C (2016)Association mapping in Salix viminalis L (Salicaceae)ndashIdentifica-tion of candidate genes associated with growth and phenologyGlobal Change Biology Bioenergy 8 670ndash685
Hanley S J amp Karp A (2014) Genetic strategies for dissectingcomplex traits in biomass willows (Salix spp) Tree Physiology34 1167ndash1180 httpsdoiorg101093treephystpt089
Harfouche A Meilan R amp Altman A (2011) Tree genetic engi-neering and applications to sustainable forestry and biomass pro-duction Trends in Biotechnology 29 9ndash17 httpsdoiorg101016jtibtech201009003
Harfouche A Meilan R amp Altman A (2014) Molecular and phys-iological responses to abiotic stress in forest trees and their rele-vance to tree improvement Tree Physiology 34 1181ndash1198httpsdoiorg101093treephystpu012
Harfouche A Meilan R Kirst M Morgante M Boerjan WSabatti M amp Mugnozza G S (2012) Accelerating the domesti-cation of forest trees in a changing world Trends in PlantScience 17 64ndash72 httpsdoiorg101016jtplants201111005
Hastings A Clifton‐Brown J Wattenbach M Mitchell C PStampfl P amp Smith P (2009) Future energy potential of Mis-canthus in Europe Global Change Biology Bioenergy 1 180ndash196
Hastings A Mos M Yesufu J AMccalmont J Schwarz KShafei RhellipClifton-Brown J (2017) Economic and environmen-tal assessment of seed and rhizome propagated Miscanthus in theUK Frontiers in Plant Science 8 16 httpsdoiorg103389fpls201701058
Heacuteder M (2017) From NASA to EU The evolution of the TRLscale in Public Sector Innovation The Innovation Journal 22 1ndash23 httpsdoiorg103389fpls201701058
Heffner E L Sorrells M E amp Jannink J L (2009) Genomicselection for crop improvement Crop Science 49 1ndash12 httpsdoiorg102135cropsci2008080512
Hodkinson T R Klaas M Jones M B Prickett R amp Barth S(2015) Miscanthus A case study for the utilization of naturalgenetic variation Plant Genetic Resources‐Characterization andUtilization 13 219ndash237 httpsdoiorg101017S147926211400094X
Hodkinson T R amp Renvoize S (2001) Nomenclature of Miscant-hus xgiganteus (Poaceae) Kew Bulletin 56 759ndash760 httpsdoiorg1023074117709
Houle D Govindaraju D R amp Omholt S (2010) Phenomics Thenext challenge Nature Reviews Genetics 11 855ndash866 httpsdoiorg101038nrg2897
Huang X H Wei X H Sang T Zhao Q Feng Q Zhao Yhellip Han B (2010) Genome‐wide association studies of 14 agro-nomic traits in rice landraces Nature Genetics 42 961ndashU976
Hwang O‐J Cho M‐A Han Y‐J Kim Y‐M Lim S‐H KimD‐S hellip Kim J‐I (2014) Agrobacterium‐mediated genetic trans-formation of Miscanthus sinensis Plant Cell Tissue and OrganCulture (PCTOC) 117 51ndash63
Hwang O‐J Lim S‐H Han Y‐J Shin A‐Y Kim D‐S ampKim J‐I (2014) Phenotypic characterization of transgenic Mis-canthus sinensis plants overexpressing Arabidopsis phytochromeB International Journal of Photoenergy 2014501016
Ilstedt B (1996) Genetics and performance of Belgian poplar clonestested in Sweden International Journal of Forest Genetics 3183ndash195
Ingvarsson P K Garcia M V Hall D Luquez V amp Jansson S(2006) Clinal variation in phyB2 a candidate gene for day‐length‐induced growth cessation and bud set across a latitudinalgradient in European aspen (Populus tremula) Genetics 1721845ndash1853
Ingvarsson P K Garcia M V Luquez V Hall D amp Jansson S(2008) Nucleotide polymorphism and phenotypic associationswithin and around the phytochrome B2 locus in European aspen(Populus tremula Salicaceae) Genetics 178 2217ndash2226 httpsdoiorg101534genetics107082354
Jannink J‐L Lorenz A J amp Iwata H (2010) Genomic selectionin plant breeding From theory to practice Briefings in FunctionalGenomics 9 166ndash177 httpsdoiorg101093bfgpelq001
Jiang J Guan Y Mccormick S Juvik J Lubberstedt T amp FeiS‐Z (2017) Gametophytic self‐incompatibility is operative inMiscanthus sinensis (Poaceae) and is affected by pistil age CropScience 57 1948ndash1956
Jones H D (2015a) Future of breeding by genome editing is in thehands of regulators GM Crops amp Food 6 223ndash232
Jones H D (2015b) Regulatory uncertainty over genome editingNature Plants 1 14011
Kalinina O Nunn C Sanderson R Hastings A F S van der Weijde TOumlzguumlven MhellipClifton‐Brown J C (2017) ExtendingMiscanthus cul-tivation with novel germplasm at six contrasting sites Frontiers in PlantScience 8563 httpsdoiorg103389fpls201700563
Karp A Hanley S J Trybush S O Macalpine W Pei M ampShield I (2011) Genetic improvement of willow for bioenergyand biofuels free access Journal of Integrative Plant Biology 53151ndash165 httpsdoiorg101111j1744-7909201001015x
Kersten B Faivre Rampant P Mader M Le Paslier M‐C Bou-non R Berard A hellip Fladung M (2016) Genome sequences ofPopulus tremula chloroplast and mitochondrion Implications forholistic poplar breeding PloS One 11 e0147209 httpsdoiorg101371journalpone0147209
Kiesel A Nunn C Iqbal Y Van der Weijde T Wagner MOumlzguumlven M hellip Lewandowski I (2017) Site‐specific manage-ment of Miscanthus genotypes for combustion and anaerobicdigestion A comparison of energy yields Frontiers in PlantScience 8 347 httpsdoiorg103389fpls201700347
Kim H Kim S T Ryu J Kang B C Kim J S amp Kim S G(2017) CRISPRCpf1‐mediated DNA‐free plant genome editingNature Communications 8 14406 httpsdoiorg101038ncomms14406
King Z R Bray A L Lafayette P R amp Parrott W A (2014)Biolistic transformation of elite genotypes of switchgrass (Pan-icum virgatum L) Plant Cell Reports 33 313ndash322 httpsdoiorg101007s00299-013-1531-1
Kopp R F Maynard C A De Niella P R Smart L B amp Abra-hamson L P (2002) Collection and storage of pollen from Salix(Salicaceae) American Journal of Botany 89 248ndash252
Krzyżak J Pogrzeba M Rusinowski S Clifton‐Brown J McCal-mont J P Kiesel A hellip Mos M (2017) Heavy metal uptake
30 | CLIFTON‐BROWN ET AL
by novel miscanthus seed‐based hybrids cultivated in heavy metalcontaminated soil Civil and Environmental Engineering Reports26 121ndash132 httpsdoiorg101515ceer-2017-0040
Kuzovkina Y A amp Quigley M F (2005) Willows beyond wet-lands Uses of Salix L species for environmental projects WaterAir and Soil Pollution 162 183ndash204 httpsdoiorg101007s11270-005-6272-5
Lazarus W Headlee W L amp Zalesny R S (2015) Impacts of sup-plyshed‐level differences in productivity and land costs on the eco-nomics of hybrid poplar production in Minnesota USA BioEnergyResearch 8 231ndash248 httpsdoiorg101007s12155-014-9520-y
Lewandowski I (2015) Securing a sustainable biomass supply in agrowing bioeconomy Global Food Security 6 34ndash42 httpsdoiorg101016jgfs201510001
Lewandowski I (2016) The role of perennial biomass crops in agrowing bioeconomy In S Barth D Murphy‐Bokern O Kalin-ina G Taylor amp M Jones (Eds) Perennial biomass crops for aresource‐constrained world (pp 3ndash13) Cham SwitzerlandSpringer Switzerland AG
Li J L Meilan R Ma C Barish M amp Strauss S H (2008)Stability of herbicide resistance over 8 years of coppice in field‐grown genetically engineered poplars Western Journal of AppliedForestry 23 89ndash93
Li R Y amp Qu R D (2011) High throughput Agrobacterium‐medi-ated switchgrass transformation Biomass amp Bioenergy 35 1046ndash1054 httpsdoiorg101016jbiombioe201011025
Lindegaard K N amp Barker J H A (1997) Breeding willows forbiomass Aspects of Applied Biology 49 155ndash162
Linde‐Laursen I B (1993) Cytogenetic analysis of MiscanthusGiganteus an interspecific hybrid Hereditas 119 297ndash300httpsdoiorg101111j1601-5223199300297x
Liu Y Merrick P Zhang Z Z Ji C H Yang B amp Fei S Z(2018) Targeted mutagenesis in tetraploid switchgrass (Panicumvirgatum L) using CRISPRCas9 Plant Biotechnology Journal16 381ndash393
Liu L L Wu Y Q Wang Y W amp Samuels T (2012) A high‐density simple sequence repeat‐based genetic linkage map ofswitchgrass G3 (Bethesda Md) 2 357ndash370
Lovett A Suumlnnenberg G amp Dockerty T (2014) The availabilityof land for perennial energy crops in Great Britain GlobalChange Biology Bioenergy 6 99ndash107 httpsdoiorg101111gcbb12147
Lozano‐Juste J amp Cutler S R (2014) Plant genome engineering infull bloom Trends in Plant Science 19 284ndash287 httpsdoiorg101016jtplants201402014
Lu F Lipka A E Glaubitz J Elshire R Cherney J H CaslerM D hellip Costich D E (2013) Switchgrass Genomic DiversityPloidy and Evolution Novel Insights from a Network‐Based SNPDiscovery Protocol Plos Genetics 9 e1003215
Ludovisi R Tauro F Salvati R Khoury S Mugnozza G S ampHarfouche A (2017) UAV‐based thermal imaging for high‐throughput field phenotyping of black poplar response to droughtFrontiers in Plant Science 81681
Macalpine W Shield I amp Karp A (2010) Seed to near market vari-ety the BEGIN willow breeding pipeline 2003ndash2010 and beyond InBioten conference proceedings Birmingham pp 21ndash23
Macalpine W J Shield I F Trybush S O Hayes C M ampKarp A (2008) Overcoming barriers to crossing in willow (Salixspp) breeding Aspects of Applied Biology 90 173ndash180
Mahfouz M M (2017) The efficient tool CRISPR‐Cpf1 NaturePlants 3 17028
Malinowska M Donnison I S amp Robson P R (2017) Phenomicsanalysis of drought responses in Miscanthus collected from differ-ent geographical locations Global Change Biology Bioenergy 978ndash91
Maroder H L Prego I A Facciuto G R amp Maldonado S B(2000) Storage behaviour of Salix alba and Salix matsudanaseeds Annals of Botany 86 1017ndash1021 httpsdoiorg101006anbo20001265
Martinez‐Reyna J amp Vogel K (2002) Incompatibility systems inswitchgrass Crop Science 42 1800ndash1805 httpsdoiorg102135cropsci20021800
Maurya J P amp Bhalerao R P (2017) Photoperiod‐and tempera-ture‐mediated control of growth cessation and dormancy in treesA molecular perspective Annals of Botany 120 351ndash360httpsdoiorg101093aobmcx061
McCalmont J P Hastings A Mcnamara N P Richter G MRobson P Donnison I S amp Clifton‐Brown J (2017) Environ-mental costs and benefits of growing Miscanthus for bioenergy inthe UK Global Change Biology Bioenergy 9 489ndash507
McCracken A R amp Dawson W M (1997) Using mixtures of wil-low clones as a means of controlling rust disease Aspects ofApplied Biology 49 97ndash103
McKown A D Klaacutepště J Guy R D Geraldes A Porth I Hanne-mann JhellipDouglas C J (2014) Genome‐wide association implicatesnumerous genes underlying ecological trait variation in natural popula-tions ofPopulus trichocarpaNew Phytologist 203 535ndash553
Merrick P amp Fei S Z (2015) Plant regeneration and genetic transforma-tion in switchgrass ndash A review Journal of Integrative Agriculture 14483ndash493 httpsdoiorg101016S2095-3119(14)60921-7
Moreno‐Mateos M A Fernandez J P Rouet R Lane M AVejnar C E Mis E hellip Giraldez A J (2017) CRISPR‐Cpf1mediates efficient homology‐directed repair and temperature‐con-trolled genome editing Nature Communications 8 2024 httpsdoiorg101038s41467-017-01836-2
Mosseler A (1990) Hybrid performance and species crossabilityrelationships in willows (Salix) Canadian Journal of Botany 682329ndash2338
Muchero W Guo J DiFazio S P Chen J‐G Ranjan P Slavov GT hellip Tuskan G A (2015) High‐resolution genetic mapping ofallelic variants associated with cell wall chemistry in Populus BMCGenomics 16 24 httpsdoiorg101186s12864-015-1215-z
Neale D B amp Kremer A (2011) Forest tree genomics Growingresources and applications Nature Reviews Genetics 12 111ndash122 httpsdoiorg101038nrg2931
Nunn C Hastings A F S J Kalinina O Oumlzguumlven M SchuumlleH Tarakanov I G hellip Clifton‐Brown J C (2017) Environmen-tal influences on the growing season duration and ripening ofdiverse Miscanthus germplasm grown in six countries Frontiersin Plant Science 8 907 httpsdoiorg103389fpls201700907
Ogawa Y Honda M Kondo Y amp Hara‐Nishimura I (2016) Anefficient Agrobacterium‐mediated transformation method forswitchgrass genotypes using Type I callus Plant Biotechnology33 19ndash26
Ogawa Y Shirakawa M Koumoto Y Honda M Asami Y KondoY ampHara‐Nishimura I (2014) A simple and reliable multi‐gene trans-formation method for switchgrass Plant Cell Reports 33 1161ndash1172httpsdoiorg101007s00299-014-1605-8
CLIFTON‐BROWN ET AL | 31
Okada M Lanzatella C Saha M C Bouton J Wu R L amp Tobias CM (2010) Complete switchgrass genetic maps reveal subgenomecollinearity preferential pairing and multilocus interactions Genetics185 745ndash760 httpsdoiorg101534genetics110113910
Palomo‐Riacuteos E Macalpine W Shield I Amey J Karaoğlu C WestJ hellip Jones H D (2015) Efficient method for rapid multiplication ofclean and healthy willow clones via in vitro propagation with broadgenotype applicability Canadian Journal of Forest Research 451662ndash1667 httpsdoiorg101139cjfr-2015-0055
Pilate G Allona I Boerjan W Deacutejardin A Fladung M Gal-lardo F hellip Halpin C (2016) Lessons from 25 years of GM treefield trials in Europe and prospects for the future In C Vettori FGallardo H Haumlggman V Kazana F Migliacci G Pilateamp MFladung (Eds) Biosafety of forest transgenic trees (pp 67ndash100)Dordrecht Netherlands Springer Switzerland AG
Pinosio S Giacomello S Faivre-Rampant P Taylor G Jorge VLe Paslier M C amp Marroni F (2016) Characterization of thepoplar pan-genome by genome-wide identification of structuralvariation Molecular Biology and Evolution 33 2706ndash2719
Porth I Klaacutepště J Skyba O Friedmann M C Hannemann JEhlting J hellip Douglas C J (2013) Network analysis reveals therelationship among wood properties gene expression levels andgenotypes of natural Populus trichocarpa accessions New Phytol-ogist 200 727ndash742
Pugesgaard S Schelde K Larsen S U Laeligrke P E amp JoslashrgensenU (2015) Comparing annual and perennial crops for bioenergyproductionndashinfluence on nitrate leaching and energy balance Glo-bal Change Biology Bioenergy 7 1136ndash1149 httpsdoiorg101111gcbb12215
Quinn L D Allen D J amp Stewart J R (2010) Invasivenesspotential of Miscanthus sinensis Implications for bioenergy pro-duction in the United States Global Change Biology Bioenergy2 310ndash320 httpsdoiorg101111j1757-1707201001062x
Rae A M Robinson K M Street N R amp Taylor G (2004)Morphological and physiological traits influencing biomass pro-ductivity in short‐rotation coppice poplar Canadian Journal ofForest Research‐Revue Canadienne De Recherche Forestiere 341488ndash1498 httpsdoiorg101139x04-033
Rambaud C Arnoult S Bluteau A Mansard M C Blassiau Camp Brancourt‐Hulmel M (2013) Shoot organogenesis in threeMiscanthus species and evaluation for genetic uniformity usingAFLP analysis Plant Cell Tissue and Organ Culture (PCTOC)113 437ndash448 httpsdoiorg101007s11240-012-0284-9
Ramstein G P Evans J Kaeppler S M Mitchell R B Vogel K PBuell C R amp Casler M D (2016) Accuracy of genomic prediction inswitchgrass (Panicum virgatum L) improved by accounting for linkagedisequilibriumG3 (Bethesda Md) 6 1049ndash1062
Rayburn A L Crawford J Rayburn C M amp Juvik J A (2009)Genome size of three Miscanthus species Plant Molecular Biol-ogy Reporter 27 184 httpsdoiorg101007s11105-008-0070-3
Richardson J Isebrands J amp Ball J (2014) Ecology and physiol-ogy of poplars and willows In J Isebrands amp J Richardson(Eds) Poplars and willows Trees for society and the environ-ment (pp 92‐123) Boston MA and Rome CABI and FAO
Robison T L Rousseau R J amp Zhang J (2006) Biomass produc-tivity improvement for eastern cottonwood Biomass amp Bioenergy30 735ndash739 httpsdoiorg101016jbiombioe200601012
Robson P R Farrar K Gay A P Jensen E F Clifton‐Brown JC amp Donnison I S (2013) Variation in canopy duration in the
perennial biofuel crop Miscanthus reveals complex associationswith yield Journal of Experimental Botany 64 2373ndash2383
Robson P Jensen E Hawkins S White S R Kenobi K Clif-ton‐Brown J hellip Farrar K (2013) Accelerating the domestica-tion of a bioenergy crop Identifying and modelling morphologicaltargets for sustainable yield increase in Miscanthus Journal ofExperimental Botany 64 4143ndash4155
Sanderson M A Adler P R Boateng A A Casler M D ampSarath G (2006) Switchgrass as a biofuels feedstock in theUSA Canadian Journal of Plant Science 86 1315ndash1325httpsdoiorg104141P06-136
Scarlat N Dallemand J F Monforti‐Ferrario F amp Nita V(2015) The role of biomass and bioenergy in a future bioecon-omy Policies and facts Environmental Development 15 3ndash34httpsdoiorg101016jenvdev201503006
Serapiglia M J Gouker F E amp Smart L B (2014) Early selec-tion of novel triploid hybrids of shrub willow with improved bio-mass yield relative to diploids BMC Plant Biology 14 74httpsdoiorg1011861471-2229-14-74
Serba D Wu L Daverdin G Bahri B A Wang X Kilian Ahellip Devos K M (2013) Linkage maps of lowland and upland tet-raploid switchgrass ecotypes BioEnergy Research 6 953ndash965httpsdoiorg101007s12155-013-9315-6
Shakoor N Lee S amp Mockler T C (2017) High throughput phe-notyping to accelerate crop breeding and monitoring of diseases inthe field Current Opinion in Plant Biology 38 184ndash192 httpsdoiorg101016jpbi201705006
Shield I Macalpine W Hanley S amp Karp A (2015) Breedingwillow for short rotation coppice energy cropping In Industrialcrops Breeding for BioEnergy and Bioproducts (pp 67ndash80) NewYork NY Springer
Sjodin A Street N R Sandberg G Gustafsson P amp Jansson S (2009)The Populus Genome Integrative Explorer (PopGenIE) A new resourcefor exploring the Populus genomeNewPhytologist 182 1013ndash1025
Slavov G amp Davey C (2017) Integrated genomic prediction inbioenergy crops In Abstracts from the International Conferenceon Developing biomass crops for future climates pp 34 24ndash27September 2017 Oriel College Oxford wwwwatbioeu
Slavov G Davey C Bosch M Robson P Donnison I ampMackay I (2018a) Genomic index selection provides a pragmaticframework for setting and refining multi‐objective breeding targetsin Miscanthus Annals of Botany (in press)
Slavov G T DiFazio S P Martin J Schackwitz W MucheroW Rodgers‐Melnick E hellip Tuskan G A (2012) Genome rese-quencing reveals multiscale geographic structure and extensivelinkage disequilibrium in the forest tree Populus trichocarpa NewPhytologist 196 713ndash725
Slavov G Davey C Robson P Donnison I amp Mackay I(2018b) Domestication of Miscanthus through genomic indexselection In Plant and Animal Genome Conference 13 January2018 San Diego CA Retrieved from httpspagconfexcompagxxvimeetingappcgiPaper30622
Slavov G T Nipper R Robson P Farrar K Allison G GBosch M hellip Jensen E (2013) Genome‐wide association studiesand prediction of 17 traits related to phenology biomass and cellwall composition in the energy grass Miscanthus sinensis NewPhytologist 201 1227ndash1239
Slavov G Robson P Jensen E Hodgson E Farrar K AllisonG hellip Donnison I (2014) Contrasting geographic patterns of
32 | CLIFTON‐BROWN ET AL
genetic variation for molecular markers vs phenotypic traits in theenergy grass Miscanthus sinensis Global Change Biology Bioen-ergy 5 562ndash571
Ślusarkiewicz‐Jarzina A Ponitka A Cerazy‐Waliszewska J Woj-ciechowicz M K Sobańska K Jeżowski S amp Pniewski T(2017) Effective and simple in vitro regeneration system of Mis-canthus sinensis Mtimes giganteus and M sacchariflorus for plant-ing and biotechnology purposes Biomass and Bioenergy 107219ndash226 httpsdoiorg101016jbiombioe201710012
Smart L amp Cameron K (2012) Shrub willow In C Kole S Joshiamp D Shonnard (Eds) Handbook of bioenergy crop plants (pp687ndash708) Boca Raton FL Taylor and Francis Group
Stanton B J (2014) The domestication and conservation of Populusand Salix genetic resources (Chapter 4) In Isebrands J ampRichardson J (Eds) Poplars and willows Trees for society andthe environment (pp 124ndash200) Rome Italy CAB InternationalFood and Agricultural Organization of the United Nations
Stanton B J Neale D B amp Li S (2010) Populus breeding Fromthe classical to the genomic approach In S Jansson R Bha-lerao amp A T Groover (Eds) Genetics and genomics of popu-lus (pp 309ndash348) New York NY Springer
Stewart J R Toma Y Fernandez F G Nishiwaki A Yamada T ampBollero G (2009) The ecology and agronomy ofMiscanthus sinensisa species important to bioenergy crop development in its native range inJapan A reviewGlobal Change Biology Bioenergy 1 126ndash153
Stott K G (1992) Willows in the service of man Proceedings of the RoyalSociety of Edinburgh Section B Biological Sciences 98 169ndash182
Strauss S H Ma C Ault K amp Klocko A L (2016) Lessonsfrom two decades of field trials with genetically modified trees inthe USA Biology and regulatory compliance In C Vettori FGallardo H Haumlggman V Kazana F Migliacci G Pilate amp MFladung (Eds) Biosafety of forest transgenic trees (pp 101ndash124)Dordrecht Netherlands Springer
Suda Y amp Argus G W (1968) Chromosome numbers of some NorthAmerican Salix Brittonia 20 191ndash197 httpsdoiorg1023072805440
Tan B (2018) Genomic selection and genome‐wide association studies todissect quantitative traits in forest trees Umearing Sweden University
Tornqvist C Taylor M Jiang Y Evans J Buell C KaepplerS amp Casler M (2018) Quantitative trait locus mapping for flow-ering time in a lowland x upland switchgrass pseudo‐F2 popula-tion The Plant Genome 11 170093
Toacuteth G Hermann T Da Silva M amp Montanarella L (2016)Heavy metals in agricultural soils of the European Union withimplications for food safety Environment International 88 299ndash309 httpsdoiorg101016jenvint201512017
Tracy W Dawson J Moore V amp Fisch J (2016) Intellectual propertyrights and public plant breeding Recommendations and proceedings ofa conference on best practices for intellectual property protection of pub-lically developed plant germplasm (p 70) University of Wisconsin‐Madison Retrieved from httpshostcalswisceduagronomywp-contentuploadssites1620162005Proceedings-IPR-Finalpdf
Trybush S Jahodovaacute Š Macalpine W amp Karp A (2008) A geneticstudy of a Salix germplasm resource reveals new insights into rela-tionships among subgenera sections and species BioEnergyResearch 1 67ndash79 httpsdoiorg101007s12155-008-9007-9
Tsai C J (2013) Next‐generation sequencing for next‐generationbreeding and more New Phytologist 198 635ndash637 httpsdoiorg101111nph12245
Tuskan G A Difazio S Jansson S Bohlmann J Grigoriev I HellstenU hellip Rokhsar D (2006) The genome of black cottonwood Populustrichocarpa (Torr ampGray) Science 313 1596ndash1604
Tuskan G A amp Walsh M E (2001) Short‐rotation woody cropsystems atmospheric carbon dioxide and carbon management AUS case study The Forestry Chronicle 77 259ndash264 httpsdoiorg105558tfc77259-2
Van Den Broek R Treffers D‐J Meeusen M Van Wijk ANieuwlaar E amp Turkenburg W (2001) Green energy or organicfood A life‐cycle assessment comparing two uses of set‐asideland Journal of Industrial Ecology 5 65ndash87 httpsdoiorg101162108819801760049477
Van Der Schoot J Pospiskova M Vosman B amp Smulders M J M(2000) Development and characterization of microsatellite markers inblack poplar (Populus nigra L) Theoretical and Applied Genetics 101317ndash322 httpsdoiorg101007s001220051485
Van Der Weijde T Dolstra O Visser R G amp Trindade L M(2017a) Stability of cell wall composition and saccharificationefficiency in Miscanthus across diverse environments Frontiers inPlant Science 7 2004
Van der Weijde T Huxley L M Hawkins S Sembiring E HFarrar K Dolstra O hellip Trindade L M (2017) Impact ofdrought stress on growth and quality of miscanthus for biofuelproduction Global Change Biology Bioenergy 9 770ndash782
Van der Weijde T Kamei C L A Severing E I Torres A FGomez L D Dolstra O hellip Trindade L M (2017) Geneticcomplexity of miscanthus cell wall composition and biomass qual-ity for biofuels BMC Genomics 18 406
Van der Weijde T Kiesel A Iqbal Y Muylle H Dolstra O Visser RG FhellipTrindade LM (2017) Evaluation ofMiscanthus sinensis bio-mass quality as feedstock for conversion into different bioenergy prod-uctsGlobal Change Biology Bioenergy 9 176ndash190
Vanholme B Cesarino I Goeminne G Kim H Marroni F VanAcker R hellip Boerjan W (2013) Breeding with rare defectivealleles (BRDA) A natural Populus nigra HCT mutant with modi-fied lignin as a case study New Phytologist 198 765ndash776
Vogel K P Mitchell R B Casler M D amp Sarath G (2014)Registration of Liberty Switchgrass Journal of Plant Registra-tions 8 242ndash247 httpsdoiorg103198jpr2013120076crc
Wang X Yamada T Kong F‐J Abe Y Hoshino Y Sato Hhellip Yamada T (2011) Establishment of an efficient in vitro cul-ture and particle bombardment‐mediated transformation systems inMiscanthus sinensis Anderss a potential bioenergy crop GlobalChange Biology Bioenergy 3 322ndash332 httpsdoiorg101111j1757-1707201101090x
Watrud L S Lee E H Fairbrother A Burdick C Reichman JR Bollman M hellip Van de Water P K (2004) Evidence forlandscape‐level pollen‐mediated gene flow from genetically modi-fied creeping bentgrass with CP4 EPSPS as a marker Proceedingsof the National Academy of Sciences of the United States of Amer-ica 101 14533ndash14538
Weisgerber H (1993) Poplar breeding for the purpose of biomassproduction in short‐rotation periods in Germany ndash Problems andfirst findings Forestry Chronicle 69 727ndash729 httpsdoiorg105558tfc69727-6
Wheeler N C Steiner K C Schlarbaum S E amp Neale D B(2015) The evolution of forest genetics and tree improvementresearch in the United States Journal of Forestry 113 500ndash510httpsdoiorg105849jof14-120
CLIFTON‐BROWN ET AL | 33
Whittaker C Macalpine W J Yates N E amp Shield I (2016)Dry matter losses and methane emissions during wood chip stor-age The impact on full life cycle greenhouse gas savings of shortrotation coppice willow for heat BioEnergy Research 9 820ndash835 httpsdoiorg101007s12155-016-9728-0
Xi Q (2000) Investigation on the distribution and potential of giant grassesin China PhD thesis Goettingen Germany Cuvillier Verlag
Xi Q amp Jezowkski S (2004) Plant resources of Triarrhena andMiscanthus species in China and its meaning for Europe PlantBreeding and Seed Science 49 63ndash77
Xue S Kalinina O amp Lewandowski I (2015) Present and future optionsfor the improvement of Miscanthus propagation techniques Renewableand Sustainable Energy Reviews 49 1233ndash1246
Yin K Q Gao C X amp Qiu J L (2017) Progress and prospectsin plant genome editing Nature Plants 3 httpsdoiorg101038nplants2017107
YookM (2016)Genetic diversity and transcriptome analysis for salt toler-ance inMiscanthus PhD thesis Seoul National University Korea
Zaidi S S E A Mahfouz M M amp Mansoor S (2017) CRISPR‐Cpf1 A new tool for plant genome editing Trends in PlantScience 22 550ndash553 httpsdoiorg101016jtplants201705001
Zalesny R S Stanturf J A Gardiner E S Perdue J H YoungT M Coyle D R hellip Hass A (2016) Ecosystem services ofwoody crop production systems BioEnergy Research 9 465ndash491 httpsdoiorg101007s12155-016-9737-z
Zhang Q X Sun Y Hu H K Chen B Hong C T Guo H Phellip Zheng B S (2012) Micropropagation and plant regenerationfrom embryogenic callus of Miscanthus sinensis Vitro Cellular ampDevelopmental Biology‐Plant 48 50ndash57 httpsdoiorg101007s11627-011-9387-y
Zhang Y Zalapa J E Jakubowski A R Price D L AcharyaA Wei Y hellip Casler M D (2011) Post‐glacial evolution of
Panicum virgatum Centers of diversity and gene pools revealedby SSR markers and cpDNA sequences Genetica 139 933ndash948httpsdoiorg101007s10709-011-9597-6
Zhao H Wang B He J Yang J Pan L Sun D amp Peng J(2013) Genetic diversity and population structure of Miscanthussinensis germplasm in China PloS One 8 e75672
Zhou X H Jacobs T B Xue L J Harding S A amp Tsai C J(2015) Exploiting SNPs for biallelic CRISPR mutations in theoutcrossing woody perennial Populus reveals 4‐coumarate CoAligase specificity and redundancy New Phytologist 208 298ndash301
Zhou R Macaya‐Sanz D Rodgers‐Melnick E Carlson C HGouker F E Evans L M hellip DiFazio S P (2018) Characteri-zation of a large sex determination region in Salix purpurea L(Salicaceae) Molecular Genetics and Genomics 1ndash16 httpsdoiorg101007s00438-018-1473-y
Zsuffa L Mosseler A amp Raj Y (1984) Prospects for interspecifichybridization in willow for biomass production In K L Perttu(Ed) Ecology and management of forest biomass production sys-tems Report 15 (pp 261ndash281) Uppsala Sweden SwedishUniversity of Agricultural Sciences
How to cite this article Clifton‐Brown J HarfoucheA Casler MD et al Breeding progress andpreparedness for mass‐scale deployment of perenniallignocellulosic biomass crops switchgrassmiscanthus willow and poplar GCB Bioenergy201800000ndash000 httpsdoiorg101111gcbb12566
34 | CLIFTON‐BROWN ET AL
Graphical AbstractThe contents of this page will be used as part of the graphical abstract of html only
It will not be published as part of main article
Plant breeding links the research effort with commercial mass upscaling The authorsrsquo assessment of development status ofthe four species is shown (poplar having two one for short rotation coppice (SRC) poplar and one for the more traditionalshort rotation forestry (SRF)) Mass scale deployment needs developments outside the breeding arenas to drive breedingactivities more rapidly and extensively
TABLE3
Preparedness
formassupscaling
currentmarketvalueandresearch
investmentin
four
perennialbiom
asscrops(PBCs)
Species
Switc
hgrass
Miscanthu
sW
illow
Pop
lar
Current
commercial
plantin
gcostsperha
USa20
0USD
bin
the
establishm
entyear
DE3
375
Eurob
(XueK
alininaamp
Lew
ando
wski20
15)
UK2
153
GBPb
(Evans201
6)redu
cedto
116
9GBPwith
Mxg
rhizom
ewhere
thefarm
erdo
estheland
preparation(w
wwterravestacom
)US
173
0ndash222
5USD
UK150
0ndash173
9GBPplus
land
preparation(Evans20
16)
US
197
6USD
(=80
0USD
acre)
IT110
0Euro
FRSE100
0ndash200
0Euro
US
863USD
ha(LazarusHeadleeamp
Zalesny
20
15)
Current
marketvalue
ofthebiom
asspert
DM
US
80ndash1
00USD
UK~8
0GBP(bales)(Terravesta
person
alcommun
ication)
US
94USD
(chipp
ed)
UK49
41GBP(chipp
ed)(Evans20
16)
US
55ndash7
0USD
IT10
0Euro
US
80ndash90USD
deliv
ered
basedon
40milesof
haulage
Scienceforg
enetic
improv
ementprojects
inthelast10
years
USgt50
projectsfund
edby
USDOEand
USD
ANIFA
CNgt
30projectsfund
edby
CN‐N
SFCandMBI
EUc 6projects(O
PTIM
ISCO
PTIM
A
WATBIO
SUNLIBBF
IBRAandGRACE)
FRB
FFSK
2projectsfund
edby
IPETandPM
BC
US
EBICABBImultip
leDOEFeedstock
Genom
icsandUSD
AAFR
Iprojects
UKB
SBECB
EGIN
(200
3ndash20
10)
US
USDOEJG
IGenom
eSequ
encing
(200
9ndash20
122
015ndash
2018
)USD
ANortheastSu
nGrant
(200
9ndash20
12)
USD
ANIFANEWBio
Con
sortium
(201
2ndash20
17)USD
ANIFAWillow
SkyC
AP(201
8ndash20
21)
EUF
P7(EnergyPo
plarN
ovelTreeWATBIO
Tree4Fu
ture)
SEC
limate‐adaptedpo
plarsandNanow
ood
SLU
andST
TUS
Bioenergy
Science(O
RNL)USD
Afeedstocks
geno
micsprog
rammeBRDIa
ndBRC‐CBI
Major
projects
supp
ortin
gcrossing
andselectioncycles
inthelast10
years
USandCA12
projects
NLRUEmiscanthu
sPP
P20
15ndash2
019
50ndash
100K
Euroyear
UKGIA
NT‐LIN
KPP
I20
11ndash2
016
13M
GBPyear
US
EBICABBI~0
5M
USD
year
UKBEGIN
2000
ndash201
0US
USD
ANortheastSu
nGrant
(200
9ndash20
12)USD
ANIFA
NEWBio
Con
sortium
(201
2ndash20
17)USD
ANIFA
Willow
SkyC
AP(201
8ndash20
21)
FR1
4region
alandnatio
nalp
rojects10
0ndash15
0k
Euroyear
SES
TTo
nebreeding
effortas
aninternalproject
in20
10with
aresulting
prog
enytrial
US
DOESu
nGrant
feedstockdevelopm
ent
partnershipprog
ramme(including
Willow
)
Current
annu
alinvestmentin
projects
fortranslationinto
commercial
hybrids
USgt20
MUSD
year
UKMUST
201
6ndash20
1905M
GBPyear
US
CABBIUSD
AAFR
I20
17ndash2
02205M
USD
year
US
USD
ANIFA
NEWBio
~04
MUSD
year
FRScienceforim
prov
ement~2
0kEuroyear
US
USD
Aun
derAFR
ICo‐ordinatedAg
Prod
ucer
projectsAnd
severaltranslational
geno
micsprojectswith
outpu
blic
fund
ing
Upscalin
gtim
e(years)
toproducesufficient
propagules
toplantgt10
0ha
US
Using
seed
2ndash3years
UKRhizomefor1ndash200
0hectares
canbe
ready
in6mon
ths
UKNL2
0ha
ofseed
hybridsplantedin
2018
In
2019
sufficientseedfor5
0ndash10
0ha
isexpected
UK3yearsusingconv
entio
nalcutting
sfaster
usingmicroprop
agation
FRIT3yearsby
vegetativ
eprop
agation
SEgt3yearsby
vegetativ
eprop
agation(cuttin
gs)
and3yearsby
microprop
agation
Note
AFR
IAgriculture
andFo
odResearchInitiativein
theUnitedStatesBEGIN
BiomassforEnergyGenetic
Improvem
entNetwork
BFF
BiomassFo
rtheFu
tureBRC‐CBI
Bioenergy
ResearchCentre‐Centrefor
Bioenergy
Innovatio
nBRDIBiomassresearch
developm
entinitiative
BSB
ECBBSR
CSu
stainableBioenergy
Centre
CABBICenterforadvanced
bioenergyandbioproductsinnovatio
nin
theUnitedStatesCN‐N
SFC
Natural
ScienceFo
undatio
nof
ChinaDOEDepartm
entof
Energyin
theUnitedStatesEBIEnergyBiosciences
Institu
teFIBRAFibrecropsas
sustainablesource
ofbiobased
materialforindustrial
productsin
Europeand
ChinaFP
7SeventhFram
eworkProgrammein
theEUGIA
NT‐LIN
KGenetic
improvem
entof
miscanthusas
asustainablefeedstockforbioenergyin
theUnitedKingdom
GRACEGRow
ingAdvancedindustrial
Crops
onmarginallandsforbiorEfineriesIPETKorea
Institu
teof
Planning
andEvaluationforTechnologyin
FoodA
gricultureFo
restry
andFisheriesJG
IJointGenom
eInstitu
teMBIMendelBiotechnology
IncMUST
Miscant-
husUpS
calin
gTechnology
NEWBioNortheast
WoodyW
arm‐seasonBiomassConsortiumNIFANationalInstitu
teof
Food
andAgriculture
intheUnitedStatesNovelTree
Novel
tree
breeding
strategiesOPT
IMAOpti-
mizationof
perennialgrassesforbiom
assproductio
nin
theMediterraneanareaOPT
IMISCOptim
izingbioenergyproductio
nfrom
Miscanthus
ORNLOak
Ridge
NationalLabPM
BCPlantMolecular
BreedingCenterof
theNextGenerationBiogreenResearchCenters
oftheRepublic
ofKoreaPP
Ipublicndashp
rivate
investmentPP
Ppublicndashp
rivate
partnership
RUEradiationuseefficiencySk
yCAP
Coordinated
AgriculturalProjectSL
U
SwedishUniversity
ofAgriculturalSciencesST
TSw
eTreeTech
SUNLIBBSu
stainableLiquidBiofuelsfrom
BiomassBiorefiningTree4Fu
tureDesigning
Trees
forthefutureUSD
AUSDepartm
entof
Agriculture
WATBIO
Developmentof
improved
perennialnonfoodbiom
assandbioproduct
cropsforwater
stressed
environm
ents
a ISO
Alpha‐2
lettercountrycodes
b Local
currencies
areused
asat
2018Euro
GBP(G
reat
Britain
Pound)USD
(USDollar)c EU
isused
forEurope
CLIFTON‐BROWN ET AL | 7
TABLE4
Prebreedingresearch
andthestatus
ofconventio
nalbreeding
infour
leadingperennialbiom
asscrops(PBCs)
Breeding
techno
logy
step
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sWillow
Poplar
1Collected
wild
accessions
or
secondarysources
availablefor
breeding
Provideabroadbase
of
useful
traits
Respect
forCBD
and
Nagoyaprotocol
on
collections
after2014
Not
allindigenous
genetic
resourcesareaccessible
under
CBD
forpolitical
reasons
USa~2
000
(181
inofficial
GRIN
gene
bank
ofwhich
96
wereavailable
others
in
variousunofficial
collections)
CNChangsha
~1000
andNanjin
g
2000from
China
DK~1
20from
JP
FRworking
collectionof
~100
(mainlyM
sin)
JP~1
500
from
JPS
KandCN
NLworking
collectionof
~300
Msin
SK~7
00from
SKC
NJP
andRU
UK~1
500
accessions
(from
~500
sites)
ofwhich
1000arein
field
nurseriesin
2018
US
14in
GRIN
global
US
Illin
ois~1
500
collections
made
inAsiawith
25
currently
available
inUnitedStates
forbreeding
UK1500with
about20
incommon
with
the
UnitedStates
US
350largelyunique
FR3370Populus
delto
ides
(650)Pnigra(2000)
Ptrichocarpa(600)and
Pmaximow
iczii(120)
IT30000P
delto
idsPnigra
Pmaximow
icziiand
Ptrichocarpa
SE13000
accessionsmainly
Ptrichocarpa
byST
TandSL
U
US
GWR150Ptrichocarpaand
300Pmaximow
icziiaccessions
forPacificNorthwestprogram
535Pdelto
ides
and~2
00
Pnigraaccessions
for
Southeastprogrammeand77
Psimoniifrom
Mongolia
UMD
NRRI550+
clonal
accessions
(primarily
Ptimescanadnesisa
long
with
Pdelto
ides
andPnigra
parents)
forUpper
Midwest
program
2Wild
accessions
which
have
undergone
phenotypic
screeningin
field
trials
Selectionof
parental
lines
with
useful
traits
Strong
partnerships
torun
multilocationfieldtrials
tophenotypeconsistently
theaccessionsgenotypes
indifferentenvironm
ents
anddatabases
Costof
runningmultilocation
US
gt500000genotypes
CN~1
250
inChangsha
~1700
in
HunanJiangsuS
handong
Hainan
NL~250genotypes
JP~1
200
SK~4
00
UK‐le
d~1
000
genotypesin
a
replicated
trial
US‐led
~1200
genotypesin
multi‐
locatio
nreplicated
trialsin
Asiaand
North
America
MsinSK
CNJP
CA
(Ontario)US(Illinoisand
Colorado)MsacSK
(Kangw
on)
CN
(Zhejiang)JP
(Sapporo)CA
(Onatario)U
S(Illinois)DK
(Aarhus)
UKgt400
US
180
FR2720
IT10000
SE~1
50
US
GWR~1
500
wild
Ptrichocarpaphenotypes
and
OPseedlots(total
of70)from
thePmaximow
icziispecies
complex
(suaveolens
cathayana
koreana
ussuriensismaximow
iczii)and
2000ndash3000Pdelto
ides
collections
(based
on104s‐
generatio
nfamilies)
UMD
NRRIscreened
seedlin
gsfrom
aPdelto
ides
(OPor
CP)
breeding
programme(1996ndash2016)
gt7200OPseedlin
gsof
PnigraunderMinnesota
clim
atic
conditions(improve
parental
populatio
n)
3Wild
germ
plasm
genotyping
Amanaged
living
collectionof
clonal
types
US
~20000genotypes
CNChangsha
~1000Nanjin
g37
FR44
bycpDNA
(Fenget
al
UK~4
00
US
225
(Con
tinues)
8 | CLIFTON‐BROWN ET AL
TABLE
4(Contin
ued)
Breeding
techno
logy
step
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sWillow
Poplar
Construct
phylogenetic
treesanddissim
ilarity
indices
The
type
ofmolecular
analysis
mdashAFL
PcpDNAR
ADseq
whole
genomesequencing
2014)
JP~1
200
NL250genotypesby
RADseq
UK~1
000
byRADseq
SK~3
00
US
Illin
oisRADseqon
allwild
accessions
FR2310
SE150
US
1500
4Exploratory
crossing
and
progenytests
Discovergood
parental
combinatio
nsmdashgeneral
combining
ability
Geographicseparatio
n
speciesor
phylogenetic
trees
Costly
long‐te
rmmultilocation
trialsforprogeny
US
gt10000
CN~1
0
FR~2
00
JP~3
0
NL3600
SK~2
0
UK~4
000
(~1500Msin)
US
500ndash1000
UK gt700since2003
gt800EWBP(1996ndash
2002)
US
800
FRCloned13
times13
factorial
matingdesign
with
3species[10
Pdelto
ides8Ptrichocarpa
8
Pnigra)
US
GWR1000exploratory
crossingsof
Pfrem
ontii
Psimoniiforim
proved
droughttolerance
UMD
NRRIcrossednativ
e
Pdelto
ides
with
non‐nativ
e
Ptrichocarpaand
Pmaximow
icziiMajority
of
progenypopulatio
nssuffered
from
clim
atic
andor
pathogenic
susceptib
ility
issues
5Wideintraspecies
hybridization
Discovergood
parental
combinatio
nsmdashgeneral
combining
ability
1to
4above
inform
atics
Flow
eringsynchronizationand
low
seed
set
US
~200
CN~1
0
NL1800
SK~1
0
UK~5
00
UK276
US
400
IT500
SE120
US
50ndash100
6With
inspecies
recurrentselection
Concentratio
nof
positiv
e
traits
Identificationof
theright
heterotic
groups
insteps
1ndash5
Difficultto
introducenew
germ
plasm
with
outdilutio
n
ofbesttraits
US
gt40
populatio
ns
undergoing
recurrentselection
NL2populatio
ns
UK6populatio
ns
US
~12populatio
nsto
beestablished
by2019
UK4
US
150
FRPdelto
ides
(gt30
FSfamilies
ndash900clones)Pnigra(gt40
FS
families
ndash1600clones)and
Ptrichocarpa(15FS
Families
minus350clones)
IT80
SEca50
parent
breeding
populatio
nswith
in
Ptrichocarpa
US
Goalsis~1
00parent
breeding
populatio
nswith
inPtrichocarpa
Pdelto
ides
PnigraandPmaximow
iczii
7Interspecies
hybrid
breeding
Com
bining
complem
entary
traitsto
producegood
morphotypes
and
heterosiseffects
Allthesteps1ndash6
Inearlystage
improvem
ents
areunpredictable
Awide
base
iscostly
tomanage
None
CNChangsha
3Nanjin
g
120MsintimesMflo
r
JP~5
with
Saccharum
spontaneum
SK5
UK420
US
250
FRPcanadensis(1800)
Pdelto
ides
timesPtrichocarpa
(800)and
PtrichocarpatimesPmaximow
iczii
(1200)
(Con
tinues)
CLIFTON‐BROWN ET AL | 9
TABLE
4(Contin
ued)
Breeding
techno
logy
step
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sWillow
Poplar
IT~1
20
US
525genotypesin
eastside
hybrid
programme
205westside
hybrid
programmeand~1
200
in
SoutheastPdelto
ides
program
8Chrom
osom
e
doublin
g
Arouteto
triploid
seeded
typesdoublin
gadiploid
parent
ofknow
n
breeding
valuefrom
recurrentselection
Needs
1ndash7to
help
identify
therightparental
lines
Doubled
plantsarenotorious
forrevertingto
diploid
US
~50plantstakenfrom
tetraploid
tooctaploid
which
arenow
infieldtrials
CN~1
0Msactriploids
DKSeveralattemptsin
mid
90s
FR5genotypesof
Msin
UK2Msin1Mflora
nd1Msac
US
~30
UKAttempted
butnot
routinelyused
US
Not
partof
theregular
breeding
becausethere
existnaturalpolyploids
NA
9DoubleHaploids
Arouteto
producing
homogeneous
progeny
andalso
forthe
introductio
nof
transgenes
orforgenome
editing
Needs
1ndash7to
help
identify
therightparental
lines
Fully
homozygousplantsare
weakandeasily
die
andmay
notflow
ersynchronously
US
Noneyetattempteddueto
poor
vigour
andviability
of
haploids
CN2genotypesof
Msin
andMflor
UKDihaploidsof
MsinMflor
and
Msacand2transgenic
Msinvia
anther
cultu
re
US
6createdandused
toidentify
miss‐identifying
paralogueloci
(causedby
recent
genome
duplicationin
miscanthus)Usedin
creatin
gthereferencegenomefor
MsinVeryweakplantsand
difficultto
retain
thelin
es
NA
NA
10Embryo
rescue
Anattractiv
etechniquefor
recovering
plantsfrom
sexual
crosses
The
majority
ofem
bryos
cannot
survivein
vivo
or
becomelong‐timedorm
ant
None
UKone3times
hybrid
UKEmbryosrescue
protocol
proved
robustat
8days
post‐pollin
ation
FRAnem
bryo
rescue
protocol
proved
robustandim
proved
hybridizationsuccessfor
Pdelto
ides
timesPtrichocarpa
crosses
11Po
llenstorage
Flexibility
tocross
interestingparents
with
outneed
of
flow
ering
synchronization
None
Nosuccessto
daten
otoriously
difficultwith
grassesincluding
sugarcane
Freshpollencommonly
used
incrossing
Pollenstoragein
some
interspecificcrosses
FRBothstored
(cryobank)
and
freshindividual
pollen
Note
AFL
Pam
plifiedfragmentlength
polymorphismCBDconventio
non
biological
diversity
CP
controlledpollinatio
nCpD
NAchloroplastDNAEWBP
Europeanwillow
breeding
programme
FSfull‐sib
GtimesE
genotype‐by‐environm
entGRIN
germ
plasm
resourcesinform
ationnetwork
GWRGreenWoodResourcesMflorMiscanthusflo
ridulusMsacMsacchariflo
rus
MsinMsinensisNAnotapplicableNRRINatural
Resources
ResearchInstitu
teOP
open
pollinatio
nRADseq
restrictionsite‐associatedDNA
sequencingSL
USw
edishUniversity
ofAgriculturalSciencesST
TSw
eTreeTech
UMDUniversity
ofMinnesota
Duluth
a ISO
Alpha‐2
lettercountrycodes
10 | CLIFTON‐BROWN ET AL
programmes mature The average rate of gain for biomassyield in long‐term switchgrass breeding programmes hasbeen 1ndash4 per year depending on ecotype populationand location of the breeding programme (Casler amp Vogel2014 Casler et al 2018) The hybrid derivative Libertyhas a biomass yield 43 higher than the better of its twoparents (Casler amp Vogel 2014 Vogel et al 2014) Thedevelopment of cold‐tolerant and late‐flowering lowland‐ecotype populations for the northern United States hasincreased biomass yields by 27 (Casler et al 2018)
Currently more than 20 recurrent selection populationsare being managed in the United States to select parentsfor improved yield yield resilience and compositional qual-ity of the biomass For the agronomic development andupscaling high seed multiplication rates need to be com-bined with lower seed dormancy to reduce both crop estab-lishment costs and risks Expresso is the first cultivar withsignificantly reduced seed dormancy which is the first steptowards development of domesticated populations (Casleret al 2015) Most phenotypic traits of interest to breeders
require a minimum of 2 years to be fully expressed whichresults in a breeding cycle that is at least two years Morecomplicated breeding programmes or traits that requiremore time to evaluate can extend the breeding cycle to 4ndash8 years per generation for example progeny testing forbiomass yield or field‐based selection for cold toleranceBreeding for a range of traits with such long cycles callsfor the development of molecular methods to reduce time-scales and improve breeding efficiency
Two association panels of switchgrass have been pheno-typically and genotypically characterized to identify quanti-tative trait loci (QTLs) that control important biomasstraits The northern panel consists of 60 populationsapproximately 65 from the upland ecotype The southernpanel consists of 48 populations approximately 65 fromthe lowland ecotype Numerous QTLs have been identifiedwithin the northern panel to date (Grabowski et al 2017)Both panels are the subject of additional studies focused onbiomass quality flowering and phenology and cold toler-ance Additionally numerous linkage maps have been
FIGURE 1 Cumulative minimum years needed for the conventional breeding cycle through the steps from wild germplasm to thecommercial hybrids in switchgrass miscanthus willow and poplar Information links between the steps are indicated by dotted arrows andhighlight the importance of long‐term informatics to maximize breeding gain
CLIFTON‐BROWN ET AL | 11
created by the pairwise crossing of individuals with diver-gent characteristics often to generate four‐way crosses thatare analysed as pseudo‐F2 crosses (Liu Wu Wang ampSamuels 2012 Okada et al 2010 Serba et al 2013Tornqvist et al 2018) Individual markers and QTLs iden-tified can be used to design marker‐assisted selection(MAS) programmes to accelerate breeding and increase itsefficiency Genomic prediction and selection (GS) holdseven more promise with the potential to double or triplethe rate of gain for biomass yield and other highly complexquantitative traits of switchgrass (Casler amp Ramstein 2018Ramstein et al 2016) The genome of switchgrass hasrecently been made public through the Joint Genome Insti-tute (httpsphytozomejgidoegov)
Transgenic approaches have been heavily relied upon togenerate unique genetic variants principally for traitsrelated to biomass quality (Merrick amp Fei 2015) Switch-grass is highly transformable using either Agrobacterium‐mediated transformation or biolistics bombardment butregeneration of plants is the bottleneck to these systemsTraditionally plants from the cultivar Alamo were the onlyregenerable genotypes but recent efforts have begun toidentify more genotypes from different populations that arecapable of both transformation and subsequent regeneration(King Bray Lafayette amp Parrott 2014 Li amp Qu 2011Ogawa et al 2014 Ogawa Honda Kondo amp Hara‐Nishi-mura 2016) Cell wall recalcitrance and improved sugarrelease are the most common targets for modification (Bis-wal et al 2018 Fu et al 2011) Transgenic approacheshave the potential to provide traits that cannot be bredusing natural genetic variability However they will stillrequire about 10ndash15 years and will cost $70ndash100 millionfor cultivar development and deployment (Harfouche Mei-lan amp Altman 2011) In addition there is commercialuncertainty due to the significant costs and unpredictabletimescales and outcomes of the regulatory approval processin the countries targeted for seed sales As seen in maizeone advantage of transgenic approaches is that they caneasily be incorporated into F1 hybrid cultivars (Casler2012 Casler et al 2012) but this does not decrease thetime required for cultivar development due to field evalua-tion and seed multiplication requirements
The potential impacts of unintentional gene flow andestablishment of non‐native transgene sequences in nativeprairie species via cross‐pollination are also major issues forthe environmental risk assessment These limit further thecommercialization of varieties made using these technolo-gies Although there is active research into switchgrass steril-ity mechanisms to curb unintended pollen‐mediated genetransfer it is likely that the first transgenic cultivars proposedfor release in the United States will be met with considerableopposition due to the potential for pollen flow to remainingwild prairie sites which account for lt1 of the original
prairie land area and are highly protected by various govern-mental and nongovernmental organizations (Casler et al2015) Evidence for landscape‐level pollen‐mediated geneflow from genetically modified Agrostis seed multiplicationfields (over a mountain range) to pollinate wild relatives(Watrud et al 2004) confirms the challenge of using trans-genic approaches Looking ahead genome editing technolo-gies hold considerable promise for creating targeted changesin phenotype (Burris Dlugosz Collins Stewart amp Lena-ghan 2016 Liu et al 2018) and at least in some jurisdic-tions it is likely that cultivars resulting from gene editingwill not need the same regulatory approval as GMOs (Jones2015a) However in July 2018 the European Court of Justice(ECJ) ruled that cultivars carrying mutations resulting fromgene editing should be regulated in the same way as GMOsThe ECJ ruled that such cultivars be distinguished from thosearising from untargeted mutation breeding which isexempted from regulation under Directive 200118EC
3 | MISCANTHUS
Miscanthus is indigenous to eastern Asia and Oceania whereit is traditionally used for forage thatching and papermaking(Xi 2000 Xi amp Jezowkski 2004) In the 1960s the highbiomass potential of a Japanese genotype introduced to Eur-ope by Danish nurseryman Aksel Olsen in 1935 was firstrecognized in Denmark (Linde‐Laursen 1993) Later thisaccession was characterized described and named asldquoM times giganteusrdquo (Greef amp Deuter 1993 Hodkinson ampRenvoize 2001) commonly abbreviated as Mxg It is a natu-rally occurring interspecies triploid hybrid between tetraploidM sacchariflorus (2n = 4x) and diploid M sinensis(2n = 2x) Despite its favourable agronomic characteristicsand ability to produce high yields in a wide range of environ-ments in Europe (Kalinina et al 2017) the risks of relianceon it as a single clone have been recognized Miscanthuslike switchgrass is outcrossing due to self‐incompatibility(Jiang et al 2017) Thus seeded hybrids are an option forcommercial breeding Miscanthus can also be vegetativelypropagated by rhizome or in vitro culture which allows thedevelopment of clones The breeding approaches are usuallybased on F1 crosses and recurrent selection cycles within thesynthetic populations There are several breeding pro-grammes that target improvement of miscanthus traitsincluding stress resilience targeted regional adaptation agro-nomic ldquoscalabilityrdquo through cheaper propagation fasterestablishment lower moisture and ash contents and greaterusable yield (Clifton‐Brown et al 2017)
Germplasm collections specifically to support breedingfor biomass started in the late 1980s and early 1990s inDenmark Germany and the United Kingdom (Clifton‐Brown Schwarz amp Hastings 2015) These collectionshave continued with successive expeditions from European
12 | CLIFTON‐BROWN ET AL
and US teams assembling diverse collections from a widegeographic range in eastern Asia including from ChinaJapan South Korea Russia and Taiwan (Hodkinson KlaasJones Prickett amp Barth 2015 Stewart et al 2009) Threekey miscanthus species for biomass production are M si-nensis M floridulus and M sacchariflorus M sinensis iswidely distributed throughout eastern Asia with an adap-tive range from the subtropics to southern Russia (Zhaoet al 2013) This species has small rhizomes and producesmany tightly packed shoots forming a ldquotuftrdquo M floridulushas a more southerly adaptive range with a rather similarmorphology to M sinensis but grows taller with thickerstems and is evergreen and less cold‐tolerant than the othermiscanthus species M sacchariflorus is the most northern‐adapted species ranging to 50 degN in eastern Russia (Clarket al 2016) Populations of diploid and tetraploid M sac-chariflorus are found in China (Xi 2000) and South Korea(Yook 2016) and eastern Russia but only tetraploids havebeen found in Japan (Clark Jin amp Petersen 2018)
Germplasm has been assembled from multiple collec-tions over the last century though some early collectionsare poorly documented This historical germplasm has beenused to initiate breeding programmes largely based on phe-notypic and genotypic characterization As many of theaccessions from these collections are ldquoorigin unknownrdquocrucial environmental envelope data are not available UK‐led expeditions started in 2006 and continued until 2011with European and Asian partners and have built up a com-prehensive collection of 1500 accessions from 500 sitesacross Eastern Asia including China Japan South Koreaand Taiwan These collections were guided using spatialclimatic data to identify variation in abiotic stress toleranceAccessions from these recent collections were planted fol-lowing quarantine in multilocation nursery trials at severallocations in Europe to examine trait expression in differentenvironments Based on the resulting phenotypic andmolecular marker data several studies (a) characterized pat-terns of population genetic structure (Slavov et al 20132014) (b) evaluated the statistical power of genomewideassociation studies (GWASs) and identified preliminarymarkerndashtrait associations (Slavov et al 2013 2014) and(c) assessed the potential of genomic prediction (Daveyet al 2017 Slavov et al 2014 2018b) Genomic indexselection in particular offers the possibility of exploringscenarios for different locations or industrial markets (Sla-vov et al 2018a 2018b)
Separately US‐led expeditions also collected about1500 accessions between 2010 and 2014 (Clark et al2014 2016 2018 2015) A comprehensive genetic analy-sis of the population structure has been produced by RAD-seq for M sinensis (Clark et al 2015 Van der WeijdeKamei et al 2017) and M sacchariflorus (Clark et al2018) Multilocation replicated field trials have also been
conducted on these materials in North America and inAsia GWAS has been conducted for both M sinensis anda subset of M sacchariflorus accessions (Clark et al2016) To date about 75 of these recent US‐led collec-tions are in nursery trials outside the United States Due tolengthy US quarantine procedures these are not yet avail-able for breeding in the United States However molecularanalyses have allowed us to identify and prioritize sets ofgenotypes that best encompass the genetic variation in eachspecies
While mostM sinensis accessions flower in northern Eur-ope very few M sacchariflorus accessions flower even inheated glasshouses For this reason the European pro-grammes in the United Kingdom the Netherlands and Francehave performed mainly M sinensis (intraspecies) hybridiza-tions (Table 4) Selected progeny become the parents of latergenerations (recurrent selection as in switchgrass) Seed setsof up to 400 seed per panicle occur inM sinensis In Aberyst-wyth and Illinois significant efforts to induce synchronousflowering in M sacchariflorus and M sinensis have beenmade because interspecies hybrids have proven higher yieldperformance and wide adaptability (Kalinina et al 2017) Ininterspecies pairwise crosses in glasshouses breathable bagsandor large crossing tubes or chambers in which two or morewhole plants fit are used for pollination control Seed sets arelower in bags than in the open air because bags restrict pollenmovement while increasing temperatures and reducinghumidity (Clifton‐Brown Senior amp Purdy 2018) About30 of attempted crosses produced 10 to 60 seeds per baggedpanicle The seed (thousand seed mass ranges from 05 to09 g) is threshed from the inflorescences and sown into mod-ular trays to produce plug plants which are then planted infield nurseries to identify key parental combinations
A breeding programme of this scale must serve theneeds of different environments accepting the commonpurpose is to optimize the interception of solar radiationAn ideal hybrid for a given environment combines adapta-tion to date of emergence with optimization of traits suchas height number of stems per plant flowering and senes-cence time to optimize solar interception to produce a highbiomass yield with low moisture content at harvest (Rob-son Farrar et al 2013 Robson Jensen et al 2013) By20132014 conventional breeding in Europe had producedintra‐ and interspecific fertile seeded hybrids When acohort (typically about 5) of outstanding crosses have beenidentified it is important to work on related upscaling mat-ters in parallel These are as follows
Assessment of the yield and critical traits in selectedhybrids using a network of field trials
Efficient cloning of the seed parents While in vitro andmacro‐cloning techniques are used some genotypes areamenable to neither technique
CLIFTON‐BROWN ET AL | 13
TABLE5
Status
ofmod
ernplantbreeding
techniqu
esin
four
leadingperennialbiom
asscrop
s(PBCs)
Breeding
techno
logy
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sW
illow
Pop
lar
MAS
Use
ofmarker
sequ
encesthat
correlatewith
atrait
allowingearly
prog
enyselectionor
rejectionlocatin
gkn
owngenesfor
useful
traits(suchas
height)from
other
speciesin
your
crop
Breeding
prog
ramme
relevant
biparental
crosses(Table
4)to
create
ldquomapping
popu
latio
nsrdquoand
QTLidentification
andestim
ationto
identifyrobu
stmarkers
alternatively
acomprehensive
panelof
unrelated
geno
typesfor
GWAS
Ineffectivewhere
traits
areaffected
bymany
geneswith
small
effects
US
Literally
dozens
ofstud
iesto
identify
SNPs
andmarkers
ofinterestNoattempts
atMASas
yetlargely
dueto
popu
latio
nspecificity
CNS
SRISS
Rmarkers
forM
sinMsacand
Mflor
FR2
Msin
mapping
popu
latio
ns(BFF
project)
NL2
Msin
mapping
popu
latio
ns(A
tienza
Ram
irezetal
2003
AtienzaSatovic
PetersenD
olstraamp
Martin
200
3Van
Der
Weijdeetal20
17a
Van
derW
eijde
Hux
ley
etal20
17
Van
derW
eijdeKam
ei
etal20
17V
ander
WeijdeKieseletal
2017
)SK
2GWASpanels(1
collectionof
Msin1
diploidbiparentalF 1
popu
latio
n)in
the
pipelin
eUK1
2families
tofind
QTLin
common
indifferentfam
ilies
ofthe
sameanddifferent
speciesandhy
brids
US
2largeGWAS
panels4
diploidbi‐
parentalF 1
popu
latio
ns
1F 2
popu
latio
n(add
ition
alin
pipelin
e)
UK16
families
for
QTLdiscov
ery
2crossesMASscreened
1po
tentialvariety
selected
US
9families
geno
typedby
GBS(8
areF 1o
neisF 2)a
numberof
prom
ising
QTLdevelopm
entof
markers
inprog
ress
FR(INRA)tested
inlargeFS
families
and
infactorialmating
design
low
efficiency
dueto
family
specificity
US
49families
GS
Metho
dto
accelerate
breeding
throug
hredu
cing
the
resourcesforcross
attemptsby
Aldquotraining
popu
latio
nrdquoof
individu
alsfrom
stages
abov
ethat
have
been
both
Risks
ofpo
orpredictio
nProg
eny
testingneedsto
becontinueddu
ring
the
timethetraining
setis
US
One
prog
rammeso
farmdash
USD
Ain
WisconsinThree
cycles
ofgeno
mic
selectioncompleted
CN~1
000
geno
types
NLprelim
inary
mod
elsforMsin
developed
UK~1
000
geno
types
Verysuitable
application
not
attempted
yet
Maybe
challeng
ingfor
interspecies
hybrids
FR(INRA)120
0geno
type
ofPop
ulus
nigraarebeingused
todevelop
intraspecificGS
(Con
tinues)
14 | CLIFTON‐BROWN ET AL
TABLE
5(Contin
ued)
Breeding
techno
logy
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sW
illow
Pop
lar
predictin
gthe
performance
ofprog
enyof
crosses
(and
inresearch
topredictbestparents
tousein
biparental
crosses)
geno
typedand
phenotyp
edto
developamod
elthattakesgeno
typic
datafrom
aldquocandidate
popu
latio
nrdquoof
untested
individu
als
andprod
uces
GEBVs(Jannink
Lorenzamp
Iwata
2010
)
beingph
enotyp
edand
geno
typed
Inthis
time
next‐generation
germ
plasm
andreadyto
beginthe
second
roun
dof
training
and
recalib
ratio
nof
geno
mic
predictio
nmod
els
US
Prelim
inary
mod
elsforMsac
and
Msin
developed
Work
isun
derw
ayto
deal
moreeffectivelywith
polyploidgenetics
calib
ratio
nsUS
125
0geno
types
arebeingused
todevelopGS
calib
ratio
ns
Traditio
nal
transgenesis
Efficient
introd
uctio
nof
ldquoforeign
rdquotraits
(possiblyfrom
other
genu
sspecies)
into
anelite
plant(ega
prov
enparent
ora
hybrid)that
needsa
simpletraitto
beim
prov
edValidate
cand
idategenesfrom
QTLstud
ies
MASand
know
ledg
eof
the
biolog
yof
thetrait
andsource
ofgenesto
confer
the
relevant
changesto
phenotyp
eWorking
transformation
protocol
IPissuescostof
regu
latory
approv
al
GMO
labelling
marketin
gissuestricky
tousetransgenes
inou
t‐breedersbecause
ofcomplexity
oftransformingandgene
flow
risks
US
Manyprog
ramsin
UnitedStates
are
creatin
gtransgenic
plantsusingAlamoas
asource
oftransformable
geno
typesManytraits
ofinterestNothing
commercial
yet
CNC
hang
shaBtg
ene
transformed
into
Msac
in20
04M
sin
transformed
with
MINAC2gene
and
Msactransform
edwith
Cry2A
ageneb
othwith
amarkerg
eneby
Agrob
acterium
JP1
Msintransgenic
forlow
temperature
tolerancewith
increased
expression
offructans
(unp
ublished)
NLImprov
ingprotocol
forM
sin
transformation
SK2
Msin
with
Arabido
psisand
Brachypod
ium
PhytochromeBgenes
(Hwang
Cho
etal
2014
Hwang
Lim
etal20
14)
UK2
Msin
and
1Mflorg
enotyp
es
UKRou
tine
transformationno
tyet
possibleResearchto
overcomerecalcitrance
ongo
ing
Currently
trying
differentspecies
andcond
ition
sPo
plar
transformationused
atpresent
US
Frequently
attempted
with
very
limitedsuccess
US
600transgenic
lines
have
been
characterized
mostly
performed
inthe
aspenhy
brids
reprod
uctiv
esterility
drou
ghttolerance
with
Kno
ckdo
wns
(Con
tinues)
CLIFTON‐BROWN ET AL | 15
TABLE
5(Contin
ued)
Breeding
techno
logy
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sW
illow
Pop
lar
variou
slytransformed
with
four
cellwall
genesiptand
uidA
genesusingtwo
selectionsystem
sby
biolisticsand
Agrob
acterium
Transform
edplantsare
beinganalysed
US
Prelim
inarywork
will
betakenforw
ardin
2018
ndash202
2in
CABBI
Genom
eediting
CRISPR
Refinem
entofexisting
traits
inusefulparents
orprom
isinghybrids
bygeneratingtargeted
mutations
ingenes
know
ntocontrolthe
traitof
interestF
irst
adouble‐stranded
breakismadeinthe
DNAwhich
isrepairedby
natural
DNArepair
machineryL
eads
tofram
eshiftSN
Por
can
usealdquorepair
templaterdquo
orcanbe
used
toinserta
transgene
intoaldquosafe
harbourlocusrdquoIt
couldbe
used
todeleterepressors(or
transcriptionfactors)
Identification
mapping
and
sequ
encing
oftarget
genes(from
DNA
sequ
ence)
avoiding
screening
outof
unintend
ededits
Manyregu
latory
authorities
have
not
decidedwhether
CRISPR
andother
geno
meediting
techno
logies
are
GMOsor
notIf
GMOsseecomment
onstage10
abov
eIf
noteditedcrop
swill
beregu
latedas
conv
entio
nalvarieties
CATechn
olog
yisstill
toonewand
switchg
rass
geno
meis
very
complexOther
labo
ratories
are
interestedbu
tno
tyet
mov
ingon
this
US
One
prog
rammeso
farmdash
USD
Ain
Albany
FRInitiated
in20
16(M
ISEDIT
project)
NLInitiatingin
2018
US
Initiatingin
2018
in
CABBI
UKNot
started
Not
possible
until
transformation
achieved
US
CRISPR
‐Cas9and
Cpf1have
been
successfully
developedin
Pop
ulus
andarehigh
lyefficientExamples
forlig
nin
biosyn
thesis
Note
BFF
BiomassFo
rtheFu
tureBtBacillus
thuringiensisCABBICenterforadvanced
bioenergyandbioproductsinnovatio
nCpf1
CRISPR
from
Prevotella
andFrancisella
1CRISPR
clusteredregularlyinterspaced
palin
drom
icrepeatsCRISPR
‐CasC
RISPR
‐associated
Cry2A
acrystaltoxins2A
asubfam
ilyproduced
byBtFS
full‐sibG
BS
genotyping
bysequencingG
EBVsgenomic‐estim
ated
breeding
valuesG
MOg
enetically
modi-
fied
organism
GS
genomicselection
GWAS
genomew
ideassociationstudy
INRAF
renchNationalInstituteforAgriculturalR
esearch
IPintellectualp
roperty
IPTisopentenyltransferase
ISSR
inter‐SSR
MflorMiscant-
husflo
ridulusMsacMsacchariflo
rusMsinMsinensisM
AS
marker‐a
ssistedselection
MISEDITm
iscanthusgene
editing
forseed‐propagatedtriploidsNACn
oapicalmeristemA
TAF1
2and
cup‐shaped
cotyledon2
‐lik
eQTLq
uantitativ
etraitlocusS
NP
singlenucleotid
epolymorphismS
SRsim
plesequence
repeatsUSD
AU
SDepartm
ento
fAgriculture
a ISO
Alpha‐2
lettercountrycodes
16 | CLIFTON‐BROWN ET AL
High seed production from field crossing trials con-ducted in locations where flowering in both seed andpollen parents is likely to happen synchronously
Scalable and adapted harvesting threshing and seed pro-cessing methods for producing high seed quality
The results of these parallel activities need to becombined to identify the upscaling pathway for eachhybrid if this cannot be achieved the hybrid will likelynot be commercially viable The UK‐led programme withpartners in Italy and Germany shows that seedbased mul-tiplication rates of 12000 are achievable several inter-specific hybrids (Clifton‐Brown et al 2017) Themultiplication rate of M sinensis is higher probably15000ndash10000 Conventional cloning from rhizome islimited to around 120 that is one ha could provide rhi-zomes for around 20 ha of new plantation
Multilocation field testing of wild and novel miscanthushybrids selected by breeding programmes in the Nether-lands and the United Kingdom was performed as part ofthe project Optimizing Miscanthus Biomass Production(OPTIMISC 2012ndash2016) These trials showed that com-mercial yields and biomass qualities (Kiesel et al 2017Van der Weijde et al 2017a Van der Weijde Kieselet al 2017) could be produced in a wide range of climatesand soil conditions from the temperate maritime climate ofwestern Wales to the continental climate of eastern Russiaand the Ukraine (Kalinina et al 2017) Extensive environ-mental measurements of soil and climate combined withgrowth monitoring are being used to understand abioticstresses (Nunn et al 2017 Van der Weijde Huxley et al2017) and develop genotype‐specific scenarios similar tothose reported earlier in Hastings et al (2009) Phenomicsexperiments on drought tolerance have been conducted onwild and improved germplasm (Malinowska Donnison ampRobson 2017 Van der Weijde Huxley et al 2017)Recently produced interspecific hybrids displaying excep-tional yield under drought (~30 greater than control Mxg)in field trials in Poland and Moldova are being furtherstudied in detail in the phenomics and genomics facility atAberystwyth to better understand genendashtrait associationswhich can be fed back into breeding
Intraspecific seeded hybrids of M sinensis produced inthe Netherlands and interspecific M sacchariflorus times M si-nensis hybrids produced by the UK‐led breeding programmehave entered yield testing in 2018 with the recently EU‐funded project ldquoGRowing Advanced industrial Crops on mar-ginal lands for biorEfineries (GRACE)rdquo (httpswwwgrace-bbieu) Substantial variation in biomass quality for sacchari-fication efficiency (glucose release as of dry matter) ashcontent and melting point has already been generated inintraspecific M sinensis hybrids (Van der Weijde Kiesel
et al 2017) across environments (Weijde Dolstra et al2017a) GRACE aims to establish more than 20 hectares ofnew inter‐ and intraspecific seeded hybrids across six Euro-pean countries This project is building the know‐how andagronomy needed to transition from small research plots tocommercial‐scale field sites and linking biomass productiondirectly to industrial applications The biomass produced byhybrids in different locations will be supplied to innovativeindustrial end‐users making a wide range of biobased prod-ucts both for chemicals and for energy In the United Statesmulti‐location yield were initiated in 2018 to evaluate new tri-ploid M times giganteus genotypes developed at Illinois Cur-rently infertile hybrids are favoured in the United Statesbecause this eliminates the risk of invasiveness from naturallydispersed viable seed The precautionary principle is appliedas fertile miscanthus has naturalized in several states (QuinnAllen amp Stewart 2010) North European multilocation fieldtrials in the EMI and OPTIMISC projects have shown thereis minimal risk of invasiveness even in years when fertileflowering hybrids produce viable seed Naturalized standshave not established here due perhaps to low dormancy pooroverwintering and low seedling competitive strength In addi-tion to breeding for nonshattering or sterile seeded hybridsQuinn et al (2010) suggest management strategies which canfurther minimize environmental opportunities to manage therisk of invasiveness
31 | Molecular breeding and biotechnology
In miscanthus new plant breeding techniques (Table 5)have focussed on developing molecular markers forbreeding in Europe the United States South Korea andJapan There are several publications on QTL mappingpopulations for key traits such as flowering (AtienzaRamirez amp Martin 2003) and compositional traits(Atienza Satovic Petersen Dolstra amp Martin 2003) Inthe United States and United Kingdom independent andinterconnected bi‐parental ldquomappingrdquo families have beenstudied (Dong et al 2018 Gifford Chae SwaminathanMoose amp Juvik 2015) alongside panels of diverse germ-plasm accessions for GWAS (Slavov et al 2013) Fur-ther developments calibrating GS with very large panelsof parents and cross progeny are underway (Daveyet al 2017) The recently completed first miscanthus ref-erence genome sequence is expected to improve the effi-ciency of MAS strategies and especially GWAS(httpsphytozomejgidoegovpzportalhtmlinfoalias=Org_Msinensis_er) For example without a referencegenome sequence Clark et al (2014) obtained 21207RADseq SNPs (single nucleotide polymorphisms) on apanel of 767 miscanthus genotypes (mostly M sinensis)but subsequent reanalysis of the RADseq data using the
CLIFTON‐BROWN ET AL | 17
new reference genome resulted in hundreds of thousandsof SNPs being called
Robust and effective in vitro regeneration systems havebeen developed for Miscanthus sinensis M times giganteusand M sacchariflorus (Dalton 2013 Guo et al 2013Hwang Cho et al 2014 Rambaud et al 2013 Ślusarkie-wicz‐Jarzina et al 2017 Wang et al 2011 Zhang et al2012) However there is still significant genotype speci-ficity and these methods need ldquoin‐houserdquo optimization anddevelopment to be used routinely They provide potentialroutes for rapid clonal propagation and also as a basis forgenetic transformation Stable transformation using bothbiolistics (Wang et al 2011) and Agrobacterium tumefa-ciens DNA delivery methods (Hwang Cho et al 2014Hwang Lim et al 2014) has been achieved in M sinen-sis The development of miscanthus transformation andgene editing to generate diplogametes for producing seed‐propagated triploid hybrids are performed as part of theFrench project MISEDIT (miscanthus gene editing forseed‐propagated triploids) There are no reports of genomeediting in any miscanthus species but new breeding inno-vations including genome editing are particularly relevantin this slow‐to‐breed nonfood bioenergy crop (Table 4)
4 | WILLOW
Willow (Salix spp) is a very diverse group of catkin‐bear-ing trees and shrubs Willow belongs to the family Sali-caceae which also includes the Populus genus There areapproximately 350 willow species (Argus 2007) foundmostly in temperate and arctic zones in the northern hemi-sphere A few are adapted to subtropical and tropicalzones The centre of diversity is believed to be in Asiawith over 200 species in China Around 120 species arefound in the former Soviet Union over 100 in NorthAmerica and around 65 species in Europe and one speciesis native to South America (Karp et al 2011) Willows aredioecious thus obligate outcrossers and highly heterozy-gous The haploid chromosome number is 19 (Hanley ampKarp 2014) Around 40 of species are polyploid (Sudaamp Argus 1968) ranging from triploids to the atypicaldodecaploid S maxxaliana with 2n=190 (Zsuffa et al1984)
Although almost exclusively native to the NorthernHemisphere willow has been grown around the globe formany thousands of years to support a wide range of appli-cations (Kuzovkina amp Quigley 2005 Stott 1992) How-ever it has been the focus of domestication for bioenergypurposes for only a relatively short period since the 1970sin North America and Europe For bioenergy breedershave focused their efforts on the shrub willows (subgenusVetix) because of their rapid juvenile growth rates as aresponse to coppicing on a 2‐ to 4‐year cycle that can be
accomplished using farm machinery rather than forestryequipment (Shield Macalpine Hanley amp Karp 2015Smart amp Cameron 2012)
Since shrub willow was not generally recognized as anagricultural crop until very recently there has been littlecommitment to building and maintaining germplasm reposi-tories of willow to support long‐term breeding One excep-tion is the United Kingdom where a large and well‐characterized Salix germplasm collection comprising over1500 accessions is held at Rothamsted Research (Stott1992 Trybush et al 2008) Originally initiated for use inbasketry in 1923 accessions have been added ever sinceIn the United States a germplasm collection of gt350accessions is located at Cornell University to support thebreeding programme there The UK and Cornell collectionshave a relatively small number of accessions in common(around 20) Taken together they represent much of thespecies diversity but only a small fraction of the overallgenetic diversity within the genus There are three activewillow breeding programmes in Europe RothamstedResearch (UK) Salixenergi Europa AB (SEE) and a pro-gramme at the University of Warmia and Mazury in Olsz-tyn (Poland) (abbreviations used in Table 1) There is oneactive US programme based at Cornell University Culti-vars are still being marketed by the European WillowBreeding Programme (EWBP) (UK) which was activelybreeding biomass varieties from 1996 to 2002 Cultivarsare protected by plant breedersrsquo rights (PBRs) in Europeand by plant patents in the United States The sharing ofgenetic resources in the willow community is generallyregulated by material transfer agreements (MTA) and tai-lored licensing agreements although the import of cuttingsinto North America is prohibited except under special quar-antine permit conditions
Efforts to augment breeding germplasm collection fromnature are continuing with phenotypic screening of wildgermplasm performed in field experiments with 177 S pur-purea genotypes in the United States (at sites in Genevaand Portland NY and Morgantown WV) that have beengenotyped using genotyping by sequencing (GBS) (Elshireet al 2011) In addition there are approximately 400accessions of S viminalis in Europe (near Pustnaumls Upp-sala Sweden and Woburn UK (Berlin et al 2014 Hal-lingbaumlck et al 2016) The S viminalis accessions wereinitially genotyped using 38 simple sequence repeats (SSR)markers to assess genetic diversity and screened with~1600 SNPs in genes of potential interest for phenologyand biomass traits Genetics and genomics combined withextensive phenotyping have substantially improved thegenetic basis of biomass‐related traits in willow and arenow being developed in targeted breeding via MAS Thisunderpinning work has been conducted on large specifi-cally developed biparental Salix mapping populations
18 | CLIFTON‐BROWN ET AL
(Hanley amp Karp 2014 Zhou et al 2018) as well asGWAS panels (Hallingbaumlck et al 2016)
Once promising parental combinations are identifiedcrosses are usually performed using fresh pollen frommaterial that has been subject to a phased removal fromcold storage (minus4degC) (Lindegaard amp Barker 1997 Macal-pine Shield Trybush Hayes amp Karp 2008 Mosseler1990) Pollen storage is useful in certain interspecific com-binations where flowering is not naturally synchronizedThis can be overcome by using pollen collection and stor-age protocol which involves extracting pollen using toluene(Kopp Maynard Niella Smart amp Abrahamson 2002)
The main breeding approach to improve willow yieldsrelies on species hybridization to capture hybrid vigour(Fabio et al 2017 Serapiglia Gouker amp Smart 2014) Inthe absence of genotypic models for heterosis breedershave extensively tested general and specific combiningability of parents to produce superior progeny The UKbreeding programmes (EWBP 1996ndash2002 and RothamstedResearch from 2003 on) have performed more than 1500exploratory cross‐pollinations The Cornell programme hassuccessfully completed about 550 crosses since 1998Investment into the characterization of genetic diversitycombined with progeny tests from exploratory crosses hasbeen used to produce hundreds of targeted intraspeciescrosses in the United Kingdom and United States respec-tively (see Table 1) To achieve long‐term gains beyond F1hybrids four intraspecific recurrent selection populationshave been created in the United Kingdom (for S dasycla-dos S viminalis and S miyabeana) and Cornell is pursu-ing recurrent selection of S purpurea Interspecifichybridizations with genotypes selected from the recurrentselection cycles are well advanced in willow with suchcrosses to date totalling 420 in the United Kingdom andover 100 in the United States
While species hybridization is common in Salix it isnot universal Of the crosses attempted about 50 hybri-dize and produce seed (Macalpine Shield amp Karp 2010)As the viability of seed from successful crosses is short (amatter of days at ambient temperatures) proper seed rear-ing and storage protocols are essential (Maroder PregoFacciuto amp Maldonado 2000)
Progeny from crosses are treated in different waysamong the breeding programmes at the seedling stage Inthe United States seedlings are planted into an irrigatedfield where plants are screened for two seasons beforebeing progressed to further field trials In the United King-dom seedlings are planted into trays of compost wherethey remain containerized in an irrigated nursery for theremainder of year one In the United Kingdom seedlingsare subject to two rounds of selection in the nursery yearThe first round takes place in September to select againstsusceptibility to rust infection (Melampsora spp) A second
round of selection in winter assesses tip damage from frostand giant willow aphid infestation In the United Stateswhere the rust pressure is lower screening for Melampsoraspp cannot be performed at the nursery stage Both pro-grammes monitor Melampsora spp pest susceptibilityyield and architecture over multiple years in field trialsSelected material is subject to two rounds of field trials fol-lowed by a final multilocation yield trial to identify vari-eties for commercialization
Promising selections (ie potential cultivars) need to beclonally propagated A rapid in vitro tissue culture propa-gation method has been developed (Palomo‐Riacuteos et al2015) This method can generate about 5000 viable trans-plantable clones from a single plant in just 24 weeks Anin vitro system can also accommodate early selection viamolecular or biochemical markers to increase selectionspeed Conventional breeding systems take 13 years viafour rounds of selection from crossing to selecting a variety(Figure 1) but this has the potential to be reduced to7 years if micropropagation and MAS selection are adopted(Hanley amp Karp 2014 Palomo‐Riacuteos et al 2015)
Willows are currently propagated commercially byplanting winter‐dormant stem cuttings in spring Commer-cial planting systems for willow use mechanical plantersthat cut and insert stem sections from whips into a well‐prepared soil One hectare of stock plants grown in specificmultiplication beds planted at 40000 plants per ha pro-duces planting material for 80 hectares of commercialshort‐rotation coppice willow annually (planted at 15000cuttings per hectare) (Whittaker et al 2016) When com-mercial plantations are established the industry standard isto plant intimate mixtures of ~5 diverse rust (Melampsoraspp)‐resistant varieties (McCracken amp Dawson 1997 VanDen Broek et al 2001)
The foundations for using new plant breeding tech-niques have been established with funding from both thepublic and the private sectors To establish QTL maps 16mapping populations from biparental crosses are understudy in the United Kingdom Nine are under study in theUnited States The average number of individuals in thesefamilies ranges from 150 to 947 (Hanley amp Karp 2014)GS is also being evaluated in S viminalis and preliminaryresults indicate that multiomic approaches combininggenomic and metabolomic data have great potential (Sla-vov amp Davey 2017) For both QTL and GS approachesthe field phenotyping demands are large as several thou-sand individuals need to be phenotyped for a wide rangeof traits These include the following dates of bud burstand growth succession stem height stem density wooddensity and disease resistance The greater the number ofindividuals the more precise the QTL marker maps andGS models are However the logistical and financial chal-lenges of phenotyping large numbers of individuals are
CLIFTON‐BROWN ET AL | 19
considerable because the willow crop is gt5 m tall in thesecond year There is tremendous potential to improve thethroughput of phenotyping using unmanned aerial systemswhich is being tested in the USDA National Institute ofFood and Agriculture (NIFA) Willow SkyCAP project atCornell Further investment in these approaches needs tobe sustained over many years fully realizes the potentialof a marker‐assisted selection programme for willow
To date despite considerable efforts in Europe and theUnited States to establish a routine transformation systemthere has not been a breakthrough in willow but attemptsare ongoing As some form of transformation is typically aprerequisite for genome editing techniques these have notyet been applied to willow
In Europe there are 53 short‐rotation coppice (SRC) bio-mass willow cultivars registered with the Community PlantVariety Office (CPVO) for PBRs of which ~25 are availablecommercially in the United Kingdom There are eightpatented cultivars commercially available in the UnitedStates In Sweden there are nine commercial cultivars regis-tered in Europe and two others which are unregistered(httpssalixenergiseplanting-material) Furthermore thereare about 20 precommercial hybrids in final yield trials inboth the United States and the United Kingdom It has beenestimated that it would take two years to produce the stockrequired to plant 50 ha commercially from the plant stock inthe final yield trials Breeding programmes have alreadydelivered rust‐resistant varieties and increases in yield to themarket The adoption of advanced breeding technologies willlikely lead to a step change in improving traits of interest
5 | POPLAR
Poplar a fast‐growing tree from the northern hemispherewith a small genome size has been adopted for commercialforestry and scientific purposes The genus Populus con-sists of about 29 species classified in six different sectionsPopulus (formerly Leuce) Tacamahaca Aigeiros AbasoTuranga and Leucoides (Eckenwalder 1996) The Populusspecies of most interest for breeding and testing in the Uni-ted States and Europe are P nigra P deltoides P maxi-mowiczii and P trichocarpa (Stanton 2014) Populusclones for biomass production are being developed byintra‐ and interspecies hybridization (DeWoody Trewin ampTaylor 2015 Richardson Isebrands amp Ball 2014 van derSchoot et al 2000) Recurrent selection approaches areused for gradual population improvement and to create eliteclonal lines for commercialization (Berguson McMahon ampRiemenschneider 2017 Neale amp Kremer 2011) Currentlypoplar breeding in the United States occurs in industrialand academic programmes located in the Southeast theMidwest and the Pacific Northwest These use six speciesand five interspecific taxa (Stanton 2014)
The southeastern programme historically focused onrecurrent selection of P deltoides from accessions made inthe lower Mississippi River alluvial plain (Robison Rous-seau amp Zhang 2006) More recently the genetic base hasbeen broadened to produce interspecific hybrids with resis-tance to the fungal infection Septoria musiva which causescankers
In the midwest of the United States populationimprovement efforts are focused on P deltoides selectionsfrom native provenances and hybrid crosses with acces-sions introduced from Europe Interspecific intercontinental(Europe and America) hybrid crosses between P nigra andP deltoides (P times canadensis) are behind many of theleading commercial hybrids which are the most advancedbreeding materials for many applications and regions InMinnesota previous breeding experience and efforts utiliz-ing P maximowiczii and P trichocarpa have been discon-tinued due to Septoria susceptibility and a lack of coldhardiness (Berguson et al 2017) Traits targeted forimprovement include yieldgrowth rate cold hardinessadventitious rooting resistance to Septoria and Melamp-sora leaf rust and stem form The Upper Midwest pro-gramme also carries out wide hybridizations within thesection Populus The P times wettsteinii (P trem-ula times P tremuloides) taxon is bred for gains in growthrate wood quality and resistance to the fungus Entoleucamammata which causes hypoxylon canker (David ampAnderson 2002)
In the Pacific Northwest GreenWood Resources Incleads poplar breeding that emphasizes interspecifichybrid improvement of P times generosa (P deltoides timesP trichocarpa and reciprocal) and P deltoides times P maxi-mowiczii taxa for coastal regions and the P times canadensistaxon for the drier continental regions Intraspecificimprovement of second‐generation breeding populations ofP deltoides P nigra P maximowiczii and P trichocarpaare also involved (Stanton et al 2010) The present focusof GreenWood Resourcesrsquo hybridization is bioenergy feed-stock improvement concentrating on coppice yield wood‐specific gravity and rate of sugar release
Industrial interest in poplar in the United States has his-torically come from the pulp and paper sector althoughveneer and dimensional lumber markets have been pursuedat times Currently the biomass market for liquid trans-portation fuels is being emphasized along with the use oftraditional and improved poplar genotypes for ecosystemservices such as phytoremediation (Tuskan amp Walsh 2001Zalesny et al 2016)
In Europe there are breeding programmes in FranceGermany Italy and Sweden These include the following(a) Alasia Franco Vivai (AFV) programme in northernItaly (b) the French programme led by the poplar Scien-tific Interest Group (GIS Peuplier) and carried out
20 | CLIFTON‐BROWN ET AL
collaboratively between the National Institute for Agricul-tural Research (INRA) the National Research Unit ofScience and Technology for Environment and Agriculture(IRSTEA) and the Forest Cellulose Wood Constructionand Furniture Technology Institute (FCBA) (c) the Ger-man programme at Northwest German Forest Research Sta-tion (NW‐FVA) at Hannoversch Muumlnden and (d) theSwedish programme at the Swedish University of Agricul-tural Sciences and SweTree Technologies AB (Table 1)
AFV leads an Italian poplar breeding programme usingextensive field‐grown germplasm collections of P albaP deltoides P nigra and P trichocarpa While interspeci-fic hybridization uses several taxa the focus is onP times canadensis The breeding programme addresses dis-ease resistance (Marssonina brunnea Melampsora larici‐populina Discosporium populeum and poplar mosaicvirus) growth rate and photoperiod adaptation AFV andGreenWood Resources collaborate in poplar improvementin Europe through the exchange of frozen pollen and seedfor reciprocal breeding projects Plantations in Poland andRomania are currently the focus of the collaboration
The ongoing French GIS Peuplier is developing a long‐term breeding programme based on intraspecific recurrentselection for the four parental species (P deltoides P tri-chocarpa P nigra and P maximowiczii) designed to betterbenefit from hybrid vigour demonstrated by the interspeci-fic crosses P canadensis P deltoides times P trichocarpaand P trichocarpa times P maximowiczii Current selectionpriorities are targeting adaptation to soil and climate condi-tions resistance and tolerance to the most economicallyimportant diseases and pests high volume production underSRC and traditional poplar cultivation regimes as well aswood quality of interest by different markets Currentlygenomic selection is under exploration to increase selectionaccuracy and selection intensity while maintaining geneticdiversity over generations
The German NW‐FVA programme is breeding intersec-tional AigeirosndashTacamahaca hybrids with a focus on resis-tance to Pollaccia elegans Xanthomonas populiDothichiza spp Marssonina brunnea and Melampsoraspp (Stanton 2014) Various cross combinations ofP maximowiczii P trichocarpa P nigra and P deltoideshave led to new cultivars suitable for deployment in vari-etal mixtures of five to ten genotypes of complementarystature high productivity and phenotypic stability (Weis-gerber 1993) The current priority is the selection of culti-vars for high‐yield short‐rotation biomass production Sixhundred P nigra genotypes are maintained in an ex situconservation programme An in situ P nigra conservationeffort involves an inventory of native stands which havebeen molecular fingerprinted for identity and diversity
The Swedish programme is concentrating on locallyadapted genotypes used for short‐rotation forestry (SRF)
because these meet the needs of the current pulping mar-kets Several field trials have shown that commercial poplarclones tested and deployed in Southern and Central Europeare not well adapted to photoperiods and low temperaturesin Sweden and in the Baltics Consequently SwedishUniversity of Agricultural Sciences and SweTree Technolo-gies AB started breeding in Sweden in 1990s to producepoplar clones better adapted to local climates and markets
51 | Molecular breeding technologies
Poplar genetic improvement cannot be rapidly achievedthrough traditional methods alone because of the longbreeding cycles outcrossing breeding systems and highheterozygosity Integrating modern genetic genomic andphenomics techniques with conventional breeding has thepotential to expedite poplar improvement
The genome of poplar has been sequenced (Tuskanet al 2006) It has an estimated genome size of485 plusmn 10 Mbp divided into 19 chromosomes This is smal-ler than other PBCs and makes poplar more amenable togenetic engineering (transgenesis) GS and genome editingPoplar has seen major investment in both the United Statesand Europe being the model system for woody perennialplant genetics and genomics research
52 | Targets for genetic modification
Traits targeted include wood properties (lignin content andcomposition) earlylate flowering male sterility to addressbiosafety regulation issues enhanced yield traits and herbi-cide tolerance These extensive transgenic experiments haveshown differences in recalcitrance to in vitro regenerationand genetic transformation in some of the most importantcommercial hybrid poplars (Alburquerque et al 2016)Further transgene expression stability is being studied Sofar China is the only country known to have commerciallyused transgenic insect‐resistant poplar A precommercialherbicide‐tolerant poplar was trialled for 8 years in the Uni-ted States (Li Meilan Ma Barish amp Strauss 2008) butcould not be released due to stringent environmental riskassessments required for regulatory approval This increasestranslation costs and delays reducing investor confidencefor commercial deployment (Harfouche et al 2011)
The first field trials of transgenic poplar were performedin France in 1987 (Fillatti Sellmer Mccown Haissig ampComai 1987) and in Belgium in 1988 (Deblock 1990)Although there have been a total of 28 research‐scale GMpoplar field trials approved in the European Union underCouncil Directive 90220EEC since October 1991 (inPoland Belgium Finland France Germany Spain Swe-den and in the United Kingdom (Pilate et al 2016) onlyauthorizations in Poland and Belgium are in place today In
CLIFTON‐BROWN ET AL | 21
the United States regulatory notifications and permits fornearly 20000 transgenic poplar trees derived from approxi-mately 600 different constructs have been issued since1995 by the USDAs Animal and Plant Health InspectionService (APHIS) (Strauss et al 2016)
53 | Genome editing CRISPR technologies
Clustered regularly interspaced palindromic repeats(CRISPR) and the CRISPR‐associated (CRISPR‐Cas)nucleases are a groundbreaking genome‐engineering toolthat complements classical plant breeding and transgenicmethods (Moreno‐Mateos et al 2017) Only two publishedstudies in poplar have applied the CRISPRCas9 technol-ogy One is in P tomentosa in which an endogenous phy-toene desaturase gene (PtoPDS) was successfully disruptedsite specifically in the first generation of transgenic plantsresulting in an albino and dwarf phenotype (Fan et al2015) The second was in P tremula times alba in whichhigh CRISPR‐Cas9 mutational efficiency was achieved forthree 4‐coumarateCoA ligase (4Cl) genes 4CL1 4CL2and 4CL5 associated with lignin and flavonoid biosynthe-sis (Zhou et al 2015) Due to its low cost precision andrapidness it is very probable that cultivars or clones pro-duced using CRISPR technology will be ready for market-ing in the near future (Yin et al 2017) Recently aCRISPR with a smaller associated endonuclease has beendiscovered from Prevotella and Francisella 1 (Cpf1) whichmay have advantages over Cas9 In addition there arereports of DNA‐free editing in plants using both CRISPRCpf1 and CRISPR Cas9 for example Ref (Kim et al2017 Mahfouz 2017 Zaidi et al 2017)
It remains unresolved whether plants modified by genomeediting will be regulated as genetically modified organisms(GMOs) by the relevant authorities in different countries(Lozano‐Juste amp Cutler 2014) Regulations to cover thesenew breeding techniques are still evolving but those coun-tries who have published specific guidance (including UnitedStates Argentina and Chile) are indicating that plants pos-sessing simple genome edits will not be regulated as conven-tional transgenesis (Jones 2015b) The first generation ofgenome‐edited crops will likely be phenocopy gene knock-outs that already exist to produce ldquonature identicalrdquo traits thatis traits that could also be derived by conventional breedingDespite this confidence in applying these new powerfulbreeding tools remains limited owing to the uncertain regula-tory environment in many parts of the world (Gao 2018)including the recent ECJ 2018 rulings mentioned earlier
54 | Genomics‐based breeding technologies
Poplar breeding programs are becoming well equipped withuseful genomics tools and resources that are critical to
explore genomewide variability and make use of the varia-tion for enhancing genetic gains Deep transcriptomesequencing resequencing of alternate genomes and GBStechnology for genomewide marker detection using next‐generation sequencing (NGS) are yielding valuable geno-mics tools GWAS with NGS‐based markers facilitatesmarker identification for MAS breeding by design and GS
GWAS approaches have provided a deeper understand-ing of genome function as well as allelic architectures ofcomplex traits (Huang et al 2010) and have been widelyimplemented in poplar for wood characteristics (Porthet al 2013) stomatal patterning carbon gain versus dis-ease resistance (McKown et al 2014) height and phenol-ogy (Evans et al 2014) cell wall chemistry (Mucheroet al 2015) growth and cell walls traits (Fahrenkroget al 2017) bark roughness (Bdeir et al 2017) andheight and diameter growth (Liu et al 2018) Using high‐throughput sequencing and genotyping platforms an enor-mous amount of SNP markers have been used to character-ize the linkage disequilibrium (LD) in poplar (eg Slavovet al 2012 discussed below)
The genetic architecture of photoperiodic traits in peren-nial trees is complex involving many loci However itshows high levels of conservation during evolution (Mau-rya amp Bhalerao 2017) These genomics tools can thereforebe used to address adaptation issues and fine‐tune themovement of elite lines into new environments For exam-ple poor timing of spring bud burst and autumn bud setcan result in frost damage resulting in yield losses (Ilstedt1996) These have been studied in P tremula genotypesalong a latitudinal cline in Sweden (~56ndash66oN) and haverevealed high nucleotide polymorphism in two nonsynony-mous SNPs within and around the phytochrome B2 locus(Ingvarsson Garcia Hall Luquez amp Jansson 2006 Ing-varsson Garcia Luquez Hall amp Jansson 2008) Rese-quencing 94 of these P tremula genotypes for GWASshowed that noncoding variation of a single genomicregion containing the PtFT2 gene described 65 ofobserved genetic variation in bud set along the latitudinalcline (Tan 2018)
Resequencing genomes is currently the most rapid andeffective method detecting genetic differences betweenvariants and for linking loci to complex and importantagronomical and biomass traits thus addressing breedingchallenges associated with long‐lived plants like poplars
To date whole genome resequencing initiatives havebeen launched for several poplar species and genotypes InPopulus LD studies based on genome resequencing sug-gested the feasibility of GWAS in undomesticated popula-tions (Slavov et al 2012) This plant population is beingused to inform breeding for bioenergy development Forexample the detection of reliable phenotypegenotype asso-ciations and molecular signatures of selection requires a
22 | CLIFTON‐BROWN ET AL
detailed knowledge about genomewide patterns of allelefrequency variation LD and recombination suggesting thatGWAS and GS in undomesticated populations may bemore feasible in Populus than previously assumed Slavovet al (2012) have resequenced 16 genomes of P tri-chocarpa and genotyped 120 trees from 10 subpopulationsusing 29213 SNPs (Geraldes et al 2013) The largest everSNP data set of genetic variations in poplar has recentlybeen released providing useful information for breedinghttpswwwbioenergycenterorgbescgwasindexcfmAlso deep sequencing of transcriptomes using RNA‐Seqhas been used for identification of functional genes andmolecular markers that is polymorphism markers andSSRs A multitissue and multiple experimental data set forP trichocarpa RNA‐Seq is publicly available (httpsjgidoegovdoe-jgi-plant-flagship-gene-atlas)
The availability of genomic information of DNA‐con-taining cell organelles (nucleus chloroplast and mitochon-dria) will also allow a holistic approach in poplarmolecular breeding in the near future (Kersten et al2016) Complete Populus genome sequences are availablefor nucleus (P trichocarpa section Tacamahaca) andchloroplasts (seven species and two clones from P trem-ula W52 and P tremula times P alba 717ndash1B4) A compara-tive approach revealed structural and functionalinformation broadening the knowledge base of PopuluscpDNA and stimulating future diagnostic marker develop-ment The availability of whole genome sequences of thesecellular compartments of P tremula holds promise forboosting marker‐assisted poplar breeding Other nucleargenome sequences from additional Populus species arenow available (eg P deltoides (httpsphytozomejgidoegovpz) and will become available in the forthcomingyears (eg P tremula and P tremuloidesmdashPopGenIE(Sjodin Street Sandberg Gustafsson amp Jansson 2009))Recently the characterization of the poplar pan‐genome bygenomewide identification of structural variation in threecrossable poplar species P nigra P deltoides and P tri-chocarpa revealed a deeper understanding of the role ofinter‐ and intraspecific structural variants in poplar pheno-type and may have important implications for breedingparticularly interspecific hybrids (Pinosio et al 2016)
GS has been proposed as an alternative to MAS in cropimprovement (Bernardo amp Yu 2007 Heffner Sorrells ampJannink 2009) GS is particularly well suited for specieswith long generation times for characteristics that displaymoderate‐to‐low heritability for traits that are expensive tomeasure and for selection of traits expressed late in the lifecycle as is the case for most traits of commercial value inforestry (Harfouche et al 2012) Current joint genomesequencing efforts to implement GS in poplar using geno-mic‐estimated breeding values for bioenergy conversiontraits from 49 P trichocarpa families and 20 full‐sib
progeny are taking place at the Oak Ridge National Labo-ratory and GreenWood Resources (Brian Stanton personalcommunication httpscbiornlgov) These data togetherwith the resequenced GWAS population data will be thebasis for developing GS algorithms Genomic breedingtools have been developed for the intraspecific programmetargeting yield resistance to Venturia shoot blightMelampsora leaf rust resistance to Cryptorhynchus lapathistem form wood‐specific gravity and wind firmness (Evanset al 2014 Guerra et al 2016) A newly developedldquobreeding with rare defective allelesrdquo (BRDA) technologyhas been developed to exploit natural variation of P nigraand identify defective variants of genes predicted by priortransgenic research to impact lignin properties Individualtrees carrying naturally defective alleles can then be incor-porated directly into breeding programs thereby bypassingthe need for transgenics (Vanholme et al 2013) Thisnovel breeding technology offers a reverse genetics com-plement to emerging GS for targeted improvement of quan-titative traits (Tsai 2013)
55 | Phenomics‐assisted breeding technology
Phenomics involves the characterization of phenomesmdashthefull set of phenotypes of given individual plants (HouleGovindaraju amp Omholt 2010) Traditional phenotypingtools which inefficiently measure a limited set of pheno-types have become a bottleneck in plant breeding studiesHigh‐throughput plant phenotyping facilities provide accu-rate screening of thousands of plant breeding lines clones orpopulations over time (Fu 2015) are critical for acceleratinggenomics‐based breeding Automated image collection andanalysis phenomics technologies allow accurate and nonde-structive measurements of a diversity of phenotypic traits inlarge breeding populations (Gegas Gay Camargo amp Doo-nan 2014 Goggin Lorence amp Topp 2015 Ludovisi et al2017 Shakoor Lee amp Mockler 2017) One important con-sideration is the identification of relevant and quantifiabletarget traits that are early diagnostic indicators of biomassyield Good progress has been made in elucidating theseunderpinning morpho‐physiological traits that are amenableto remote sensing in Populus (Harfouche Meilan amp Altman2014 Rae Robinson Street amp Taylor 2004) Morerecently Ludovisi et al (2017) developed a novel methodol-ogy for field phenomics of drought stress in a P nigra F2partially inbred population using thermal infrared imagesrecorded from an unmanned aerial vehicle‐based platform
Energy is the current main market for poplar biomass butthe market return provided is not sufficient to support pro-duction expansion even with added demand for environmen-tal and land management ldquoecosystem servicesrdquo such as thetreatment of effluent phytoremediation riparian buffer zonesand agro‐forestry plantings Aviation fuel is a significant
CLIFTON‐BROWN ET AL | 23
target market (Crawford et al 2016) To serve this marketand to reduce current carbon costs of production (Budsberget al 2016) key improvement traits in addition to yield(eg coppice regeneration pestdisease resistance water‐and nutrient‐use efficiencies) will be trace greenhouse gas(GHG) emissions (eg isoprene volatiles) site adaptabilityand biomass conversion efficiency Efforts are underway tohave national environmental protection agenciesrsquo approvalfor poplar hybrids qualifying for renewable energy credits
6 | REFLECTIONS ON THECOMMERCIALIZATIONCHALLENGE
The research and innovation activities reviewed in thispaper aim to advance the genetic improvement of speciesthat can provide feedstocks for bioenergy applicationsshould those markets eventually develop These marketsneed to generate sufficient revenue and adequately dis-tribute it to the actors along the value chain The work onall four crops shares one thing in common long‐termefforts to integrate fundamental knowledge into breedingand crop development along a research and development(RampD) pipeline The development of miscanthus led byAberystwyth University exemplifies the concerted researcheffort that has integrated the RampD activities from eight pro-jects over 14 years with background core research fundingalong an emerging innovation chain (Figure 2) This pro-gramme has produced a first range of conventionally bredseeded interspecies hybrids which are now in upscaling tri-als (Table 3) The application of molecular approaches(Table 5) with further conventional breeding (Table 4)offers the prospect of a second range of improved seededhybrids This example shows that research‐based support ofthe development of new crops or crop types requires along‐term commitment that goes beyond that normallyavailable from project‐based funding (Figure 1) Innovationin this sector requires continuous resourcing of conven-tional breeding operations and capability to minimize timeand investment losses caused by funding discontinuities
This challenge is increased further by the well‐knownmarket failure in the breeding of many agricultural cropspecies The UK Department for Environment Food andRural Affairs (Defra) examined the role of geneticimprovement in relation to nonmarket outcomes such asenvironmental protection and concluded that public invest-ment in breeding was required if profound market failure isto be addressed (Defra 2002) With the exception ofwidely grown hybrid crops such as maize and some high‐value horticultural crops royalties arising from plant breed-ersrsquo rights or other returns to breeders fail to adequatelycompensate for the full cost for research‐based plant breed-ing The result even for major crops such as wheat is sub‐
optimal investment and suboptimal returns for society Thismarket failure is especially acute for perennial crops devel-oped for improved sustainability rather than consumerappeal (Tracy et al 2016) Figure 3 illustrates the underly-ing challenge of capturing value for the breeding effortThe ldquovalley of deathrdquo that results from the low and delayedreturns to investment applies generally to the research‐to‐product innovation pipeline (Beard Ford Koutsky amp Spi-wak 2009) and certainly to most agricultural crop speciesHowever this schematic is particularly relevant to PBCsMost of the value for society from the improved breedingof these crops comes from changes in how agricultural landis used that is it depends on the increased production ofthese crops The value for society includes many ecosys-tems benefits the effects of a return to seminatural peren-nial crop cover that protects soils the increase in soilcarbon storage the protection of vulnerable land or the cul-tivation of polluted soils and the reductions in GHG emis-sions (Lewandowski 2016) By its very nature theproduction of biomass on agricultural land marginal forfood production challenges farm‐level profitability Thecosts of planting material and one‐time nature of cropestablishment are major early‐stage costs and therefore theopportunities for conventional royalty capture by breedersthat are manifold for annual crops are limited for PBCs(Hastings et al 2017) Public investment in developingPBCs for the nonfood biobased sector needs to providemore long‐term support for this critical foundation to a sus-tainable bioeconomy
7 | CONCLUSIONS
This paper provides an overview of research‐based plantbreeding in four leading PBCs For all four PBC generasignificant progress has been made in genetic improvementthrough collaboration between research scientists and thoseoperating ongoing breeding programmes Compared withthe main food crops most PBC breeding programmes dateback only a few decades (Table 1) This breeding efforthas thus co‐evolved with molecular biology and the result-ing ‐omics technologies that can support breeding Thedevelopment of all four PBCs has depended strongly onpublic investment in research and innovation The natureand driver of the investment varied In close associationwith public research organizations or universities all theseprogrammes started with germplasm collection and charac-terization which underpin the selection of parents forexploratory wide crosses for progeny testing (Figure 1Table 4)
Public support for switchgrass in North America wasexplicitly linked to plant breeding with 12 breeding pro-grammes supported in the United States and CanadaSwitchgrass breeding efforts to date using conventional
24 | CLIFTON‐BROWN ET AL
breeding have resulted in over 36 registered cultivars inthe United States (Table 1) with the development of dedi-cated biomass‐type cultivars coming within the past fewyears While ‐omics technologies have been incorporatedinto several of these breeding programmes they have notyet led to commercial deployment in either conventional orhybrid cultivars
Willow genetic improvement was led by the researchcommunity closely linked to plant breeding programmesWillow and poplar have the longest record of public invest-ment in genetic improvement that can be traced back to1920s in the United Kingdom and United States respec-tively Like switchgrass breeding programmes for willoware connected to public research efforts The United King-dom in partnership with the programme based at CornellUniversity remains the European leader in willowimprovement with a long‐term breeding effort closelylinked to supporting biological research at Rothamsted Inwillow F1 hybrids have produced impressive yield gainsover parental germplasm by capturing hybrid vigour Over30 willow clones are commercially available in the UnitedStates and Europe and a further ~90 are under precommer-cial testing (Table 1)
Compared with willow and poplar miscanthus is a rela-tive newcomer with all the current breeding programmesstarting in the 2000s Clonal M times giganteus propagated byrhizomes is expected to be replaced by more readily scal-able seeded hybrids from intra‐ (M sinensis) and inter‐(M sacchariflorus times M sinensis) species crosses with highseed multiplication rates (of gt2000) The first group ofhybrid cultivars is expected to be market‐ready around2022
Of the four genera used as PBCs Populus is the mostadvanced in terms of achievements in biological researchas a result of its use as a model for basic research of treesMuch of this biological research is not directly connectedto plant breeding Nevertheless reflecting the fact thatpoplar is widely grown as a single‐stem tree in SRF thereare about 60 commercially available clones and an addi-tional 80 clones in commercial pipelines (Table 1) Trans-genic poplar hybrids have moved beyond proof of conceptto commercial reality in China
Many PBC programmes have initiated long‐term con-ventional recurrent selection breeding cycles for populationimprovement which is a key process in increasing yieldthrough hybrid vigour As this approach requires manyyears most programmes are experimenting with molecularbreeding methods as these have the potential to accelerateprecision breeding For all four PBCs investments in basicgenetic and genomic resources including the developmentof mapping populations for QTLs and whole genomesequences are available to support long‐term advancesMore recently association genetics with panels of diversegermplasm are being used as training populations for GSmodels (Table 5) These efforts are benefitting from pub-licly available DNA sequences and whole genome assem-blies in crop databases Key to these accelerated breedingtechnologies are developments in novel phenomics tech-nologies to bridge the genotypephenotype gap In poplarnovel remote sensing field phenotyping is now beingdeployed to assist breeders These advances are being com-bined with in vitro and in planta modern molecular breed-ing techniques such as CRISPR (Table 5) CRISPRtechnology for genome editing has been proven in poplar
FIGURE 2 A schematic developmentpathway for miscanthus in the UnitedKingdom related to the investment in RampDprojects at Aberystwyth (top coloured areasfor projects in the three categories basicresearch breeding and commercialupscaling) leading to a projected croppingarea of 350000 ha by 2030 with clonal andsuccessive ranges of improved seed‐basedhybrids Purple represents theBiotechnology and Biological SciencesResearch Council (BBSRC) and brown theDepartment for Environment Food andRural Affairs (Defra) (UK Nationalfunding) blue bars represent EU fundingand green private sector funding (Terravestaand CERES) and GIANT‐LINK andMiscanthus Upscaling Technology (MUST)are publicndashprivate‐initiatives (PPI)
CLIFTON‐BROWN ET AL | 25
This technology is also being applied in switchgrass andmiscanthus (Table 5) but the future of CRISPR in com-mercial breeding for the European market is uncertain inthe light of recent ECJ 2018 rulings
There is integration of research and plant breeding itselfin all four PBCs Therefore estimating the ongoing costs ofmaintaining these breeding programmes is difficult Invest-ment in research also seeks wider benefits associated withtechnological advances in plant science rather than cultivardevelopment per se However in all cases the conventionalbreeding cycle shown in Figure 1 is the basic ldquoenginerdquo withmolecular technologies (‐omics) serving to accelerate thisengine The history of the development shows that the exis-tence of these breeding programmes is essential to gain ben-efits from the biological research Despite this it is thisessential step that is at most risk from reductions in invest-ment A conventional breeding programme typicallyrequires a breeder and several technicians who are supported
over the long term (20ndash30 years Figure 1) at costs of about05 to 10 million Euro per year (as of 2018) The analysisreported here shows that the time needed to perform onecycle of conventional breeding bringing germplasm fromthe wild to a commercial hybrid ranged from 11 years inswitchgrass to 26 years in poplar (Figure 1) In a maturecrop grown on over 100000 ha with effective cultivar pro-tection and a suitable business model this level of revenuecould come from royalties Until such levels are reachedPBCs lie in the innovation valley of death (Figure 3) andneed public support
Applying industrial ldquotechnology readiness levelsrdquo (TRL)originally developed for aerospace (Heacuteder 2017) to ourplant breeding efforts we estimate many promising hybridscultivars are at TRL levels of 3ndash4 In Table 3 experts ineach crop estimate that it would take 3 years from now toupscale planting material from leading cultivars in plot tri-als to 100 ha
FIGURE 3 A schematic relating some of the steps in the innovation chain from relatively basic crop science research through to thedeployment in commercial cropping systems and value chains The shape of the funnel above the expanding development and deploymentrepresents the availability of investment along the development chain from relatively basic research at the top to the upscaled deployment at thebottom Plant breeding links the research effort with the development of cropping systems The constriction represents the constrained fundingfor breeding that links conventional public research investment and the potential returns from commercial development The handover pointsbetween publicly funded work to develop the germplasm resources (often known as prebreeding) the breeding and the subsequent cropdevelopment are shown on the left The constriction point is aggravated by the lack academic rewards for this essential breeding activity Theoutcome is such that this innovation system is constrained by the precarious resourcing of plant breeding The authorsrsquo assessment ofdevelopment status of the four species is shown (poplar having two one for short‐rotation coppice (SRC) poplar and one for the more traditionalshort‐rotation forestry (SRF)) The four new perennial biomass crops (PBCs) are now in the critical phase of depending of plant breedingprogress without the income stream from a large crop production base
26 | CLIFTON‐BROWN ET AL
Taking the UK example mentioned in the introductionplanting rates of ~35000 ha per year from 2022 onwardsare needed to reach over 1 m ha by 2050 Ongoing workin the UK‐funded Miscanthus Upscaling Technology(MUST) project shows that ramping annual hybrid seedproduction from the current level of sufficient seed for10 ha in 2018 to 35000 ha would take about 5 yearsassuming no setbacks If current hybrids of any of the fourPBCs in the upscaling pipeline fail on any step for exam-ple lower than expected multiplication rates or unforeseenagronomic barriers then further selections from ongoingbreeding are needed to replace earlier candidates
In conclusion the breeding foundations have been laidwell for switchgrass miscanthus willow and poplar owingto public funding over the long time periods necessaryImproved cultivars or genotypes are available that could bescaled up over a few years if real sustained market oppor-tunities emerged in response to sustained favourable poli-cies and industrial market pull The potential contributionsof growing and using these PBCs for socioeconomic andenvironmental benefits are clear but how farmers andothers in commercial value chains are rewarded for mass‐scale deployment as is necessary is not obvious at presentTherefore mass‐scale deployment of these lignocellulosecrops needs developments outside the breeding arenas todrive breeding activities more rapidly and extensively
8 | UNCITED REFERENCE
Huang et al (2019)
ACKNOWLEDGEMENTS
UK The UK‐led miscanthus research and breeding wasmainly supported by the Biotechnology and BiologicalSciences Research Council (BBSRC) Department for Envi-ronment Food and Rural Affairs (Defra) the BBSRC CSPstrategic funding grant BBCSP17301 Innovate UKBBSRC ldquoMUSTrdquo BBN0161491 CERES Inc and Ter-ravesta Ltd through the GIANT‐LINK project (LK0863)Genomic selection and genomewide association study activi-ties were supported by BBSRC grant BBK01711X1 theBBSRC strategic programme grant on Energy Grasses ampBio‐refining BBSEW10963A01 The UK‐led willow RampDwork reported here was supported by BBSRC (BBSEC00005199 BBSEC00005201 BBG0162161 BBE0068331 BBG00580X1 and BBSEC000I0410) Defra(NF0424 NF0426) and the Department of Trade and Indus-try (DTI) (BW6005990000) IT The Brain Gain Program(Rientro dei cervelli) of the Italian Ministry of EducationUniversity and Research supports Antoine Harfouche USContributions by Gerald Tuskan to this manuscript were sup-ported by the Center for Bioenergy Innovation a US
Department of Energy Bioenergy Research Center supportedby the Office of Biological and Environmental Research inthe DOE Office of Science under contract number DE‐AC05‐00OR22725 Willow breeding efforts at CornellUniversity have been supported by grants from the USDepartment of Agriculture National Institute of Food andAgriculture Contributions by the University of Illinois weresupported primarily by the DOE Office of Science Office ofBiological and Environmental Research (BER) grant nosDE‐SC0006634 DE‐SC0012379 and DE‐SC0018420 (Cen-ter for Advanced Bioenergy and Bioproducts Innovation)and the Energy Biosciences Institute EU We would like tofurther acknowledge contributions from the EU projectsldquoOPTIMISCrdquo FP7‐289159 on miscanthus and ldquoWATBIOrdquoFP7‐311929 on poplar and miscanthus as well as ldquoGRACErdquoH2020‐EU326 Biobased Industries Joint Technology Ini-tiative (BBI‐JTI) Project ID 745012 on miscanthus
CONFLICT OF INTEREST
The authors declare that progress reported in this paperwhich includes input from industrial partners is not biasedby their business interests
ORCID
John Clifton‐Brown httporcidorg0000-0001-6477-5452Antoine Harfouche httporcidorg0000-0003-3998-0694Lawrence B Smart httporcidorg0000-0002-7812-7736Danny Awty‐Carroll httporcidorg0000-0001-5855-0775VasileBotnari httpsorcidorg0000-0002-0470-0384Lindsay V Clark httporcidorg0000-0002-3881-9252SalvatoreCosentino httporcidorg0000-0001-7076-8777Lin SHuang httporcidorg0000-0003-3079-326XJon P McCalmont httporcidorg0000-0002-5978-9574Paul Robson httporcidorg0000-0003-1841-3594AnatoliiSandu httporcidorg0000-0002-7900-448XDaniloScordia httporcidorg0000-0002-3822-788XRezaShafiei httporcidorg0000-0003-0891-0730Gail Taylor httporcidorg0000-0001-8470-6390IrisLewandowski httporcidorg0000-0002-0388-4521
REFERENCES
Alburquerque N Baldacci‐Cresp F Baucher M Casacuberta JM Collonnier C ElJaziri M hellip Burgos L (2016) New trans-formation technologies for trees In C Vettori F Gallardo H
CLIFTON‐BROWN ET AL | 27
Haumlggman V Kazana F Migliacci G Pilateamp M Fladung(Eds) Biosafety of forest transgenic trees (pp 31ndash66) DordrechtNetherlands Springer
Argus G W (2007) Salix (Salicaceae) distribution maps and a syn-opsis of their classification in North America north of MexicoHarvard Papers in Botany 12 335ndash368 httpsdoiorg1031001043-4534(2007)12[335SSDMAA]20CO2
Atienza S G Ramirez M C amp Martin A (2003) Mapping‐QTLscontrolling flowering date in Miscanthus sinensis Anderss CerealResearch Communications 31 265ndash271
Atienza S G Satovic Z Petersen K K Dolstra O amp Martin A(2003) Identification of QTLs influencing agronomic traits inMiscanthus sinensis Anderss I Total height flag‐leaf height andstem diameter Theoretical and Applied Genetics 107 123ndash129
Atienza S G Satovic Z Petersen K K Dolstra O amp Martin A(2003) Influencing combustion quality in Miscanthus sinensisAnderss Identification of QTLs for calcium phosphorus and sul-phur content Plant Breeding 122 141ndash145
Bdeir R Muchero W Yordanov Y Tuskan G A Busov V ampGailing O (2017) Quantitative trait locus mapping of Populusbark features and stem diameter BMC Plant Biology 17 224httpsdoiorg101186s12870-017-1166-4
Beard T R Ford G S Koutsky T M amp Spiwak L J (2009) AValley of Death in the innovation sequence An economic investi-gation Research Evaluation 18 343ndash356 httpsdoiorg103152095820209X481057
Berguson W McMahon B amp Riemenschneider D (2017) Addi-tive and non‐additive genetic variances for tree growth in severalhybrid poplar populations and implications regarding breedingstrategy Silvae Genetica 66(1) 33ndash39 httpsdoiorg101515sg-2017-0005
Berlin S Trybush S O Fogelqvist J Gyllenstrand N Halling-baumlck H R Aringhman I hellip Hanley S J (2014) Genetic diversitypopulation structure and phenotypic variation in European Salixviminalis L (Salicaceae) Tree Genetics amp Genomes 10 1595ndash1610 httpsdoiorg101007s11295-014-0782-5
Bernardo R amp Yu J (2007) Prospects for genomewide selectionfor quantitative traits in maize Crop Science 47 1082ndash1090httpsdoiorg102135cropsci2006110690
Biswal A K Atmodjo M A Li M Baxter H L Yoo C GPu Y hellip Mohnen D (2018) Sugar release and growth of bio-fuel crops are improved by downregulation of pectin biosynthesisNature Biotechnology 36 249ndash257
Bmu (2009) National biomass action plan for Germany (p 17) Bun-desministerium fuumlr Umwelt Naturschutz und ReaktorsicherheitBundesministerium fuumlr Ernaumlhrung (BMU) amp Landwirtschaft undVerbraucherschutz (BMELV) 11055 Berlin | Germany Retrievedfrom httpwwwbmeldeSharedDocsDownloadsENPublicationsBiomassActionPlanpdf__blob=publicationFile
Brummer E C (1999) Capturing heterosis in forage crop cultivardevelopment Crop Science 39 943ndash954 httpsdoiorg102135cropsci19990011183X003900040001x
Budsberg E Crawford J T Morgan H Chin W S Bura R ampGustafson R (2016) Hydrocarbon bio‐jet fuel from bioconver-sion of poplar biomass Life cycle assessment Biotechnology forBiofuels 9 170 httpsdoiorg101186s13068-016-0582-2
Burris K P Dlugosz E M Collins A G Stewart C N amp Lena-ghan S C (2016) Development of a rapid low‐cost protoplasttransfection system for switchgrass (Panicum virgatum L) Plant
Cell Reports 35 693ndash704 httpsdoiorg101007s00299-015-1913-7
Casler M (2012) Switchgrass breeding genetics and genomics InA Monti (Ed) Switchgrass (pp 29ndash54) London UK Springer
Casler M Mitchell R amp Vogel K (2012) Switchgrass In CKole C P Joshi amp D R Shonnard (Eds) Handbook of bioen-ergy crop plants Vol 2 New York NY Taylor amp Francis
Casler M D amp Ramstein G P (2018) Breeding for biomass yieldin switchgrass using surrogate measures of yield BioEnergyResearch 11 6ndash12
Casler M D Sosa S Hoffman L Mayton H Ernst C Adler PR hellip Bonos S A (2017) Biomass Yield of Switchgrass Culti-vars under High‐ versus Low‐Input Conditions Crop Science 57821ndash832 httpsdoiorg102135cropsci2016080698
Casler M D amp Vogel K P (2014) Selection for biomass yield inupland lowland and hybrid switchgrass Crop Science 54 626ndash636 httpsdoiorg102135cropsci2013040239
Casler M D Vogel K P amp Harrison M (2015) Switchgrassgermplasm resources Crop Science 55 2463ndash2478 httpsdoiorg102135cropsci2015020076
Casler M D Vogel K P Lee D K Mitchell R B Adler P RSulc R M hellip Moore K J (2018) 30 Years of progress towardincreasing biomass yield of switchgrass and big bluestem CropScience 58 1242
Clark L V Brummer J E Głowacka K Hall M C Heo K PengJ hellip Sacks E J (2014) A footprint of past climate change on thediversity and population structure of Miscanthus sinensis Annals ofBotany 114 97ndash107 httpsdoiorg101093aobmcu084
Clark L V Dzyubenko E Dzyubenko N Bagmet L SabitovA Chebukin P hellip Sacks E J (2016) Ecological characteristicsand in situ genetic associations for yield‐component traits of wildMiscanthus from eastern Russia Annals of Botany 118 941ndash955
Clark L V Jin X Petersen K K Anzoua K G Bagmet LChebukin P hellip Sacks E J(2018) Population structure of Mis-canthus sacchariflorus reveals two major polyploidization eventstetraploid‐mediated unidirectional introgression from diploid Msinensis and diversity centred around the Yellow Sea Annals ofBotany 20 1ndash18 httpsdoiorg101093aobmcy1161
Clark L V Stewart J R Nishiwaki A Toma Y Kjeldsen J BJoslashrgensen U hellip Sacks E J (2015) Genetic structure ofMiscanthus sinensis and Miscanthus sacchariflorus in Japan indi-cates a gradient of bidirectional but asymmetric introgressionJournal of Experimental Botany 66 4213ndash4225
Clifton‐Brown J Hastings A Mos M McCalmont J P Ashman CAwty‐Carroll DhellipCracroft‐EleyW (2017) Progress in upscalingMis-canthus biomass production for the European bio‐economy with seed‐based hybridsGlobal Change Biology Bioenergy 9 6ndash17
Clifton‐Brown J Schwarz K U amp Hastings A (2015) History ofthe development of Miscanthus as a bioenergy crop From smallbeginnings to potential realisation Biology and Environment‐Pro-ceedings of the Royal Irish Academy 115B 45ndash57
Clifton‐Brown J Senior H Purdy S Horsnell R Lankamp BMuumlennekhoff A-K hellip Bentley A R (2018) Investigating thepotential of novel non‐woven fabrics to increase pollination effi-ciency in plant breeding PloS One (Accepted)
Crawford J T Shan CW Budsberg E Morgan H Bura R ampGus-tafson R (2016) Hydrocarbon bio‐jet fuel from bioconversion ofpoplar biomass Techno‐economic assessment Biotechnology forBiofuels 9 141 httpsdoiorg101186s13068-016-0545-7
28 | CLIFTON‐BROWN ET AL
Dalton S (2013) Biotechnology of Miscanthus In S M Jain amp SD Gupta (Eds) Biotechnology of neglected and underutilizedcrops (pp 243ndash294) Dordrecht Netherlands Springer
Davey C L Robson P Hawkins S Farrar K Clifton‐Brown J CDonnison I S amp Slavov G T (2017) Genetic relationshipsbetween spring emergence canopy phenology and biomass yieldincrease the accuracy of genomic prediction in Miscanthus Journalof Experimental Botany 68 5093ndash5102 httpsdoiorg101093jxberx339
David X amp Anderson X (2002) Aspen and larch genetics cooper-ative annual report 13 St Paul MN Department of ForestResources University of Minnesota
Deblock M (1990) Factors influencing the tissue‐culture and theAgrobacterium‐Tumefaciens‐mediated transformation of hybridaspen and poplar clones Plant Physiology 93 1110ndash1116httpsdoiorg101104pp9331110
Defra (2002) The role of future public research investment in thegenetic improvement of UK‐grown crops (p 220) LondonRetrieved from sciencesearchdefragovukDefaultaspxMenu=MenuampModule=MoreampLocation=NoneampCompleted=220ampProjectID=10412
Dewoody J Trewin H amp Taylor G (2015) Genetic and morpho-logical differentiation in Populus nigra L Isolation by coloniza-tion or isolation by adaptation Molecular Ecology 24 2641ndash2655
Dong H Liu S Clark L V Sharma S Gifford J M Juvik JA hellip Sacks E J (2018) Genetic mapping of biomass yield inthree interconnected Miscanthus populations Global Change Biol-ogy Bioenergy 10 165ndash185
Dukes (2017) Digest of UK Energy Statistics (DUKES) (p 264) Lon-don UK Department for Business Energy amp Industrial StrategyRetrieved from wwwgovukgovernmentcollectionsdigest-of-uk-energy-statistics-dukes
Eckenwalder J (1996) Systematics and evolution of Populus In RStettler H Bradshaw J Heilman amp T Hinckley (Eds) Biologyof Populus and its implications for management and conservation(pp 7‐32) Ottawa ON National Research Council of Canada
Elshire R J Glaubitz J C Sun Q Poland J A Kawamoto KBuckler E S amp Mitchell S E (2011) A robust simple geno-typing‐by‐sequencing (GBS) approach for high diversity speciesPloS One 6 e19379
Evans H (2016) Bioenergy crops in the UK Case studies of suc-cessful whole farm integration (p 23) Loughborough UKEnergy Technologies Institute Retrieved from httpswwweticouklibrarybioenergy-crops-in-the-uk-case-studies-on-successful-whole-farm-integration-evidence-pack
Evans H (2017) Increasing UK biomass production through moreproductive use of land (p 7) Loughborough UK Energy Tech-nologies Institute Retrieved from httpswwweticouklibraryan-eti-perspective-increasing-uk-biomass-production-through-more-productive-use-of-land
Evans J Sanciangco M D Lau K H Crisovan E Barry KDaum C hellip Buell C R (2018) Extensive genetic diversity ispresent within North American switchgrass germplasm The PlantGenome 11 1ndash16 httpsdoiorg103835plantgenome2017060055
Evans L M Slavov G T Rodgers‐Melnick E Martin J RanjanP Muchero W hellip DiFazio S P (2014) Population genomicsof Populus trichocarpa identifies signatures of selection and
adaptive trait associations Nature Genetics 46 1089ndash1096httpsdoiorg101038ng3075
Fabio E S Volk T A Miller R O Serapiglia M J Gauch HG Van Rees K C J hellip Smart L B (2017) Genotypetimes envi-ronment interaction analysis of North American shrub willowyield trials confirms superior performance of triploid hybrids Glo-bal Change Biology Bioenergy 9 445ndash459 httpsdoiorg101111gcbb12344
Fahrenkrog A M Neves L G Resende M F R Vazquez A Ide los Campos G Dervinis C hellip Kirst M (2017) Genome‐wide association study reveals putative regulators of bioenergytraits in Populus deltoides New Phytologist 213 799ndash811
Fan D Liu T T Li C F Jiao B Li S Hou Y S amp Luo KM (2015) Efficient CRISPRCas9‐mediated targeted mutagenesisin Populus in the first generation Scientific Reports 5 12217
Feng X P Lourgant K Castric V Saumitou-Laprade P ZhengB S Jiang D M amp Brancourt-Hulmel M (2014) The discov-ery of natural accessions related to Miscanthus x giganteus usingchloroplast DNA Crop Science 54 1645ndash1655
Fillatti J J Sellmer J Mccown B Haissig B amp Comai L(1987) Agrobacterium-mediated transformation and regenerationof Populus Molecular amp General Genetics 206 192ndash199
Fu Y B (2015) Understanding crop genetic diversity under modernplant breeding Theoretical and Applied Genetics 128 2131ndash2142 httpsdoiorg101007s00122-015-2585-y
Fu C Mielenz J R Xiao X Ge Y Hamilton C Y RodriguezM hellip Wang Z‐Y (2011) Genetic manipulation of ligninreduces recalcitrance and improves ethanol production fromswitchgrass Proceedings of the National Academy of Sciences ofthe United States of America 108 3803ndash3808 httpsdoiorg101073pnas1100310108
Gao C (2018) The future of CRISPR technologies in agricultureNature Reviews Molecular Cell Biology 19(5) 275ndash276
Gegas V Gay A Camargo A amp Doonan J (2014) Challengesof Crop Phenomics in the Post-genomic Era In J Hancock (Ed)Phenomics (pp 142ndash171) Boca Raton FL CRC Press
Geraldes A DiFazio S P Slavov G T Ranjan P Muchero WHannemann J hellip Tuskan G A (2013) A 34K SNP genotypingarray for Populus trichocarpa Design application to the study ofnatural populations and transferability to other Populus speciesMolecular Ecology Resources 13 306ndash323
Gifford J M Chae W B Swaminathan K Moose S P amp JuvikJ A (2015) Mapping the genome of Miscanthus sinensis forQTL associated with biomass productivity Global Change Biol-ogy Bioenergy 7 797ndash810
Goggin F L Lorence A amp Topp C N (2015) Applying high‐throughput phenotyping to plant‐insect interactions Picturingmore resistant crops Current Opinion in Insect Science 9 69ndash76httpsdoiorg101016jcois201503002
Grabowski P P Evans J Daum C Deshpande S Barry K WKennedy M hellip Casler M D (2017) Genome‐wide associationswith flowering time in switchgrass using exome‐capture sequenc-ing data New Phytologist 213 154ndash169 httpsdoiorg101111nph14101
Greef J M amp Deuter M (1993) Syntaxonomy of Miscanthus xgiganteus GREEF et DEU Angewandte Botanik 67 87ndash90
Guerra F P Richards J H Fiehn O Famula R Stanton B JShuren R hellip Neale D B (2016) Analysis of the genetic varia-tion in growth ecophysiology and chemical and metabolomic
CLIFTON‐BROWN ET AL | 29
composition of wood of Populus trichocarpa provenances TreeGenetics amp Genomes 12 6 httpsdoiorg101007s11295-015-0965-8
Guo H P Shao R X Hong C T Hu H K Zheng B S ampZhang Q X (2013) Rapid in vitro propagation of bioenergy cropMiscanthus sacchariflorus Applied Mechanics and Materials 260181ndash186
Hallingbaumlck H R Fogelqvist J Powers S J Turrion‐Gomez JRossiter R Amey J hellip Roumlnnberg‐Waumlstljung A‐C (2016)Association mapping in Salix viminalis L (Salicaceae)ndashIdentifica-tion of candidate genes associated with growth and phenologyGlobal Change Biology Bioenergy 8 670ndash685
Hanley S J amp Karp A (2014) Genetic strategies for dissectingcomplex traits in biomass willows (Salix spp) Tree Physiology34 1167ndash1180 httpsdoiorg101093treephystpt089
Harfouche A Meilan R amp Altman A (2011) Tree genetic engi-neering and applications to sustainable forestry and biomass pro-duction Trends in Biotechnology 29 9ndash17 httpsdoiorg101016jtibtech201009003
Harfouche A Meilan R amp Altman A (2014) Molecular and phys-iological responses to abiotic stress in forest trees and their rele-vance to tree improvement Tree Physiology 34 1181ndash1198httpsdoiorg101093treephystpu012
Harfouche A Meilan R Kirst M Morgante M Boerjan WSabatti M amp Mugnozza G S (2012) Accelerating the domesti-cation of forest trees in a changing world Trends in PlantScience 17 64ndash72 httpsdoiorg101016jtplants201111005
Hastings A Clifton‐Brown J Wattenbach M Mitchell C PStampfl P amp Smith P (2009) Future energy potential of Mis-canthus in Europe Global Change Biology Bioenergy 1 180ndash196
Hastings A Mos M Yesufu J AMccalmont J Schwarz KShafei RhellipClifton-Brown J (2017) Economic and environmen-tal assessment of seed and rhizome propagated Miscanthus in theUK Frontiers in Plant Science 8 16 httpsdoiorg103389fpls201701058
Heacuteder M (2017) From NASA to EU The evolution of the TRLscale in Public Sector Innovation The Innovation Journal 22 1ndash23 httpsdoiorg103389fpls201701058
Heffner E L Sorrells M E amp Jannink J L (2009) Genomicselection for crop improvement Crop Science 49 1ndash12 httpsdoiorg102135cropsci2008080512
Hodkinson T R Klaas M Jones M B Prickett R amp Barth S(2015) Miscanthus A case study for the utilization of naturalgenetic variation Plant Genetic Resources‐Characterization andUtilization 13 219ndash237 httpsdoiorg101017S147926211400094X
Hodkinson T R amp Renvoize S (2001) Nomenclature of Miscant-hus xgiganteus (Poaceae) Kew Bulletin 56 759ndash760 httpsdoiorg1023074117709
Houle D Govindaraju D R amp Omholt S (2010) Phenomics Thenext challenge Nature Reviews Genetics 11 855ndash866 httpsdoiorg101038nrg2897
Huang X H Wei X H Sang T Zhao Q Feng Q Zhao Yhellip Han B (2010) Genome‐wide association studies of 14 agro-nomic traits in rice landraces Nature Genetics 42 961ndashU976
Hwang O‐J Cho M‐A Han Y‐J Kim Y‐M Lim S‐H KimD‐S hellip Kim J‐I (2014) Agrobacterium‐mediated genetic trans-formation of Miscanthus sinensis Plant Cell Tissue and OrganCulture (PCTOC) 117 51ndash63
Hwang O‐J Lim S‐H Han Y‐J Shin A‐Y Kim D‐S ampKim J‐I (2014) Phenotypic characterization of transgenic Mis-canthus sinensis plants overexpressing Arabidopsis phytochromeB International Journal of Photoenergy 2014501016
Ilstedt B (1996) Genetics and performance of Belgian poplar clonestested in Sweden International Journal of Forest Genetics 3183ndash195
Ingvarsson P K Garcia M V Hall D Luquez V amp Jansson S(2006) Clinal variation in phyB2 a candidate gene for day‐length‐induced growth cessation and bud set across a latitudinalgradient in European aspen (Populus tremula) Genetics 1721845ndash1853
Ingvarsson P K Garcia M V Luquez V Hall D amp Jansson S(2008) Nucleotide polymorphism and phenotypic associationswithin and around the phytochrome B2 locus in European aspen(Populus tremula Salicaceae) Genetics 178 2217ndash2226 httpsdoiorg101534genetics107082354
Jannink J‐L Lorenz A J amp Iwata H (2010) Genomic selectionin plant breeding From theory to practice Briefings in FunctionalGenomics 9 166ndash177 httpsdoiorg101093bfgpelq001
Jiang J Guan Y Mccormick S Juvik J Lubberstedt T amp FeiS‐Z (2017) Gametophytic self‐incompatibility is operative inMiscanthus sinensis (Poaceae) and is affected by pistil age CropScience 57 1948ndash1956
Jones H D (2015a) Future of breeding by genome editing is in thehands of regulators GM Crops amp Food 6 223ndash232
Jones H D (2015b) Regulatory uncertainty over genome editingNature Plants 1 14011
Kalinina O Nunn C Sanderson R Hastings A F S van der Weijde TOumlzguumlven MhellipClifton‐Brown J C (2017) ExtendingMiscanthus cul-tivation with novel germplasm at six contrasting sites Frontiers in PlantScience 8563 httpsdoiorg103389fpls201700563
Karp A Hanley S J Trybush S O Macalpine W Pei M ampShield I (2011) Genetic improvement of willow for bioenergyand biofuels free access Journal of Integrative Plant Biology 53151ndash165 httpsdoiorg101111j1744-7909201001015x
Kersten B Faivre Rampant P Mader M Le Paslier M‐C Bou-non R Berard A hellip Fladung M (2016) Genome sequences ofPopulus tremula chloroplast and mitochondrion Implications forholistic poplar breeding PloS One 11 e0147209 httpsdoiorg101371journalpone0147209
Kiesel A Nunn C Iqbal Y Van der Weijde T Wagner MOumlzguumlven M hellip Lewandowski I (2017) Site‐specific manage-ment of Miscanthus genotypes for combustion and anaerobicdigestion A comparison of energy yields Frontiers in PlantScience 8 347 httpsdoiorg103389fpls201700347
Kim H Kim S T Ryu J Kang B C Kim J S amp Kim S G(2017) CRISPRCpf1‐mediated DNA‐free plant genome editingNature Communications 8 14406 httpsdoiorg101038ncomms14406
King Z R Bray A L Lafayette P R amp Parrott W A (2014)Biolistic transformation of elite genotypes of switchgrass (Pan-icum virgatum L) Plant Cell Reports 33 313ndash322 httpsdoiorg101007s00299-013-1531-1
Kopp R F Maynard C A De Niella P R Smart L B amp Abra-hamson L P (2002) Collection and storage of pollen from Salix(Salicaceae) American Journal of Botany 89 248ndash252
Krzyżak J Pogrzeba M Rusinowski S Clifton‐Brown J McCal-mont J P Kiesel A hellip Mos M (2017) Heavy metal uptake
30 | CLIFTON‐BROWN ET AL
by novel miscanthus seed‐based hybrids cultivated in heavy metalcontaminated soil Civil and Environmental Engineering Reports26 121ndash132 httpsdoiorg101515ceer-2017-0040
Kuzovkina Y A amp Quigley M F (2005) Willows beyond wet-lands Uses of Salix L species for environmental projects WaterAir and Soil Pollution 162 183ndash204 httpsdoiorg101007s11270-005-6272-5
Lazarus W Headlee W L amp Zalesny R S (2015) Impacts of sup-plyshed‐level differences in productivity and land costs on the eco-nomics of hybrid poplar production in Minnesota USA BioEnergyResearch 8 231ndash248 httpsdoiorg101007s12155-014-9520-y
Lewandowski I (2015) Securing a sustainable biomass supply in agrowing bioeconomy Global Food Security 6 34ndash42 httpsdoiorg101016jgfs201510001
Lewandowski I (2016) The role of perennial biomass crops in agrowing bioeconomy In S Barth D Murphy‐Bokern O Kalin-ina G Taylor amp M Jones (Eds) Perennial biomass crops for aresource‐constrained world (pp 3ndash13) Cham SwitzerlandSpringer Switzerland AG
Li J L Meilan R Ma C Barish M amp Strauss S H (2008)Stability of herbicide resistance over 8 years of coppice in field‐grown genetically engineered poplars Western Journal of AppliedForestry 23 89ndash93
Li R Y amp Qu R D (2011) High throughput Agrobacterium‐medi-ated switchgrass transformation Biomass amp Bioenergy 35 1046ndash1054 httpsdoiorg101016jbiombioe201011025
Lindegaard K N amp Barker J H A (1997) Breeding willows forbiomass Aspects of Applied Biology 49 155ndash162
Linde‐Laursen I B (1993) Cytogenetic analysis of MiscanthusGiganteus an interspecific hybrid Hereditas 119 297ndash300httpsdoiorg101111j1601-5223199300297x
Liu Y Merrick P Zhang Z Z Ji C H Yang B amp Fei S Z(2018) Targeted mutagenesis in tetraploid switchgrass (Panicumvirgatum L) using CRISPRCas9 Plant Biotechnology Journal16 381ndash393
Liu L L Wu Y Q Wang Y W amp Samuels T (2012) A high‐density simple sequence repeat‐based genetic linkage map ofswitchgrass G3 (Bethesda Md) 2 357ndash370
Lovett A Suumlnnenberg G amp Dockerty T (2014) The availabilityof land for perennial energy crops in Great Britain GlobalChange Biology Bioenergy 6 99ndash107 httpsdoiorg101111gcbb12147
Lozano‐Juste J amp Cutler S R (2014) Plant genome engineering infull bloom Trends in Plant Science 19 284ndash287 httpsdoiorg101016jtplants201402014
Lu F Lipka A E Glaubitz J Elshire R Cherney J H CaslerM D hellip Costich D E (2013) Switchgrass Genomic DiversityPloidy and Evolution Novel Insights from a Network‐Based SNPDiscovery Protocol Plos Genetics 9 e1003215
Ludovisi R Tauro F Salvati R Khoury S Mugnozza G S ampHarfouche A (2017) UAV‐based thermal imaging for high‐throughput field phenotyping of black poplar response to droughtFrontiers in Plant Science 81681
Macalpine W Shield I amp Karp A (2010) Seed to near market vari-ety the BEGIN willow breeding pipeline 2003ndash2010 and beyond InBioten conference proceedings Birmingham pp 21ndash23
Macalpine W J Shield I F Trybush S O Hayes C M ampKarp A (2008) Overcoming barriers to crossing in willow (Salixspp) breeding Aspects of Applied Biology 90 173ndash180
Mahfouz M M (2017) The efficient tool CRISPR‐Cpf1 NaturePlants 3 17028
Malinowska M Donnison I S amp Robson P R (2017) Phenomicsanalysis of drought responses in Miscanthus collected from differ-ent geographical locations Global Change Biology Bioenergy 978ndash91
Maroder H L Prego I A Facciuto G R amp Maldonado S B(2000) Storage behaviour of Salix alba and Salix matsudanaseeds Annals of Botany 86 1017ndash1021 httpsdoiorg101006anbo20001265
Martinez‐Reyna J amp Vogel K (2002) Incompatibility systems inswitchgrass Crop Science 42 1800ndash1805 httpsdoiorg102135cropsci20021800
Maurya J P amp Bhalerao R P (2017) Photoperiod‐and tempera-ture‐mediated control of growth cessation and dormancy in treesA molecular perspective Annals of Botany 120 351ndash360httpsdoiorg101093aobmcx061
McCalmont J P Hastings A Mcnamara N P Richter G MRobson P Donnison I S amp Clifton‐Brown J (2017) Environ-mental costs and benefits of growing Miscanthus for bioenergy inthe UK Global Change Biology Bioenergy 9 489ndash507
McCracken A R amp Dawson W M (1997) Using mixtures of wil-low clones as a means of controlling rust disease Aspects ofApplied Biology 49 97ndash103
McKown A D Klaacutepště J Guy R D Geraldes A Porth I Hanne-mann JhellipDouglas C J (2014) Genome‐wide association implicatesnumerous genes underlying ecological trait variation in natural popula-tions ofPopulus trichocarpaNew Phytologist 203 535ndash553
Merrick P amp Fei S Z (2015) Plant regeneration and genetic transforma-tion in switchgrass ndash A review Journal of Integrative Agriculture 14483ndash493 httpsdoiorg101016S2095-3119(14)60921-7
Moreno‐Mateos M A Fernandez J P Rouet R Lane M AVejnar C E Mis E hellip Giraldez A J (2017) CRISPR‐Cpf1mediates efficient homology‐directed repair and temperature‐con-trolled genome editing Nature Communications 8 2024 httpsdoiorg101038s41467-017-01836-2
Mosseler A (1990) Hybrid performance and species crossabilityrelationships in willows (Salix) Canadian Journal of Botany 682329ndash2338
Muchero W Guo J DiFazio S P Chen J‐G Ranjan P Slavov GT hellip Tuskan G A (2015) High‐resolution genetic mapping ofallelic variants associated with cell wall chemistry in Populus BMCGenomics 16 24 httpsdoiorg101186s12864-015-1215-z
Neale D B amp Kremer A (2011) Forest tree genomics Growingresources and applications Nature Reviews Genetics 12 111ndash122 httpsdoiorg101038nrg2931
Nunn C Hastings A F S J Kalinina O Oumlzguumlven M SchuumlleH Tarakanov I G hellip Clifton‐Brown J C (2017) Environmen-tal influences on the growing season duration and ripening ofdiverse Miscanthus germplasm grown in six countries Frontiersin Plant Science 8 907 httpsdoiorg103389fpls201700907
Ogawa Y Honda M Kondo Y amp Hara‐Nishimura I (2016) Anefficient Agrobacterium‐mediated transformation method forswitchgrass genotypes using Type I callus Plant Biotechnology33 19ndash26
Ogawa Y Shirakawa M Koumoto Y Honda M Asami Y KondoY ampHara‐Nishimura I (2014) A simple and reliable multi‐gene trans-formation method for switchgrass Plant Cell Reports 33 1161ndash1172httpsdoiorg101007s00299-014-1605-8
CLIFTON‐BROWN ET AL | 31
Okada M Lanzatella C Saha M C Bouton J Wu R L amp Tobias CM (2010) Complete switchgrass genetic maps reveal subgenomecollinearity preferential pairing and multilocus interactions Genetics185 745ndash760 httpsdoiorg101534genetics110113910
Palomo‐Riacuteos E Macalpine W Shield I Amey J Karaoğlu C WestJ hellip Jones H D (2015) Efficient method for rapid multiplication ofclean and healthy willow clones via in vitro propagation with broadgenotype applicability Canadian Journal of Forest Research 451662ndash1667 httpsdoiorg101139cjfr-2015-0055
Pilate G Allona I Boerjan W Deacutejardin A Fladung M Gal-lardo F hellip Halpin C (2016) Lessons from 25 years of GM treefield trials in Europe and prospects for the future In C Vettori FGallardo H Haumlggman V Kazana F Migliacci G Pilateamp MFladung (Eds) Biosafety of forest transgenic trees (pp 67ndash100)Dordrecht Netherlands Springer Switzerland AG
Pinosio S Giacomello S Faivre-Rampant P Taylor G Jorge VLe Paslier M C amp Marroni F (2016) Characterization of thepoplar pan-genome by genome-wide identification of structuralvariation Molecular Biology and Evolution 33 2706ndash2719
Porth I Klaacutepště J Skyba O Friedmann M C Hannemann JEhlting J hellip Douglas C J (2013) Network analysis reveals therelationship among wood properties gene expression levels andgenotypes of natural Populus trichocarpa accessions New Phytol-ogist 200 727ndash742
Pugesgaard S Schelde K Larsen S U Laeligrke P E amp JoslashrgensenU (2015) Comparing annual and perennial crops for bioenergyproductionndashinfluence on nitrate leaching and energy balance Glo-bal Change Biology Bioenergy 7 1136ndash1149 httpsdoiorg101111gcbb12215
Quinn L D Allen D J amp Stewart J R (2010) Invasivenesspotential of Miscanthus sinensis Implications for bioenergy pro-duction in the United States Global Change Biology Bioenergy2 310ndash320 httpsdoiorg101111j1757-1707201001062x
Rae A M Robinson K M Street N R amp Taylor G (2004)Morphological and physiological traits influencing biomass pro-ductivity in short‐rotation coppice poplar Canadian Journal ofForest Research‐Revue Canadienne De Recherche Forestiere 341488ndash1498 httpsdoiorg101139x04-033
Rambaud C Arnoult S Bluteau A Mansard M C Blassiau Camp Brancourt‐Hulmel M (2013) Shoot organogenesis in threeMiscanthus species and evaluation for genetic uniformity usingAFLP analysis Plant Cell Tissue and Organ Culture (PCTOC)113 437ndash448 httpsdoiorg101007s11240-012-0284-9
Ramstein G P Evans J Kaeppler S M Mitchell R B Vogel K PBuell C R amp Casler M D (2016) Accuracy of genomic prediction inswitchgrass (Panicum virgatum L) improved by accounting for linkagedisequilibriumG3 (Bethesda Md) 6 1049ndash1062
Rayburn A L Crawford J Rayburn C M amp Juvik J A (2009)Genome size of three Miscanthus species Plant Molecular Biol-ogy Reporter 27 184 httpsdoiorg101007s11105-008-0070-3
Richardson J Isebrands J amp Ball J (2014) Ecology and physiol-ogy of poplars and willows In J Isebrands amp J Richardson(Eds) Poplars and willows Trees for society and the environ-ment (pp 92‐123) Boston MA and Rome CABI and FAO
Robison T L Rousseau R J amp Zhang J (2006) Biomass produc-tivity improvement for eastern cottonwood Biomass amp Bioenergy30 735ndash739 httpsdoiorg101016jbiombioe200601012
Robson P R Farrar K Gay A P Jensen E F Clifton‐Brown JC amp Donnison I S (2013) Variation in canopy duration in the
perennial biofuel crop Miscanthus reveals complex associationswith yield Journal of Experimental Botany 64 2373ndash2383
Robson P Jensen E Hawkins S White S R Kenobi K Clif-ton‐Brown J hellip Farrar K (2013) Accelerating the domestica-tion of a bioenergy crop Identifying and modelling morphologicaltargets for sustainable yield increase in Miscanthus Journal ofExperimental Botany 64 4143ndash4155
Sanderson M A Adler P R Boateng A A Casler M D ampSarath G (2006) Switchgrass as a biofuels feedstock in theUSA Canadian Journal of Plant Science 86 1315ndash1325httpsdoiorg104141P06-136
Scarlat N Dallemand J F Monforti‐Ferrario F amp Nita V(2015) The role of biomass and bioenergy in a future bioecon-omy Policies and facts Environmental Development 15 3ndash34httpsdoiorg101016jenvdev201503006
Serapiglia M J Gouker F E amp Smart L B (2014) Early selec-tion of novel triploid hybrids of shrub willow with improved bio-mass yield relative to diploids BMC Plant Biology 14 74httpsdoiorg1011861471-2229-14-74
Serba D Wu L Daverdin G Bahri B A Wang X Kilian Ahellip Devos K M (2013) Linkage maps of lowland and upland tet-raploid switchgrass ecotypes BioEnergy Research 6 953ndash965httpsdoiorg101007s12155-013-9315-6
Shakoor N Lee S amp Mockler T C (2017) High throughput phe-notyping to accelerate crop breeding and monitoring of diseases inthe field Current Opinion in Plant Biology 38 184ndash192 httpsdoiorg101016jpbi201705006
Shield I Macalpine W Hanley S amp Karp A (2015) Breedingwillow for short rotation coppice energy cropping In Industrialcrops Breeding for BioEnergy and Bioproducts (pp 67ndash80) NewYork NY Springer
Sjodin A Street N R Sandberg G Gustafsson P amp Jansson S (2009)The Populus Genome Integrative Explorer (PopGenIE) A new resourcefor exploring the Populus genomeNewPhytologist 182 1013ndash1025
Slavov G amp Davey C (2017) Integrated genomic prediction inbioenergy crops In Abstracts from the International Conferenceon Developing biomass crops for future climates pp 34 24ndash27September 2017 Oriel College Oxford wwwwatbioeu
Slavov G Davey C Bosch M Robson P Donnison I ampMackay I (2018a) Genomic index selection provides a pragmaticframework for setting and refining multi‐objective breeding targetsin Miscanthus Annals of Botany (in press)
Slavov G T DiFazio S P Martin J Schackwitz W MucheroW Rodgers‐Melnick E hellip Tuskan G A (2012) Genome rese-quencing reveals multiscale geographic structure and extensivelinkage disequilibrium in the forest tree Populus trichocarpa NewPhytologist 196 713ndash725
Slavov G Davey C Robson P Donnison I amp Mackay I(2018b) Domestication of Miscanthus through genomic indexselection In Plant and Animal Genome Conference 13 January2018 San Diego CA Retrieved from httpspagconfexcompagxxvimeetingappcgiPaper30622
Slavov G T Nipper R Robson P Farrar K Allison G GBosch M hellip Jensen E (2013) Genome‐wide association studiesand prediction of 17 traits related to phenology biomass and cellwall composition in the energy grass Miscanthus sinensis NewPhytologist 201 1227ndash1239
Slavov G Robson P Jensen E Hodgson E Farrar K AllisonG hellip Donnison I (2014) Contrasting geographic patterns of
32 | CLIFTON‐BROWN ET AL
genetic variation for molecular markers vs phenotypic traits in theenergy grass Miscanthus sinensis Global Change Biology Bioen-ergy 5 562ndash571
Ślusarkiewicz‐Jarzina A Ponitka A Cerazy‐Waliszewska J Woj-ciechowicz M K Sobańska K Jeżowski S amp Pniewski T(2017) Effective and simple in vitro regeneration system of Mis-canthus sinensis Mtimes giganteus and M sacchariflorus for plant-ing and biotechnology purposes Biomass and Bioenergy 107219ndash226 httpsdoiorg101016jbiombioe201710012
Smart L amp Cameron K (2012) Shrub willow In C Kole S Joshiamp D Shonnard (Eds) Handbook of bioenergy crop plants (pp687ndash708) Boca Raton FL Taylor and Francis Group
Stanton B J (2014) The domestication and conservation of Populusand Salix genetic resources (Chapter 4) In Isebrands J ampRichardson J (Eds) Poplars and willows Trees for society andthe environment (pp 124ndash200) Rome Italy CAB InternationalFood and Agricultural Organization of the United Nations
Stanton B J Neale D B amp Li S (2010) Populus breeding Fromthe classical to the genomic approach In S Jansson R Bha-lerao amp A T Groover (Eds) Genetics and genomics of popu-lus (pp 309ndash348) New York NY Springer
Stewart J R Toma Y Fernandez F G Nishiwaki A Yamada T ampBollero G (2009) The ecology and agronomy ofMiscanthus sinensisa species important to bioenergy crop development in its native range inJapan A reviewGlobal Change Biology Bioenergy 1 126ndash153
Stott K G (1992) Willows in the service of man Proceedings of the RoyalSociety of Edinburgh Section B Biological Sciences 98 169ndash182
Strauss S H Ma C Ault K amp Klocko A L (2016) Lessonsfrom two decades of field trials with genetically modified trees inthe USA Biology and regulatory compliance In C Vettori FGallardo H Haumlggman V Kazana F Migliacci G Pilate amp MFladung (Eds) Biosafety of forest transgenic trees (pp 101ndash124)Dordrecht Netherlands Springer
Suda Y amp Argus G W (1968) Chromosome numbers of some NorthAmerican Salix Brittonia 20 191ndash197 httpsdoiorg1023072805440
Tan B (2018) Genomic selection and genome‐wide association studies todissect quantitative traits in forest trees Umearing Sweden University
Tornqvist C Taylor M Jiang Y Evans J Buell C KaepplerS amp Casler M (2018) Quantitative trait locus mapping for flow-ering time in a lowland x upland switchgrass pseudo‐F2 popula-tion The Plant Genome 11 170093
Toacuteth G Hermann T Da Silva M amp Montanarella L (2016)Heavy metals in agricultural soils of the European Union withimplications for food safety Environment International 88 299ndash309 httpsdoiorg101016jenvint201512017
Tracy W Dawson J Moore V amp Fisch J (2016) Intellectual propertyrights and public plant breeding Recommendations and proceedings ofa conference on best practices for intellectual property protection of pub-lically developed plant germplasm (p 70) University of Wisconsin‐Madison Retrieved from httpshostcalswisceduagronomywp-contentuploadssites1620162005Proceedings-IPR-Finalpdf
Trybush S Jahodovaacute Š Macalpine W amp Karp A (2008) A geneticstudy of a Salix germplasm resource reveals new insights into rela-tionships among subgenera sections and species BioEnergyResearch 1 67ndash79 httpsdoiorg101007s12155-008-9007-9
Tsai C J (2013) Next‐generation sequencing for next‐generationbreeding and more New Phytologist 198 635ndash637 httpsdoiorg101111nph12245
Tuskan G A Difazio S Jansson S Bohlmann J Grigoriev I HellstenU hellip Rokhsar D (2006) The genome of black cottonwood Populustrichocarpa (Torr ampGray) Science 313 1596ndash1604
Tuskan G A amp Walsh M E (2001) Short‐rotation woody cropsystems atmospheric carbon dioxide and carbon management AUS case study The Forestry Chronicle 77 259ndash264 httpsdoiorg105558tfc77259-2
Van Den Broek R Treffers D‐J Meeusen M Van Wijk ANieuwlaar E amp Turkenburg W (2001) Green energy or organicfood A life‐cycle assessment comparing two uses of set‐asideland Journal of Industrial Ecology 5 65ndash87 httpsdoiorg101162108819801760049477
Van Der Schoot J Pospiskova M Vosman B amp Smulders M J M(2000) Development and characterization of microsatellite markers inblack poplar (Populus nigra L) Theoretical and Applied Genetics 101317ndash322 httpsdoiorg101007s001220051485
Van Der Weijde T Dolstra O Visser R G amp Trindade L M(2017a) Stability of cell wall composition and saccharificationefficiency in Miscanthus across diverse environments Frontiers inPlant Science 7 2004
Van der Weijde T Huxley L M Hawkins S Sembiring E HFarrar K Dolstra O hellip Trindade L M (2017) Impact ofdrought stress on growth and quality of miscanthus for biofuelproduction Global Change Biology Bioenergy 9 770ndash782
Van der Weijde T Kamei C L A Severing E I Torres A FGomez L D Dolstra O hellip Trindade L M (2017) Geneticcomplexity of miscanthus cell wall composition and biomass qual-ity for biofuels BMC Genomics 18 406
Van der Weijde T Kiesel A Iqbal Y Muylle H Dolstra O Visser RG FhellipTrindade LM (2017) Evaluation ofMiscanthus sinensis bio-mass quality as feedstock for conversion into different bioenergy prod-uctsGlobal Change Biology Bioenergy 9 176ndash190
Vanholme B Cesarino I Goeminne G Kim H Marroni F VanAcker R hellip Boerjan W (2013) Breeding with rare defectivealleles (BRDA) A natural Populus nigra HCT mutant with modi-fied lignin as a case study New Phytologist 198 765ndash776
Vogel K P Mitchell R B Casler M D amp Sarath G (2014)Registration of Liberty Switchgrass Journal of Plant Registra-tions 8 242ndash247 httpsdoiorg103198jpr2013120076crc
Wang X Yamada T Kong F‐J Abe Y Hoshino Y Sato Hhellip Yamada T (2011) Establishment of an efficient in vitro cul-ture and particle bombardment‐mediated transformation systems inMiscanthus sinensis Anderss a potential bioenergy crop GlobalChange Biology Bioenergy 3 322ndash332 httpsdoiorg101111j1757-1707201101090x
Watrud L S Lee E H Fairbrother A Burdick C Reichman JR Bollman M hellip Van de Water P K (2004) Evidence forlandscape‐level pollen‐mediated gene flow from genetically modi-fied creeping bentgrass with CP4 EPSPS as a marker Proceedingsof the National Academy of Sciences of the United States of Amer-ica 101 14533ndash14538
Weisgerber H (1993) Poplar breeding for the purpose of biomassproduction in short‐rotation periods in Germany ndash Problems andfirst findings Forestry Chronicle 69 727ndash729 httpsdoiorg105558tfc69727-6
Wheeler N C Steiner K C Schlarbaum S E amp Neale D B(2015) The evolution of forest genetics and tree improvementresearch in the United States Journal of Forestry 113 500ndash510httpsdoiorg105849jof14-120
CLIFTON‐BROWN ET AL | 33
Whittaker C Macalpine W J Yates N E amp Shield I (2016)Dry matter losses and methane emissions during wood chip stor-age The impact on full life cycle greenhouse gas savings of shortrotation coppice willow for heat BioEnergy Research 9 820ndash835 httpsdoiorg101007s12155-016-9728-0
Xi Q (2000) Investigation on the distribution and potential of giant grassesin China PhD thesis Goettingen Germany Cuvillier Verlag
Xi Q amp Jezowkski S (2004) Plant resources of Triarrhena andMiscanthus species in China and its meaning for Europe PlantBreeding and Seed Science 49 63ndash77
Xue S Kalinina O amp Lewandowski I (2015) Present and future optionsfor the improvement of Miscanthus propagation techniques Renewableand Sustainable Energy Reviews 49 1233ndash1246
Yin K Q Gao C X amp Qiu J L (2017) Progress and prospectsin plant genome editing Nature Plants 3 httpsdoiorg101038nplants2017107
YookM (2016)Genetic diversity and transcriptome analysis for salt toler-ance inMiscanthus PhD thesis Seoul National University Korea
Zaidi S S E A Mahfouz M M amp Mansoor S (2017) CRISPR‐Cpf1 A new tool for plant genome editing Trends in PlantScience 22 550ndash553 httpsdoiorg101016jtplants201705001
Zalesny R S Stanturf J A Gardiner E S Perdue J H YoungT M Coyle D R hellip Hass A (2016) Ecosystem services ofwoody crop production systems BioEnergy Research 9 465ndash491 httpsdoiorg101007s12155-016-9737-z
Zhang Q X Sun Y Hu H K Chen B Hong C T Guo H Phellip Zheng B S (2012) Micropropagation and plant regenerationfrom embryogenic callus of Miscanthus sinensis Vitro Cellular ampDevelopmental Biology‐Plant 48 50ndash57 httpsdoiorg101007s11627-011-9387-y
Zhang Y Zalapa J E Jakubowski A R Price D L AcharyaA Wei Y hellip Casler M D (2011) Post‐glacial evolution of
Panicum virgatum Centers of diversity and gene pools revealedby SSR markers and cpDNA sequences Genetica 139 933ndash948httpsdoiorg101007s10709-011-9597-6
Zhao H Wang B He J Yang J Pan L Sun D amp Peng J(2013) Genetic diversity and population structure of Miscanthussinensis germplasm in China PloS One 8 e75672
Zhou X H Jacobs T B Xue L J Harding S A amp Tsai C J(2015) Exploiting SNPs for biallelic CRISPR mutations in theoutcrossing woody perennial Populus reveals 4‐coumarate CoAligase specificity and redundancy New Phytologist 208 298ndash301
Zhou R Macaya‐Sanz D Rodgers‐Melnick E Carlson C HGouker F E Evans L M hellip DiFazio S P (2018) Characteri-zation of a large sex determination region in Salix purpurea L(Salicaceae) Molecular Genetics and Genomics 1ndash16 httpsdoiorg101007s00438-018-1473-y
Zsuffa L Mosseler A amp Raj Y (1984) Prospects for interspecifichybridization in willow for biomass production In K L Perttu(Ed) Ecology and management of forest biomass production sys-tems Report 15 (pp 261ndash281) Uppsala Sweden SwedishUniversity of Agricultural Sciences
How to cite this article Clifton‐Brown J HarfoucheA Casler MD et al Breeding progress andpreparedness for mass‐scale deployment of perenniallignocellulosic biomass crops switchgrassmiscanthus willow and poplar GCB Bioenergy201800000ndash000 httpsdoiorg101111gcbb12566
34 | CLIFTON‐BROWN ET AL
Graphical AbstractThe contents of this page will be used as part of the graphical abstract of html only
It will not be published as part of main article
Plant breeding links the research effort with commercial mass upscaling The authorsrsquo assessment of development status ofthe four species is shown (poplar having two one for short rotation coppice (SRC) poplar and one for the more traditionalshort rotation forestry (SRF)) Mass scale deployment needs developments outside the breeding arenas to drive breedingactivities more rapidly and extensively
TABLE4
Prebreedingresearch
andthestatus
ofconventio
nalbreeding
infour
leadingperennialbiom
asscrops(PBCs)
Breeding
techno
logy
step
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sWillow
Poplar
1Collected
wild
accessions
or
secondarysources
availablefor
breeding
Provideabroadbase
of
useful
traits
Respect
forCBD
and
Nagoyaprotocol
on
collections
after2014
Not
allindigenous
genetic
resourcesareaccessible
under
CBD
forpolitical
reasons
USa~2
000
(181
inofficial
GRIN
gene
bank
ofwhich
96
wereavailable
others
in
variousunofficial
collections)
CNChangsha
~1000
andNanjin
g
2000from
China
DK~1
20from
JP
FRworking
collectionof
~100
(mainlyM
sin)
JP~1
500
from
JPS
KandCN
NLworking
collectionof
~300
Msin
SK~7
00from
SKC
NJP
andRU
UK~1
500
accessions
(from
~500
sites)
ofwhich
1000arein
field
nurseriesin
2018
US
14in
GRIN
global
US
Illin
ois~1
500
collections
made
inAsiawith
25
currently
available
inUnitedStates
forbreeding
UK1500with
about20
incommon
with
the
UnitedStates
US
350largelyunique
FR3370Populus
delto
ides
(650)Pnigra(2000)
Ptrichocarpa(600)and
Pmaximow
iczii(120)
IT30000P
delto
idsPnigra
Pmaximow
icziiand
Ptrichocarpa
SE13000
accessionsmainly
Ptrichocarpa
byST
TandSL
U
US
GWR150Ptrichocarpaand
300Pmaximow
icziiaccessions
forPacificNorthwestprogram
535Pdelto
ides
and~2
00
Pnigraaccessions
for
Southeastprogrammeand77
Psimoniifrom
Mongolia
UMD
NRRI550+
clonal
accessions
(primarily
Ptimescanadnesisa
long
with
Pdelto
ides
andPnigra
parents)
forUpper
Midwest
program
2Wild
accessions
which
have
undergone
phenotypic
screeningin
field
trials
Selectionof
parental
lines
with
useful
traits
Strong
partnerships
torun
multilocationfieldtrials
tophenotypeconsistently
theaccessionsgenotypes
indifferentenvironm
ents
anddatabases
Costof
runningmultilocation
US
gt500000genotypes
CN~1
250
inChangsha
~1700
in
HunanJiangsuS
handong
Hainan
NL~250genotypes
JP~1
200
SK~4
00
UK‐le
d~1
000
genotypesin
a
replicated
trial
US‐led
~1200
genotypesin
multi‐
locatio
nreplicated
trialsin
Asiaand
North
America
MsinSK
CNJP
CA
(Ontario)US(Illinoisand
Colorado)MsacSK
(Kangw
on)
CN
(Zhejiang)JP
(Sapporo)CA
(Onatario)U
S(Illinois)DK
(Aarhus)
UKgt400
US
180
FR2720
IT10000
SE~1
50
US
GWR~1
500
wild
Ptrichocarpaphenotypes
and
OPseedlots(total
of70)from
thePmaximow
icziispecies
complex
(suaveolens
cathayana
koreana
ussuriensismaximow
iczii)and
2000ndash3000Pdelto
ides
collections
(based
on104s‐
generatio
nfamilies)
UMD
NRRIscreened
seedlin
gsfrom
aPdelto
ides
(OPor
CP)
breeding
programme(1996ndash2016)
gt7200OPseedlin
gsof
PnigraunderMinnesota
clim
atic
conditions(improve
parental
populatio
n)
3Wild
germ
plasm
genotyping
Amanaged
living
collectionof
clonal
types
US
~20000genotypes
CNChangsha
~1000Nanjin
g37
FR44
bycpDNA
(Fenget
al
UK~4
00
US
225
(Con
tinues)
8 | CLIFTON‐BROWN ET AL
TABLE
4(Contin
ued)
Breeding
techno
logy
step
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sWillow
Poplar
Construct
phylogenetic
treesanddissim
ilarity
indices
The
type
ofmolecular
analysis
mdashAFL
PcpDNAR
ADseq
whole
genomesequencing
2014)
JP~1
200
NL250genotypesby
RADseq
UK~1
000
byRADseq
SK~3
00
US
Illin
oisRADseqon
allwild
accessions
FR2310
SE150
US
1500
4Exploratory
crossing
and
progenytests
Discovergood
parental
combinatio
nsmdashgeneral
combining
ability
Geographicseparatio
n
speciesor
phylogenetic
trees
Costly
long‐te
rmmultilocation
trialsforprogeny
US
gt10000
CN~1
0
FR~2
00
JP~3
0
NL3600
SK~2
0
UK~4
000
(~1500Msin)
US
500ndash1000
UK gt700since2003
gt800EWBP(1996ndash
2002)
US
800
FRCloned13
times13
factorial
matingdesign
with
3species[10
Pdelto
ides8Ptrichocarpa
8
Pnigra)
US
GWR1000exploratory
crossingsof
Pfrem
ontii
Psimoniiforim
proved
droughttolerance
UMD
NRRIcrossednativ
e
Pdelto
ides
with
non‐nativ
e
Ptrichocarpaand
Pmaximow
icziiMajority
of
progenypopulatio
nssuffered
from
clim
atic
andor
pathogenic
susceptib
ility
issues
5Wideintraspecies
hybridization
Discovergood
parental
combinatio
nsmdashgeneral
combining
ability
1to
4above
inform
atics
Flow
eringsynchronizationand
low
seed
set
US
~200
CN~1
0
NL1800
SK~1
0
UK~5
00
UK276
US
400
IT500
SE120
US
50ndash100
6With
inspecies
recurrentselection
Concentratio
nof
positiv
e
traits
Identificationof
theright
heterotic
groups
insteps
1ndash5
Difficultto
introducenew
germ
plasm
with
outdilutio
n
ofbesttraits
US
gt40
populatio
ns
undergoing
recurrentselection
NL2populatio
ns
UK6populatio
ns
US
~12populatio
nsto
beestablished
by2019
UK4
US
150
FRPdelto
ides
(gt30
FSfamilies
ndash900clones)Pnigra(gt40
FS
families
ndash1600clones)and
Ptrichocarpa(15FS
Families
minus350clones)
IT80
SEca50
parent
breeding
populatio
nswith
in
Ptrichocarpa
US
Goalsis~1
00parent
breeding
populatio
nswith
inPtrichocarpa
Pdelto
ides
PnigraandPmaximow
iczii
7Interspecies
hybrid
breeding
Com
bining
complem
entary
traitsto
producegood
morphotypes
and
heterosiseffects
Allthesteps1ndash6
Inearlystage
improvem
ents
areunpredictable
Awide
base
iscostly
tomanage
None
CNChangsha
3Nanjin
g
120MsintimesMflo
r
JP~5
with
Saccharum
spontaneum
SK5
UK420
US
250
FRPcanadensis(1800)
Pdelto
ides
timesPtrichocarpa
(800)and
PtrichocarpatimesPmaximow
iczii
(1200)
(Con
tinues)
CLIFTON‐BROWN ET AL | 9
TABLE
4(Contin
ued)
Breeding
techno
logy
step
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sWillow
Poplar
IT~1
20
US
525genotypesin
eastside
hybrid
programme
205westside
hybrid
programmeand~1
200
in
SoutheastPdelto
ides
program
8Chrom
osom
e
doublin
g
Arouteto
triploid
seeded
typesdoublin
gadiploid
parent
ofknow
n
breeding
valuefrom
recurrentselection
Needs
1ndash7to
help
identify
therightparental
lines
Doubled
plantsarenotorious
forrevertingto
diploid
US
~50plantstakenfrom
tetraploid
tooctaploid
which
arenow
infieldtrials
CN~1
0Msactriploids
DKSeveralattemptsin
mid
90s
FR5genotypesof
Msin
UK2Msin1Mflora
nd1Msac
US
~30
UKAttempted
butnot
routinelyused
US
Not
partof
theregular
breeding
becausethere
existnaturalpolyploids
NA
9DoubleHaploids
Arouteto
producing
homogeneous
progeny
andalso
forthe
introductio
nof
transgenes
orforgenome
editing
Needs
1ndash7to
help
identify
therightparental
lines
Fully
homozygousplantsare
weakandeasily
die
andmay
notflow
ersynchronously
US
Noneyetattempteddueto
poor
vigour
andviability
of
haploids
CN2genotypesof
Msin
andMflor
UKDihaploidsof
MsinMflor
and
Msacand2transgenic
Msinvia
anther
cultu
re
US
6createdandused
toidentify
miss‐identifying
paralogueloci
(causedby
recent
genome
duplicationin
miscanthus)Usedin
creatin
gthereferencegenomefor
MsinVeryweakplantsand
difficultto
retain
thelin
es
NA
NA
10Embryo
rescue
Anattractiv
etechniquefor
recovering
plantsfrom
sexual
crosses
The
majority
ofem
bryos
cannot
survivein
vivo
or
becomelong‐timedorm
ant
None
UKone3times
hybrid
UKEmbryosrescue
protocol
proved
robustat
8days
post‐pollin
ation
FRAnem
bryo
rescue
protocol
proved
robustandim
proved
hybridizationsuccessfor
Pdelto
ides
timesPtrichocarpa
crosses
11Po
llenstorage
Flexibility
tocross
interestingparents
with
outneed
of
flow
ering
synchronization
None
Nosuccessto
daten
otoriously
difficultwith
grassesincluding
sugarcane
Freshpollencommonly
used
incrossing
Pollenstoragein
some
interspecificcrosses
FRBothstored
(cryobank)
and
freshindividual
pollen
Note
AFL
Pam
plifiedfragmentlength
polymorphismCBDconventio
non
biological
diversity
CP
controlledpollinatio
nCpD
NAchloroplastDNAEWBP
Europeanwillow
breeding
programme
FSfull‐sib
GtimesE
genotype‐by‐environm
entGRIN
germ
plasm
resourcesinform
ationnetwork
GWRGreenWoodResourcesMflorMiscanthusflo
ridulusMsacMsacchariflo
rus
MsinMsinensisNAnotapplicableNRRINatural
Resources
ResearchInstitu
teOP
open
pollinatio
nRADseq
restrictionsite‐associatedDNA
sequencingSL
USw
edishUniversity
ofAgriculturalSciencesST
TSw
eTreeTech
UMDUniversity
ofMinnesota
Duluth
a ISO
Alpha‐2
lettercountrycodes
10 | CLIFTON‐BROWN ET AL
programmes mature The average rate of gain for biomassyield in long‐term switchgrass breeding programmes hasbeen 1ndash4 per year depending on ecotype populationand location of the breeding programme (Casler amp Vogel2014 Casler et al 2018) The hybrid derivative Libertyhas a biomass yield 43 higher than the better of its twoparents (Casler amp Vogel 2014 Vogel et al 2014) Thedevelopment of cold‐tolerant and late‐flowering lowland‐ecotype populations for the northern United States hasincreased biomass yields by 27 (Casler et al 2018)
Currently more than 20 recurrent selection populationsare being managed in the United States to select parentsfor improved yield yield resilience and compositional qual-ity of the biomass For the agronomic development andupscaling high seed multiplication rates need to be com-bined with lower seed dormancy to reduce both crop estab-lishment costs and risks Expresso is the first cultivar withsignificantly reduced seed dormancy which is the first steptowards development of domesticated populations (Casleret al 2015) Most phenotypic traits of interest to breeders
require a minimum of 2 years to be fully expressed whichresults in a breeding cycle that is at least two years Morecomplicated breeding programmes or traits that requiremore time to evaluate can extend the breeding cycle to 4ndash8 years per generation for example progeny testing forbiomass yield or field‐based selection for cold toleranceBreeding for a range of traits with such long cycles callsfor the development of molecular methods to reduce time-scales and improve breeding efficiency
Two association panels of switchgrass have been pheno-typically and genotypically characterized to identify quanti-tative trait loci (QTLs) that control important biomasstraits The northern panel consists of 60 populationsapproximately 65 from the upland ecotype The southernpanel consists of 48 populations approximately 65 fromthe lowland ecotype Numerous QTLs have been identifiedwithin the northern panel to date (Grabowski et al 2017)Both panels are the subject of additional studies focused onbiomass quality flowering and phenology and cold toler-ance Additionally numerous linkage maps have been
FIGURE 1 Cumulative minimum years needed for the conventional breeding cycle through the steps from wild germplasm to thecommercial hybrids in switchgrass miscanthus willow and poplar Information links between the steps are indicated by dotted arrows andhighlight the importance of long‐term informatics to maximize breeding gain
CLIFTON‐BROWN ET AL | 11
created by the pairwise crossing of individuals with diver-gent characteristics often to generate four‐way crosses thatare analysed as pseudo‐F2 crosses (Liu Wu Wang ampSamuels 2012 Okada et al 2010 Serba et al 2013Tornqvist et al 2018) Individual markers and QTLs iden-tified can be used to design marker‐assisted selection(MAS) programmes to accelerate breeding and increase itsefficiency Genomic prediction and selection (GS) holdseven more promise with the potential to double or triplethe rate of gain for biomass yield and other highly complexquantitative traits of switchgrass (Casler amp Ramstein 2018Ramstein et al 2016) The genome of switchgrass hasrecently been made public through the Joint Genome Insti-tute (httpsphytozomejgidoegov)
Transgenic approaches have been heavily relied upon togenerate unique genetic variants principally for traitsrelated to biomass quality (Merrick amp Fei 2015) Switch-grass is highly transformable using either Agrobacterium‐mediated transformation or biolistics bombardment butregeneration of plants is the bottleneck to these systemsTraditionally plants from the cultivar Alamo were the onlyregenerable genotypes but recent efforts have begun toidentify more genotypes from different populations that arecapable of both transformation and subsequent regeneration(King Bray Lafayette amp Parrott 2014 Li amp Qu 2011Ogawa et al 2014 Ogawa Honda Kondo amp Hara‐Nishi-mura 2016) Cell wall recalcitrance and improved sugarrelease are the most common targets for modification (Bis-wal et al 2018 Fu et al 2011) Transgenic approacheshave the potential to provide traits that cannot be bredusing natural genetic variability However they will stillrequire about 10ndash15 years and will cost $70ndash100 millionfor cultivar development and deployment (Harfouche Mei-lan amp Altman 2011) In addition there is commercialuncertainty due to the significant costs and unpredictabletimescales and outcomes of the regulatory approval processin the countries targeted for seed sales As seen in maizeone advantage of transgenic approaches is that they caneasily be incorporated into F1 hybrid cultivars (Casler2012 Casler et al 2012) but this does not decrease thetime required for cultivar development due to field evalua-tion and seed multiplication requirements
The potential impacts of unintentional gene flow andestablishment of non‐native transgene sequences in nativeprairie species via cross‐pollination are also major issues forthe environmental risk assessment These limit further thecommercialization of varieties made using these technolo-gies Although there is active research into switchgrass steril-ity mechanisms to curb unintended pollen‐mediated genetransfer it is likely that the first transgenic cultivars proposedfor release in the United States will be met with considerableopposition due to the potential for pollen flow to remainingwild prairie sites which account for lt1 of the original
prairie land area and are highly protected by various govern-mental and nongovernmental organizations (Casler et al2015) Evidence for landscape‐level pollen‐mediated geneflow from genetically modified Agrostis seed multiplicationfields (over a mountain range) to pollinate wild relatives(Watrud et al 2004) confirms the challenge of using trans-genic approaches Looking ahead genome editing technolo-gies hold considerable promise for creating targeted changesin phenotype (Burris Dlugosz Collins Stewart amp Lena-ghan 2016 Liu et al 2018) and at least in some jurisdic-tions it is likely that cultivars resulting from gene editingwill not need the same regulatory approval as GMOs (Jones2015a) However in July 2018 the European Court of Justice(ECJ) ruled that cultivars carrying mutations resulting fromgene editing should be regulated in the same way as GMOsThe ECJ ruled that such cultivars be distinguished from thosearising from untargeted mutation breeding which isexempted from regulation under Directive 200118EC
3 | MISCANTHUS
Miscanthus is indigenous to eastern Asia and Oceania whereit is traditionally used for forage thatching and papermaking(Xi 2000 Xi amp Jezowkski 2004) In the 1960s the highbiomass potential of a Japanese genotype introduced to Eur-ope by Danish nurseryman Aksel Olsen in 1935 was firstrecognized in Denmark (Linde‐Laursen 1993) Later thisaccession was characterized described and named asldquoM times giganteusrdquo (Greef amp Deuter 1993 Hodkinson ampRenvoize 2001) commonly abbreviated as Mxg It is a natu-rally occurring interspecies triploid hybrid between tetraploidM sacchariflorus (2n = 4x) and diploid M sinensis(2n = 2x) Despite its favourable agronomic characteristicsand ability to produce high yields in a wide range of environ-ments in Europe (Kalinina et al 2017) the risks of relianceon it as a single clone have been recognized Miscanthuslike switchgrass is outcrossing due to self‐incompatibility(Jiang et al 2017) Thus seeded hybrids are an option forcommercial breeding Miscanthus can also be vegetativelypropagated by rhizome or in vitro culture which allows thedevelopment of clones The breeding approaches are usuallybased on F1 crosses and recurrent selection cycles within thesynthetic populations There are several breeding pro-grammes that target improvement of miscanthus traitsincluding stress resilience targeted regional adaptation agro-nomic ldquoscalabilityrdquo through cheaper propagation fasterestablishment lower moisture and ash contents and greaterusable yield (Clifton‐Brown et al 2017)
Germplasm collections specifically to support breedingfor biomass started in the late 1980s and early 1990s inDenmark Germany and the United Kingdom (Clifton‐Brown Schwarz amp Hastings 2015) These collectionshave continued with successive expeditions from European
12 | CLIFTON‐BROWN ET AL
and US teams assembling diverse collections from a widegeographic range in eastern Asia including from ChinaJapan South Korea Russia and Taiwan (Hodkinson KlaasJones Prickett amp Barth 2015 Stewart et al 2009) Threekey miscanthus species for biomass production are M si-nensis M floridulus and M sacchariflorus M sinensis iswidely distributed throughout eastern Asia with an adap-tive range from the subtropics to southern Russia (Zhaoet al 2013) This species has small rhizomes and producesmany tightly packed shoots forming a ldquotuftrdquo M floridulushas a more southerly adaptive range with a rather similarmorphology to M sinensis but grows taller with thickerstems and is evergreen and less cold‐tolerant than the othermiscanthus species M sacchariflorus is the most northern‐adapted species ranging to 50 degN in eastern Russia (Clarket al 2016) Populations of diploid and tetraploid M sac-chariflorus are found in China (Xi 2000) and South Korea(Yook 2016) and eastern Russia but only tetraploids havebeen found in Japan (Clark Jin amp Petersen 2018)
Germplasm has been assembled from multiple collec-tions over the last century though some early collectionsare poorly documented This historical germplasm has beenused to initiate breeding programmes largely based on phe-notypic and genotypic characterization As many of theaccessions from these collections are ldquoorigin unknownrdquocrucial environmental envelope data are not available UK‐led expeditions started in 2006 and continued until 2011with European and Asian partners and have built up a com-prehensive collection of 1500 accessions from 500 sitesacross Eastern Asia including China Japan South Koreaand Taiwan These collections were guided using spatialclimatic data to identify variation in abiotic stress toleranceAccessions from these recent collections were planted fol-lowing quarantine in multilocation nursery trials at severallocations in Europe to examine trait expression in differentenvironments Based on the resulting phenotypic andmolecular marker data several studies (a) characterized pat-terns of population genetic structure (Slavov et al 20132014) (b) evaluated the statistical power of genomewideassociation studies (GWASs) and identified preliminarymarkerndashtrait associations (Slavov et al 2013 2014) and(c) assessed the potential of genomic prediction (Daveyet al 2017 Slavov et al 2014 2018b) Genomic indexselection in particular offers the possibility of exploringscenarios for different locations or industrial markets (Sla-vov et al 2018a 2018b)
Separately US‐led expeditions also collected about1500 accessions between 2010 and 2014 (Clark et al2014 2016 2018 2015) A comprehensive genetic analy-sis of the population structure has been produced by RAD-seq for M sinensis (Clark et al 2015 Van der WeijdeKamei et al 2017) and M sacchariflorus (Clark et al2018) Multilocation replicated field trials have also been
conducted on these materials in North America and inAsia GWAS has been conducted for both M sinensis anda subset of M sacchariflorus accessions (Clark et al2016) To date about 75 of these recent US‐led collec-tions are in nursery trials outside the United States Due tolengthy US quarantine procedures these are not yet avail-able for breeding in the United States However molecularanalyses have allowed us to identify and prioritize sets ofgenotypes that best encompass the genetic variation in eachspecies
While mostM sinensis accessions flower in northern Eur-ope very few M sacchariflorus accessions flower even inheated glasshouses For this reason the European pro-grammes in the United Kingdom the Netherlands and Francehave performed mainly M sinensis (intraspecies) hybridiza-tions (Table 4) Selected progeny become the parents of latergenerations (recurrent selection as in switchgrass) Seed setsof up to 400 seed per panicle occur inM sinensis In Aberyst-wyth and Illinois significant efforts to induce synchronousflowering in M sacchariflorus and M sinensis have beenmade because interspecies hybrids have proven higher yieldperformance and wide adaptability (Kalinina et al 2017) Ininterspecies pairwise crosses in glasshouses breathable bagsandor large crossing tubes or chambers in which two or morewhole plants fit are used for pollination control Seed sets arelower in bags than in the open air because bags restrict pollenmovement while increasing temperatures and reducinghumidity (Clifton‐Brown Senior amp Purdy 2018) About30 of attempted crosses produced 10 to 60 seeds per baggedpanicle The seed (thousand seed mass ranges from 05 to09 g) is threshed from the inflorescences and sown into mod-ular trays to produce plug plants which are then planted infield nurseries to identify key parental combinations
A breeding programme of this scale must serve theneeds of different environments accepting the commonpurpose is to optimize the interception of solar radiationAn ideal hybrid for a given environment combines adapta-tion to date of emergence with optimization of traits suchas height number of stems per plant flowering and senes-cence time to optimize solar interception to produce a highbiomass yield with low moisture content at harvest (Rob-son Farrar et al 2013 Robson Jensen et al 2013) By20132014 conventional breeding in Europe had producedintra‐ and interspecific fertile seeded hybrids When acohort (typically about 5) of outstanding crosses have beenidentified it is important to work on related upscaling mat-ters in parallel These are as follows
Assessment of the yield and critical traits in selectedhybrids using a network of field trials
Efficient cloning of the seed parents While in vitro andmacro‐cloning techniques are used some genotypes areamenable to neither technique
CLIFTON‐BROWN ET AL | 13
TABLE5
Status
ofmod
ernplantbreeding
techniqu
esin
four
leadingperennialbiom
asscrop
s(PBCs)
Breeding
techno
logy
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sW
illow
Pop
lar
MAS
Use
ofmarker
sequ
encesthat
correlatewith
atrait
allowingearly
prog
enyselectionor
rejectionlocatin
gkn
owngenesfor
useful
traits(suchas
height)from
other
speciesin
your
crop
Breeding
prog
ramme
relevant
biparental
crosses(Table
4)to
create
ldquomapping
popu
latio
nsrdquoand
QTLidentification
andestim
ationto
identifyrobu
stmarkers
alternatively
acomprehensive
panelof
unrelated
geno
typesfor
GWAS
Ineffectivewhere
traits
areaffected
bymany
geneswith
small
effects
US
Literally
dozens
ofstud
iesto
identify
SNPs
andmarkers
ofinterestNoattempts
atMASas
yetlargely
dueto
popu
latio
nspecificity
CNS
SRISS
Rmarkers
forM
sinMsacand
Mflor
FR2
Msin
mapping
popu
latio
ns(BFF
project)
NL2
Msin
mapping
popu
latio
ns(A
tienza
Ram
irezetal
2003
AtienzaSatovic
PetersenD
olstraamp
Martin
200
3Van
Der
Weijdeetal20
17a
Van
derW
eijde
Hux
ley
etal20
17
Van
derW
eijdeKam
ei
etal20
17V
ander
WeijdeKieseletal
2017
)SK
2GWASpanels(1
collectionof
Msin1
diploidbiparentalF 1
popu
latio
n)in
the
pipelin
eUK1
2families
tofind
QTLin
common
indifferentfam
ilies
ofthe
sameanddifferent
speciesandhy
brids
US
2largeGWAS
panels4
diploidbi‐
parentalF 1
popu
latio
ns
1F 2
popu
latio
n(add
ition
alin
pipelin
e)
UK16
families
for
QTLdiscov
ery
2crossesMASscreened
1po
tentialvariety
selected
US
9families
geno
typedby
GBS(8
areF 1o
neisF 2)a
numberof
prom
ising
QTLdevelopm
entof
markers
inprog
ress
FR(INRA)tested
inlargeFS
families
and
infactorialmating
design
low
efficiency
dueto
family
specificity
US
49families
GS
Metho
dto
accelerate
breeding
throug
hredu
cing
the
resourcesforcross
attemptsby
Aldquotraining
popu
latio
nrdquoof
individu
alsfrom
stages
abov
ethat
have
been
both
Risks
ofpo
orpredictio
nProg
eny
testingneedsto
becontinueddu
ring
the
timethetraining
setis
US
One
prog
rammeso
farmdash
USD
Ain
WisconsinThree
cycles
ofgeno
mic
selectioncompleted
CN~1
000
geno
types
NLprelim
inary
mod
elsforMsin
developed
UK~1
000
geno
types
Verysuitable
application
not
attempted
yet
Maybe
challeng
ingfor
interspecies
hybrids
FR(INRA)120
0geno
type
ofPop
ulus
nigraarebeingused
todevelop
intraspecificGS
(Con
tinues)
14 | CLIFTON‐BROWN ET AL
TABLE
5(Contin
ued)
Breeding
techno
logy
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sW
illow
Pop
lar
predictin
gthe
performance
ofprog
enyof
crosses
(and
inresearch
topredictbestparents
tousein
biparental
crosses)
geno
typedand
phenotyp
edto
developamod
elthattakesgeno
typic
datafrom
aldquocandidate
popu
latio
nrdquoof
untested
individu
als
andprod
uces
GEBVs(Jannink
Lorenzamp
Iwata
2010
)
beingph
enotyp
edand
geno
typed
Inthis
time
next‐generation
germ
plasm
andreadyto
beginthe
second
roun
dof
training
and
recalib
ratio
nof
geno
mic
predictio
nmod
els
US
Prelim
inary
mod
elsforMsac
and
Msin
developed
Work
isun
derw
ayto
deal
moreeffectivelywith
polyploidgenetics
calib
ratio
nsUS
125
0geno
types
arebeingused
todevelopGS
calib
ratio
ns
Traditio
nal
transgenesis
Efficient
introd
uctio
nof
ldquoforeign
rdquotraits
(possiblyfrom
other
genu
sspecies)
into
anelite
plant(ega
prov
enparent
ora
hybrid)that
needsa
simpletraitto
beim
prov
edValidate
cand
idategenesfrom
QTLstud
ies
MASand
know
ledg
eof
the
biolog
yof
thetrait
andsource
ofgenesto
confer
the
relevant
changesto
phenotyp
eWorking
transformation
protocol
IPissuescostof
regu
latory
approv
al
GMO
labelling
marketin
gissuestricky
tousetransgenes
inou
t‐breedersbecause
ofcomplexity
oftransformingandgene
flow
risks
US
Manyprog
ramsin
UnitedStates
are
creatin
gtransgenic
plantsusingAlamoas
asource
oftransformable
geno
typesManytraits
ofinterestNothing
commercial
yet
CNC
hang
shaBtg
ene
transformed
into
Msac
in20
04M
sin
transformed
with
MINAC2gene
and
Msactransform
edwith
Cry2A
ageneb
othwith
amarkerg
eneby
Agrob
acterium
JP1
Msintransgenic
forlow
temperature
tolerancewith
increased
expression
offructans
(unp
ublished)
NLImprov
ingprotocol
forM
sin
transformation
SK2
Msin
with
Arabido
psisand
Brachypod
ium
PhytochromeBgenes
(Hwang
Cho
etal
2014
Hwang
Lim
etal20
14)
UK2
Msin
and
1Mflorg
enotyp
es
UKRou
tine
transformationno
tyet
possibleResearchto
overcomerecalcitrance
ongo
ing
Currently
trying
differentspecies
andcond
ition
sPo
plar
transformationused
atpresent
US
Frequently
attempted
with
very
limitedsuccess
US
600transgenic
lines
have
been
characterized
mostly
performed
inthe
aspenhy
brids
reprod
uctiv
esterility
drou
ghttolerance
with
Kno
ckdo
wns
(Con
tinues)
CLIFTON‐BROWN ET AL | 15
TABLE
5(Contin
ued)
Breeding
techno
logy
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sW
illow
Pop
lar
variou
slytransformed
with
four
cellwall
genesiptand
uidA
genesusingtwo
selectionsystem
sby
biolisticsand
Agrob
acterium
Transform
edplantsare
beinganalysed
US
Prelim
inarywork
will
betakenforw
ardin
2018
ndash202
2in
CABBI
Genom
eediting
CRISPR
Refinem
entofexisting
traits
inusefulparents
orprom
isinghybrids
bygeneratingtargeted
mutations
ingenes
know
ntocontrolthe
traitof
interestF
irst
adouble‐stranded
breakismadeinthe
DNAwhich
isrepairedby
natural
DNArepair
machineryL
eads
tofram
eshiftSN
Por
can
usealdquorepair
templaterdquo
orcanbe
used
toinserta
transgene
intoaldquosafe
harbourlocusrdquoIt
couldbe
used
todeleterepressors(or
transcriptionfactors)
Identification
mapping
and
sequ
encing
oftarget
genes(from
DNA
sequ
ence)
avoiding
screening
outof
unintend
ededits
Manyregu
latory
authorities
have
not
decidedwhether
CRISPR
andother
geno
meediting
techno
logies
are
GMOsor
notIf
GMOsseecomment
onstage10
abov
eIf
noteditedcrop
swill
beregu
latedas
conv
entio
nalvarieties
CATechn
olog
yisstill
toonewand
switchg
rass
geno
meis
very
complexOther
labo
ratories
are
interestedbu
tno
tyet
mov
ingon
this
US
One
prog
rammeso
farmdash
USD
Ain
Albany
FRInitiated
in20
16(M
ISEDIT
project)
NLInitiatingin
2018
US
Initiatingin
2018
in
CABBI
UKNot
started
Not
possible
until
transformation
achieved
US
CRISPR
‐Cas9and
Cpf1have
been
successfully
developedin
Pop
ulus
andarehigh
lyefficientExamples
forlig
nin
biosyn
thesis
Note
BFF
BiomassFo
rtheFu
tureBtBacillus
thuringiensisCABBICenterforadvanced
bioenergyandbioproductsinnovatio
nCpf1
CRISPR
from
Prevotella
andFrancisella
1CRISPR
clusteredregularlyinterspaced
palin
drom
icrepeatsCRISPR
‐CasC
RISPR
‐associated
Cry2A
acrystaltoxins2A
asubfam
ilyproduced
byBtFS
full‐sibG
BS
genotyping
bysequencingG
EBVsgenomic‐estim
ated
breeding
valuesG
MOg
enetically
modi-
fied
organism
GS
genomicselection
GWAS
genomew
ideassociationstudy
INRAF
renchNationalInstituteforAgriculturalR
esearch
IPintellectualp
roperty
IPTisopentenyltransferase
ISSR
inter‐SSR
MflorMiscant-
husflo
ridulusMsacMsacchariflo
rusMsinMsinensisM
AS
marker‐a
ssistedselection
MISEDITm
iscanthusgene
editing
forseed‐propagatedtriploidsNACn
oapicalmeristemA
TAF1
2and
cup‐shaped
cotyledon2
‐lik
eQTLq
uantitativ
etraitlocusS
NP
singlenucleotid
epolymorphismS
SRsim
plesequence
repeatsUSD
AU
SDepartm
ento
fAgriculture
a ISO
Alpha‐2
lettercountrycodes
16 | CLIFTON‐BROWN ET AL
High seed production from field crossing trials con-ducted in locations where flowering in both seed andpollen parents is likely to happen synchronously
Scalable and adapted harvesting threshing and seed pro-cessing methods for producing high seed quality
The results of these parallel activities need to becombined to identify the upscaling pathway for eachhybrid if this cannot be achieved the hybrid will likelynot be commercially viable The UK‐led programme withpartners in Italy and Germany shows that seedbased mul-tiplication rates of 12000 are achievable several inter-specific hybrids (Clifton‐Brown et al 2017) Themultiplication rate of M sinensis is higher probably15000ndash10000 Conventional cloning from rhizome islimited to around 120 that is one ha could provide rhi-zomes for around 20 ha of new plantation
Multilocation field testing of wild and novel miscanthushybrids selected by breeding programmes in the Nether-lands and the United Kingdom was performed as part ofthe project Optimizing Miscanthus Biomass Production(OPTIMISC 2012ndash2016) These trials showed that com-mercial yields and biomass qualities (Kiesel et al 2017Van der Weijde et al 2017a Van der Weijde Kieselet al 2017) could be produced in a wide range of climatesand soil conditions from the temperate maritime climate ofwestern Wales to the continental climate of eastern Russiaand the Ukraine (Kalinina et al 2017) Extensive environ-mental measurements of soil and climate combined withgrowth monitoring are being used to understand abioticstresses (Nunn et al 2017 Van der Weijde Huxley et al2017) and develop genotype‐specific scenarios similar tothose reported earlier in Hastings et al (2009) Phenomicsexperiments on drought tolerance have been conducted onwild and improved germplasm (Malinowska Donnison ampRobson 2017 Van der Weijde Huxley et al 2017)Recently produced interspecific hybrids displaying excep-tional yield under drought (~30 greater than control Mxg)in field trials in Poland and Moldova are being furtherstudied in detail in the phenomics and genomics facility atAberystwyth to better understand genendashtrait associationswhich can be fed back into breeding
Intraspecific seeded hybrids of M sinensis produced inthe Netherlands and interspecific M sacchariflorus times M si-nensis hybrids produced by the UK‐led breeding programmehave entered yield testing in 2018 with the recently EU‐funded project ldquoGRowing Advanced industrial Crops on mar-ginal lands for biorEfineries (GRACE)rdquo (httpswwwgrace-bbieu) Substantial variation in biomass quality for sacchari-fication efficiency (glucose release as of dry matter) ashcontent and melting point has already been generated inintraspecific M sinensis hybrids (Van der Weijde Kiesel
et al 2017) across environments (Weijde Dolstra et al2017a) GRACE aims to establish more than 20 hectares ofnew inter‐ and intraspecific seeded hybrids across six Euro-pean countries This project is building the know‐how andagronomy needed to transition from small research plots tocommercial‐scale field sites and linking biomass productiondirectly to industrial applications The biomass produced byhybrids in different locations will be supplied to innovativeindustrial end‐users making a wide range of biobased prod-ucts both for chemicals and for energy In the United Statesmulti‐location yield were initiated in 2018 to evaluate new tri-ploid M times giganteus genotypes developed at Illinois Cur-rently infertile hybrids are favoured in the United Statesbecause this eliminates the risk of invasiveness from naturallydispersed viable seed The precautionary principle is appliedas fertile miscanthus has naturalized in several states (QuinnAllen amp Stewart 2010) North European multilocation fieldtrials in the EMI and OPTIMISC projects have shown thereis minimal risk of invasiveness even in years when fertileflowering hybrids produce viable seed Naturalized standshave not established here due perhaps to low dormancy pooroverwintering and low seedling competitive strength In addi-tion to breeding for nonshattering or sterile seeded hybridsQuinn et al (2010) suggest management strategies which canfurther minimize environmental opportunities to manage therisk of invasiveness
31 | Molecular breeding and biotechnology
In miscanthus new plant breeding techniques (Table 5)have focussed on developing molecular markers forbreeding in Europe the United States South Korea andJapan There are several publications on QTL mappingpopulations for key traits such as flowering (AtienzaRamirez amp Martin 2003) and compositional traits(Atienza Satovic Petersen Dolstra amp Martin 2003) Inthe United States and United Kingdom independent andinterconnected bi‐parental ldquomappingrdquo families have beenstudied (Dong et al 2018 Gifford Chae SwaminathanMoose amp Juvik 2015) alongside panels of diverse germ-plasm accessions for GWAS (Slavov et al 2013) Fur-ther developments calibrating GS with very large panelsof parents and cross progeny are underway (Daveyet al 2017) The recently completed first miscanthus ref-erence genome sequence is expected to improve the effi-ciency of MAS strategies and especially GWAS(httpsphytozomejgidoegovpzportalhtmlinfoalias=Org_Msinensis_er) For example without a referencegenome sequence Clark et al (2014) obtained 21207RADseq SNPs (single nucleotide polymorphisms) on apanel of 767 miscanthus genotypes (mostly M sinensis)but subsequent reanalysis of the RADseq data using the
CLIFTON‐BROWN ET AL | 17
new reference genome resulted in hundreds of thousandsof SNPs being called
Robust and effective in vitro regeneration systems havebeen developed for Miscanthus sinensis M times giganteusand M sacchariflorus (Dalton 2013 Guo et al 2013Hwang Cho et al 2014 Rambaud et al 2013 Ślusarkie-wicz‐Jarzina et al 2017 Wang et al 2011 Zhang et al2012) However there is still significant genotype speci-ficity and these methods need ldquoin‐houserdquo optimization anddevelopment to be used routinely They provide potentialroutes for rapid clonal propagation and also as a basis forgenetic transformation Stable transformation using bothbiolistics (Wang et al 2011) and Agrobacterium tumefa-ciens DNA delivery methods (Hwang Cho et al 2014Hwang Lim et al 2014) has been achieved in M sinen-sis The development of miscanthus transformation andgene editing to generate diplogametes for producing seed‐propagated triploid hybrids are performed as part of theFrench project MISEDIT (miscanthus gene editing forseed‐propagated triploids) There are no reports of genomeediting in any miscanthus species but new breeding inno-vations including genome editing are particularly relevantin this slow‐to‐breed nonfood bioenergy crop (Table 4)
4 | WILLOW
Willow (Salix spp) is a very diverse group of catkin‐bear-ing trees and shrubs Willow belongs to the family Sali-caceae which also includes the Populus genus There areapproximately 350 willow species (Argus 2007) foundmostly in temperate and arctic zones in the northern hemi-sphere A few are adapted to subtropical and tropicalzones The centre of diversity is believed to be in Asiawith over 200 species in China Around 120 species arefound in the former Soviet Union over 100 in NorthAmerica and around 65 species in Europe and one speciesis native to South America (Karp et al 2011) Willows aredioecious thus obligate outcrossers and highly heterozy-gous The haploid chromosome number is 19 (Hanley ampKarp 2014) Around 40 of species are polyploid (Sudaamp Argus 1968) ranging from triploids to the atypicaldodecaploid S maxxaliana with 2n=190 (Zsuffa et al1984)
Although almost exclusively native to the NorthernHemisphere willow has been grown around the globe formany thousands of years to support a wide range of appli-cations (Kuzovkina amp Quigley 2005 Stott 1992) How-ever it has been the focus of domestication for bioenergypurposes for only a relatively short period since the 1970sin North America and Europe For bioenergy breedershave focused their efforts on the shrub willows (subgenusVetix) because of their rapid juvenile growth rates as aresponse to coppicing on a 2‐ to 4‐year cycle that can be
accomplished using farm machinery rather than forestryequipment (Shield Macalpine Hanley amp Karp 2015Smart amp Cameron 2012)
Since shrub willow was not generally recognized as anagricultural crop until very recently there has been littlecommitment to building and maintaining germplasm reposi-tories of willow to support long‐term breeding One excep-tion is the United Kingdom where a large and well‐characterized Salix germplasm collection comprising over1500 accessions is held at Rothamsted Research (Stott1992 Trybush et al 2008) Originally initiated for use inbasketry in 1923 accessions have been added ever sinceIn the United States a germplasm collection of gt350accessions is located at Cornell University to support thebreeding programme there The UK and Cornell collectionshave a relatively small number of accessions in common(around 20) Taken together they represent much of thespecies diversity but only a small fraction of the overallgenetic diversity within the genus There are three activewillow breeding programmes in Europe RothamstedResearch (UK) Salixenergi Europa AB (SEE) and a pro-gramme at the University of Warmia and Mazury in Olsz-tyn (Poland) (abbreviations used in Table 1) There is oneactive US programme based at Cornell University Culti-vars are still being marketed by the European WillowBreeding Programme (EWBP) (UK) which was activelybreeding biomass varieties from 1996 to 2002 Cultivarsare protected by plant breedersrsquo rights (PBRs) in Europeand by plant patents in the United States The sharing ofgenetic resources in the willow community is generallyregulated by material transfer agreements (MTA) and tai-lored licensing agreements although the import of cuttingsinto North America is prohibited except under special quar-antine permit conditions
Efforts to augment breeding germplasm collection fromnature are continuing with phenotypic screening of wildgermplasm performed in field experiments with 177 S pur-purea genotypes in the United States (at sites in Genevaand Portland NY and Morgantown WV) that have beengenotyped using genotyping by sequencing (GBS) (Elshireet al 2011) In addition there are approximately 400accessions of S viminalis in Europe (near Pustnaumls Upp-sala Sweden and Woburn UK (Berlin et al 2014 Hal-lingbaumlck et al 2016) The S viminalis accessions wereinitially genotyped using 38 simple sequence repeats (SSR)markers to assess genetic diversity and screened with~1600 SNPs in genes of potential interest for phenologyand biomass traits Genetics and genomics combined withextensive phenotyping have substantially improved thegenetic basis of biomass‐related traits in willow and arenow being developed in targeted breeding via MAS Thisunderpinning work has been conducted on large specifi-cally developed biparental Salix mapping populations
18 | CLIFTON‐BROWN ET AL
(Hanley amp Karp 2014 Zhou et al 2018) as well asGWAS panels (Hallingbaumlck et al 2016)
Once promising parental combinations are identifiedcrosses are usually performed using fresh pollen frommaterial that has been subject to a phased removal fromcold storage (minus4degC) (Lindegaard amp Barker 1997 Macal-pine Shield Trybush Hayes amp Karp 2008 Mosseler1990) Pollen storage is useful in certain interspecific com-binations where flowering is not naturally synchronizedThis can be overcome by using pollen collection and stor-age protocol which involves extracting pollen using toluene(Kopp Maynard Niella Smart amp Abrahamson 2002)
The main breeding approach to improve willow yieldsrelies on species hybridization to capture hybrid vigour(Fabio et al 2017 Serapiglia Gouker amp Smart 2014) Inthe absence of genotypic models for heterosis breedershave extensively tested general and specific combiningability of parents to produce superior progeny The UKbreeding programmes (EWBP 1996ndash2002 and RothamstedResearch from 2003 on) have performed more than 1500exploratory cross‐pollinations The Cornell programme hassuccessfully completed about 550 crosses since 1998Investment into the characterization of genetic diversitycombined with progeny tests from exploratory crosses hasbeen used to produce hundreds of targeted intraspeciescrosses in the United Kingdom and United States respec-tively (see Table 1) To achieve long‐term gains beyond F1hybrids four intraspecific recurrent selection populationshave been created in the United Kingdom (for S dasycla-dos S viminalis and S miyabeana) and Cornell is pursu-ing recurrent selection of S purpurea Interspecifichybridizations with genotypes selected from the recurrentselection cycles are well advanced in willow with suchcrosses to date totalling 420 in the United Kingdom andover 100 in the United States
While species hybridization is common in Salix it isnot universal Of the crosses attempted about 50 hybri-dize and produce seed (Macalpine Shield amp Karp 2010)As the viability of seed from successful crosses is short (amatter of days at ambient temperatures) proper seed rear-ing and storage protocols are essential (Maroder PregoFacciuto amp Maldonado 2000)
Progeny from crosses are treated in different waysamong the breeding programmes at the seedling stage Inthe United States seedlings are planted into an irrigatedfield where plants are screened for two seasons beforebeing progressed to further field trials In the United King-dom seedlings are planted into trays of compost wherethey remain containerized in an irrigated nursery for theremainder of year one In the United Kingdom seedlingsare subject to two rounds of selection in the nursery yearThe first round takes place in September to select againstsusceptibility to rust infection (Melampsora spp) A second
round of selection in winter assesses tip damage from frostand giant willow aphid infestation In the United Stateswhere the rust pressure is lower screening for Melampsoraspp cannot be performed at the nursery stage Both pro-grammes monitor Melampsora spp pest susceptibilityyield and architecture over multiple years in field trialsSelected material is subject to two rounds of field trials fol-lowed by a final multilocation yield trial to identify vari-eties for commercialization
Promising selections (ie potential cultivars) need to beclonally propagated A rapid in vitro tissue culture propa-gation method has been developed (Palomo‐Riacuteos et al2015) This method can generate about 5000 viable trans-plantable clones from a single plant in just 24 weeks Anin vitro system can also accommodate early selection viamolecular or biochemical markers to increase selectionspeed Conventional breeding systems take 13 years viafour rounds of selection from crossing to selecting a variety(Figure 1) but this has the potential to be reduced to7 years if micropropagation and MAS selection are adopted(Hanley amp Karp 2014 Palomo‐Riacuteos et al 2015)
Willows are currently propagated commercially byplanting winter‐dormant stem cuttings in spring Commer-cial planting systems for willow use mechanical plantersthat cut and insert stem sections from whips into a well‐prepared soil One hectare of stock plants grown in specificmultiplication beds planted at 40000 plants per ha pro-duces planting material for 80 hectares of commercialshort‐rotation coppice willow annually (planted at 15000cuttings per hectare) (Whittaker et al 2016) When com-mercial plantations are established the industry standard isto plant intimate mixtures of ~5 diverse rust (Melampsoraspp)‐resistant varieties (McCracken amp Dawson 1997 VanDen Broek et al 2001)
The foundations for using new plant breeding tech-niques have been established with funding from both thepublic and the private sectors To establish QTL maps 16mapping populations from biparental crosses are understudy in the United Kingdom Nine are under study in theUnited States The average number of individuals in thesefamilies ranges from 150 to 947 (Hanley amp Karp 2014)GS is also being evaluated in S viminalis and preliminaryresults indicate that multiomic approaches combininggenomic and metabolomic data have great potential (Sla-vov amp Davey 2017) For both QTL and GS approachesthe field phenotyping demands are large as several thou-sand individuals need to be phenotyped for a wide rangeof traits These include the following dates of bud burstand growth succession stem height stem density wooddensity and disease resistance The greater the number ofindividuals the more precise the QTL marker maps andGS models are However the logistical and financial chal-lenges of phenotyping large numbers of individuals are
CLIFTON‐BROWN ET AL | 19
considerable because the willow crop is gt5 m tall in thesecond year There is tremendous potential to improve thethroughput of phenotyping using unmanned aerial systemswhich is being tested in the USDA National Institute ofFood and Agriculture (NIFA) Willow SkyCAP project atCornell Further investment in these approaches needs tobe sustained over many years fully realizes the potentialof a marker‐assisted selection programme for willow
To date despite considerable efforts in Europe and theUnited States to establish a routine transformation systemthere has not been a breakthrough in willow but attemptsare ongoing As some form of transformation is typically aprerequisite for genome editing techniques these have notyet been applied to willow
In Europe there are 53 short‐rotation coppice (SRC) bio-mass willow cultivars registered with the Community PlantVariety Office (CPVO) for PBRs of which ~25 are availablecommercially in the United Kingdom There are eightpatented cultivars commercially available in the UnitedStates In Sweden there are nine commercial cultivars regis-tered in Europe and two others which are unregistered(httpssalixenergiseplanting-material) Furthermore thereare about 20 precommercial hybrids in final yield trials inboth the United States and the United Kingdom It has beenestimated that it would take two years to produce the stockrequired to plant 50 ha commercially from the plant stock inthe final yield trials Breeding programmes have alreadydelivered rust‐resistant varieties and increases in yield to themarket The adoption of advanced breeding technologies willlikely lead to a step change in improving traits of interest
5 | POPLAR
Poplar a fast‐growing tree from the northern hemispherewith a small genome size has been adopted for commercialforestry and scientific purposes The genus Populus con-sists of about 29 species classified in six different sectionsPopulus (formerly Leuce) Tacamahaca Aigeiros AbasoTuranga and Leucoides (Eckenwalder 1996) The Populusspecies of most interest for breeding and testing in the Uni-ted States and Europe are P nigra P deltoides P maxi-mowiczii and P trichocarpa (Stanton 2014) Populusclones for biomass production are being developed byintra‐ and interspecies hybridization (DeWoody Trewin ampTaylor 2015 Richardson Isebrands amp Ball 2014 van derSchoot et al 2000) Recurrent selection approaches areused for gradual population improvement and to create eliteclonal lines for commercialization (Berguson McMahon ampRiemenschneider 2017 Neale amp Kremer 2011) Currentlypoplar breeding in the United States occurs in industrialand academic programmes located in the Southeast theMidwest and the Pacific Northwest These use six speciesand five interspecific taxa (Stanton 2014)
The southeastern programme historically focused onrecurrent selection of P deltoides from accessions made inthe lower Mississippi River alluvial plain (Robison Rous-seau amp Zhang 2006) More recently the genetic base hasbeen broadened to produce interspecific hybrids with resis-tance to the fungal infection Septoria musiva which causescankers
In the midwest of the United States populationimprovement efforts are focused on P deltoides selectionsfrom native provenances and hybrid crosses with acces-sions introduced from Europe Interspecific intercontinental(Europe and America) hybrid crosses between P nigra andP deltoides (P times canadensis) are behind many of theleading commercial hybrids which are the most advancedbreeding materials for many applications and regions InMinnesota previous breeding experience and efforts utiliz-ing P maximowiczii and P trichocarpa have been discon-tinued due to Septoria susceptibility and a lack of coldhardiness (Berguson et al 2017) Traits targeted forimprovement include yieldgrowth rate cold hardinessadventitious rooting resistance to Septoria and Melamp-sora leaf rust and stem form The Upper Midwest pro-gramme also carries out wide hybridizations within thesection Populus The P times wettsteinii (P trem-ula times P tremuloides) taxon is bred for gains in growthrate wood quality and resistance to the fungus Entoleucamammata which causes hypoxylon canker (David ampAnderson 2002)
In the Pacific Northwest GreenWood Resources Incleads poplar breeding that emphasizes interspecifichybrid improvement of P times generosa (P deltoides timesP trichocarpa and reciprocal) and P deltoides times P maxi-mowiczii taxa for coastal regions and the P times canadensistaxon for the drier continental regions Intraspecificimprovement of second‐generation breeding populations ofP deltoides P nigra P maximowiczii and P trichocarpaare also involved (Stanton et al 2010) The present focusof GreenWood Resourcesrsquo hybridization is bioenergy feed-stock improvement concentrating on coppice yield wood‐specific gravity and rate of sugar release
Industrial interest in poplar in the United States has his-torically come from the pulp and paper sector althoughveneer and dimensional lumber markets have been pursuedat times Currently the biomass market for liquid trans-portation fuels is being emphasized along with the use oftraditional and improved poplar genotypes for ecosystemservices such as phytoremediation (Tuskan amp Walsh 2001Zalesny et al 2016)
In Europe there are breeding programmes in FranceGermany Italy and Sweden These include the following(a) Alasia Franco Vivai (AFV) programme in northernItaly (b) the French programme led by the poplar Scien-tific Interest Group (GIS Peuplier) and carried out
20 | CLIFTON‐BROWN ET AL
collaboratively between the National Institute for Agricul-tural Research (INRA) the National Research Unit ofScience and Technology for Environment and Agriculture(IRSTEA) and the Forest Cellulose Wood Constructionand Furniture Technology Institute (FCBA) (c) the Ger-man programme at Northwest German Forest Research Sta-tion (NW‐FVA) at Hannoversch Muumlnden and (d) theSwedish programme at the Swedish University of Agricul-tural Sciences and SweTree Technologies AB (Table 1)
AFV leads an Italian poplar breeding programme usingextensive field‐grown germplasm collections of P albaP deltoides P nigra and P trichocarpa While interspeci-fic hybridization uses several taxa the focus is onP times canadensis The breeding programme addresses dis-ease resistance (Marssonina brunnea Melampsora larici‐populina Discosporium populeum and poplar mosaicvirus) growth rate and photoperiod adaptation AFV andGreenWood Resources collaborate in poplar improvementin Europe through the exchange of frozen pollen and seedfor reciprocal breeding projects Plantations in Poland andRomania are currently the focus of the collaboration
The ongoing French GIS Peuplier is developing a long‐term breeding programme based on intraspecific recurrentselection for the four parental species (P deltoides P tri-chocarpa P nigra and P maximowiczii) designed to betterbenefit from hybrid vigour demonstrated by the interspeci-fic crosses P canadensis P deltoides times P trichocarpaand P trichocarpa times P maximowiczii Current selectionpriorities are targeting adaptation to soil and climate condi-tions resistance and tolerance to the most economicallyimportant diseases and pests high volume production underSRC and traditional poplar cultivation regimes as well aswood quality of interest by different markets Currentlygenomic selection is under exploration to increase selectionaccuracy and selection intensity while maintaining geneticdiversity over generations
The German NW‐FVA programme is breeding intersec-tional AigeirosndashTacamahaca hybrids with a focus on resis-tance to Pollaccia elegans Xanthomonas populiDothichiza spp Marssonina brunnea and Melampsoraspp (Stanton 2014) Various cross combinations ofP maximowiczii P trichocarpa P nigra and P deltoideshave led to new cultivars suitable for deployment in vari-etal mixtures of five to ten genotypes of complementarystature high productivity and phenotypic stability (Weis-gerber 1993) The current priority is the selection of culti-vars for high‐yield short‐rotation biomass production Sixhundred P nigra genotypes are maintained in an ex situconservation programme An in situ P nigra conservationeffort involves an inventory of native stands which havebeen molecular fingerprinted for identity and diversity
The Swedish programme is concentrating on locallyadapted genotypes used for short‐rotation forestry (SRF)
because these meet the needs of the current pulping mar-kets Several field trials have shown that commercial poplarclones tested and deployed in Southern and Central Europeare not well adapted to photoperiods and low temperaturesin Sweden and in the Baltics Consequently SwedishUniversity of Agricultural Sciences and SweTree Technolo-gies AB started breeding in Sweden in 1990s to producepoplar clones better adapted to local climates and markets
51 | Molecular breeding technologies
Poplar genetic improvement cannot be rapidly achievedthrough traditional methods alone because of the longbreeding cycles outcrossing breeding systems and highheterozygosity Integrating modern genetic genomic andphenomics techniques with conventional breeding has thepotential to expedite poplar improvement
The genome of poplar has been sequenced (Tuskanet al 2006) It has an estimated genome size of485 plusmn 10 Mbp divided into 19 chromosomes This is smal-ler than other PBCs and makes poplar more amenable togenetic engineering (transgenesis) GS and genome editingPoplar has seen major investment in both the United Statesand Europe being the model system for woody perennialplant genetics and genomics research
52 | Targets for genetic modification
Traits targeted include wood properties (lignin content andcomposition) earlylate flowering male sterility to addressbiosafety regulation issues enhanced yield traits and herbi-cide tolerance These extensive transgenic experiments haveshown differences in recalcitrance to in vitro regenerationand genetic transformation in some of the most importantcommercial hybrid poplars (Alburquerque et al 2016)Further transgene expression stability is being studied Sofar China is the only country known to have commerciallyused transgenic insect‐resistant poplar A precommercialherbicide‐tolerant poplar was trialled for 8 years in the Uni-ted States (Li Meilan Ma Barish amp Strauss 2008) butcould not be released due to stringent environmental riskassessments required for regulatory approval This increasestranslation costs and delays reducing investor confidencefor commercial deployment (Harfouche et al 2011)
The first field trials of transgenic poplar were performedin France in 1987 (Fillatti Sellmer Mccown Haissig ampComai 1987) and in Belgium in 1988 (Deblock 1990)Although there have been a total of 28 research‐scale GMpoplar field trials approved in the European Union underCouncil Directive 90220EEC since October 1991 (inPoland Belgium Finland France Germany Spain Swe-den and in the United Kingdom (Pilate et al 2016) onlyauthorizations in Poland and Belgium are in place today In
CLIFTON‐BROWN ET AL | 21
the United States regulatory notifications and permits fornearly 20000 transgenic poplar trees derived from approxi-mately 600 different constructs have been issued since1995 by the USDAs Animal and Plant Health InspectionService (APHIS) (Strauss et al 2016)
53 | Genome editing CRISPR technologies
Clustered regularly interspaced palindromic repeats(CRISPR) and the CRISPR‐associated (CRISPR‐Cas)nucleases are a groundbreaking genome‐engineering toolthat complements classical plant breeding and transgenicmethods (Moreno‐Mateos et al 2017) Only two publishedstudies in poplar have applied the CRISPRCas9 technol-ogy One is in P tomentosa in which an endogenous phy-toene desaturase gene (PtoPDS) was successfully disruptedsite specifically in the first generation of transgenic plantsresulting in an albino and dwarf phenotype (Fan et al2015) The second was in P tremula times alba in whichhigh CRISPR‐Cas9 mutational efficiency was achieved forthree 4‐coumarateCoA ligase (4Cl) genes 4CL1 4CL2and 4CL5 associated with lignin and flavonoid biosynthe-sis (Zhou et al 2015) Due to its low cost precision andrapidness it is very probable that cultivars or clones pro-duced using CRISPR technology will be ready for market-ing in the near future (Yin et al 2017) Recently aCRISPR with a smaller associated endonuclease has beendiscovered from Prevotella and Francisella 1 (Cpf1) whichmay have advantages over Cas9 In addition there arereports of DNA‐free editing in plants using both CRISPRCpf1 and CRISPR Cas9 for example Ref (Kim et al2017 Mahfouz 2017 Zaidi et al 2017)
It remains unresolved whether plants modified by genomeediting will be regulated as genetically modified organisms(GMOs) by the relevant authorities in different countries(Lozano‐Juste amp Cutler 2014) Regulations to cover thesenew breeding techniques are still evolving but those coun-tries who have published specific guidance (including UnitedStates Argentina and Chile) are indicating that plants pos-sessing simple genome edits will not be regulated as conven-tional transgenesis (Jones 2015b) The first generation ofgenome‐edited crops will likely be phenocopy gene knock-outs that already exist to produce ldquonature identicalrdquo traits thatis traits that could also be derived by conventional breedingDespite this confidence in applying these new powerfulbreeding tools remains limited owing to the uncertain regula-tory environment in many parts of the world (Gao 2018)including the recent ECJ 2018 rulings mentioned earlier
54 | Genomics‐based breeding technologies
Poplar breeding programs are becoming well equipped withuseful genomics tools and resources that are critical to
explore genomewide variability and make use of the varia-tion for enhancing genetic gains Deep transcriptomesequencing resequencing of alternate genomes and GBStechnology for genomewide marker detection using next‐generation sequencing (NGS) are yielding valuable geno-mics tools GWAS with NGS‐based markers facilitatesmarker identification for MAS breeding by design and GS
GWAS approaches have provided a deeper understand-ing of genome function as well as allelic architectures ofcomplex traits (Huang et al 2010) and have been widelyimplemented in poplar for wood characteristics (Porthet al 2013) stomatal patterning carbon gain versus dis-ease resistance (McKown et al 2014) height and phenol-ogy (Evans et al 2014) cell wall chemistry (Mucheroet al 2015) growth and cell walls traits (Fahrenkroget al 2017) bark roughness (Bdeir et al 2017) andheight and diameter growth (Liu et al 2018) Using high‐throughput sequencing and genotyping platforms an enor-mous amount of SNP markers have been used to character-ize the linkage disequilibrium (LD) in poplar (eg Slavovet al 2012 discussed below)
The genetic architecture of photoperiodic traits in peren-nial trees is complex involving many loci However itshows high levels of conservation during evolution (Mau-rya amp Bhalerao 2017) These genomics tools can thereforebe used to address adaptation issues and fine‐tune themovement of elite lines into new environments For exam-ple poor timing of spring bud burst and autumn bud setcan result in frost damage resulting in yield losses (Ilstedt1996) These have been studied in P tremula genotypesalong a latitudinal cline in Sweden (~56ndash66oN) and haverevealed high nucleotide polymorphism in two nonsynony-mous SNPs within and around the phytochrome B2 locus(Ingvarsson Garcia Hall Luquez amp Jansson 2006 Ing-varsson Garcia Luquez Hall amp Jansson 2008) Rese-quencing 94 of these P tremula genotypes for GWASshowed that noncoding variation of a single genomicregion containing the PtFT2 gene described 65 ofobserved genetic variation in bud set along the latitudinalcline (Tan 2018)
Resequencing genomes is currently the most rapid andeffective method detecting genetic differences betweenvariants and for linking loci to complex and importantagronomical and biomass traits thus addressing breedingchallenges associated with long‐lived plants like poplars
To date whole genome resequencing initiatives havebeen launched for several poplar species and genotypes InPopulus LD studies based on genome resequencing sug-gested the feasibility of GWAS in undomesticated popula-tions (Slavov et al 2012) This plant population is beingused to inform breeding for bioenergy development Forexample the detection of reliable phenotypegenotype asso-ciations and molecular signatures of selection requires a
22 | CLIFTON‐BROWN ET AL
detailed knowledge about genomewide patterns of allelefrequency variation LD and recombination suggesting thatGWAS and GS in undomesticated populations may bemore feasible in Populus than previously assumed Slavovet al (2012) have resequenced 16 genomes of P tri-chocarpa and genotyped 120 trees from 10 subpopulationsusing 29213 SNPs (Geraldes et al 2013) The largest everSNP data set of genetic variations in poplar has recentlybeen released providing useful information for breedinghttpswwwbioenergycenterorgbescgwasindexcfmAlso deep sequencing of transcriptomes using RNA‐Seqhas been used for identification of functional genes andmolecular markers that is polymorphism markers andSSRs A multitissue and multiple experimental data set forP trichocarpa RNA‐Seq is publicly available (httpsjgidoegovdoe-jgi-plant-flagship-gene-atlas)
The availability of genomic information of DNA‐con-taining cell organelles (nucleus chloroplast and mitochon-dria) will also allow a holistic approach in poplarmolecular breeding in the near future (Kersten et al2016) Complete Populus genome sequences are availablefor nucleus (P trichocarpa section Tacamahaca) andchloroplasts (seven species and two clones from P trem-ula W52 and P tremula times P alba 717ndash1B4) A compara-tive approach revealed structural and functionalinformation broadening the knowledge base of PopuluscpDNA and stimulating future diagnostic marker develop-ment The availability of whole genome sequences of thesecellular compartments of P tremula holds promise forboosting marker‐assisted poplar breeding Other nucleargenome sequences from additional Populus species arenow available (eg P deltoides (httpsphytozomejgidoegovpz) and will become available in the forthcomingyears (eg P tremula and P tremuloidesmdashPopGenIE(Sjodin Street Sandberg Gustafsson amp Jansson 2009))Recently the characterization of the poplar pan‐genome bygenomewide identification of structural variation in threecrossable poplar species P nigra P deltoides and P tri-chocarpa revealed a deeper understanding of the role ofinter‐ and intraspecific structural variants in poplar pheno-type and may have important implications for breedingparticularly interspecific hybrids (Pinosio et al 2016)
GS has been proposed as an alternative to MAS in cropimprovement (Bernardo amp Yu 2007 Heffner Sorrells ampJannink 2009) GS is particularly well suited for specieswith long generation times for characteristics that displaymoderate‐to‐low heritability for traits that are expensive tomeasure and for selection of traits expressed late in the lifecycle as is the case for most traits of commercial value inforestry (Harfouche et al 2012) Current joint genomesequencing efforts to implement GS in poplar using geno-mic‐estimated breeding values for bioenergy conversiontraits from 49 P trichocarpa families and 20 full‐sib
progeny are taking place at the Oak Ridge National Labo-ratory and GreenWood Resources (Brian Stanton personalcommunication httpscbiornlgov) These data togetherwith the resequenced GWAS population data will be thebasis for developing GS algorithms Genomic breedingtools have been developed for the intraspecific programmetargeting yield resistance to Venturia shoot blightMelampsora leaf rust resistance to Cryptorhynchus lapathistem form wood‐specific gravity and wind firmness (Evanset al 2014 Guerra et al 2016) A newly developedldquobreeding with rare defective allelesrdquo (BRDA) technologyhas been developed to exploit natural variation of P nigraand identify defective variants of genes predicted by priortransgenic research to impact lignin properties Individualtrees carrying naturally defective alleles can then be incor-porated directly into breeding programs thereby bypassingthe need for transgenics (Vanholme et al 2013) Thisnovel breeding technology offers a reverse genetics com-plement to emerging GS for targeted improvement of quan-titative traits (Tsai 2013)
55 | Phenomics‐assisted breeding technology
Phenomics involves the characterization of phenomesmdashthefull set of phenotypes of given individual plants (HouleGovindaraju amp Omholt 2010) Traditional phenotypingtools which inefficiently measure a limited set of pheno-types have become a bottleneck in plant breeding studiesHigh‐throughput plant phenotyping facilities provide accu-rate screening of thousands of plant breeding lines clones orpopulations over time (Fu 2015) are critical for acceleratinggenomics‐based breeding Automated image collection andanalysis phenomics technologies allow accurate and nonde-structive measurements of a diversity of phenotypic traits inlarge breeding populations (Gegas Gay Camargo amp Doo-nan 2014 Goggin Lorence amp Topp 2015 Ludovisi et al2017 Shakoor Lee amp Mockler 2017) One important con-sideration is the identification of relevant and quantifiabletarget traits that are early diagnostic indicators of biomassyield Good progress has been made in elucidating theseunderpinning morpho‐physiological traits that are amenableto remote sensing in Populus (Harfouche Meilan amp Altman2014 Rae Robinson Street amp Taylor 2004) Morerecently Ludovisi et al (2017) developed a novel methodol-ogy for field phenomics of drought stress in a P nigra F2partially inbred population using thermal infrared imagesrecorded from an unmanned aerial vehicle‐based platform
Energy is the current main market for poplar biomass butthe market return provided is not sufficient to support pro-duction expansion even with added demand for environmen-tal and land management ldquoecosystem servicesrdquo such as thetreatment of effluent phytoremediation riparian buffer zonesand agro‐forestry plantings Aviation fuel is a significant
CLIFTON‐BROWN ET AL | 23
target market (Crawford et al 2016) To serve this marketand to reduce current carbon costs of production (Budsberget al 2016) key improvement traits in addition to yield(eg coppice regeneration pestdisease resistance water‐and nutrient‐use efficiencies) will be trace greenhouse gas(GHG) emissions (eg isoprene volatiles) site adaptabilityand biomass conversion efficiency Efforts are underway tohave national environmental protection agenciesrsquo approvalfor poplar hybrids qualifying for renewable energy credits
6 | REFLECTIONS ON THECOMMERCIALIZATIONCHALLENGE
The research and innovation activities reviewed in thispaper aim to advance the genetic improvement of speciesthat can provide feedstocks for bioenergy applicationsshould those markets eventually develop These marketsneed to generate sufficient revenue and adequately dis-tribute it to the actors along the value chain The work onall four crops shares one thing in common long‐termefforts to integrate fundamental knowledge into breedingand crop development along a research and development(RampD) pipeline The development of miscanthus led byAberystwyth University exemplifies the concerted researcheffort that has integrated the RampD activities from eight pro-jects over 14 years with background core research fundingalong an emerging innovation chain (Figure 2) This pro-gramme has produced a first range of conventionally bredseeded interspecies hybrids which are now in upscaling tri-als (Table 3) The application of molecular approaches(Table 5) with further conventional breeding (Table 4)offers the prospect of a second range of improved seededhybrids This example shows that research‐based support ofthe development of new crops or crop types requires along‐term commitment that goes beyond that normallyavailable from project‐based funding (Figure 1) Innovationin this sector requires continuous resourcing of conven-tional breeding operations and capability to minimize timeand investment losses caused by funding discontinuities
This challenge is increased further by the well‐knownmarket failure in the breeding of many agricultural cropspecies The UK Department for Environment Food andRural Affairs (Defra) examined the role of geneticimprovement in relation to nonmarket outcomes such asenvironmental protection and concluded that public invest-ment in breeding was required if profound market failure isto be addressed (Defra 2002) With the exception ofwidely grown hybrid crops such as maize and some high‐value horticultural crops royalties arising from plant breed-ersrsquo rights or other returns to breeders fail to adequatelycompensate for the full cost for research‐based plant breed-ing The result even for major crops such as wheat is sub‐
optimal investment and suboptimal returns for society Thismarket failure is especially acute for perennial crops devel-oped for improved sustainability rather than consumerappeal (Tracy et al 2016) Figure 3 illustrates the underly-ing challenge of capturing value for the breeding effortThe ldquovalley of deathrdquo that results from the low and delayedreturns to investment applies generally to the research‐to‐product innovation pipeline (Beard Ford Koutsky amp Spi-wak 2009) and certainly to most agricultural crop speciesHowever this schematic is particularly relevant to PBCsMost of the value for society from the improved breedingof these crops comes from changes in how agricultural landis used that is it depends on the increased production ofthese crops The value for society includes many ecosys-tems benefits the effects of a return to seminatural peren-nial crop cover that protects soils the increase in soilcarbon storage the protection of vulnerable land or the cul-tivation of polluted soils and the reductions in GHG emis-sions (Lewandowski 2016) By its very nature theproduction of biomass on agricultural land marginal forfood production challenges farm‐level profitability Thecosts of planting material and one‐time nature of cropestablishment are major early‐stage costs and therefore theopportunities for conventional royalty capture by breedersthat are manifold for annual crops are limited for PBCs(Hastings et al 2017) Public investment in developingPBCs for the nonfood biobased sector needs to providemore long‐term support for this critical foundation to a sus-tainable bioeconomy
7 | CONCLUSIONS
This paper provides an overview of research‐based plantbreeding in four leading PBCs For all four PBC generasignificant progress has been made in genetic improvementthrough collaboration between research scientists and thoseoperating ongoing breeding programmes Compared withthe main food crops most PBC breeding programmes dateback only a few decades (Table 1) This breeding efforthas thus co‐evolved with molecular biology and the result-ing ‐omics technologies that can support breeding Thedevelopment of all four PBCs has depended strongly onpublic investment in research and innovation The natureand driver of the investment varied In close associationwith public research organizations or universities all theseprogrammes started with germplasm collection and charac-terization which underpin the selection of parents forexploratory wide crosses for progeny testing (Figure 1Table 4)
Public support for switchgrass in North America wasexplicitly linked to plant breeding with 12 breeding pro-grammes supported in the United States and CanadaSwitchgrass breeding efforts to date using conventional
24 | CLIFTON‐BROWN ET AL
breeding have resulted in over 36 registered cultivars inthe United States (Table 1) with the development of dedi-cated biomass‐type cultivars coming within the past fewyears While ‐omics technologies have been incorporatedinto several of these breeding programmes they have notyet led to commercial deployment in either conventional orhybrid cultivars
Willow genetic improvement was led by the researchcommunity closely linked to plant breeding programmesWillow and poplar have the longest record of public invest-ment in genetic improvement that can be traced back to1920s in the United Kingdom and United States respec-tively Like switchgrass breeding programmes for willoware connected to public research efforts The United King-dom in partnership with the programme based at CornellUniversity remains the European leader in willowimprovement with a long‐term breeding effort closelylinked to supporting biological research at Rothamsted Inwillow F1 hybrids have produced impressive yield gainsover parental germplasm by capturing hybrid vigour Over30 willow clones are commercially available in the UnitedStates and Europe and a further ~90 are under precommer-cial testing (Table 1)
Compared with willow and poplar miscanthus is a rela-tive newcomer with all the current breeding programmesstarting in the 2000s Clonal M times giganteus propagated byrhizomes is expected to be replaced by more readily scal-able seeded hybrids from intra‐ (M sinensis) and inter‐(M sacchariflorus times M sinensis) species crosses with highseed multiplication rates (of gt2000) The first group ofhybrid cultivars is expected to be market‐ready around2022
Of the four genera used as PBCs Populus is the mostadvanced in terms of achievements in biological researchas a result of its use as a model for basic research of treesMuch of this biological research is not directly connectedto plant breeding Nevertheless reflecting the fact thatpoplar is widely grown as a single‐stem tree in SRF thereare about 60 commercially available clones and an addi-tional 80 clones in commercial pipelines (Table 1) Trans-genic poplar hybrids have moved beyond proof of conceptto commercial reality in China
Many PBC programmes have initiated long‐term con-ventional recurrent selection breeding cycles for populationimprovement which is a key process in increasing yieldthrough hybrid vigour As this approach requires manyyears most programmes are experimenting with molecularbreeding methods as these have the potential to accelerateprecision breeding For all four PBCs investments in basicgenetic and genomic resources including the developmentof mapping populations for QTLs and whole genomesequences are available to support long‐term advancesMore recently association genetics with panels of diversegermplasm are being used as training populations for GSmodels (Table 5) These efforts are benefitting from pub-licly available DNA sequences and whole genome assem-blies in crop databases Key to these accelerated breedingtechnologies are developments in novel phenomics tech-nologies to bridge the genotypephenotype gap In poplarnovel remote sensing field phenotyping is now beingdeployed to assist breeders These advances are being com-bined with in vitro and in planta modern molecular breed-ing techniques such as CRISPR (Table 5) CRISPRtechnology for genome editing has been proven in poplar
FIGURE 2 A schematic developmentpathway for miscanthus in the UnitedKingdom related to the investment in RampDprojects at Aberystwyth (top coloured areasfor projects in the three categories basicresearch breeding and commercialupscaling) leading to a projected croppingarea of 350000 ha by 2030 with clonal andsuccessive ranges of improved seed‐basedhybrids Purple represents theBiotechnology and Biological SciencesResearch Council (BBSRC) and brown theDepartment for Environment Food andRural Affairs (Defra) (UK Nationalfunding) blue bars represent EU fundingand green private sector funding (Terravestaand CERES) and GIANT‐LINK andMiscanthus Upscaling Technology (MUST)are publicndashprivate‐initiatives (PPI)
CLIFTON‐BROWN ET AL | 25
This technology is also being applied in switchgrass andmiscanthus (Table 5) but the future of CRISPR in com-mercial breeding for the European market is uncertain inthe light of recent ECJ 2018 rulings
There is integration of research and plant breeding itselfin all four PBCs Therefore estimating the ongoing costs ofmaintaining these breeding programmes is difficult Invest-ment in research also seeks wider benefits associated withtechnological advances in plant science rather than cultivardevelopment per se However in all cases the conventionalbreeding cycle shown in Figure 1 is the basic ldquoenginerdquo withmolecular technologies (‐omics) serving to accelerate thisengine The history of the development shows that the exis-tence of these breeding programmes is essential to gain ben-efits from the biological research Despite this it is thisessential step that is at most risk from reductions in invest-ment A conventional breeding programme typicallyrequires a breeder and several technicians who are supported
over the long term (20ndash30 years Figure 1) at costs of about05 to 10 million Euro per year (as of 2018) The analysisreported here shows that the time needed to perform onecycle of conventional breeding bringing germplasm fromthe wild to a commercial hybrid ranged from 11 years inswitchgrass to 26 years in poplar (Figure 1) In a maturecrop grown on over 100000 ha with effective cultivar pro-tection and a suitable business model this level of revenuecould come from royalties Until such levels are reachedPBCs lie in the innovation valley of death (Figure 3) andneed public support
Applying industrial ldquotechnology readiness levelsrdquo (TRL)originally developed for aerospace (Heacuteder 2017) to ourplant breeding efforts we estimate many promising hybridscultivars are at TRL levels of 3ndash4 In Table 3 experts ineach crop estimate that it would take 3 years from now toupscale planting material from leading cultivars in plot tri-als to 100 ha
FIGURE 3 A schematic relating some of the steps in the innovation chain from relatively basic crop science research through to thedeployment in commercial cropping systems and value chains The shape of the funnel above the expanding development and deploymentrepresents the availability of investment along the development chain from relatively basic research at the top to the upscaled deployment at thebottom Plant breeding links the research effort with the development of cropping systems The constriction represents the constrained fundingfor breeding that links conventional public research investment and the potential returns from commercial development The handover pointsbetween publicly funded work to develop the germplasm resources (often known as prebreeding) the breeding and the subsequent cropdevelopment are shown on the left The constriction point is aggravated by the lack academic rewards for this essential breeding activity Theoutcome is such that this innovation system is constrained by the precarious resourcing of plant breeding The authorsrsquo assessment ofdevelopment status of the four species is shown (poplar having two one for short‐rotation coppice (SRC) poplar and one for the more traditionalshort‐rotation forestry (SRF)) The four new perennial biomass crops (PBCs) are now in the critical phase of depending of plant breedingprogress without the income stream from a large crop production base
26 | CLIFTON‐BROWN ET AL
Taking the UK example mentioned in the introductionplanting rates of ~35000 ha per year from 2022 onwardsare needed to reach over 1 m ha by 2050 Ongoing workin the UK‐funded Miscanthus Upscaling Technology(MUST) project shows that ramping annual hybrid seedproduction from the current level of sufficient seed for10 ha in 2018 to 35000 ha would take about 5 yearsassuming no setbacks If current hybrids of any of the fourPBCs in the upscaling pipeline fail on any step for exam-ple lower than expected multiplication rates or unforeseenagronomic barriers then further selections from ongoingbreeding are needed to replace earlier candidates
In conclusion the breeding foundations have been laidwell for switchgrass miscanthus willow and poplar owingto public funding over the long time periods necessaryImproved cultivars or genotypes are available that could bescaled up over a few years if real sustained market oppor-tunities emerged in response to sustained favourable poli-cies and industrial market pull The potential contributionsof growing and using these PBCs for socioeconomic andenvironmental benefits are clear but how farmers andothers in commercial value chains are rewarded for mass‐scale deployment as is necessary is not obvious at presentTherefore mass‐scale deployment of these lignocellulosecrops needs developments outside the breeding arenas todrive breeding activities more rapidly and extensively
8 | UNCITED REFERENCE
Huang et al (2019)
ACKNOWLEDGEMENTS
UK The UK‐led miscanthus research and breeding wasmainly supported by the Biotechnology and BiologicalSciences Research Council (BBSRC) Department for Envi-ronment Food and Rural Affairs (Defra) the BBSRC CSPstrategic funding grant BBCSP17301 Innovate UKBBSRC ldquoMUSTrdquo BBN0161491 CERES Inc and Ter-ravesta Ltd through the GIANT‐LINK project (LK0863)Genomic selection and genomewide association study activi-ties were supported by BBSRC grant BBK01711X1 theBBSRC strategic programme grant on Energy Grasses ampBio‐refining BBSEW10963A01 The UK‐led willow RampDwork reported here was supported by BBSRC (BBSEC00005199 BBSEC00005201 BBG0162161 BBE0068331 BBG00580X1 and BBSEC000I0410) Defra(NF0424 NF0426) and the Department of Trade and Indus-try (DTI) (BW6005990000) IT The Brain Gain Program(Rientro dei cervelli) of the Italian Ministry of EducationUniversity and Research supports Antoine Harfouche USContributions by Gerald Tuskan to this manuscript were sup-ported by the Center for Bioenergy Innovation a US
Department of Energy Bioenergy Research Center supportedby the Office of Biological and Environmental Research inthe DOE Office of Science under contract number DE‐AC05‐00OR22725 Willow breeding efforts at CornellUniversity have been supported by grants from the USDepartment of Agriculture National Institute of Food andAgriculture Contributions by the University of Illinois weresupported primarily by the DOE Office of Science Office ofBiological and Environmental Research (BER) grant nosDE‐SC0006634 DE‐SC0012379 and DE‐SC0018420 (Cen-ter for Advanced Bioenergy and Bioproducts Innovation)and the Energy Biosciences Institute EU We would like tofurther acknowledge contributions from the EU projectsldquoOPTIMISCrdquo FP7‐289159 on miscanthus and ldquoWATBIOrdquoFP7‐311929 on poplar and miscanthus as well as ldquoGRACErdquoH2020‐EU326 Biobased Industries Joint Technology Ini-tiative (BBI‐JTI) Project ID 745012 on miscanthus
CONFLICT OF INTEREST
The authors declare that progress reported in this paperwhich includes input from industrial partners is not biasedby their business interests
ORCID
John Clifton‐Brown httporcidorg0000-0001-6477-5452Antoine Harfouche httporcidorg0000-0003-3998-0694Lawrence B Smart httporcidorg0000-0002-7812-7736Danny Awty‐Carroll httporcidorg0000-0001-5855-0775VasileBotnari httpsorcidorg0000-0002-0470-0384Lindsay V Clark httporcidorg0000-0002-3881-9252SalvatoreCosentino httporcidorg0000-0001-7076-8777Lin SHuang httporcidorg0000-0003-3079-326XJon P McCalmont httporcidorg0000-0002-5978-9574Paul Robson httporcidorg0000-0003-1841-3594AnatoliiSandu httporcidorg0000-0002-7900-448XDaniloScordia httporcidorg0000-0002-3822-788XRezaShafiei httporcidorg0000-0003-0891-0730Gail Taylor httporcidorg0000-0001-8470-6390IrisLewandowski httporcidorg0000-0002-0388-4521
REFERENCES
Alburquerque N Baldacci‐Cresp F Baucher M Casacuberta JM Collonnier C ElJaziri M hellip Burgos L (2016) New trans-formation technologies for trees In C Vettori F Gallardo H
CLIFTON‐BROWN ET AL | 27
Haumlggman V Kazana F Migliacci G Pilateamp M Fladung(Eds) Biosafety of forest transgenic trees (pp 31ndash66) DordrechtNetherlands Springer
Argus G W (2007) Salix (Salicaceae) distribution maps and a syn-opsis of their classification in North America north of MexicoHarvard Papers in Botany 12 335ndash368 httpsdoiorg1031001043-4534(2007)12[335SSDMAA]20CO2
Atienza S G Ramirez M C amp Martin A (2003) Mapping‐QTLscontrolling flowering date in Miscanthus sinensis Anderss CerealResearch Communications 31 265ndash271
Atienza S G Satovic Z Petersen K K Dolstra O amp Martin A(2003) Identification of QTLs influencing agronomic traits inMiscanthus sinensis Anderss I Total height flag‐leaf height andstem diameter Theoretical and Applied Genetics 107 123ndash129
Atienza S G Satovic Z Petersen K K Dolstra O amp Martin A(2003) Influencing combustion quality in Miscanthus sinensisAnderss Identification of QTLs for calcium phosphorus and sul-phur content Plant Breeding 122 141ndash145
Bdeir R Muchero W Yordanov Y Tuskan G A Busov V ampGailing O (2017) Quantitative trait locus mapping of Populusbark features and stem diameter BMC Plant Biology 17 224httpsdoiorg101186s12870-017-1166-4
Beard T R Ford G S Koutsky T M amp Spiwak L J (2009) AValley of Death in the innovation sequence An economic investi-gation Research Evaluation 18 343ndash356 httpsdoiorg103152095820209X481057
Berguson W McMahon B amp Riemenschneider D (2017) Addi-tive and non‐additive genetic variances for tree growth in severalhybrid poplar populations and implications regarding breedingstrategy Silvae Genetica 66(1) 33ndash39 httpsdoiorg101515sg-2017-0005
Berlin S Trybush S O Fogelqvist J Gyllenstrand N Halling-baumlck H R Aringhman I hellip Hanley S J (2014) Genetic diversitypopulation structure and phenotypic variation in European Salixviminalis L (Salicaceae) Tree Genetics amp Genomes 10 1595ndash1610 httpsdoiorg101007s11295-014-0782-5
Bernardo R amp Yu J (2007) Prospects for genomewide selectionfor quantitative traits in maize Crop Science 47 1082ndash1090httpsdoiorg102135cropsci2006110690
Biswal A K Atmodjo M A Li M Baxter H L Yoo C GPu Y hellip Mohnen D (2018) Sugar release and growth of bio-fuel crops are improved by downregulation of pectin biosynthesisNature Biotechnology 36 249ndash257
Bmu (2009) National biomass action plan for Germany (p 17) Bun-desministerium fuumlr Umwelt Naturschutz und ReaktorsicherheitBundesministerium fuumlr Ernaumlhrung (BMU) amp Landwirtschaft undVerbraucherschutz (BMELV) 11055 Berlin | Germany Retrievedfrom httpwwwbmeldeSharedDocsDownloadsENPublicationsBiomassActionPlanpdf__blob=publicationFile
Brummer E C (1999) Capturing heterosis in forage crop cultivardevelopment Crop Science 39 943ndash954 httpsdoiorg102135cropsci19990011183X003900040001x
Budsberg E Crawford J T Morgan H Chin W S Bura R ampGustafson R (2016) Hydrocarbon bio‐jet fuel from bioconver-sion of poplar biomass Life cycle assessment Biotechnology forBiofuels 9 170 httpsdoiorg101186s13068-016-0582-2
Burris K P Dlugosz E M Collins A G Stewart C N amp Lena-ghan S C (2016) Development of a rapid low‐cost protoplasttransfection system for switchgrass (Panicum virgatum L) Plant
Cell Reports 35 693ndash704 httpsdoiorg101007s00299-015-1913-7
Casler M (2012) Switchgrass breeding genetics and genomics InA Monti (Ed) Switchgrass (pp 29ndash54) London UK Springer
Casler M Mitchell R amp Vogel K (2012) Switchgrass In CKole C P Joshi amp D R Shonnard (Eds) Handbook of bioen-ergy crop plants Vol 2 New York NY Taylor amp Francis
Casler M D amp Ramstein G P (2018) Breeding for biomass yieldin switchgrass using surrogate measures of yield BioEnergyResearch 11 6ndash12
Casler M D Sosa S Hoffman L Mayton H Ernst C Adler PR hellip Bonos S A (2017) Biomass Yield of Switchgrass Culti-vars under High‐ versus Low‐Input Conditions Crop Science 57821ndash832 httpsdoiorg102135cropsci2016080698
Casler M D amp Vogel K P (2014) Selection for biomass yield inupland lowland and hybrid switchgrass Crop Science 54 626ndash636 httpsdoiorg102135cropsci2013040239
Casler M D Vogel K P amp Harrison M (2015) Switchgrassgermplasm resources Crop Science 55 2463ndash2478 httpsdoiorg102135cropsci2015020076
Casler M D Vogel K P Lee D K Mitchell R B Adler P RSulc R M hellip Moore K J (2018) 30 Years of progress towardincreasing biomass yield of switchgrass and big bluestem CropScience 58 1242
Clark L V Brummer J E Głowacka K Hall M C Heo K PengJ hellip Sacks E J (2014) A footprint of past climate change on thediversity and population structure of Miscanthus sinensis Annals ofBotany 114 97ndash107 httpsdoiorg101093aobmcu084
Clark L V Dzyubenko E Dzyubenko N Bagmet L SabitovA Chebukin P hellip Sacks E J (2016) Ecological characteristicsand in situ genetic associations for yield‐component traits of wildMiscanthus from eastern Russia Annals of Botany 118 941ndash955
Clark L V Jin X Petersen K K Anzoua K G Bagmet LChebukin P hellip Sacks E J(2018) Population structure of Mis-canthus sacchariflorus reveals two major polyploidization eventstetraploid‐mediated unidirectional introgression from diploid Msinensis and diversity centred around the Yellow Sea Annals ofBotany 20 1ndash18 httpsdoiorg101093aobmcy1161
Clark L V Stewart J R Nishiwaki A Toma Y Kjeldsen J BJoslashrgensen U hellip Sacks E J (2015) Genetic structure ofMiscanthus sinensis and Miscanthus sacchariflorus in Japan indi-cates a gradient of bidirectional but asymmetric introgressionJournal of Experimental Botany 66 4213ndash4225
Clifton‐Brown J Hastings A Mos M McCalmont J P Ashman CAwty‐Carroll DhellipCracroft‐EleyW (2017) Progress in upscalingMis-canthus biomass production for the European bio‐economy with seed‐based hybridsGlobal Change Biology Bioenergy 9 6ndash17
Clifton‐Brown J Schwarz K U amp Hastings A (2015) History ofthe development of Miscanthus as a bioenergy crop From smallbeginnings to potential realisation Biology and Environment‐Pro-ceedings of the Royal Irish Academy 115B 45ndash57
Clifton‐Brown J Senior H Purdy S Horsnell R Lankamp BMuumlennekhoff A-K hellip Bentley A R (2018) Investigating thepotential of novel non‐woven fabrics to increase pollination effi-ciency in plant breeding PloS One (Accepted)
Crawford J T Shan CW Budsberg E Morgan H Bura R ampGus-tafson R (2016) Hydrocarbon bio‐jet fuel from bioconversion ofpoplar biomass Techno‐economic assessment Biotechnology forBiofuels 9 141 httpsdoiorg101186s13068-016-0545-7
28 | CLIFTON‐BROWN ET AL
Dalton S (2013) Biotechnology of Miscanthus In S M Jain amp SD Gupta (Eds) Biotechnology of neglected and underutilizedcrops (pp 243ndash294) Dordrecht Netherlands Springer
Davey C L Robson P Hawkins S Farrar K Clifton‐Brown J CDonnison I S amp Slavov G T (2017) Genetic relationshipsbetween spring emergence canopy phenology and biomass yieldincrease the accuracy of genomic prediction in Miscanthus Journalof Experimental Botany 68 5093ndash5102 httpsdoiorg101093jxberx339
David X amp Anderson X (2002) Aspen and larch genetics cooper-ative annual report 13 St Paul MN Department of ForestResources University of Minnesota
Deblock M (1990) Factors influencing the tissue‐culture and theAgrobacterium‐Tumefaciens‐mediated transformation of hybridaspen and poplar clones Plant Physiology 93 1110ndash1116httpsdoiorg101104pp9331110
Defra (2002) The role of future public research investment in thegenetic improvement of UK‐grown crops (p 220) LondonRetrieved from sciencesearchdefragovukDefaultaspxMenu=MenuampModule=MoreampLocation=NoneampCompleted=220ampProjectID=10412
Dewoody J Trewin H amp Taylor G (2015) Genetic and morpho-logical differentiation in Populus nigra L Isolation by coloniza-tion or isolation by adaptation Molecular Ecology 24 2641ndash2655
Dong H Liu S Clark L V Sharma S Gifford J M Juvik JA hellip Sacks E J (2018) Genetic mapping of biomass yield inthree interconnected Miscanthus populations Global Change Biol-ogy Bioenergy 10 165ndash185
Dukes (2017) Digest of UK Energy Statistics (DUKES) (p 264) Lon-don UK Department for Business Energy amp Industrial StrategyRetrieved from wwwgovukgovernmentcollectionsdigest-of-uk-energy-statistics-dukes
Eckenwalder J (1996) Systematics and evolution of Populus In RStettler H Bradshaw J Heilman amp T Hinckley (Eds) Biologyof Populus and its implications for management and conservation(pp 7‐32) Ottawa ON National Research Council of Canada
Elshire R J Glaubitz J C Sun Q Poland J A Kawamoto KBuckler E S amp Mitchell S E (2011) A robust simple geno-typing‐by‐sequencing (GBS) approach for high diversity speciesPloS One 6 e19379
Evans H (2016) Bioenergy crops in the UK Case studies of suc-cessful whole farm integration (p 23) Loughborough UKEnergy Technologies Institute Retrieved from httpswwweticouklibrarybioenergy-crops-in-the-uk-case-studies-on-successful-whole-farm-integration-evidence-pack
Evans H (2017) Increasing UK biomass production through moreproductive use of land (p 7) Loughborough UK Energy Tech-nologies Institute Retrieved from httpswwweticouklibraryan-eti-perspective-increasing-uk-biomass-production-through-more-productive-use-of-land
Evans J Sanciangco M D Lau K H Crisovan E Barry KDaum C hellip Buell C R (2018) Extensive genetic diversity ispresent within North American switchgrass germplasm The PlantGenome 11 1ndash16 httpsdoiorg103835plantgenome2017060055
Evans L M Slavov G T Rodgers‐Melnick E Martin J RanjanP Muchero W hellip DiFazio S P (2014) Population genomicsof Populus trichocarpa identifies signatures of selection and
adaptive trait associations Nature Genetics 46 1089ndash1096httpsdoiorg101038ng3075
Fabio E S Volk T A Miller R O Serapiglia M J Gauch HG Van Rees K C J hellip Smart L B (2017) Genotypetimes envi-ronment interaction analysis of North American shrub willowyield trials confirms superior performance of triploid hybrids Glo-bal Change Biology Bioenergy 9 445ndash459 httpsdoiorg101111gcbb12344
Fahrenkrog A M Neves L G Resende M F R Vazquez A Ide los Campos G Dervinis C hellip Kirst M (2017) Genome‐wide association study reveals putative regulators of bioenergytraits in Populus deltoides New Phytologist 213 799ndash811
Fan D Liu T T Li C F Jiao B Li S Hou Y S amp Luo KM (2015) Efficient CRISPRCas9‐mediated targeted mutagenesisin Populus in the first generation Scientific Reports 5 12217
Feng X P Lourgant K Castric V Saumitou-Laprade P ZhengB S Jiang D M amp Brancourt-Hulmel M (2014) The discov-ery of natural accessions related to Miscanthus x giganteus usingchloroplast DNA Crop Science 54 1645ndash1655
Fillatti J J Sellmer J Mccown B Haissig B amp Comai L(1987) Agrobacterium-mediated transformation and regenerationof Populus Molecular amp General Genetics 206 192ndash199
Fu Y B (2015) Understanding crop genetic diversity under modernplant breeding Theoretical and Applied Genetics 128 2131ndash2142 httpsdoiorg101007s00122-015-2585-y
Fu C Mielenz J R Xiao X Ge Y Hamilton C Y RodriguezM hellip Wang Z‐Y (2011) Genetic manipulation of ligninreduces recalcitrance and improves ethanol production fromswitchgrass Proceedings of the National Academy of Sciences ofthe United States of America 108 3803ndash3808 httpsdoiorg101073pnas1100310108
Gao C (2018) The future of CRISPR technologies in agricultureNature Reviews Molecular Cell Biology 19(5) 275ndash276
Gegas V Gay A Camargo A amp Doonan J (2014) Challengesof Crop Phenomics in the Post-genomic Era In J Hancock (Ed)Phenomics (pp 142ndash171) Boca Raton FL CRC Press
Geraldes A DiFazio S P Slavov G T Ranjan P Muchero WHannemann J hellip Tuskan G A (2013) A 34K SNP genotypingarray for Populus trichocarpa Design application to the study ofnatural populations and transferability to other Populus speciesMolecular Ecology Resources 13 306ndash323
Gifford J M Chae W B Swaminathan K Moose S P amp JuvikJ A (2015) Mapping the genome of Miscanthus sinensis forQTL associated with biomass productivity Global Change Biol-ogy Bioenergy 7 797ndash810
Goggin F L Lorence A amp Topp C N (2015) Applying high‐throughput phenotyping to plant‐insect interactions Picturingmore resistant crops Current Opinion in Insect Science 9 69ndash76httpsdoiorg101016jcois201503002
Grabowski P P Evans J Daum C Deshpande S Barry K WKennedy M hellip Casler M D (2017) Genome‐wide associationswith flowering time in switchgrass using exome‐capture sequenc-ing data New Phytologist 213 154ndash169 httpsdoiorg101111nph14101
Greef J M amp Deuter M (1993) Syntaxonomy of Miscanthus xgiganteus GREEF et DEU Angewandte Botanik 67 87ndash90
Guerra F P Richards J H Fiehn O Famula R Stanton B JShuren R hellip Neale D B (2016) Analysis of the genetic varia-tion in growth ecophysiology and chemical and metabolomic
CLIFTON‐BROWN ET AL | 29
composition of wood of Populus trichocarpa provenances TreeGenetics amp Genomes 12 6 httpsdoiorg101007s11295-015-0965-8
Guo H P Shao R X Hong C T Hu H K Zheng B S ampZhang Q X (2013) Rapid in vitro propagation of bioenergy cropMiscanthus sacchariflorus Applied Mechanics and Materials 260181ndash186
Hallingbaumlck H R Fogelqvist J Powers S J Turrion‐Gomez JRossiter R Amey J hellip Roumlnnberg‐Waumlstljung A‐C (2016)Association mapping in Salix viminalis L (Salicaceae)ndashIdentifica-tion of candidate genes associated with growth and phenologyGlobal Change Biology Bioenergy 8 670ndash685
Hanley S J amp Karp A (2014) Genetic strategies for dissectingcomplex traits in biomass willows (Salix spp) Tree Physiology34 1167ndash1180 httpsdoiorg101093treephystpt089
Harfouche A Meilan R amp Altman A (2011) Tree genetic engi-neering and applications to sustainable forestry and biomass pro-duction Trends in Biotechnology 29 9ndash17 httpsdoiorg101016jtibtech201009003
Harfouche A Meilan R amp Altman A (2014) Molecular and phys-iological responses to abiotic stress in forest trees and their rele-vance to tree improvement Tree Physiology 34 1181ndash1198httpsdoiorg101093treephystpu012
Harfouche A Meilan R Kirst M Morgante M Boerjan WSabatti M amp Mugnozza G S (2012) Accelerating the domesti-cation of forest trees in a changing world Trends in PlantScience 17 64ndash72 httpsdoiorg101016jtplants201111005
Hastings A Clifton‐Brown J Wattenbach M Mitchell C PStampfl P amp Smith P (2009) Future energy potential of Mis-canthus in Europe Global Change Biology Bioenergy 1 180ndash196
Hastings A Mos M Yesufu J AMccalmont J Schwarz KShafei RhellipClifton-Brown J (2017) Economic and environmen-tal assessment of seed and rhizome propagated Miscanthus in theUK Frontiers in Plant Science 8 16 httpsdoiorg103389fpls201701058
Heacuteder M (2017) From NASA to EU The evolution of the TRLscale in Public Sector Innovation The Innovation Journal 22 1ndash23 httpsdoiorg103389fpls201701058
Heffner E L Sorrells M E amp Jannink J L (2009) Genomicselection for crop improvement Crop Science 49 1ndash12 httpsdoiorg102135cropsci2008080512
Hodkinson T R Klaas M Jones M B Prickett R amp Barth S(2015) Miscanthus A case study for the utilization of naturalgenetic variation Plant Genetic Resources‐Characterization andUtilization 13 219ndash237 httpsdoiorg101017S147926211400094X
Hodkinson T R amp Renvoize S (2001) Nomenclature of Miscant-hus xgiganteus (Poaceae) Kew Bulletin 56 759ndash760 httpsdoiorg1023074117709
Houle D Govindaraju D R amp Omholt S (2010) Phenomics Thenext challenge Nature Reviews Genetics 11 855ndash866 httpsdoiorg101038nrg2897
Huang X H Wei X H Sang T Zhao Q Feng Q Zhao Yhellip Han B (2010) Genome‐wide association studies of 14 agro-nomic traits in rice landraces Nature Genetics 42 961ndashU976
Hwang O‐J Cho M‐A Han Y‐J Kim Y‐M Lim S‐H KimD‐S hellip Kim J‐I (2014) Agrobacterium‐mediated genetic trans-formation of Miscanthus sinensis Plant Cell Tissue and OrganCulture (PCTOC) 117 51ndash63
Hwang O‐J Lim S‐H Han Y‐J Shin A‐Y Kim D‐S ampKim J‐I (2014) Phenotypic characterization of transgenic Mis-canthus sinensis plants overexpressing Arabidopsis phytochromeB International Journal of Photoenergy 2014501016
Ilstedt B (1996) Genetics and performance of Belgian poplar clonestested in Sweden International Journal of Forest Genetics 3183ndash195
Ingvarsson P K Garcia M V Hall D Luquez V amp Jansson S(2006) Clinal variation in phyB2 a candidate gene for day‐length‐induced growth cessation and bud set across a latitudinalgradient in European aspen (Populus tremula) Genetics 1721845ndash1853
Ingvarsson P K Garcia M V Luquez V Hall D amp Jansson S(2008) Nucleotide polymorphism and phenotypic associationswithin and around the phytochrome B2 locus in European aspen(Populus tremula Salicaceae) Genetics 178 2217ndash2226 httpsdoiorg101534genetics107082354
Jannink J‐L Lorenz A J amp Iwata H (2010) Genomic selectionin plant breeding From theory to practice Briefings in FunctionalGenomics 9 166ndash177 httpsdoiorg101093bfgpelq001
Jiang J Guan Y Mccormick S Juvik J Lubberstedt T amp FeiS‐Z (2017) Gametophytic self‐incompatibility is operative inMiscanthus sinensis (Poaceae) and is affected by pistil age CropScience 57 1948ndash1956
Jones H D (2015a) Future of breeding by genome editing is in thehands of regulators GM Crops amp Food 6 223ndash232
Jones H D (2015b) Regulatory uncertainty over genome editingNature Plants 1 14011
Kalinina O Nunn C Sanderson R Hastings A F S van der Weijde TOumlzguumlven MhellipClifton‐Brown J C (2017) ExtendingMiscanthus cul-tivation with novel germplasm at six contrasting sites Frontiers in PlantScience 8563 httpsdoiorg103389fpls201700563
Karp A Hanley S J Trybush S O Macalpine W Pei M ampShield I (2011) Genetic improvement of willow for bioenergyand biofuels free access Journal of Integrative Plant Biology 53151ndash165 httpsdoiorg101111j1744-7909201001015x
Kersten B Faivre Rampant P Mader M Le Paslier M‐C Bou-non R Berard A hellip Fladung M (2016) Genome sequences ofPopulus tremula chloroplast and mitochondrion Implications forholistic poplar breeding PloS One 11 e0147209 httpsdoiorg101371journalpone0147209
Kiesel A Nunn C Iqbal Y Van der Weijde T Wagner MOumlzguumlven M hellip Lewandowski I (2017) Site‐specific manage-ment of Miscanthus genotypes for combustion and anaerobicdigestion A comparison of energy yields Frontiers in PlantScience 8 347 httpsdoiorg103389fpls201700347
Kim H Kim S T Ryu J Kang B C Kim J S amp Kim S G(2017) CRISPRCpf1‐mediated DNA‐free plant genome editingNature Communications 8 14406 httpsdoiorg101038ncomms14406
King Z R Bray A L Lafayette P R amp Parrott W A (2014)Biolistic transformation of elite genotypes of switchgrass (Pan-icum virgatum L) Plant Cell Reports 33 313ndash322 httpsdoiorg101007s00299-013-1531-1
Kopp R F Maynard C A De Niella P R Smart L B amp Abra-hamson L P (2002) Collection and storage of pollen from Salix(Salicaceae) American Journal of Botany 89 248ndash252
Krzyżak J Pogrzeba M Rusinowski S Clifton‐Brown J McCal-mont J P Kiesel A hellip Mos M (2017) Heavy metal uptake
30 | CLIFTON‐BROWN ET AL
by novel miscanthus seed‐based hybrids cultivated in heavy metalcontaminated soil Civil and Environmental Engineering Reports26 121ndash132 httpsdoiorg101515ceer-2017-0040
Kuzovkina Y A amp Quigley M F (2005) Willows beyond wet-lands Uses of Salix L species for environmental projects WaterAir and Soil Pollution 162 183ndash204 httpsdoiorg101007s11270-005-6272-5
Lazarus W Headlee W L amp Zalesny R S (2015) Impacts of sup-plyshed‐level differences in productivity and land costs on the eco-nomics of hybrid poplar production in Minnesota USA BioEnergyResearch 8 231ndash248 httpsdoiorg101007s12155-014-9520-y
Lewandowski I (2015) Securing a sustainable biomass supply in agrowing bioeconomy Global Food Security 6 34ndash42 httpsdoiorg101016jgfs201510001
Lewandowski I (2016) The role of perennial biomass crops in agrowing bioeconomy In S Barth D Murphy‐Bokern O Kalin-ina G Taylor amp M Jones (Eds) Perennial biomass crops for aresource‐constrained world (pp 3ndash13) Cham SwitzerlandSpringer Switzerland AG
Li J L Meilan R Ma C Barish M amp Strauss S H (2008)Stability of herbicide resistance over 8 years of coppice in field‐grown genetically engineered poplars Western Journal of AppliedForestry 23 89ndash93
Li R Y amp Qu R D (2011) High throughput Agrobacterium‐medi-ated switchgrass transformation Biomass amp Bioenergy 35 1046ndash1054 httpsdoiorg101016jbiombioe201011025
Lindegaard K N amp Barker J H A (1997) Breeding willows forbiomass Aspects of Applied Biology 49 155ndash162
Linde‐Laursen I B (1993) Cytogenetic analysis of MiscanthusGiganteus an interspecific hybrid Hereditas 119 297ndash300httpsdoiorg101111j1601-5223199300297x
Liu Y Merrick P Zhang Z Z Ji C H Yang B amp Fei S Z(2018) Targeted mutagenesis in tetraploid switchgrass (Panicumvirgatum L) using CRISPRCas9 Plant Biotechnology Journal16 381ndash393
Liu L L Wu Y Q Wang Y W amp Samuels T (2012) A high‐density simple sequence repeat‐based genetic linkage map ofswitchgrass G3 (Bethesda Md) 2 357ndash370
Lovett A Suumlnnenberg G amp Dockerty T (2014) The availabilityof land for perennial energy crops in Great Britain GlobalChange Biology Bioenergy 6 99ndash107 httpsdoiorg101111gcbb12147
Lozano‐Juste J amp Cutler S R (2014) Plant genome engineering infull bloom Trends in Plant Science 19 284ndash287 httpsdoiorg101016jtplants201402014
Lu F Lipka A E Glaubitz J Elshire R Cherney J H CaslerM D hellip Costich D E (2013) Switchgrass Genomic DiversityPloidy and Evolution Novel Insights from a Network‐Based SNPDiscovery Protocol Plos Genetics 9 e1003215
Ludovisi R Tauro F Salvati R Khoury S Mugnozza G S ampHarfouche A (2017) UAV‐based thermal imaging for high‐throughput field phenotyping of black poplar response to droughtFrontiers in Plant Science 81681
Macalpine W Shield I amp Karp A (2010) Seed to near market vari-ety the BEGIN willow breeding pipeline 2003ndash2010 and beyond InBioten conference proceedings Birmingham pp 21ndash23
Macalpine W J Shield I F Trybush S O Hayes C M ampKarp A (2008) Overcoming barriers to crossing in willow (Salixspp) breeding Aspects of Applied Biology 90 173ndash180
Mahfouz M M (2017) The efficient tool CRISPR‐Cpf1 NaturePlants 3 17028
Malinowska M Donnison I S amp Robson P R (2017) Phenomicsanalysis of drought responses in Miscanthus collected from differ-ent geographical locations Global Change Biology Bioenergy 978ndash91
Maroder H L Prego I A Facciuto G R amp Maldonado S B(2000) Storage behaviour of Salix alba and Salix matsudanaseeds Annals of Botany 86 1017ndash1021 httpsdoiorg101006anbo20001265
Martinez‐Reyna J amp Vogel K (2002) Incompatibility systems inswitchgrass Crop Science 42 1800ndash1805 httpsdoiorg102135cropsci20021800
Maurya J P amp Bhalerao R P (2017) Photoperiod‐and tempera-ture‐mediated control of growth cessation and dormancy in treesA molecular perspective Annals of Botany 120 351ndash360httpsdoiorg101093aobmcx061
McCalmont J P Hastings A Mcnamara N P Richter G MRobson P Donnison I S amp Clifton‐Brown J (2017) Environ-mental costs and benefits of growing Miscanthus for bioenergy inthe UK Global Change Biology Bioenergy 9 489ndash507
McCracken A R amp Dawson W M (1997) Using mixtures of wil-low clones as a means of controlling rust disease Aspects ofApplied Biology 49 97ndash103
McKown A D Klaacutepště J Guy R D Geraldes A Porth I Hanne-mann JhellipDouglas C J (2014) Genome‐wide association implicatesnumerous genes underlying ecological trait variation in natural popula-tions ofPopulus trichocarpaNew Phytologist 203 535ndash553
Merrick P amp Fei S Z (2015) Plant regeneration and genetic transforma-tion in switchgrass ndash A review Journal of Integrative Agriculture 14483ndash493 httpsdoiorg101016S2095-3119(14)60921-7
Moreno‐Mateos M A Fernandez J P Rouet R Lane M AVejnar C E Mis E hellip Giraldez A J (2017) CRISPR‐Cpf1mediates efficient homology‐directed repair and temperature‐con-trolled genome editing Nature Communications 8 2024 httpsdoiorg101038s41467-017-01836-2
Mosseler A (1990) Hybrid performance and species crossabilityrelationships in willows (Salix) Canadian Journal of Botany 682329ndash2338
Muchero W Guo J DiFazio S P Chen J‐G Ranjan P Slavov GT hellip Tuskan G A (2015) High‐resolution genetic mapping ofallelic variants associated with cell wall chemistry in Populus BMCGenomics 16 24 httpsdoiorg101186s12864-015-1215-z
Neale D B amp Kremer A (2011) Forest tree genomics Growingresources and applications Nature Reviews Genetics 12 111ndash122 httpsdoiorg101038nrg2931
Nunn C Hastings A F S J Kalinina O Oumlzguumlven M SchuumlleH Tarakanov I G hellip Clifton‐Brown J C (2017) Environmen-tal influences on the growing season duration and ripening ofdiverse Miscanthus germplasm grown in six countries Frontiersin Plant Science 8 907 httpsdoiorg103389fpls201700907
Ogawa Y Honda M Kondo Y amp Hara‐Nishimura I (2016) Anefficient Agrobacterium‐mediated transformation method forswitchgrass genotypes using Type I callus Plant Biotechnology33 19ndash26
Ogawa Y Shirakawa M Koumoto Y Honda M Asami Y KondoY ampHara‐Nishimura I (2014) A simple and reliable multi‐gene trans-formation method for switchgrass Plant Cell Reports 33 1161ndash1172httpsdoiorg101007s00299-014-1605-8
CLIFTON‐BROWN ET AL | 31
Okada M Lanzatella C Saha M C Bouton J Wu R L amp Tobias CM (2010) Complete switchgrass genetic maps reveal subgenomecollinearity preferential pairing and multilocus interactions Genetics185 745ndash760 httpsdoiorg101534genetics110113910
Palomo‐Riacuteos E Macalpine W Shield I Amey J Karaoğlu C WestJ hellip Jones H D (2015) Efficient method for rapid multiplication ofclean and healthy willow clones via in vitro propagation with broadgenotype applicability Canadian Journal of Forest Research 451662ndash1667 httpsdoiorg101139cjfr-2015-0055
Pilate G Allona I Boerjan W Deacutejardin A Fladung M Gal-lardo F hellip Halpin C (2016) Lessons from 25 years of GM treefield trials in Europe and prospects for the future In C Vettori FGallardo H Haumlggman V Kazana F Migliacci G Pilateamp MFladung (Eds) Biosafety of forest transgenic trees (pp 67ndash100)Dordrecht Netherlands Springer Switzerland AG
Pinosio S Giacomello S Faivre-Rampant P Taylor G Jorge VLe Paslier M C amp Marroni F (2016) Characterization of thepoplar pan-genome by genome-wide identification of structuralvariation Molecular Biology and Evolution 33 2706ndash2719
Porth I Klaacutepště J Skyba O Friedmann M C Hannemann JEhlting J hellip Douglas C J (2013) Network analysis reveals therelationship among wood properties gene expression levels andgenotypes of natural Populus trichocarpa accessions New Phytol-ogist 200 727ndash742
Pugesgaard S Schelde K Larsen S U Laeligrke P E amp JoslashrgensenU (2015) Comparing annual and perennial crops for bioenergyproductionndashinfluence on nitrate leaching and energy balance Glo-bal Change Biology Bioenergy 7 1136ndash1149 httpsdoiorg101111gcbb12215
Quinn L D Allen D J amp Stewart J R (2010) Invasivenesspotential of Miscanthus sinensis Implications for bioenergy pro-duction in the United States Global Change Biology Bioenergy2 310ndash320 httpsdoiorg101111j1757-1707201001062x
Rae A M Robinson K M Street N R amp Taylor G (2004)Morphological and physiological traits influencing biomass pro-ductivity in short‐rotation coppice poplar Canadian Journal ofForest Research‐Revue Canadienne De Recherche Forestiere 341488ndash1498 httpsdoiorg101139x04-033
Rambaud C Arnoult S Bluteau A Mansard M C Blassiau Camp Brancourt‐Hulmel M (2013) Shoot organogenesis in threeMiscanthus species and evaluation for genetic uniformity usingAFLP analysis Plant Cell Tissue and Organ Culture (PCTOC)113 437ndash448 httpsdoiorg101007s11240-012-0284-9
Ramstein G P Evans J Kaeppler S M Mitchell R B Vogel K PBuell C R amp Casler M D (2016) Accuracy of genomic prediction inswitchgrass (Panicum virgatum L) improved by accounting for linkagedisequilibriumG3 (Bethesda Md) 6 1049ndash1062
Rayburn A L Crawford J Rayburn C M amp Juvik J A (2009)Genome size of three Miscanthus species Plant Molecular Biol-ogy Reporter 27 184 httpsdoiorg101007s11105-008-0070-3
Richardson J Isebrands J amp Ball J (2014) Ecology and physiol-ogy of poplars and willows In J Isebrands amp J Richardson(Eds) Poplars and willows Trees for society and the environ-ment (pp 92‐123) Boston MA and Rome CABI and FAO
Robison T L Rousseau R J amp Zhang J (2006) Biomass produc-tivity improvement for eastern cottonwood Biomass amp Bioenergy30 735ndash739 httpsdoiorg101016jbiombioe200601012
Robson P R Farrar K Gay A P Jensen E F Clifton‐Brown JC amp Donnison I S (2013) Variation in canopy duration in the
perennial biofuel crop Miscanthus reveals complex associationswith yield Journal of Experimental Botany 64 2373ndash2383
Robson P Jensen E Hawkins S White S R Kenobi K Clif-ton‐Brown J hellip Farrar K (2013) Accelerating the domestica-tion of a bioenergy crop Identifying and modelling morphologicaltargets for sustainable yield increase in Miscanthus Journal ofExperimental Botany 64 4143ndash4155
Sanderson M A Adler P R Boateng A A Casler M D ampSarath G (2006) Switchgrass as a biofuels feedstock in theUSA Canadian Journal of Plant Science 86 1315ndash1325httpsdoiorg104141P06-136
Scarlat N Dallemand J F Monforti‐Ferrario F amp Nita V(2015) The role of biomass and bioenergy in a future bioecon-omy Policies and facts Environmental Development 15 3ndash34httpsdoiorg101016jenvdev201503006
Serapiglia M J Gouker F E amp Smart L B (2014) Early selec-tion of novel triploid hybrids of shrub willow with improved bio-mass yield relative to diploids BMC Plant Biology 14 74httpsdoiorg1011861471-2229-14-74
Serba D Wu L Daverdin G Bahri B A Wang X Kilian Ahellip Devos K M (2013) Linkage maps of lowland and upland tet-raploid switchgrass ecotypes BioEnergy Research 6 953ndash965httpsdoiorg101007s12155-013-9315-6
Shakoor N Lee S amp Mockler T C (2017) High throughput phe-notyping to accelerate crop breeding and monitoring of diseases inthe field Current Opinion in Plant Biology 38 184ndash192 httpsdoiorg101016jpbi201705006
Shield I Macalpine W Hanley S amp Karp A (2015) Breedingwillow for short rotation coppice energy cropping In Industrialcrops Breeding for BioEnergy and Bioproducts (pp 67ndash80) NewYork NY Springer
Sjodin A Street N R Sandberg G Gustafsson P amp Jansson S (2009)The Populus Genome Integrative Explorer (PopGenIE) A new resourcefor exploring the Populus genomeNewPhytologist 182 1013ndash1025
Slavov G amp Davey C (2017) Integrated genomic prediction inbioenergy crops In Abstracts from the International Conferenceon Developing biomass crops for future climates pp 34 24ndash27September 2017 Oriel College Oxford wwwwatbioeu
Slavov G Davey C Bosch M Robson P Donnison I ampMackay I (2018a) Genomic index selection provides a pragmaticframework for setting and refining multi‐objective breeding targetsin Miscanthus Annals of Botany (in press)
Slavov G T DiFazio S P Martin J Schackwitz W MucheroW Rodgers‐Melnick E hellip Tuskan G A (2012) Genome rese-quencing reveals multiscale geographic structure and extensivelinkage disequilibrium in the forest tree Populus trichocarpa NewPhytologist 196 713ndash725
Slavov G Davey C Robson P Donnison I amp Mackay I(2018b) Domestication of Miscanthus through genomic indexselection In Plant and Animal Genome Conference 13 January2018 San Diego CA Retrieved from httpspagconfexcompagxxvimeetingappcgiPaper30622
Slavov G T Nipper R Robson P Farrar K Allison G GBosch M hellip Jensen E (2013) Genome‐wide association studiesand prediction of 17 traits related to phenology biomass and cellwall composition in the energy grass Miscanthus sinensis NewPhytologist 201 1227ndash1239
Slavov G Robson P Jensen E Hodgson E Farrar K AllisonG hellip Donnison I (2014) Contrasting geographic patterns of
32 | CLIFTON‐BROWN ET AL
genetic variation for molecular markers vs phenotypic traits in theenergy grass Miscanthus sinensis Global Change Biology Bioen-ergy 5 562ndash571
Ślusarkiewicz‐Jarzina A Ponitka A Cerazy‐Waliszewska J Woj-ciechowicz M K Sobańska K Jeżowski S amp Pniewski T(2017) Effective and simple in vitro regeneration system of Mis-canthus sinensis Mtimes giganteus and M sacchariflorus for plant-ing and biotechnology purposes Biomass and Bioenergy 107219ndash226 httpsdoiorg101016jbiombioe201710012
Smart L amp Cameron K (2012) Shrub willow In C Kole S Joshiamp D Shonnard (Eds) Handbook of bioenergy crop plants (pp687ndash708) Boca Raton FL Taylor and Francis Group
Stanton B J (2014) The domestication and conservation of Populusand Salix genetic resources (Chapter 4) In Isebrands J ampRichardson J (Eds) Poplars and willows Trees for society andthe environment (pp 124ndash200) Rome Italy CAB InternationalFood and Agricultural Organization of the United Nations
Stanton B J Neale D B amp Li S (2010) Populus breeding Fromthe classical to the genomic approach In S Jansson R Bha-lerao amp A T Groover (Eds) Genetics and genomics of popu-lus (pp 309ndash348) New York NY Springer
Stewart J R Toma Y Fernandez F G Nishiwaki A Yamada T ampBollero G (2009) The ecology and agronomy ofMiscanthus sinensisa species important to bioenergy crop development in its native range inJapan A reviewGlobal Change Biology Bioenergy 1 126ndash153
Stott K G (1992) Willows in the service of man Proceedings of the RoyalSociety of Edinburgh Section B Biological Sciences 98 169ndash182
Strauss S H Ma C Ault K amp Klocko A L (2016) Lessonsfrom two decades of field trials with genetically modified trees inthe USA Biology and regulatory compliance In C Vettori FGallardo H Haumlggman V Kazana F Migliacci G Pilate amp MFladung (Eds) Biosafety of forest transgenic trees (pp 101ndash124)Dordrecht Netherlands Springer
Suda Y amp Argus G W (1968) Chromosome numbers of some NorthAmerican Salix Brittonia 20 191ndash197 httpsdoiorg1023072805440
Tan B (2018) Genomic selection and genome‐wide association studies todissect quantitative traits in forest trees Umearing Sweden University
Tornqvist C Taylor M Jiang Y Evans J Buell C KaepplerS amp Casler M (2018) Quantitative trait locus mapping for flow-ering time in a lowland x upland switchgrass pseudo‐F2 popula-tion The Plant Genome 11 170093
Toacuteth G Hermann T Da Silva M amp Montanarella L (2016)Heavy metals in agricultural soils of the European Union withimplications for food safety Environment International 88 299ndash309 httpsdoiorg101016jenvint201512017
Tracy W Dawson J Moore V amp Fisch J (2016) Intellectual propertyrights and public plant breeding Recommendations and proceedings ofa conference on best practices for intellectual property protection of pub-lically developed plant germplasm (p 70) University of Wisconsin‐Madison Retrieved from httpshostcalswisceduagronomywp-contentuploadssites1620162005Proceedings-IPR-Finalpdf
Trybush S Jahodovaacute Š Macalpine W amp Karp A (2008) A geneticstudy of a Salix germplasm resource reveals new insights into rela-tionships among subgenera sections and species BioEnergyResearch 1 67ndash79 httpsdoiorg101007s12155-008-9007-9
Tsai C J (2013) Next‐generation sequencing for next‐generationbreeding and more New Phytologist 198 635ndash637 httpsdoiorg101111nph12245
Tuskan G A Difazio S Jansson S Bohlmann J Grigoriev I HellstenU hellip Rokhsar D (2006) The genome of black cottonwood Populustrichocarpa (Torr ampGray) Science 313 1596ndash1604
Tuskan G A amp Walsh M E (2001) Short‐rotation woody cropsystems atmospheric carbon dioxide and carbon management AUS case study The Forestry Chronicle 77 259ndash264 httpsdoiorg105558tfc77259-2
Van Den Broek R Treffers D‐J Meeusen M Van Wijk ANieuwlaar E amp Turkenburg W (2001) Green energy or organicfood A life‐cycle assessment comparing two uses of set‐asideland Journal of Industrial Ecology 5 65ndash87 httpsdoiorg101162108819801760049477
Van Der Schoot J Pospiskova M Vosman B amp Smulders M J M(2000) Development and characterization of microsatellite markers inblack poplar (Populus nigra L) Theoretical and Applied Genetics 101317ndash322 httpsdoiorg101007s001220051485
Van Der Weijde T Dolstra O Visser R G amp Trindade L M(2017a) Stability of cell wall composition and saccharificationefficiency in Miscanthus across diverse environments Frontiers inPlant Science 7 2004
Van der Weijde T Huxley L M Hawkins S Sembiring E HFarrar K Dolstra O hellip Trindade L M (2017) Impact ofdrought stress on growth and quality of miscanthus for biofuelproduction Global Change Biology Bioenergy 9 770ndash782
Van der Weijde T Kamei C L A Severing E I Torres A FGomez L D Dolstra O hellip Trindade L M (2017) Geneticcomplexity of miscanthus cell wall composition and biomass qual-ity for biofuels BMC Genomics 18 406
Van der Weijde T Kiesel A Iqbal Y Muylle H Dolstra O Visser RG FhellipTrindade LM (2017) Evaluation ofMiscanthus sinensis bio-mass quality as feedstock for conversion into different bioenergy prod-uctsGlobal Change Biology Bioenergy 9 176ndash190
Vanholme B Cesarino I Goeminne G Kim H Marroni F VanAcker R hellip Boerjan W (2013) Breeding with rare defectivealleles (BRDA) A natural Populus nigra HCT mutant with modi-fied lignin as a case study New Phytologist 198 765ndash776
Vogel K P Mitchell R B Casler M D amp Sarath G (2014)Registration of Liberty Switchgrass Journal of Plant Registra-tions 8 242ndash247 httpsdoiorg103198jpr2013120076crc
Wang X Yamada T Kong F‐J Abe Y Hoshino Y Sato Hhellip Yamada T (2011) Establishment of an efficient in vitro cul-ture and particle bombardment‐mediated transformation systems inMiscanthus sinensis Anderss a potential bioenergy crop GlobalChange Biology Bioenergy 3 322ndash332 httpsdoiorg101111j1757-1707201101090x
Watrud L S Lee E H Fairbrother A Burdick C Reichman JR Bollman M hellip Van de Water P K (2004) Evidence forlandscape‐level pollen‐mediated gene flow from genetically modi-fied creeping bentgrass with CP4 EPSPS as a marker Proceedingsof the National Academy of Sciences of the United States of Amer-ica 101 14533ndash14538
Weisgerber H (1993) Poplar breeding for the purpose of biomassproduction in short‐rotation periods in Germany ndash Problems andfirst findings Forestry Chronicle 69 727ndash729 httpsdoiorg105558tfc69727-6
Wheeler N C Steiner K C Schlarbaum S E amp Neale D B(2015) The evolution of forest genetics and tree improvementresearch in the United States Journal of Forestry 113 500ndash510httpsdoiorg105849jof14-120
CLIFTON‐BROWN ET AL | 33
Whittaker C Macalpine W J Yates N E amp Shield I (2016)Dry matter losses and methane emissions during wood chip stor-age The impact on full life cycle greenhouse gas savings of shortrotation coppice willow for heat BioEnergy Research 9 820ndash835 httpsdoiorg101007s12155-016-9728-0
Xi Q (2000) Investigation on the distribution and potential of giant grassesin China PhD thesis Goettingen Germany Cuvillier Verlag
Xi Q amp Jezowkski S (2004) Plant resources of Triarrhena andMiscanthus species in China and its meaning for Europe PlantBreeding and Seed Science 49 63ndash77
Xue S Kalinina O amp Lewandowski I (2015) Present and future optionsfor the improvement of Miscanthus propagation techniques Renewableand Sustainable Energy Reviews 49 1233ndash1246
Yin K Q Gao C X amp Qiu J L (2017) Progress and prospectsin plant genome editing Nature Plants 3 httpsdoiorg101038nplants2017107
YookM (2016)Genetic diversity and transcriptome analysis for salt toler-ance inMiscanthus PhD thesis Seoul National University Korea
Zaidi S S E A Mahfouz M M amp Mansoor S (2017) CRISPR‐Cpf1 A new tool for plant genome editing Trends in PlantScience 22 550ndash553 httpsdoiorg101016jtplants201705001
Zalesny R S Stanturf J A Gardiner E S Perdue J H YoungT M Coyle D R hellip Hass A (2016) Ecosystem services ofwoody crop production systems BioEnergy Research 9 465ndash491 httpsdoiorg101007s12155-016-9737-z
Zhang Q X Sun Y Hu H K Chen B Hong C T Guo H Phellip Zheng B S (2012) Micropropagation and plant regenerationfrom embryogenic callus of Miscanthus sinensis Vitro Cellular ampDevelopmental Biology‐Plant 48 50ndash57 httpsdoiorg101007s11627-011-9387-y
Zhang Y Zalapa J E Jakubowski A R Price D L AcharyaA Wei Y hellip Casler M D (2011) Post‐glacial evolution of
Panicum virgatum Centers of diversity and gene pools revealedby SSR markers and cpDNA sequences Genetica 139 933ndash948httpsdoiorg101007s10709-011-9597-6
Zhao H Wang B He J Yang J Pan L Sun D amp Peng J(2013) Genetic diversity and population structure of Miscanthussinensis germplasm in China PloS One 8 e75672
Zhou X H Jacobs T B Xue L J Harding S A amp Tsai C J(2015) Exploiting SNPs for biallelic CRISPR mutations in theoutcrossing woody perennial Populus reveals 4‐coumarate CoAligase specificity and redundancy New Phytologist 208 298ndash301
Zhou R Macaya‐Sanz D Rodgers‐Melnick E Carlson C HGouker F E Evans L M hellip DiFazio S P (2018) Characteri-zation of a large sex determination region in Salix purpurea L(Salicaceae) Molecular Genetics and Genomics 1ndash16 httpsdoiorg101007s00438-018-1473-y
Zsuffa L Mosseler A amp Raj Y (1984) Prospects for interspecifichybridization in willow for biomass production In K L Perttu(Ed) Ecology and management of forest biomass production sys-tems Report 15 (pp 261ndash281) Uppsala Sweden SwedishUniversity of Agricultural Sciences
How to cite this article Clifton‐Brown J HarfoucheA Casler MD et al Breeding progress andpreparedness for mass‐scale deployment of perenniallignocellulosic biomass crops switchgrassmiscanthus willow and poplar GCB Bioenergy201800000ndash000 httpsdoiorg101111gcbb12566
34 | CLIFTON‐BROWN ET AL
Graphical AbstractThe contents of this page will be used as part of the graphical abstract of html only
It will not be published as part of main article
Plant breeding links the research effort with commercial mass upscaling The authorsrsquo assessment of development status ofthe four species is shown (poplar having two one for short rotation coppice (SRC) poplar and one for the more traditionalshort rotation forestry (SRF)) Mass scale deployment needs developments outside the breeding arenas to drive breedingactivities more rapidly and extensively
TABLE
4(Contin
ued)
Breeding
techno
logy
step
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sWillow
Poplar
Construct
phylogenetic
treesanddissim
ilarity
indices
The
type
ofmolecular
analysis
mdashAFL
PcpDNAR
ADseq
whole
genomesequencing
2014)
JP~1
200
NL250genotypesby
RADseq
UK~1
000
byRADseq
SK~3
00
US
Illin
oisRADseqon
allwild
accessions
FR2310
SE150
US
1500
4Exploratory
crossing
and
progenytests
Discovergood
parental
combinatio
nsmdashgeneral
combining
ability
Geographicseparatio
n
speciesor
phylogenetic
trees
Costly
long‐te
rmmultilocation
trialsforprogeny
US
gt10000
CN~1
0
FR~2
00
JP~3
0
NL3600
SK~2
0
UK~4
000
(~1500Msin)
US
500ndash1000
UK gt700since2003
gt800EWBP(1996ndash
2002)
US
800
FRCloned13
times13
factorial
matingdesign
with
3species[10
Pdelto
ides8Ptrichocarpa
8
Pnigra)
US
GWR1000exploratory
crossingsof
Pfrem
ontii
Psimoniiforim
proved
droughttolerance
UMD
NRRIcrossednativ
e
Pdelto
ides
with
non‐nativ
e
Ptrichocarpaand
Pmaximow
icziiMajority
of
progenypopulatio
nssuffered
from
clim
atic
andor
pathogenic
susceptib
ility
issues
5Wideintraspecies
hybridization
Discovergood
parental
combinatio
nsmdashgeneral
combining
ability
1to
4above
inform
atics
Flow
eringsynchronizationand
low
seed
set
US
~200
CN~1
0
NL1800
SK~1
0
UK~5
00
UK276
US
400
IT500
SE120
US
50ndash100
6With
inspecies
recurrentselection
Concentratio
nof
positiv
e
traits
Identificationof
theright
heterotic
groups
insteps
1ndash5
Difficultto
introducenew
germ
plasm
with
outdilutio
n
ofbesttraits
US
gt40
populatio
ns
undergoing
recurrentselection
NL2populatio
ns
UK6populatio
ns
US
~12populatio
nsto
beestablished
by2019
UK4
US
150
FRPdelto
ides
(gt30
FSfamilies
ndash900clones)Pnigra(gt40
FS
families
ndash1600clones)and
Ptrichocarpa(15FS
Families
minus350clones)
IT80
SEca50
parent
breeding
populatio
nswith
in
Ptrichocarpa
US
Goalsis~1
00parent
breeding
populatio
nswith
inPtrichocarpa
Pdelto
ides
PnigraandPmaximow
iczii
7Interspecies
hybrid
breeding
Com
bining
complem
entary
traitsto
producegood
morphotypes
and
heterosiseffects
Allthesteps1ndash6
Inearlystage
improvem
ents
areunpredictable
Awide
base
iscostly
tomanage
None
CNChangsha
3Nanjin
g
120MsintimesMflo
r
JP~5
with
Saccharum
spontaneum
SK5
UK420
US
250
FRPcanadensis(1800)
Pdelto
ides
timesPtrichocarpa
(800)and
PtrichocarpatimesPmaximow
iczii
(1200)
(Con
tinues)
CLIFTON‐BROWN ET AL | 9
TABLE
4(Contin
ued)
Breeding
techno
logy
step
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sWillow
Poplar
IT~1
20
US
525genotypesin
eastside
hybrid
programme
205westside
hybrid
programmeand~1
200
in
SoutheastPdelto
ides
program
8Chrom
osom
e
doublin
g
Arouteto
triploid
seeded
typesdoublin
gadiploid
parent
ofknow
n
breeding
valuefrom
recurrentselection
Needs
1ndash7to
help
identify
therightparental
lines
Doubled
plantsarenotorious
forrevertingto
diploid
US
~50plantstakenfrom
tetraploid
tooctaploid
which
arenow
infieldtrials
CN~1
0Msactriploids
DKSeveralattemptsin
mid
90s
FR5genotypesof
Msin
UK2Msin1Mflora
nd1Msac
US
~30
UKAttempted
butnot
routinelyused
US
Not
partof
theregular
breeding
becausethere
existnaturalpolyploids
NA
9DoubleHaploids
Arouteto
producing
homogeneous
progeny
andalso
forthe
introductio
nof
transgenes
orforgenome
editing
Needs
1ndash7to
help
identify
therightparental
lines
Fully
homozygousplantsare
weakandeasily
die
andmay
notflow
ersynchronously
US
Noneyetattempteddueto
poor
vigour
andviability
of
haploids
CN2genotypesof
Msin
andMflor
UKDihaploidsof
MsinMflor
and
Msacand2transgenic
Msinvia
anther
cultu
re
US
6createdandused
toidentify
miss‐identifying
paralogueloci
(causedby
recent
genome
duplicationin
miscanthus)Usedin
creatin
gthereferencegenomefor
MsinVeryweakplantsand
difficultto
retain
thelin
es
NA
NA
10Embryo
rescue
Anattractiv
etechniquefor
recovering
plantsfrom
sexual
crosses
The
majority
ofem
bryos
cannot
survivein
vivo
or
becomelong‐timedorm
ant
None
UKone3times
hybrid
UKEmbryosrescue
protocol
proved
robustat
8days
post‐pollin
ation
FRAnem
bryo
rescue
protocol
proved
robustandim
proved
hybridizationsuccessfor
Pdelto
ides
timesPtrichocarpa
crosses
11Po
llenstorage
Flexibility
tocross
interestingparents
with
outneed
of
flow
ering
synchronization
None
Nosuccessto
daten
otoriously
difficultwith
grassesincluding
sugarcane
Freshpollencommonly
used
incrossing
Pollenstoragein
some
interspecificcrosses
FRBothstored
(cryobank)
and
freshindividual
pollen
Note
AFL
Pam
plifiedfragmentlength
polymorphismCBDconventio
non
biological
diversity
CP
controlledpollinatio
nCpD
NAchloroplastDNAEWBP
Europeanwillow
breeding
programme
FSfull‐sib
GtimesE
genotype‐by‐environm
entGRIN
germ
plasm
resourcesinform
ationnetwork
GWRGreenWoodResourcesMflorMiscanthusflo
ridulusMsacMsacchariflo
rus
MsinMsinensisNAnotapplicableNRRINatural
Resources
ResearchInstitu
teOP
open
pollinatio
nRADseq
restrictionsite‐associatedDNA
sequencingSL
USw
edishUniversity
ofAgriculturalSciencesST
TSw
eTreeTech
UMDUniversity
ofMinnesota
Duluth
a ISO
Alpha‐2
lettercountrycodes
10 | CLIFTON‐BROWN ET AL
programmes mature The average rate of gain for biomassyield in long‐term switchgrass breeding programmes hasbeen 1ndash4 per year depending on ecotype populationand location of the breeding programme (Casler amp Vogel2014 Casler et al 2018) The hybrid derivative Libertyhas a biomass yield 43 higher than the better of its twoparents (Casler amp Vogel 2014 Vogel et al 2014) Thedevelopment of cold‐tolerant and late‐flowering lowland‐ecotype populations for the northern United States hasincreased biomass yields by 27 (Casler et al 2018)
Currently more than 20 recurrent selection populationsare being managed in the United States to select parentsfor improved yield yield resilience and compositional qual-ity of the biomass For the agronomic development andupscaling high seed multiplication rates need to be com-bined with lower seed dormancy to reduce both crop estab-lishment costs and risks Expresso is the first cultivar withsignificantly reduced seed dormancy which is the first steptowards development of domesticated populations (Casleret al 2015) Most phenotypic traits of interest to breeders
require a minimum of 2 years to be fully expressed whichresults in a breeding cycle that is at least two years Morecomplicated breeding programmes or traits that requiremore time to evaluate can extend the breeding cycle to 4ndash8 years per generation for example progeny testing forbiomass yield or field‐based selection for cold toleranceBreeding for a range of traits with such long cycles callsfor the development of molecular methods to reduce time-scales and improve breeding efficiency
Two association panels of switchgrass have been pheno-typically and genotypically characterized to identify quanti-tative trait loci (QTLs) that control important biomasstraits The northern panel consists of 60 populationsapproximately 65 from the upland ecotype The southernpanel consists of 48 populations approximately 65 fromthe lowland ecotype Numerous QTLs have been identifiedwithin the northern panel to date (Grabowski et al 2017)Both panels are the subject of additional studies focused onbiomass quality flowering and phenology and cold toler-ance Additionally numerous linkage maps have been
FIGURE 1 Cumulative minimum years needed for the conventional breeding cycle through the steps from wild germplasm to thecommercial hybrids in switchgrass miscanthus willow and poplar Information links between the steps are indicated by dotted arrows andhighlight the importance of long‐term informatics to maximize breeding gain
CLIFTON‐BROWN ET AL | 11
created by the pairwise crossing of individuals with diver-gent characteristics often to generate four‐way crosses thatare analysed as pseudo‐F2 crosses (Liu Wu Wang ampSamuels 2012 Okada et al 2010 Serba et al 2013Tornqvist et al 2018) Individual markers and QTLs iden-tified can be used to design marker‐assisted selection(MAS) programmes to accelerate breeding and increase itsefficiency Genomic prediction and selection (GS) holdseven more promise with the potential to double or triplethe rate of gain for biomass yield and other highly complexquantitative traits of switchgrass (Casler amp Ramstein 2018Ramstein et al 2016) The genome of switchgrass hasrecently been made public through the Joint Genome Insti-tute (httpsphytozomejgidoegov)
Transgenic approaches have been heavily relied upon togenerate unique genetic variants principally for traitsrelated to biomass quality (Merrick amp Fei 2015) Switch-grass is highly transformable using either Agrobacterium‐mediated transformation or biolistics bombardment butregeneration of plants is the bottleneck to these systemsTraditionally plants from the cultivar Alamo were the onlyregenerable genotypes but recent efforts have begun toidentify more genotypes from different populations that arecapable of both transformation and subsequent regeneration(King Bray Lafayette amp Parrott 2014 Li amp Qu 2011Ogawa et al 2014 Ogawa Honda Kondo amp Hara‐Nishi-mura 2016) Cell wall recalcitrance and improved sugarrelease are the most common targets for modification (Bis-wal et al 2018 Fu et al 2011) Transgenic approacheshave the potential to provide traits that cannot be bredusing natural genetic variability However they will stillrequire about 10ndash15 years and will cost $70ndash100 millionfor cultivar development and deployment (Harfouche Mei-lan amp Altman 2011) In addition there is commercialuncertainty due to the significant costs and unpredictabletimescales and outcomes of the regulatory approval processin the countries targeted for seed sales As seen in maizeone advantage of transgenic approaches is that they caneasily be incorporated into F1 hybrid cultivars (Casler2012 Casler et al 2012) but this does not decrease thetime required for cultivar development due to field evalua-tion and seed multiplication requirements
The potential impacts of unintentional gene flow andestablishment of non‐native transgene sequences in nativeprairie species via cross‐pollination are also major issues forthe environmental risk assessment These limit further thecommercialization of varieties made using these technolo-gies Although there is active research into switchgrass steril-ity mechanisms to curb unintended pollen‐mediated genetransfer it is likely that the first transgenic cultivars proposedfor release in the United States will be met with considerableopposition due to the potential for pollen flow to remainingwild prairie sites which account for lt1 of the original
prairie land area and are highly protected by various govern-mental and nongovernmental organizations (Casler et al2015) Evidence for landscape‐level pollen‐mediated geneflow from genetically modified Agrostis seed multiplicationfields (over a mountain range) to pollinate wild relatives(Watrud et al 2004) confirms the challenge of using trans-genic approaches Looking ahead genome editing technolo-gies hold considerable promise for creating targeted changesin phenotype (Burris Dlugosz Collins Stewart amp Lena-ghan 2016 Liu et al 2018) and at least in some jurisdic-tions it is likely that cultivars resulting from gene editingwill not need the same regulatory approval as GMOs (Jones2015a) However in July 2018 the European Court of Justice(ECJ) ruled that cultivars carrying mutations resulting fromgene editing should be regulated in the same way as GMOsThe ECJ ruled that such cultivars be distinguished from thosearising from untargeted mutation breeding which isexempted from regulation under Directive 200118EC
3 | MISCANTHUS
Miscanthus is indigenous to eastern Asia and Oceania whereit is traditionally used for forage thatching and papermaking(Xi 2000 Xi amp Jezowkski 2004) In the 1960s the highbiomass potential of a Japanese genotype introduced to Eur-ope by Danish nurseryman Aksel Olsen in 1935 was firstrecognized in Denmark (Linde‐Laursen 1993) Later thisaccession was characterized described and named asldquoM times giganteusrdquo (Greef amp Deuter 1993 Hodkinson ampRenvoize 2001) commonly abbreviated as Mxg It is a natu-rally occurring interspecies triploid hybrid between tetraploidM sacchariflorus (2n = 4x) and diploid M sinensis(2n = 2x) Despite its favourable agronomic characteristicsand ability to produce high yields in a wide range of environ-ments in Europe (Kalinina et al 2017) the risks of relianceon it as a single clone have been recognized Miscanthuslike switchgrass is outcrossing due to self‐incompatibility(Jiang et al 2017) Thus seeded hybrids are an option forcommercial breeding Miscanthus can also be vegetativelypropagated by rhizome or in vitro culture which allows thedevelopment of clones The breeding approaches are usuallybased on F1 crosses and recurrent selection cycles within thesynthetic populations There are several breeding pro-grammes that target improvement of miscanthus traitsincluding stress resilience targeted regional adaptation agro-nomic ldquoscalabilityrdquo through cheaper propagation fasterestablishment lower moisture and ash contents and greaterusable yield (Clifton‐Brown et al 2017)
Germplasm collections specifically to support breedingfor biomass started in the late 1980s and early 1990s inDenmark Germany and the United Kingdom (Clifton‐Brown Schwarz amp Hastings 2015) These collectionshave continued with successive expeditions from European
12 | CLIFTON‐BROWN ET AL
and US teams assembling diverse collections from a widegeographic range in eastern Asia including from ChinaJapan South Korea Russia and Taiwan (Hodkinson KlaasJones Prickett amp Barth 2015 Stewart et al 2009) Threekey miscanthus species for biomass production are M si-nensis M floridulus and M sacchariflorus M sinensis iswidely distributed throughout eastern Asia with an adap-tive range from the subtropics to southern Russia (Zhaoet al 2013) This species has small rhizomes and producesmany tightly packed shoots forming a ldquotuftrdquo M floridulushas a more southerly adaptive range with a rather similarmorphology to M sinensis but grows taller with thickerstems and is evergreen and less cold‐tolerant than the othermiscanthus species M sacchariflorus is the most northern‐adapted species ranging to 50 degN in eastern Russia (Clarket al 2016) Populations of diploid and tetraploid M sac-chariflorus are found in China (Xi 2000) and South Korea(Yook 2016) and eastern Russia but only tetraploids havebeen found in Japan (Clark Jin amp Petersen 2018)
Germplasm has been assembled from multiple collec-tions over the last century though some early collectionsare poorly documented This historical germplasm has beenused to initiate breeding programmes largely based on phe-notypic and genotypic characterization As many of theaccessions from these collections are ldquoorigin unknownrdquocrucial environmental envelope data are not available UK‐led expeditions started in 2006 and continued until 2011with European and Asian partners and have built up a com-prehensive collection of 1500 accessions from 500 sitesacross Eastern Asia including China Japan South Koreaand Taiwan These collections were guided using spatialclimatic data to identify variation in abiotic stress toleranceAccessions from these recent collections were planted fol-lowing quarantine in multilocation nursery trials at severallocations in Europe to examine trait expression in differentenvironments Based on the resulting phenotypic andmolecular marker data several studies (a) characterized pat-terns of population genetic structure (Slavov et al 20132014) (b) evaluated the statistical power of genomewideassociation studies (GWASs) and identified preliminarymarkerndashtrait associations (Slavov et al 2013 2014) and(c) assessed the potential of genomic prediction (Daveyet al 2017 Slavov et al 2014 2018b) Genomic indexselection in particular offers the possibility of exploringscenarios for different locations or industrial markets (Sla-vov et al 2018a 2018b)
Separately US‐led expeditions also collected about1500 accessions between 2010 and 2014 (Clark et al2014 2016 2018 2015) A comprehensive genetic analy-sis of the population structure has been produced by RAD-seq for M sinensis (Clark et al 2015 Van der WeijdeKamei et al 2017) and M sacchariflorus (Clark et al2018) Multilocation replicated field trials have also been
conducted on these materials in North America and inAsia GWAS has been conducted for both M sinensis anda subset of M sacchariflorus accessions (Clark et al2016) To date about 75 of these recent US‐led collec-tions are in nursery trials outside the United States Due tolengthy US quarantine procedures these are not yet avail-able for breeding in the United States However molecularanalyses have allowed us to identify and prioritize sets ofgenotypes that best encompass the genetic variation in eachspecies
While mostM sinensis accessions flower in northern Eur-ope very few M sacchariflorus accessions flower even inheated glasshouses For this reason the European pro-grammes in the United Kingdom the Netherlands and Francehave performed mainly M sinensis (intraspecies) hybridiza-tions (Table 4) Selected progeny become the parents of latergenerations (recurrent selection as in switchgrass) Seed setsof up to 400 seed per panicle occur inM sinensis In Aberyst-wyth and Illinois significant efforts to induce synchronousflowering in M sacchariflorus and M sinensis have beenmade because interspecies hybrids have proven higher yieldperformance and wide adaptability (Kalinina et al 2017) Ininterspecies pairwise crosses in glasshouses breathable bagsandor large crossing tubes or chambers in which two or morewhole plants fit are used for pollination control Seed sets arelower in bags than in the open air because bags restrict pollenmovement while increasing temperatures and reducinghumidity (Clifton‐Brown Senior amp Purdy 2018) About30 of attempted crosses produced 10 to 60 seeds per baggedpanicle The seed (thousand seed mass ranges from 05 to09 g) is threshed from the inflorescences and sown into mod-ular trays to produce plug plants which are then planted infield nurseries to identify key parental combinations
A breeding programme of this scale must serve theneeds of different environments accepting the commonpurpose is to optimize the interception of solar radiationAn ideal hybrid for a given environment combines adapta-tion to date of emergence with optimization of traits suchas height number of stems per plant flowering and senes-cence time to optimize solar interception to produce a highbiomass yield with low moisture content at harvest (Rob-son Farrar et al 2013 Robson Jensen et al 2013) By20132014 conventional breeding in Europe had producedintra‐ and interspecific fertile seeded hybrids When acohort (typically about 5) of outstanding crosses have beenidentified it is important to work on related upscaling mat-ters in parallel These are as follows
Assessment of the yield and critical traits in selectedhybrids using a network of field trials
Efficient cloning of the seed parents While in vitro andmacro‐cloning techniques are used some genotypes areamenable to neither technique
CLIFTON‐BROWN ET AL | 13
TABLE5
Status
ofmod
ernplantbreeding
techniqu
esin
four
leadingperennialbiom
asscrop
s(PBCs)
Breeding
techno
logy
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sW
illow
Pop
lar
MAS
Use
ofmarker
sequ
encesthat
correlatewith
atrait
allowingearly
prog
enyselectionor
rejectionlocatin
gkn
owngenesfor
useful
traits(suchas
height)from
other
speciesin
your
crop
Breeding
prog
ramme
relevant
biparental
crosses(Table
4)to
create
ldquomapping
popu
latio
nsrdquoand
QTLidentification
andestim
ationto
identifyrobu
stmarkers
alternatively
acomprehensive
panelof
unrelated
geno
typesfor
GWAS
Ineffectivewhere
traits
areaffected
bymany
geneswith
small
effects
US
Literally
dozens
ofstud
iesto
identify
SNPs
andmarkers
ofinterestNoattempts
atMASas
yetlargely
dueto
popu
latio
nspecificity
CNS
SRISS
Rmarkers
forM
sinMsacand
Mflor
FR2
Msin
mapping
popu
latio
ns(BFF
project)
NL2
Msin
mapping
popu
latio
ns(A
tienza
Ram
irezetal
2003
AtienzaSatovic
PetersenD
olstraamp
Martin
200
3Van
Der
Weijdeetal20
17a
Van
derW
eijde
Hux
ley
etal20
17
Van
derW
eijdeKam
ei
etal20
17V
ander
WeijdeKieseletal
2017
)SK
2GWASpanels(1
collectionof
Msin1
diploidbiparentalF 1
popu
latio
n)in
the
pipelin
eUK1
2families
tofind
QTLin
common
indifferentfam
ilies
ofthe
sameanddifferent
speciesandhy
brids
US
2largeGWAS
panels4
diploidbi‐
parentalF 1
popu
latio
ns
1F 2
popu
latio
n(add
ition
alin
pipelin
e)
UK16
families
for
QTLdiscov
ery
2crossesMASscreened
1po
tentialvariety
selected
US
9families
geno
typedby
GBS(8
areF 1o
neisF 2)a
numberof
prom
ising
QTLdevelopm
entof
markers
inprog
ress
FR(INRA)tested
inlargeFS
families
and
infactorialmating
design
low
efficiency
dueto
family
specificity
US
49families
GS
Metho
dto
accelerate
breeding
throug
hredu
cing
the
resourcesforcross
attemptsby
Aldquotraining
popu
latio
nrdquoof
individu
alsfrom
stages
abov
ethat
have
been
both
Risks
ofpo
orpredictio
nProg
eny
testingneedsto
becontinueddu
ring
the
timethetraining
setis
US
One
prog
rammeso
farmdash
USD
Ain
WisconsinThree
cycles
ofgeno
mic
selectioncompleted
CN~1
000
geno
types
NLprelim
inary
mod
elsforMsin
developed
UK~1
000
geno
types
Verysuitable
application
not
attempted
yet
Maybe
challeng
ingfor
interspecies
hybrids
FR(INRA)120
0geno
type
ofPop
ulus
nigraarebeingused
todevelop
intraspecificGS
(Con
tinues)
14 | CLIFTON‐BROWN ET AL
TABLE
5(Contin
ued)
Breeding
techno
logy
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sW
illow
Pop
lar
predictin
gthe
performance
ofprog
enyof
crosses
(and
inresearch
topredictbestparents
tousein
biparental
crosses)
geno
typedand
phenotyp
edto
developamod
elthattakesgeno
typic
datafrom
aldquocandidate
popu
latio
nrdquoof
untested
individu
als
andprod
uces
GEBVs(Jannink
Lorenzamp
Iwata
2010
)
beingph
enotyp
edand
geno
typed
Inthis
time
next‐generation
germ
plasm
andreadyto
beginthe
second
roun
dof
training
and
recalib
ratio
nof
geno
mic
predictio
nmod
els
US
Prelim
inary
mod
elsforMsac
and
Msin
developed
Work
isun
derw
ayto
deal
moreeffectivelywith
polyploidgenetics
calib
ratio
nsUS
125
0geno
types
arebeingused
todevelopGS
calib
ratio
ns
Traditio
nal
transgenesis
Efficient
introd
uctio
nof
ldquoforeign
rdquotraits
(possiblyfrom
other
genu
sspecies)
into
anelite
plant(ega
prov
enparent
ora
hybrid)that
needsa
simpletraitto
beim
prov
edValidate
cand
idategenesfrom
QTLstud
ies
MASand
know
ledg
eof
the
biolog
yof
thetrait
andsource
ofgenesto
confer
the
relevant
changesto
phenotyp
eWorking
transformation
protocol
IPissuescostof
regu
latory
approv
al
GMO
labelling
marketin
gissuestricky
tousetransgenes
inou
t‐breedersbecause
ofcomplexity
oftransformingandgene
flow
risks
US
Manyprog
ramsin
UnitedStates
are
creatin
gtransgenic
plantsusingAlamoas
asource
oftransformable
geno
typesManytraits
ofinterestNothing
commercial
yet
CNC
hang
shaBtg
ene
transformed
into
Msac
in20
04M
sin
transformed
with
MINAC2gene
and
Msactransform
edwith
Cry2A
ageneb
othwith
amarkerg
eneby
Agrob
acterium
JP1
Msintransgenic
forlow
temperature
tolerancewith
increased
expression
offructans
(unp
ublished)
NLImprov
ingprotocol
forM
sin
transformation
SK2
Msin
with
Arabido
psisand
Brachypod
ium
PhytochromeBgenes
(Hwang
Cho
etal
2014
Hwang
Lim
etal20
14)
UK2
Msin
and
1Mflorg
enotyp
es
UKRou
tine
transformationno
tyet
possibleResearchto
overcomerecalcitrance
ongo
ing
Currently
trying
differentspecies
andcond
ition
sPo
plar
transformationused
atpresent
US
Frequently
attempted
with
very
limitedsuccess
US
600transgenic
lines
have
been
characterized
mostly
performed
inthe
aspenhy
brids
reprod
uctiv
esterility
drou
ghttolerance
with
Kno
ckdo
wns
(Con
tinues)
CLIFTON‐BROWN ET AL | 15
TABLE
5(Contin
ued)
Breeding
techno
logy
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sW
illow
Pop
lar
variou
slytransformed
with
four
cellwall
genesiptand
uidA
genesusingtwo
selectionsystem
sby
biolisticsand
Agrob
acterium
Transform
edplantsare
beinganalysed
US
Prelim
inarywork
will
betakenforw
ardin
2018
ndash202
2in
CABBI
Genom
eediting
CRISPR
Refinem
entofexisting
traits
inusefulparents
orprom
isinghybrids
bygeneratingtargeted
mutations
ingenes
know
ntocontrolthe
traitof
interestF
irst
adouble‐stranded
breakismadeinthe
DNAwhich
isrepairedby
natural
DNArepair
machineryL
eads
tofram
eshiftSN
Por
can
usealdquorepair
templaterdquo
orcanbe
used
toinserta
transgene
intoaldquosafe
harbourlocusrdquoIt
couldbe
used
todeleterepressors(or
transcriptionfactors)
Identification
mapping
and
sequ
encing
oftarget
genes(from
DNA
sequ
ence)
avoiding
screening
outof
unintend
ededits
Manyregu
latory
authorities
have
not
decidedwhether
CRISPR
andother
geno
meediting
techno
logies
are
GMOsor
notIf
GMOsseecomment
onstage10
abov
eIf
noteditedcrop
swill
beregu
latedas
conv
entio
nalvarieties
CATechn
olog
yisstill
toonewand
switchg
rass
geno
meis
very
complexOther
labo
ratories
are
interestedbu
tno
tyet
mov
ingon
this
US
One
prog
rammeso
farmdash
USD
Ain
Albany
FRInitiated
in20
16(M
ISEDIT
project)
NLInitiatingin
2018
US
Initiatingin
2018
in
CABBI
UKNot
started
Not
possible
until
transformation
achieved
US
CRISPR
‐Cas9and
Cpf1have
been
successfully
developedin
Pop
ulus
andarehigh
lyefficientExamples
forlig
nin
biosyn
thesis
Note
BFF
BiomassFo
rtheFu
tureBtBacillus
thuringiensisCABBICenterforadvanced
bioenergyandbioproductsinnovatio
nCpf1
CRISPR
from
Prevotella
andFrancisella
1CRISPR
clusteredregularlyinterspaced
palin
drom
icrepeatsCRISPR
‐CasC
RISPR
‐associated
Cry2A
acrystaltoxins2A
asubfam
ilyproduced
byBtFS
full‐sibG
BS
genotyping
bysequencingG
EBVsgenomic‐estim
ated
breeding
valuesG
MOg
enetically
modi-
fied
organism
GS
genomicselection
GWAS
genomew
ideassociationstudy
INRAF
renchNationalInstituteforAgriculturalR
esearch
IPintellectualp
roperty
IPTisopentenyltransferase
ISSR
inter‐SSR
MflorMiscant-
husflo
ridulusMsacMsacchariflo
rusMsinMsinensisM
AS
marker‐a
ssistedselection
MISEDITm
iscanthusgene
editing
forseed‐propagatedtriploidsNACn
oapicalmeristemA
TAF1
2and
cup‐shaped
cotyledon2
‐lik
eQTLq
uantitativ
etraitlocusS
NP
singlenucleotid
epolymorphismS
SRsim
plesequence
repeatsUSD
AU
SDepartm
ento
fAgriculture
a ISO
Alpha‐2
lettercountrycodes
16 | CLIFTON‐BROWN ET AL
High seed production from field crossing trials con-ducted in locations where flowering in both seed andpollen parents is likely to happen synchronously
Scalable and adapted harvesting threshing and seed pro-cessing methods for producing high seed quality
The results of these parallel activities need to becombined to identify the upscaling pathway for eachhybrid if this cannot be achieved the hybrid will likelynot be commercially viable The UK‐led programme withpartners in Italy and Germany shows that seedbased mul-tiplication rates of 12000 are achievable several inter-specific hybrids (Clifton‐Brown et al 2017) Themultiplication rate of M sinensis is higher probably15000ndash10000 Conventional cloning from rhizome islimited to around 120 that is one ha could provide rhi-zomes for around 20 ha of new plantation
Multilocation field testing of wild and novel miscanthushybrids selected by breeding programmes in the Nether-lands and the United Kingdom was performed as part ofthe project Optimizing Miscanthus Biomass Production(OPTIMISC 2012ndash2016) These trials showed that com-mercial yields and biomass qualities (Kiesel et al 2017Van der Weijde et al 2017a Van der Weijde Kieselet al 2017) could be produced in a wide range of climatesand soil conditions from the temperate maritime climate ofwestern Wales to the continental climate of eastern Russiaand the Ukraine (Kalinina et al 2017) Extensive environ-mental measurements of soil and climate combined withgrowth monitoring are being used to understand abioticstresses (Nunn et al 2017 Van der Weijde Huxley et al2017) and develop genotype‐specific scenarios similar tothose reported earlier in Hastings et al (2009) Phenomicsexperiments on drought tolerance have been conducted onwild and improved germplasm (Malinowska Donnison ampRobson 2017 Van der Weijde Huxley et al 2017)Recently produced interspecific hybrids displaying excep-tional yield under drought (~30 greater than control Mxg)in field trials in Poland and Moldova are being furtherstudied in detail in the phenomics and genomics facility atAberystwyth to better understand genendashtrait associationswhich can be fed back into breeding
Intraspecific seeded hybrids of M sinensis produced inthe Netherlands and interspecific M sacchariflorus times M si-nensis hybrids produced by the UK‐led breeding programmehave entered yield testing in 2018 with the recently EU‐funded project ldquoGRowing Advanced industrial Crops on mar-ginal lands for biorEfineries (GRACE)rdquo (httpswwwgrace-bbieu) Substantial variation in biomass quality for sacchari-fication efficiency (glucose release as of dry matter) ashcontent and melting point has already been generated inintraspecific M sinensis hybrids (Van der Weijde Kiesel
et al 2017) across environments (Weijde Dolstra et al2017a) GRACE aims to establish more than 20 hectares ofnew inter‐ and intraspecific seeded hybrids across six Euro-pean countries This project is building the know‐how andagronomy needed to transition from small research plots tocommercial‐scale field sites and linking biomass productiondirectly to industrial applications The biomass produced byhybrids in different locations will be supplied to innovativeindustrial end‐users making a wide range of biobased prod-ucts both for chemicals and for energy In the United Statesmulti‐location yield were initiated in 2018 to evaluate new tri-ploid M times giganteus genotypes developed at Illinois Cur-rently infertile hybrids are favoured in the United Statesbecause this eliminates the risk of invasiveness from naturallydispersed viable seed The precautionary principle is appliedas fertile miscanthus has naturalized in several states (QuinnAllen amp Stewart 2010) North European multilocation fieldtrials in the EMI and OPTIMISC projects have shown thereis minimal risk of invasiveness even in years when fertileflowering hybrids produce viable seed Naturalized standshave not established here due perhaps to low dormancy pooroverwintering and low seedling competitive strength In addi-tion to breeding for nonshattering or sterile seeded hybridsQuinn et al (2010) suggest management strategies which canfurther minimize environmental opportunities to manage therisk of invasiveness
31 | Molecular breeding and biotechnology
In miscanthus new plant breeding techniques (Table 5)have focussed on developing molecular markers forbreeding in Europe the United States South Korea andJapan There are several publications on QTL mappingpopulations for key traits such as flowering (AtienzaRamirez amp Martin 2003) and compositional traits(Atienza Satovic Petersen Dolstra amp Martin 2003) Inthe United States and United Kingdom independent andinterconnected bi‐parental ldquomappingrdquo families have beenstudied (Dong et al 2018 Gifford Chae SwaminathanMoose amp Juvik 2015) alongside panels of diverse germ-plasm accessions for GWAS (Slavov et al 2013) Fur-ther developments calibrating GS with very large panelsof parents and cross progeny are underway (Daveyet al 2017) The recently completed first miscanthus ref-erence genome sequence is expected to improve the effi-ciency of MAS strategies and especially GWAS(httpsphytozomejgidoegovpzportalhtmlinfoalias=Org_Msinensis_er) For example without a referencegenome sequence Clark et al (2014) obtained 21207RADseq SNPs (single nucleotide polymorphisms) on apanel of 767 miscanthus genotypes (mostly M sinensis)but subsequent reanalysis of the RADseq data using the
CLIFTON‐BROWN ET AL | 17
new reference genome resulted in hundreds of thousandsof SNPs being called
Robust and effective in vitro regeneration systems havebeen developed for Miscanthus sinensis M times giganteusand M sacchariflorus (Dalton 2013 Guo et al 2013Hwang Cho et al 2014 Rambaud et al 2013 Ślusarkie-wicz‐Jarzina et al 2017 Wang et al 2011 Zhang et al2012) However there is still significant genotype speci-ficity and these methods need ldquoin‐houserdquo optimization anddevelopment to be used routinely They provide potentialroutes for rapid clonal propagation and also as a basis forgenetic transformation Stable transformation using bothbiolistics (Wang et al 2011) and Agrobacterium tumefa-ciens DNA delivery methods (Hwang Cho et al 2014Hwang Lim et al 2014) has been achieved in M sinen-sis The development of miscanthus transformation andgene editing to generate diplogametes for producing seed‐propagated triploid hybrids are performed as part of theFrench project MISEDIT (miscanthus gene editing forseed‐propagated triploids) There are no reports of genomeediting in any miscanthus species but new breeding inno-vations including genome editing are particularly relevantin this slow‐to‐breed nonfood bioenergy crop (Table 4)
4 | WILLOW
Willow (Salix spp) is a very diverse group of catkin‐bear-ing trees and shrubs Willow belongs to the family Sali-caceae which also includes the Populus genus There areapproximately 350 willow species (Argus 2007) foundmostly in temperate and arctic zones in the northern hemi-sphere A few are adapted to subtropical and tropicalzones The centre of diversity is believed to be in Asiawith over 200 species in China Around 120 species arefound in the former Soviet Union over 100 in NorthAmerica and around 65 species in Europe and one speciesis native to South America (Karp et al 2011) Willows aredioecious thus obligate outcrossers and highly heterozy-gous The haploid chromosome number is 19 (Hanley ampKarp 2014) Around 40 of species are polyploid (Sudaamp Argus 1968) ranging from triploids to the atypicaldodecaploid S maxxaliana with 2n=190 (Zsuffa et al1984)
Although almost exclusively native to the NorthernHemisphere willow has been grown around the globe formany thousands of years to support a wide range of appli-cations (Kuzovkina amp Quigley 2005 Stott 1992) How-ever it has been the focus of domestication for bioenergypurposes for only a relatively short period since the 1970sin North America and Europe For bioenergy breedershave focused their efforts on the shrub willows (subgenusVetix) because of their rapid juvenile growth rates as aresponse to coppicing on a 2‐ to 4‐year cycle that can be
accomplished using farm machinery rather than forestryequipment (Shield Macalpine Hanley amp Karp 2015Smart amp Cameron 2012)
Since shrub willow was not generally recognized as anagricultural crop until very recently there has been littlecommitment to building and maintaining germplasm reposi-tories of willow to support long‐term breeding One excep-tion is the United Kingdom where a large and well‐characterized Salix germplasm collection comprising over1500 accessions is held at Rothamsted Research (Stott1992 Trybush et al 2008) Originally initiated for use inbasketry in 1923 accessions have been added ever sinceIn the United States a germplasm collection of gt350accessions is located at Cornell University to support thebreeding programme there The UK and Cornell collectionshave a relatively small number of accessions in common(around 20) Taken together they represent much of thespecies diversity but only a small fraction of the overallgenetic diversity within the genus There are three activewillow breeding programmes in Europe RothamstedResearch (UK) Salixenergi Europa AB (SEE) and a pro-gramme at the University of Warmia and Mazury in Olsz-tyn (Poland) (abbreviations used in Table 1) There is oneactive US programme based at Cornell University Culti-vars are still being marketed by the European WillowBreeding Programme (EWBP) (UK) which was activelybreeding biomass varieties from 1996 to 2002 Cultivarsare protected by plant breedersrsquo rights (PBRs) in Europeand by plant patents in the United States The sharing ofgenetic resources in the willow community is generallyregulated by material transfer agreements (MTA) and tai-lored licensing agreements although the import of cuttingsinto North America is prohibited except under special quar-antine permit conditions
Efforts to augment breeding germplasm collection fromnature are continuing with phenotypic screening of wildgermplasm performed in field experiments with 177 S pur-purea genotypes in the United States (at sites in Genevaand Portland NY and Morgantown WV) that have beengenotyped using genotyping by sequencing (GBS) (Elshireet al 2011) In addition there are approximately 400accessions of S viminalis in Europe (near Pustnaumls Upp-sala Sweden and Woburn UK (Berlin et al 2014 Hal-lingbaumlck et al 2016) The S viminalis accessions wereinitially genotyped using 38 simple sequence repeats (SSR)markers to assess genetic diversity and screened with~1600 SNPs in genes of potential interest for phenologyand biomass traits Genetics and genomics combined withextensive phenotyping have substantially improved thegenetic basis of biomass‐related traits in willow and arenow being developed in targeted breeding via MAS Thisunderpinning work has been conducted on large specifi-cally developed biparental Salix mapping populations
18 | CLIFTON‐BROWN ET AL
(Hanley amp Karp 2014 Zhou et al 2018) as well asGWAS panels (Hallingbaumlck et al 2016)
Once promising parental combinations are identifiedcrosses are usually performed using fresh pollen frommaterial that has been subject to a phased removal fromcold storage (minus4degC) (Lindegaard amp Barker 1997 Macal-pine Shield Trybush Hayes amp Karp 2008 Mosseler1990) Pollen storage is useful in certain interspecific com-binations where flowering is not naturally synchronizedThis can be overcome by using pollen collection and stor-age protocol which involves extracting pollen using toluene(Kopp Maynard Niella Smart amp Abrahamson 2002)
The main breeding approach to improve willow yieldsrelies on species hybridization to capture hybrid vigour(Fabio et al 2017 Serapiglia Gouker amp Smart 2014) Inthe absence of genotypic models for heterosis breedershave extensively tested general and specific combiningability of parents to produce superior progeny The UKbreeding programmes (EWBP 1996ndash2002 and RothamstedResearch from 2003 on) have performed more than 1500exploratory cross‐pollinations The Cornell programme hassuccessfully completed about 550 crosses since 1998Investment into the characterization of genetic diversitycombined with progeny tests from exploratory crosses hasbeen used to produce hundreds of targeted intraspeciescrosses in the United Kingdom and United States respec-tively (see Table 1) To achieve long‐term gains beyond F1hybrids four intraspecific recurrent selection populationshave been created in the United Kingdom (for S dasycla-dos S viminalis and S miyabeana) and Cornell is pursu-ing recurrent selection of S purpurea Interspecifichybridizations with genotypes selected from the recurrentselection cycles are well advanced in willow with suchcrosses to date totalling 420 in the United Kingdom andover 100 in the United States
While species hybridization is common in Salix it isnot universal Of the crosses attempted about 50 hybri-dize and produce seed (Macalpine Shield amp Karp 2010)As the viability of seed from successful crosses is short (amatter of days at ambient temperatures) proper seed rear-ing and storage protocols are essential (Maroder PregoFacciuto amp Maldonado 2000)
Progeny from crosses are treated in different waysamong the breeding programmes at the seedling stage Inthe United States seedlings are planted into an irrigatedfield where plants are screened for two seasons beforebeing progressed to further field trials In the United King-dom seedlings are planted into trays of compost wherethey remain containerized in an irrigated nursery for theremainder of year one In the United Kingdom seedlingsare subject to two rounds of selection in the nursery yearThe first round takes place in September to select againstsusceptibility to rust infection (Melampsora spp) A second
round of selection in winter assesses tip damage from frostand giant willow aphid infestation In the United Stateswhere the rust pressure is lower screening for Melampsoraspp cannot be performed at the nursery stage Both pro-grammes monitor Melampsora spp pest susceptibilityyield and architecture over multiple years in field trialsSelected material is subject to two rounds of field trials fol-lowed by a final multilocation yield trial to identify vari-eties for commercialization
Promising selections (ie potential cultivars) need to beclonally propagated A rapid in vitro tissue culture propa-gation method has been developed (Palomo‐Riacuteos et al2015) This method can generate about 5000 viable trans-plantable clones from a single plant in just 24 weeks Anin vitro system can also accommodate early selection viamolecular or biochemical markers to increase selectionspeed Conventional breeding systems take 13 years viafour rounds of selection from crossing to selecting a variety(Figure 1) but this has the potential to be reduced to7 years if micropropagation and MAS selection are adopted(Hanley amp Karp 2014 Palomo‐Riacuteos et al 2015)
Willows are currently propagated commercially byplanting winter‐dormant stem cuttings in spring Commer-cial planting systems for willow use mechanical plantersthat cut and insert stem sections from whips into a well‐prepared soil One hectare of stock plants grown in specificmultiplication beds planted at 40000 plants per ha pro-duces planting material for 80 hectares of commercialshort‐rotation coppice willow annually (planted at 15000cuttings per hectare) (Whittaker et al 2016) When com-mercial plantations are established the industry standard isto plant intimate mixtures of ~5 diverse rust (Melampsoraspp)‐resistant varieties (McCracken amp Dawson 1997 VanDen Broek et al 2001)
The foundations for using new plant breeding tech-niques have been established with funding from both thepublic and the private sectors To establish QTL maps 16mapping populations from biparental crosses are understudy in the United Kingdom Nine are under study in theUnited States The average number of individuals in thesefamilies ranges from 150 to 947 (Hanley amp Karp 2014)GS is also being evaluated in S viminalis and preliminaryresults indicate that multiomic approaches combininggenomic and metabolomic data have great potential (Sla-vov amp Davey 2017) For both QTL and GS approachesthe field phenotyping demands are large as several thou-sand individuals need to be phenotyped for a wide rangeof traits These include the following dates of bud burstand growth succession stem height stem density wooddensity and disease resistance The greater the number ofindividuals the more precise the QTL marker maps andGS models are However the logistical and financial chal-lenges of phenotyping large numbers of individuals are
CLIFTON‐BROWN ET AL | 19
considerable because the willow crop is gt5 m tall in thesecond year There is tremendous potential to improve thethroughput of phenotyping using unmanned aerial systemswhich is being tested in the USDA National Institute ofFood and Agriculture (NIFA) Willow SkyCAP project atCornell Further investment in these approaches needs tobe sustained over many years fully realizes the potentialof a marker‐assisted selection programme for willow
To date despite considerable efforts in Europe and theUnited States to establish a routine transformation systemthere has not been a breakthrough in willow but attemptsare ongoing As some form of transformation is typically aprerequisite for genome editing techniques these have notyet been applied to willow
In Europe there are 53 short‐rotation coppice (SRC) bio-mass willow cultivars registered with the Community PlantVariety Office (CPVO) for PBRs of which ~25 are availablecommercially in the United Kingdom There are eightpatented cultivars commercially available in the UnitedStates In Sweden there are nine commercial cultivars regis-tered in Europe and two others which are unregistered(httpssalixenergiseplanting-material) Furthermore thereare about 20 precommercial hybrids in final yield trials inboth the United States and the United Kingdom It has beenestimated that it would take two years to produce the stockrequired to plant 50 ha commercially from the plant stock inthe final yield trials Breeding programmes have alreadydelivered rust‐resistant varieties and increases in yield to themarket The adoption of advanced breeding technologies willlikely lead to a step change in improving traits of interest
5 | POPLAR
Poplar a fast‐growing tree from the northern hemispherewith a small genome size has been adopted for commercialforestry and scientific purposes The genus Populus con-sists of about 29 species classified in six different sectionsPopulus (formerly Leuce) Tacamahaca Aigeiros AbasoTuranga and Leucoides (Eckenwalder 1996) The Populusspecies of most interest for breeding and testing in the Uni-ted States and Europe are P nigra P deltoides P maxi-mowiczii and P trichocarpa (Stanton 2014) Populusclones for biomass production are being developed byintra‐ and interspecies hybridization (DeWoody Trewin ampTaylor 2015 Richardson Isebrands amp Ball 2014 van derSchoot et al 2000) Recurrent selection approaches areused for gradual population improvement and to create eliteclonal lines for commercialization (Berguson McMahon ampRiemenschneider 2017 Neale amp Kremer 2011) Currentlypoplar breeding in the United States occurs in industrialand academic programmes located in the Southeast theMidwest and the Pacific Northwest These use six speciesand five interspecific taxa (Stanton 2014)
The southeastern programme historically focused onrecurrent selection of P deltoides from accessions made inthe lower Mississippi River alluvial plain (Robison Rous-seau amp Zhang 2006) More recently the genetic base hasbeen broadened to produce interspecific hybrids with resis-tance to the fungal infection Septoria musiva which causescankers
In the midwest of the United States populationimprovement efforts are focused on P deltoides selectionsfrom native provenances and hybrid crosses with acces-sions introduced from Europe Interspecific intercontinental(Europe and America) hybrid crosses between P nigra andP deltoides (P times canadensis) are behind many of theleading commercial hybrids which are the most advancedbreeding materials for many applications and regions InMinnesota previous breeding experience and efforts utiliz-ing P maximowiczii and P trichocarpa have been discon-tinued due to Septoria susceptibility and a lack of coldhardiness (Berguson et al 2017) Traits targeted forimprovement include yieldgrowth rate cold hardinessadventitious rooting resistance to Septoria and Melamp-sora leaf rust and stem form The Upper Midwest pro-gramme also carries out wide hybridizations within thesection Populus The P times wettsteinii (P trem-ula times P tremuloides) taxon is bred for gains in growthrate wood quality and resistance to the fungus Entoleucamammata which causes hypoxylon canker (David ampAnderson 2002)
In the Pacific Northwest GreenWood Resources Incleads poplar breeding that emphasizes interspecifichybrid improvement of P times generosa (P deltoides timesP trichocarpa and reciprocal) and P deltoides times P maxi-mowiczii taxa for coastal regions and the P times canadensistaxon for the drier continental regions Intraspecificimprovement of second‐generation breeding populations ofP deltoides P nigra P maximowiczii and P trichocarpaare also involved (Stanton et al 2010) The present focusof GreenWood Resourcesrsquo hybridization is bioenergy feed-stock improvement concentrating on coppice yield wood‐specific gravity and rate of sugar release
Industrial interest in poplar in the United States has his-torically come from the pulp and paper sector althoughveneer and dimensional lumber markets have been pursuedat times Currently the biomass market for liquid trans-portation fuels is being emphasized along with the use oftraditional and improved poplar genotypes for ecosystemservices such as phytoremediation (Tuskan amp Walsh 2001Zalesny et al 2016)
In Europe there are breeding programmes in FranceGermany Italy and Sweden These include the following(a) Alasia Franco Vivai (AFV) programme in northernItaly (b) the French programme led by the poplar Scien-tific Interest Group (GIS Peuplier) and carried out
20 | CLIFTON‐BROWN ET AL
collaboratively between the National Institute for Agricul-tural Research (INRA) the National Research Unit ofScience and Technology for Environment and Agriculture(IRSTEA) and the Forest Cellulose Wood Constructionand Furniture Technology Institute (FCBA) (c) the Ger-man programme at Northwest German Forest Research Sta-tion (NW‐FVA) at Hannoversch Muumlnden and (d) theSwedish programme at the Swedish University of Agricul-tural Sciences and SweTree Technologies AB (Table 1)
AFV leads an Italian poplar breeding programme usingextensive field‐grown germplasm collections of P albaP deltoides P nigra and P trichocarpa While interspeci-fic hybridization uses several taxa the focus is onP times canadensis The breeding programme addresses dis-ease resistance (Marssonina brunnea Melampsora larici‐populina Discosporium populeum and poplar mosaicvirus) growth rate and photoperiod adaptation AFV andGreenWood Resources collaborate in poplar improvementin Europe through the exchange of frozen pollen and seedfor reciprocal breeding projects Plantations in Poland andRomania are currently the focus of the collaboration
The ongoing French GIS Peuplier is developing a long‐term breeding programme based on intraspecific recurrentselection for the four parental species (P deltoides P tri-chocarpa P nigra and P maximowiczii) designed to betterbenefit from hybrid vigour demonstrated by the interspeci-fic crosses P canadensis P deltoides times P trichocarpaand P trichocarpa times P maximowiczii Current selectionpriorities are targeting adaptation to soil and climate condi-tions resistance and tolerance to the most economicallyimportant diseases and pests high volume production underSRC and traditional poplar cultivation regimes as well aswood quality of interest by different markets Currentlygenomic selection is under exploration to increase selectionaccuracy and selection intensity while maintaining geneticdiversity over generations
The German NW‐FVA programme is breeding intersec-tional AigeirosndashTacamahaca hybrids with a focus on resis-tance to Pollaccia elegans Xanthomonas populiDothichiza spp Marssonina brunnea and Melampsoraspp (Stanton 2014) Various cross combinations ofP maximowiczii P trichocarpa P nigra and P deltoideshave led to new cultivars suitable for deployment in vari-etal mixtures of five to ten genotypes of complementarystature high productivity and phenotypic stability (Weis-gerber 1993) The current priority is the selection of culti-vars for high‐yield short‐rotation biomass production Sixhundred P nigra genotypes are maintained in an ex situconservation programme An in situ P nigra conservationeffort involves an inventory of native stands which havebeen molecular fingerprinted for identity and diversity
The Swedish programme is concentrating on locallyadapted genotypes used for short‐rotation forestry (SRF)
because these meet the needs of the current pulping mar-kets Several field trials have shown that commercial poplarclones tested and deployed in Southern and Central Europeare not well adapted to photoperiods and low temperaturesin Sweden and in the Baltics Consequently SwedishUniversity of Agricultural Sciences and SweTree Technolo-gies AB started breeding in Sweden in 1990s to producepoplar clones better adapted to local climates and markets
51 | Molecular breeding technologies
Poplar genetic improvement cannot be rapidly achievedthrough traditional methods alone because of the longbreeding cycles outcrossing breeding systems and highheterozygosity Integrating modern genetic genomic andphenomics techniques with conventional breeding has thepotential to expedite poplar improvement
The genome of poplar has been sequenced (Tuskanet al 2006) It has an estimated genome size of485 plusmn 10 Mbp divided into 19 chromosomes This is smal-ler than other PBCs and makes poplar more amenable togenetic engineering (transgenesis) GS and genome editingPoplar has seen major investment in both the United Statesand Europe being the model system for woody perennialplant genetics and genomics research
52 | Targets for genetic modification
Traits targeted include wood properties (lignin content andcomposition) earlylate flowering male sterility to addressbiosafety regulation issues enhanced yield traits and herbi-cide tolerance These extensive transgenic experiments haveshown differences in recalcitrance to in vitro regenerationand genetic transformation in some of the most importantcommercial hybrid poplars (Alburquerque et al 2016)Further transgene expression stability is being studied Sofar China is the only country known to have commerciallyused transgenic insect‐resistant poplar A precommercialherbicide‐tolerant poplar was trialled for 8 years in the Uni-ted States (Li Meilan Ma Barish amp Strauss 2008) butcould not be released due to stringent environmental riskassessments required for regulatory approval This increasestranslation costs and delays reducing investor confidencefor commercial deployment (Harfouche et al 2011)
The first field trials of transgenic poplar were performedin France in 1987 (Fillatti Sellmer Mccown Haissig ampComai 1987) and in Belgium in 1988 (Deblock 1990)Although there have been a total of 28 research‐scale GMpoplar field trials approved in the European Union underCouncil Directive 90220EEC since October 1991 (inPoland Belgium Finland France Germany Spain Swe-den and in the United Kingdom (Pilate et al 2016) onlyauthorizations in Poland and Belgium are in place today In
CLIFTON‐BROWN ET AL | 21
the United States regulatory notifications and permits fornearly 20000 transgenic poplar trees derived from approxi-mately 600 different constructs have been issued since1995 by the USDAs Animal and Plant Health InspectionService (APHIS) (Strauss et al 2016)
53 | Genome editing CRISPR technologies
Clustered regularly interspaced palindromic repeats(CRISPR) and the CRISPR‐associated (CRISPR‐Cas)nucleases are a groundbreaking genome‐engineering toolthat complements classical plant breeding and transgenicmethods (Moreno‐Mateos et al 2017) Only two publishedstudies in poplar have applied the CRISPRCas9 technol-ogy One is in P tomentosa in which an endogenous phy-toene desaturase gene (PtoPDS) was successfully disruptedsite specifically in the first generation of transgenic plantsresulting in an albino and dwarf phenotype (Fan et al2015) The second was in P tremula times alba in whichhigh CRISPR‐Cas9 mutational efficiency was achieved forthree 4‐coumarateCoA ligase (4Cl) genes 4CL1 4CL2and 4CL5 associated with lignin and flavonoid biosynthe-sis (Zhou et al 2015) Due to its low cost precision andrapidness it is very probable that cultivars or clones pro-duced using CRISPR technology will be ready for market-ing in the near future (Yin et al 2017) Recently aCRISPR with a smaller associated endonuclease has beendiscovered from Prevotella and Francisella 1 (Cpf1) whichmay have advantages over Cas9 In addition there arereports of DNA‐free editing in plants using both CRISPRCpf1 and CRISPR Cas9 for example Ref (Kim et al2017 Mahfouz 2017 Zaidi et al 2017)
It remains unresolved whether plants modified by genomeediting will be regulated as genetically modified organisms(GMOs) by the relevant authorities in different countries(Lozano‐Juste amp Cutler 2014) Regulations to cover thesenew breeding techniques are still evolving but those coun-tries who have published specific guidance (including UnitedStates Argentina and Chile) are indicating that plants pos-sessing simple genome edits will not be regulated as conven-tional transgenesis (Jones 2015b) The first generation ofgenome‐edited crops will likely be phenocopy gene knock-outs that already exist to produce ldquonature identicalrdquo traits thatis traits that could also be derived by conventional breedingDespite this confidence in applying these new powerfulbreeding tools remains limited owing to the uncertain regula-tory environment in many parts of the world (Gao 2018)including the recent ECJ 2018 rulings mentioned earlier
54 | Genomics‐based breeding technologies
Poplar breeding programs are becoming well equipped withuseful genomics tools and resources that are critical to
explore genomewide variability and make use of the varia-tion for enhancing genetic gains Deep transcriptomesequencing resequencing of alternate genomes and GBStechnology for genomewide marker detection using next‐generation sequencing (NGS) are yielding valuable geno-mics tools GWAS with NGS‐based markers facilitatesmarker identification for MAS breeding by design and GS
GWAS approaches have provided a deeper understand-ing of genome function as well as allelic architectures ofcomplex traits (Huang et al 2010) and have been widelyimplemented in poplar for wood characteristics (Porthet al 2013) stomatal patterning carbon gain versus dis-ease resistance (McKown et al 2014) height and phenol-ogy (Evans et al 2014) cell wall chemistry (Mucheroet al 2015) growth and cell walls traits (Fahrenkroget al 2017) bark roughness (Bdeir et al 2017) andheight and diameter growth (Liu et al 2018) Using high‐throughput sequencing and genotyping platforms an enor-mous amount of SNP markers have been used to character-ize the linkage disequilibrium (LD) in poplar (eg Slavovet al 2012 discussed below)
The genetic architecture of photoperiodic traits in peren-nial trees is complex involving many loci However itshows high levels of conservation during evolution (Mau-rya amp Bhalerao 2017) These genomics tools can thereforebe used to address adaptation issues and fine‐tune themovement of elite lines into new environments For exam-ple poor timing of spring bud burst and autumn bud setcan result in frost damage resulting in yield losses (Ilstedt1996) These have been studied in P tremula genotypesalong a latitudinal cline in Sweden (~56ndash66oN) and haverevealed high nucleotide polymorphism in two nonsynony-mous SNPs within and around the phytochrome B2 locus(Ingvarsson Garcia Hall Luquez amp Jansson 2006 Ing-varsson Garcia Luquez Hall amp Jansson 2008) Rese-quencing 94 of these P tremula genotypes for GWASshowed that noncoding variation of a single genomicregion containing the PtFT2 gene described 65 ofobserved genetic variation in bud set along the latitudinalcline (Tan 2018)
Resequencing genomes is currently the most rapid andeffective method detecting genetic differences betweenvariants and for linking loci to complex and importantagronomical and biomass traits thus addressing breedingchallenges associated with long‐lived plants like poplars
To date whole genome resequencing initiatives havebeen launched for several poplar species and genotypes InPopulus LD studies based on genome resequencing sug-gested the feasibility of GWAS in undomesticated popula-tions (Slavov et al 2012) This plant population is beingused to inform breeding for bioenergy development Forexample the detection of reliable phenotypegenotype asso-ciations and molecular signatures of selection requires a
22 | CLIFTON‐BROWN ET AL
detailed knowledge about genomewide patterns of allelefrequency variation LD and recombination suggesting thatGWAS and GS in undomesticated populations may bemore feasible in Populus than previously assumed Slavovet al (2012) have resequenced 16 genomes of P tri-chocarpa and genotyped 120 trees from 10 subpopulationsusing 29213 SNPs (Geraldes et al 2013) The largest everSNP data set of genetic variations in poplar has recentlybeen released providing useful information for breedinghttpswwwbioenergycenterorgbescgwasindexcfmAlso deep sequencing of transcriptomes using RNA‐Seqhas been used for identification of functional genes andmolecular markers that is polymorphism markers andSSRs A multitissue and multiple experimental data set forP trichocarpa RNA‐Seq is publicly available (httpsjgidoegovdoe-jgi-plant-flagship-gene-atlas)
The availability of genomic information of DNA‐con-taining cell organelles (nucleus chloroplast and mitochon-dria) will also allow a holistic approach in poplarmolecular breeding in the near future (Kersten et al2016) Complete Populus genome sequences are availablefor nucleus (P trichocarpa section Tacamahaca) andchloroplasts (seven species and two clones from P trem-ula W52 and P tremula times P alba 717ndash1B4) A compara-tive approach revealed structural and functionalinformation broadening the knowledge base of PopuluscpDNA and stimulating future diagnostic marker develop-ment The availability of whole genome sequences of thesecellular compartments of P tremula holds promise forboosting marker‐assisted poplar breeding Other nucleargenome sequences from additional Populus species arenow available (eg P deltoides (httpsphytozomejgidoegovpz) and will become available in the forthcomingyears (eg P tremula and P tremuloidesmdashPopGenIE(Sjodin Street Sandberg Gustafsson amp Jansson 2009))Recently the characterization of the poplar pan‐genome bygenomewide identification of structural variation in threecrossable poplar species P nigra P deltoides and P tri-chocarpa revealed a deeper understanding of the role ofinter‐ and intraspecific structural variants in poplar pheno-type and may have important implications for breedingparticularly interspecific hybrids (Pinosio et al 2016)
GS has been proposed as an alternative to MAS in cropimprovement (Bernardo amp Yu 2007 Heffner Sorrells ampJannink 2009) GS is particularly well suited for specieswith long generation times for characteristics that displaymoderate‐to‐low heritability for traits that are expensive tomeasure and for selection of traits expressed late in the lifecycle as is the case for most traits of commercial value inforestry (Harfouche et al 2012) Current joint genomesequencing efforts to implement GS in poplar using geno-mic‐estimated breeding values for bioenergy conversiontraits from 49 P trichocarpa families and 20 full‐sib
progeny are taking place at the Oak Ridge National Labo-ratory and GreenWood Resources (Brian Stanton personalcommunication httpscbiornlgov) These data togetherwith the resequenced GWAS population data will be thebasis for developing GS algorithms Genomic breedingtools have been developed for the intraspecific programmetargeting yield resistance to Venturia shoot blightMelampsora leaf rust resistance to Cryptorhynchus lapathistem form wood‐specific gravity and wind firmness (Evanset al 2014 Guerra et al 2016) A newly developedldquobreeding with rare defective allelesrdquo (BRDA) technologyhas been developed to exploit natural variation of P nigraand identify defective variants of genes predicted by priortransgenic research to impact lignin properties Individualtrees carrying naturally defective alleles can then be incor-porated directly into breeding programs thereby bypassingthe need for transgenics (Vanholme et al 2013) Thisnovel breeding technology offers a reverse genetics com-plement to emerging GS for targeted improvement of quan-titative traits (Tsai 2013)
55 | Phenomics‐assisted breeding technology
Phenomics involves the characterization of phenomesmdashthefull set of phenotypes of given individual plants (HouleGovindaraju amp Omholt 2010) Traditional phenotypingtools which inefficiently measure a limited set of pheno-types have become a bottleneck in plant breeding studiesHigh‐throughput plant phenotyping facilities provide accu-rate screening of thousands of plant breeding lines clones orpopulations over time (Fu 2015) are critical for acceleratinggenomics‐based breeding Automated image collection andanalysis phenomics technologies allow accurate and nonde-structive measurements of a diversity of phenotypic traits inlarge breeding populations (Gegas Gay Camargo amp Doo-nan 2014 Goggin Lorence amp Topp 2015 Ludovisi et al2017 Shakoor Lee amp Mockler 2017) One important con-sideration is the identification of relevant and quantifiabletarget traits that are early diagnostic indicators of biomassyield Good progress has been made in elucidating theseunderpinning morpho‐physiological traits that are amenableto remote sensing in Populus (Harfouche Meilan amp Altman2014 Rae Robinson Street amp Taylor 2004) Morerecently Ludovisi et al (2017) developed a novel methodol-ogy for field phenomics of drought stress in a P nigra F2partially inbred population using thermal infrared imagesrecorded from an unmanned aerial vehicle‐based platform
Energy is the current main market for poplar biomass butthe market return provided is not sufficient to support pro-duction expansion even with added demand for environmen-tal and land management ldquoecosystem servicesrdquo such as thetreatment of effluent phytoremediation riparian buffer zonesand agro‐forestry plantings Aviation fuel is a significant
CLIFTON‐BROWN ET AL | 23
target market (Crawford et al 2016) To serve this marketand to reduce current carbon costs of production (Budsberget al 2016) key improvement traits in addition to yield(eg coppice regeneration pestdisease resistance water‐and nutrient‐use efficiencies) will be trace greenhouse gas(GHG) emissions (eg isoprene volatiles) site adaptabilityand biomass conversion efficiency Efforts are underway tohave national environmental protection agenciesrsquo approvalfor poplar hybrids qualifying for renewable energy credits
6 | REFLECTIONS ON THECOMMERCIALIZATIONCHALLENGE
The research and innovation activities reviewed in thispaper aim to advance the genetic improvement of speciesthat can provide feedstocks for bioenergy applicationsshould those markets eventually develop These marketsneed to generate sufficient revenue and adequately dis-tribute it to the actors along the value chain The work onall four crops shares one thing in common long‐termefforts to integrate fundamental knowledge into breedingand crop development along a research and development(RampD) pipeline The development of miscanthus led byAberystwyth University exemplifies the concerted researcheffort that has integrated the RampD activities from eight pro-jects over 14 years with background core research fundingalong an emerging innovation chain (Figure 2) This pro-gramme has produced a first range of conventionally bredseeded interspecies hybrids which are now in upscaling tri-als (Table 3) The application of molecular approaches(Table 5) with further conventional breeding (Table 4)offers the prospect of a second range of improved seededhybrids This example shows that research‐based support ofthe development of new crops or crop types requires along‐term commitment that goes beyond that normallyavailable from project‐based funding (Figure 1) Innovationin this sector requires continuous resourcing of conven-tional breeding operations and capability to minimize timeand investment losses caused by funding discontinuities
This challenge is increased further by the well‐knownmarket failure in the breeding of many agricultural cropspecies The UK Department for Environment Food andRural Affairs (Defra) examined the role of geneticimprovement in relation to nonmarket outcomes such asenvironmental protection and concluded that public invest-ment in breeding was required if profound market failure isto be addressed (Defra 2002) With the exception ofwidely grown hybrid crops such as maize and some high‐value horticultural crops royalties arising from plant breed-ersrsquo rights or other returns to breeders fail to adequatelycompensate for the full cost for research‐based plant breed-ing The result even for major crops such as wheat is sub‐
optimal investment and suboptimal returns for society Thismarket failure is especially acute for perennial crops devel-oped for improved sustainability rather than consumerappeal (Tracy et al 2016) Figure 3 illustrates the underly-ing challenge of capturing value for the breeding effortThe ldquovalley of deathrdquo that results from the low and delayedreturns to investment applies generally to the research‐to‐product innovation pipeline (Beard Ford Koutsky amp Spi-wak 2009) and certainly to most agricultural crop speciesHowever this schematic is particularly relevant to PBCsMost of the value for society from the improved breedingof these crops comes from changes in how agricultural landis used that is it depends on the increased production ofthese crops The value for society includes many ecosys-tems benefits the effects of a return to seminatural peren-nial crop cover that protects soils the increase in soilcarbon storage the protection of vulnerable land or the cul-tivation of polluted soils and the reductions in GHG emis-sions (Lewandowski 2016) By its very nature theproduction of biomass on agricultural land marginal forfood production challenges farm‐level profitability Thecosts of planting material and one‐time nature of cropestablishment are major early‐stage costs and therefore theopportunities for conventional royalty capture by breedersthat are manifold for annual crops are limited for PBCs(Hastings et al 2017) Public investment in developingPBCs for the nonfood biobased sector needs to providemore long‐term support for this critical foundation to a sus-tainable bioeconomy
7 | CONCLUSIONS
This paper provides an overview of research‐based plantbreeding in four leading PBCs For all four PBC generasignificant progress has been made in genetic improvementthrough collaboration between research scientists and thoseoperating ongoing breeding programmes Compared withthe main food crops most PBC breeding programmes dateback only a few decades (Table 1) This breeding efforthas thus co‐evolved with molecular biology and the result-ing ‐omics technologies that can support breeding Thedevelopment of all four PBCs has depended strongly onpublic investment in research and innovation The natureand driver of the investment varied In close associationwith public research organizations or universities all theseprogrammes started with germplasm collection and charac-terization which underpin the selection of parents forexploratory wide crosses for progeny testing (Figure 1Table 4)
Public support for switchgrass in North America wasexplicitly linked to plant breeding with 12 breeding pro-grammes supported in the United States and CanadaSwitchgrass breeding efforts to date using conventional
24 | CLIFTON‐BROWN ET AL
breeding have resulted in over 36 registered cultivars inthe United States (Table 1) with the development of dedi-cated biomass‐type cultivars coming within the past fewyears While ‐omics technologies have been incorporatedinto several of these breeding programmes they have notyet led to commercial deployment in either conventional orhybrid cultivars
Willow genetic improvement was led by the researchcommunity closely linked to plant breeding programmesWillow and poplar have the longest record of public invest-ment in genetic improvement that can be traced back to1920s in the United Kingdom and United States respec-tively Like switchgrass breeding programmes for willoware connected to public research efforts The United King-dom in partnership with the programme based at CornellUniversity remains the European leader in willowimprovement with a long‐term breeding effort closelylinked to supporting biological research at Rothamsted Inwillow F1 hybrids have produced impressive yield gainsover parental germplasm by capturing hybrid vigour Over30 willow clones are commercially available in the UnitedStates and Europe and a further ~90 are under precommer-cial testing (Table 1)
Compared with willow and poplar miscanthus is a rela-tive newcomer with all the current breeding programmesstarting in the 2000s Clonal M times giganteus propagated byrhizomes is expected to be replaced by more readily scal-able seeded hybrids from intra‐ (M sinensis) and inter‐(M sacchariflorus times M sinensis) species crosses with highseed multiplication rates (of gt2000) The first group ofhybrid cultivars is expected to be market‐ready around2022
Of the four genera used as PBCs Populus is the mostadvanced in terms of achievements in biological researchas a result of its use as a model for basic research of treesMuch of this biological research is not directly connectedto plant breeding Nevertheless reflecting the fact thatpoplar is widely grown as a single‐stem tree in SRF thereare about 60 commercially available clones and an addi-tional 80 clones in commercial pipelines (Table 1) Trans-genic poplar hybrids have moved beyond proof of conceptto commercial reality in China
Many PBC programmes have initiated long‐term con-ventional recurrent selection breeding cycles for populationimprovement which is a key process in increasing yieldthrough hybrid vigour As this approach requires manyyears most programmes are experimenting with molecularbreeding methods as these have the potential to accelerateprecision breeding For all four PBCs investments in basicgenetic and genomic resources including the developmentof mapping populations for QTLs and whole genomesequences are available to support long‐term advancesMore recently association genetics with panels of diversegermplasm are being used as training populations for GSmodels (Table 5) These efforts are benefitting from pub-licly available DNA sequences and whole genome assem-blies in crop databases Key to these accelerated breedingtechnologies are developments in novel phenomics tech-nologies to bridge the genotypephenotype gap In poplarnovel remote sensing field phenotyping is now beingdeployed to assist breeders These advances are being com-bined with in vitro and in planta modern molecular breed-ing techniques such as CRISPR (Table 5) CRISPRtechnology for genome editing has been proven in poplar
FIGURE 2 A schematic developmentpathway for miscanthus in the UnitedKingdom related to the investment in RampDprojects at Aberystwyth (top coloured areasfor projects in the three categories basicresearch breeding and commercialupscaling) leading to a projected croppingarea of 350000 ha by 2030 with clonal andsuccessive ranges of improved seed‐basedhybrids Purple represents theBiotechnology and Biological SciencesResearch Council (BBSRC) and brown theDepartment for Environment Food andRural Affairs (Defra) (UK Nationalfunding) blue bars represent EU fundingand green private sector funding (Terravestaand CERES) and GIANT‐LINK andMiscanthus Upscaling Technology (MUST)are publicndashprivate‐initiatives (PPI)
CLIFTON‐BROWN ET AL | 25
This technology is also being applied in switchgrass andmiscanthus (Table 5) but the future of CRISPR in com-mercial breeding for the European market is uncertain inthe light of recent ECJ 2018 rulings
There is integration of research and plant breeding itselfin all four PBCs Therefore estimating the ongoing costs ofmaintaining these breeding programmes is difficult Invest-ment in research also seeks wider benefits associated withtechnological advances in plant science rather than cultivardevelopment per se However in all cases the conventionalbreeding cycle shown in Figure 1 is the basic ldquoenginerdquo withmolecular technologies (‐omics) serving to accelerate thisengine The history of the development shows that the exis-tence of these breeding programmes is essential to gain ben-efits from the biological research Despite this it is thisessential step that is at most risk from reductions in invest-ment A conventional breeding programme typicallyrequires a breeder and several technicians who are supported
over the long term (20ndash30 years Figure 1) at costs of about05 to 10 million Euro per year (as of 2018) The analysisreported here shows that the time needed to perform onecycle of conventional breeding bringing germplasm fromthe wild to a commercial hybrid ranged from 11 years inswitchgrass to 26 years in poplar (Figure 1) In a maturecrop grown on over 100000 ha with effective cultivar pro-tection and a suitable business model this level of revenuecould come from royalties Until such levels are reachedPBCs lie in the innovation valley of death (Figure 3) andneed public support
Applying industrial ldquotechnology readiness levelsrdquo (TRL)originally developed for aerospace (Heacuteder 2017) to ourplant breeding efforts we estimate many promising hybridscultivars are at TRL levels of 3ndash4 In Table 3 experts ineach crop estimate that it would take 3 years from now toupscale planting material from leading cultivars in plot tri-als to 100 ha
FIGURE 3 A schematic relating some of the steps in the innovation chain from relatively basic crop science research through to thedeployment in commercial cropping systems and value chains The shape of the funnel above the expanding development and deploymentrepresents the availability of investment along the development chain from relatively basic research at the top to the upscaled deployment at thebottom Plant breeding links the research effort with the development of cropping systems The constriction represents the constrained fundingfor breeding that links conventional public research investment and the potential returns from commercial development The handover pointsbetween publicly funded work to develop the germplasm resources (often known as prebreeding) the breeding and the subsequent cropdevelopment are shown on the left The constriction point is aggravated by the lack academic rewards for this essential breeding activity Theoutcome is such that this innovation system is constrained by the precarious resourcing of plant breeding The authorsrsquo assessment ofdevelopment status of the four species is shown (poplar having two one for short‐rotation coppice (SRC) poplar and one for the more traditionalshort‐rotation forestry (SRF)) The four new perennial biomass crops (PBCs) are now in the critical phase of depending of plant breedingprogress without the income stream from a large crop production base
26 | CLIFTON‐BROWN ET AL
Taking the UK example mentioned in the introductionplanting rates of ~35000 ha per year from 2022 onwardsare needed to reach over 1 m ha by 2050 Ongoing workin the UK‐funded Miscanthus Upscaling Technology(MUST) project shows that ramping annual hybrid seedproduction from the current level of sufficient seed for10 ha in 2018 to 35000 ha would take about 5 yearsassuming no setbacks If current hybrids of any of the fourPBCs in the upscaling pipeline fail on any step for exam-ple lower than expected multiplication rates or unforeseenagronomic barriers then further selections from ongoingbreeding are needed to replace earlier candidates
In conclusion the breeding foundations have been laidwell for switchgrass miscanthus willow and poplar owingto public funding over the long time periods necessaryImproved cultivars or genotypes are available that could bescaled up over a few years if real sustained market oppor-tunities emerged in response to sustained favourable poli-cies and industrial market pull The potential contributionsof growing and using these PBCs for socioeconomic andenvironmental benefits are clear but how farmers andothers in commercial value chains are rewarded for mass‐scale deployment as is necessary is not obvious at presentTherefore mass‐scale deployment of these lignocellulosecrops needs developments outside the breeding arenas todrive breeding activities more rapidly and extensively
8 | UNCITED REFERENCE
Huang et al (2019)
ACKNOWLEDGEMENTS
UK The UK‐led miscanthus research and breeding wasmainly supported by the Biotechnology and BiologicalSciences Research Council (BBSRC) Department for Envi-ronment Food and Rural Affairs (Defra) the BBSRC CSPstrategic funding grant BBCSP17301 Innovate UKBBSRC ldquoMUSTrdquo BBN0161491 CERES Inc and Ter-ravesta Ltd through the GIANT‐LINK project (LK0863)Genomic selection and genomewide association study activi-ties were supported by BBSRC grant BBK01711X1 theBBSRC strategic programme grant on Energy Grasses ampBio‐refining BBSEW10963A01 The UK‐led willow RampDwork reported here was supported by BBSRC (BBSEC00005199 BBSEC00005201 BBG0162161 BBE0068331 BBG00580X1 and BBSEC000I0410) Defra(NF0424 NF0426) and the Department of Trade and Indus-try (DTI) (BW6005990000) IT The Brain Gain Program(Rientro dei cervelli) of the Italian Ministry of EducationUniversity and Research supports Antoine Harfouche USContributions by Gerald Tuskan to this manuscript were sup-ported by the Center for Bioenergy Innovation a US
Department of Energy Bioenergy Research Center supportedby the Office of Biological and Environmental Research inthe DOE Office of Science under contract number DE‐AC05‐00OR22725 Willow breeding efforts at CornellUniversity have been supported by grants from the USDepartment of Agriculture National Institute of Food andAgriculture Contributions by the University of Illinois weresupported primarily by the DOE Office of Science Office ofBiological and Environmental Research (BER) grant nosDE‐SC0006634 DE‐SC0012379 and DE‐SC0018420 (Cen-ter for Advanced Bioenergy and Bioproducts Innovation)and the Energy Biosciences Institute EU We would like tofurther acknowledge contributions from the EU projectsldquoOPTIMISCrdquo FP7‐289159 on miscanthus and ldquoWATBIOrdquoFP7‐311929 on poplar and miscanthus as well as ldquoGRACErdquoH2020‐EU326 Biobased Industries Joint Technology Ini-tiative (BBI‐JTI) Project ID 745012 on miscanthus
CONFLICT OF INTEREST
The authors declare that progress reported in this paperwhich includes input from industrial partners is not biasedby their business interests
ORCID
John Clifton‐Brown httporcidorg0000-0001-6477-5452Antoine Harfouche httporcidorg0000-0003-3998-0694Lawrence B Smart httporcidorg0000-0002-7812-7736Danny Awty‐Carroll httporcidorg0000-0001-5855-0775VasileBotnari httpsorcidorg0000-0002-0470-0384Lindsay V Clark httporcidorg0000-0002-3881-9252SalvatoreCosentino httporcidorg0000-0001-7076-8777Lin SHuang httporcidorg0000-0003-3079-326XJon P McCalmont httporcidorg0000-0002-5978-9574Paul Robson httporcidorg0000-0003-1841-3594AnatoliiSandu httporcidorg0000-0002-7900-448XDaniloScordia httporcidorg0000-0002-3822-788XRezaShafiei httporcidorg0000-0003-0891-0730Gail Taylor httporcidorg0000-0001-8470-6390IrisLewandowski httporcidorg0000-0002-0388-4521
REFERENCES
Alburquerque N Baldacci‐Cresp F Baucher M Casacuberta JM Collonnier C ElJaziri M hellip Burgos L (2016) New trans-formation technologies for trees In C Vettori F Gallardo H
CLIFTON‐BROWN ET AL | 27
Haumlggman V Kazana F Migliacci G Pilateamp M Fladung(Eds) Biosafety of forest transgenic trees (pp 31ndash66) DordrechtNetherlands Springer
Argus G W (2007) Salix (Salicaceae) distribution maps and a syn-opsis of their classification in North America north of MexicoHarvard Papers in Botany 12 335ndash368 httpsdoiorg1031001043-4534(2007)12[335SSDMAA]20CO2
Atienza S G Ramirez M C amp Martin A (2003) Mapping‐QTLscontrolling flowering date in Miscanthus sinensis Anderss CerealResearch Communications 31 265ndash271
Atienza S G Satovic Z Petersen K K Dolstra O amp Martin A(2003) Identification of QTLs influencing agronomic traits inMiscanthus sinensis Anderss I Total height flag‐leaf height andstem diameter Theoretical and Applied Genetics 107 123ndash129
Atienza S G Satovic Z Petersen K K Dolstra O amp Martin A(2003) Influencing combustion quality in Miscanthus sinensisAnderss Identification of QTLs for calcium phosphorus and sul-phur content Plant Breeding 122 141ndash145
Bdeir R Muchero W Yordanov Y Tuskan G A Busov V ampGailing O (2017) Quantitative trait locus mapping of Populusbark features and stem diameter BMC Plant Biology 17 224httpsdoiorg101186s12870-017-1166-4
Beard T R Ford G S Koutsky T M amp Spiwak L J (2009) AValley of Death in the innovation sequence An economic investi-gation Research Evaluation 18 343ndash356 httpsdoiorg103152095820209X481057
Berguson W McMahon B amp Riemenschneider D (2017) Addi-tive and non‐additive genetic variances for tree growth in severalhybrid poplar populations and implications regarding breedingstrategy Silvae Genetica 66(1) 33ndash39 httpsdoiorg101515sg-2017-0005
Berlin S Trybush S O Fogelqvist J Gyllenstrand N Halling-baumlck H R Aringhman I hellip Hanley S J (2014) Genetic diversitypopulation structure and phenotypic variation in European Salixviminalis L (Salicaceae) Tree Genetics amp Genomes 10 1595ndash1610 httpsdoiorg101007s11295-014-0782-5
Bernardo R amp Yu J (2007) Prospects for genomewide selectionfor quantitative traits in maize Crop Science 47 1082ndash1090httpsdoiorg102135cropsci2006110690
Biswal A K Atmodjo M A Li M Baxter H L Yoo C GPu Y hellip Mohnen D (2018) Sugar release and growth of bio-fuel crops are improved by downregulation of pectin biosynthesisNature Biotechnology 36 249ndash257
Bmu (2009) National biomass action plan for Germany (p 17) Bun-desministerium fuumlr Umwelt Naturschutz und ReaktorsicherheitBundesministerium fuumlr Ernaumlhrung (BMU) amp Landwirtschaft undVerbraucherschutz (BMELV) 11055 Berlin | Germany Retrievedfrom httpwwwbmeldeSharedDocsDownloadsENPublicationsBiomassActionPlanpdf__blob=publicationFile
Brummer E C (1999) Capturing heterosis in forage crop cultivardevelopment Crop Science 39 943ndash954 httpsdoiorg102135cropsci19990011183X003900040001x
Budsberg E Crawford J T Morgan H Chin W S Bura R ampGustafson R (2016) Hydrocarbon bio‐jet fuel from bioconver-sion of poplar biomass Life cycle assessment Biotechnology forBiofuels 9 170 httpsdoiorg101186s13068-016-0582-2
Burris K P Dlugosz E M Collins A G Stewart C N amp Lena-ghan S C (2016) Development of a rapid low‐cost protoplasttransfection system for switchgrass (Panicum virgatum L) Plant
Cell Reports 35 693ndash704 httpsdoiorg101007s00299-015-1913-7
Casler M (2012) Switchgrass breeding genetics and genomics InA Monti (Ed) Switchgrass (pp 29ndash54) London UK Springer
Casler M Mitchell R amp Vogel K (2012) Switchgrass In CKole C P Joshi amp D R Shonnard (Eds) Handbook of bioen-ergy crop plants Vol 2 New York NY Taylor amp Francis
Casler M D amp Ramstein G P (2018) Breeding for biomass yieldin switchgrass using surrogate measures of yield BioEnergyResearch 11 6ndash12
Casler M D Sosa S Hoffman L Mayton H Ernst C Adler PR hellip Bonos S A (2017) Biomass Yield of Switchgrass Culti-vars under High‐ versus Low‐Input Conditions Crop Science 57821ndash832 httpsdoiorg102135cropsci2016080698
Casler M D amp Vogel K P (2014) Selection for biomass yield inupland lowland and hybrid switchgrass Crop Science 54 626ndash636 httpsdoiorg102135cropsci2013040239
Casler M D Vogel K P amp Harrison M (2015) Switchgrassgermplasm resources Crop Science 55 2463ndash2478 httpsdoiorg102135cropsci2015020076
Casler M D Vogel K P Lee D K Mitchell R B Adler P RSulc R M hellip Moore K J (2018) 30 Years of progress towardincreasing biomass yield of switchgrass and big bluestem CropScience 58 1242
Clark L V Brummer J E Głowacka K Hall M C Heo K PengJ hellip Sacks E J (2014) A footprint of past climate change on thediversity and population structure of Miscanthus sinensis Annals ofBotany 114 97ndash107 httpsdoiorg101093aobmcu084
Clark L V Dzyubenko E Dzyubenko N Bagmet L SabitovA Chebukin P hellip Sacks E J (2016) Ecological characteristicsand in situ genetic associations for yield‐component traits of wildMiscanthus from eastern Russia Annals of Botany 118 941ndash955
Clark L V Jin X Petersen K K Anzoua K G Bagmet LChebukin P hellip Sacks E J(2018) Population structure of Mis-canthus sacchariflorus reveals two major polyploidization eventstetraploid‐mediated unidirectional introgression from diploid Msinensis and diversity centred around the Yellow Sea Annals ofBotany 20 1ndash18 httpsdoiorg101093aobmcy1161
Clark L V Stewart J R Nishiwaki A Toma Y Kjeldsen J BJoslashrgensen U hellip Sacks E J (2015) Genetic structure ofMiscanthus sinensis and Miscanthus sacchariflorus in Japan indi-cates a gradient of bidirectional but asymmetric introgressionJournal of Experimental Botany 66 4213ndash4225
Clifton‐Brown J Hastings A Mos M McCalmont J P Ashman CAwty‐Carroll DhellipCracroft‐EleyW (2017) Progress in upscalingMis-canthus biomass production for the European bio‐economy with seed‐based hybridsGlobal Change Biology Bioenergy 9 6ndash17
Clifton‐Brown J Schwarz K U amp Hastings A (2015) History ofthe development of Miscanthus as a bioenergy crop From smallbeginnings to potential realisation Biology and Environment‐Pro-ceedings of the Royal Irish Academy 115B 45ndash57
Clifton‐Brown J Senior H Purdy S Horsnell R Lankamp BMuumlennekhoff A-K hellip Bentley A R (2018) Investigating thepotential of novel non‐woven fabrics to increase pollination effi-ciency in plant breeding PloS One (Accepted)
Crawford J T Shan CW Budsberg E Morgan H Bura R ampGus-tafson R (2016) Hydrocarbon bio‐jet fuel from bioconversion ofpoplar biomass Techno‐economic assessment Biotechnology forBiofuels 9 141 httpsdoiorg101186s13068-016-0545-7
28 | CLIFTON‐BROWN ET AL
Dalton S (2013) Biotechnology of Miscanthus In S M Jain amp SD Gupta (Eds) Biotechnology of neglected and underutilizedcrops (pp 243ndash294) Dordrecht Netherlands Springer
Davey C L Robson P Hawkins S Farrar K Clifton‐Brown J CDonnison I S amp Slavov G T (2017) Genetic relationshipsbetween spring emergence canopy phenology and biomass yieldincrease the accuracy of genomic prediction in Miscanthus Journalof Experimental Botany 68 5093ndash5102 httpsdoiorg101093jxberx339
David X amp Anderson X (2002) Aspen and larch genetics cooper-ative annual report 13 St Paul MN Department of ForestResources University of Minnesota
Deblock M (1990) Factors influencing the tissue‐culture and theAgrobacterium‐Tumefaciens‐mediated transformation of hybridaspen and poplar clones Plant Physiology 93 1110ndash1116httpsdoiorg101104pp9331110
Defra (2002) The role of future public research investment in thegenetic improvement of UK‐grown crops (p 220) LondonRetrieved from sciencesearchdefragovukDefaultaspxMenu=MenuampModule=MoreampLocation=NoneampCompleted=220ampProjectID=10412
Dewoody J Trewin H amp Taylor G (2015) Genetic and morpho-logical differentiation in Populus nigra L Isolation by coloniza-tion or isolation by adaptation Molecular Ecology 24 2641ndash2655
Dong H Liu S Clark L V Sharma S Gifford J M Juvik JA hellip Sacks E J (2018) Genetic mapping of biomass yield inthree interconnected Miscanthus populations Global Change Biol-ogy Bioenergy 10 165ndash185
Dukes (2017) Digest of UK Energy Statistics (DUKES) (p 264) Lon-don UK Department for Business Energy amp Industrial StrategyRetrieved from wwwgovukgovernmentcollectionsdigest-of-uk-energy-statistics-dukes
Eckenwalder J (1996) Systematics and evolution of Populus In RStettler H Bradshaw J Heilman amp T Hinckley (Eds) Biologyof Populus and its implications for management and conservation(pp 7‐32) Ottawa ON National Research Council of Canada
Elshire R J Glaubitz J C Sun Q Poland J A Kawamoto KBuckler E S amp Mitchell S E (2011) A robust simple geno-typing‐by‐sequencing (GBS) approach for high diversity speciesPloS One 6 e19379
Evans H (2016) Bioenergy crops in the UK Case studies of suc-cessful whole farm integration (p 23) Loughborough UKEnergy Technologies Institute Retrieved from httpswwweticouklibrarybioenergy-crops-in-the-uk-case-studies-on-successful-whole-farm-integration-evidence-pack
Evans H (2017) Increasing UK biomass production through moreproductive use of land (p 7) Loughborough UK Energy Tech-nologies Institute Retrieved from httpswwweticouklibraryan-eti-perspective-increasing-uk-biomass-production-through-more-productive-use-of-land
Evans J Sanciangco M D Lau K H Crisovan E Barry KDaum C hellip Buell C R (2018) Extensive genetic diversity ispresent within North American switchgrass germplasm The PlantGenome 11 1ndash16 httpsdoiorg103835plantgenome2017060055
Evans L M Slavov G T Rodgers‐Melnick E Martin J RanjanP Muchero W hellip DiFazio S P (2014) Population genomicsof Populus trichocarpa identifies signatures of selection and
adaptive trait associations Nature Genetics 46 1089ndash1096httpsdoiorg101038ng3075
Fabio E S Volk T A Miller R O Serapiglia M J Gauch HG Van Rees K C J hellip Smart L B (2017) Genotypetimes envi-ronment interaction analysis of North American shrub willowyield trials confirms superior performance of triploid hybrids Glo-bal Change Biology Bioenergy 9 445ndash459 httpsdoiorg101111gcbb12344
Fahrenkrog A M Neves L G Resende M F R Vazquez A Ide los Campos G Dervinis C hellip Kirst M (2017) Genome‐wide association study reveals putative regulators of bioenergytraits in Populus deltoides New Phytologist 213 799ndash811
Fan D Liu T T Li C F Jiao B Li S Hou Y S amp Luo KM (2015) Efficient CRISPRCas9‐mediated targeted mutagenesisin Populus in the first generation Scientific Reports 5 12217
Feng X P Lourgant K Castric V Saumitou-Laprade P ZhengB S Jiang D M amp Brancourt-Hulmel M (2014) The discov-ery of natural accessions related to Miscanthus x giganteus usingchloroplast DNA Crop Science 54 1645ndash1655
Fillatti J J Sellmer J Mccown B Haissig B amp Comai L(1987) Agrobacterium-mediated transformation and regenerationof Populus Molecular amp General Genetics 206 192ndash199
Fu Y B (2015) Understanding crop genetic diversity under modernplant breeding Theoretical and Applied Genetics 128 2131ndash2142 httpsdoiorg101007s00122-015-2585-y
Fu C Mielenz J R Xiao X Ge Y Hamilton C Y RodriguezM hellip Wang Z‐Y (2011) Genetic manipulation of ligninreduces recalcitrance and improves ethanol production fromswitchgrass Proceedings of the National Academy of Sciences ofthe United States of America 108 3803ndash3808 httpsdoiorg101073pnas1100310108
Gao C (2018) The future of CRISPR technologies in agricultureNature Reviews Molecular Cell Biology 19(5) 275ndash276
Gegas V Gay A Camargo A amp Doonan J (2014) Challengesof Crop Phenomics in the Post-genomic Era In J Hancock (Ed)Phenomics (pp 142ndash171) Boca Raton FL CRC Press
Geraldes A DiFazio S P Slavov G T Ranjan P Muchero WHannemann J hellip Tuskan G A (2013) A 34K SNP genotypingarray for Populus trichocarpa Design application to the study ofnatural populations and transferability to other Populus speciesMolecular Ecology Resources 13 306ndash323
Gifford J M Chae W B Swaminathan K Moose S P amp JuvikJ A (2015) Mapping the genome of Miscanthus sinensis forQTL associated with biomass productivity Global Change Biol-ogy Bioenergy 7 797ndash810
Goggin F L Lorence A amp Topp C N (2015) Applying high‐throughput phenotyping to plant‐insect interactions Picturingmore resistant crops Current Opinion in Insect Science 9 69ndash76httpsdoiorg101016jcois201503002
Grabowski P P Evans J Daum C Deshpande S Barry K WKennedy M hellip Casler M D (2017) Genome‐wide associationswith flowering time in switchgrass using exome‐capture sequenc-ing data New Phytologist 213 154ndash169 httpsdoiorg101111nph14101
Greef J M amp Deuter M (1993) Syntaxonomy of Miscanthus xgiganteus GREEF et DEU Angewandte Botanik 67 87ndash90
Guerra F P Richards J H Fiehn O Famula R Stanton B JShuren R hellip Neale D B (2016) Analysis of the genetic varia-tion in growth ecophysiology and chemical and metabolomic
CLIFTON‐BROWN ET AL | 29
composition of wood of Populus trichocarpa provenances TreeGenetics amp Genomes 12 6 httpsdoiorg101007s11295-015-0965-8
Guo H P Shao R X Hong C T Hu H K Zheng B S ampZhang Q X (2013) Rapid in vitro propagation of bioenergy cropMiscanthus sacchariflorus Applied Mechanics and Materials 260181ndash186
Hallingbaumlck H R Fogelqvist J Powers S J Turrion‐Gomez JRossiter R Amey J hellip Roumlnnberg‐Waumlstljung A‐C (2016)Association mapping in Salix viminalis L (Salicaceae)ndashIdentifica-tion of candidate genes associated with growth and phenologyGlobal Change Biology Bioenergy 8 670ndash685
Hanley S J amp Karp A (2014) Genetic strategies for dissectingcomplex traits in biomass willows (Salix spp) Tree Physiology34 1167ndash1180 httpsdoiorg101093treephystpt089
Harfouche A Meilan R amp Altman A (2011) Tree genetic engi-neering and applications to sustainable forestry and biomass pro-duction Trends in Biotechnology 29 9ndash17 httpsdoiorg101016jtibtech201009003
Harfouche A Meilan R amp Altman A (2014) Molecular and phys-iological responses to abiotic stress in forest trees and their rele-vance to tree improvement Tree Physiology 34 1181ndash1198httpsdoiorg101093treephystpu012
Harfouche A Meilan R Kirst M Morgante M Boerjan WSabatti M amp Mugnozza G S (2012) Accelerating the domesti-cation of forest trees in a changing world Trends in PlantScience 17 64ndash72 httpsdoiorg101016jtplants201111005
Hastings A Clifton‐Brown J Wattenbach M Mitchell C PStampfl P amp Smith P (2009) Future energy potential of Mis-canthus in Europe Global Change Biology Bioenergy 1 180ndash196
Hastings A Mos M Yesufu J AMccalmont J Schwarz KShafei RhellipClifton-Brown J (2017) Economic and environmen-tal assessment of seed and rhizome propagated Miscanthus in theUK Frontiers in Plant Science 8 16 httpsdoiorg103389fpls201701058
Heacuteder M (2017) From NASA to EU The evolution of the TRLscale in Public Sector Innovation The Innovation Journal 22 1ndash23 httpsdoiorg103389fpls201701058
Heffner E L Sorrells M E amp Jannink J L (2009) Genomicselection for crop improvement Crop Science 49 1ndash12 httpsdoiorg102135cropsci2008080512
Hodkinson T R Klaas M Jones M B Prickett R amp Barth S(2015) Miscanthus A case study for the utilization of naturalgenetic variation Plant Genetic Resources‐Characterization andUtilization 13 219ndash237 httpsdoiorg101017S147926211400094X
Hodkinson T R amp Renvoize S (2001) Nomenclature of Miscant-hus xgiganteus (Poaceae) Kew Bulletin 56 759ndash760 httpsdoiorg1023074117709
Houle D Govindaraju D R amp Omholt S (2010) Phenomics Thenext challenge Nature Reviews Genetics 11 855ndash866 httpsdoiorg101038nrg2897
Huang X H Wei X H Sang T Zhao Q Feng Q Zhao Yhellip Han B (2010) Genome‐wide association studies of 14 agro-nomic traits in rice landraces Nature Genetics 42 961ndashU976
Hwang O‐J Cho M‐A Han Y‐J Kim Y‐M Lim S‐H KimD‐S hellip Kim J‐I (2014) Agrobacterium‐mediated genetic trans-formation of Miscanthus sinensis Plant Cell Tissue and OrganCulture (PCTOC) 117 51ndash63
Hwang O‐J Lim S‐H Han Y‐J Shin A‐Y Kim D‐S ampKim J‐I (2014) Phenotypic characterization of transgenic Mis-canthus sinensis plants overexpressing Arabidopsis phytochromeB International Journal of Photoenergy 2014501016
Ilstedt B (1996) Genetics and performance of Belgian poplar clonestested in Sweden International Journal of Forest Genetics 3183ndash195
Ingvarsson P K Garcia M V Hall D Luquez V amp Jansson S(2006) Clinal variation in phyB2 a candidate gene for day‐length‐induced growth cessation and bud set across a latitudinalgradient in European aspen (Populus tremula) Genetics 1721845ndash1853
Ingvarsson P K Garcia M V Luquez V Hall D amp Jansson S(2008) Nucleotide polymorphism and phenotypic associationswithin and around the phytochrome B2 locus in European aspen(Populus tremula Salicaceae) Genetics 178 2217ndash2226 httpsdoiorg101534genetics107082354
Jannink J‐L Lorenz A J amp Iwata H (2010) Genomic selectionin plant breeding From theory to practice Briefings in FunctionalGenomics 9 166ndash177 httpsdoiorg101093bfgpelq001
Jiang J Guan Y Mccormick S Juvik J Lubberstedt T amp FeiS‐Z (2017) Gametophytic self‐incompatibility is operative inMiscanthus sinensis (Poaceae) and is affected by pistil age CropScience 57 1948ndash1956
Jones H D (2015a) Future of breeding by genome editing is in thehands of regulators GM Crops amp Food 6 223ndash232
Jones H D (2015b) Regulatory uncertainty over genome editingNature Plants 1 14011
Kalinina O Nunn C Sanderson R Hastings A F S van der Weijde TOumlzguumlven MhellipClifton‐Brown J C (2017) ExtendingMiscanthus cul-tivation with novel germplasm at six contrasting sites Frontiers in PlantScience 8563 httpsdoiorg103389fpls201700563
Karp A Hanley S J Trybush S O Macalpine W Pei M ampShield I (2011) Genetic improvement of willow for bioenergyand biofuels free access Journal of Integrative Plant Biology 53151ndash165 httpsdoiorg101111j1744-7909201001015x
Kersten B Faivre Rampant P Mader M Le Paslier M‐C Bou-non R Berard A hellip Fladung M (2016) Genome sequences ofPopulus tremula chloroplast and mitochondrion Implications forholistic poplar breeding PloS One 11 e0147209 httpsdoiorg101371journalpone0147209
Kiesel A Nunn C Iqbal Y Van der Weijde T Wagner MOumlzguumlven M hellip Lewandowski I (2017) Site‐specific manage-ment of Miscanthus genotypes for combustion and anaerobicdigestion A comparison of energy yields Frontiers in PlantScience 8 347 httpsdoiorg103389fpls201700347
Kim H Kim S T Ryu J Kang B C Kim J S amp Kim S G(2017) CRISPRCpf1‐mediated DNA‐free plant genome editingNature Communications 8 14406 httpsdoiorg101038ncomms14406
King Z R Bray A L Lafayette P R amp Parrott W A (2014)Biolistic transformation of elite genotypes of switchgrass (Pan-icum virgatum L) Plant Cell Reports 33 313ndash322 httpsdoiorg101007s00299-013-1531-1
Kopp R F Maynard C A De Niella P R Smart L B amp Abra-hamson L P (2002) Collection and storage of pollen from Salix(Salicaceae) American Journal of Botany 89 248ndash252
Krzyżak J Pogrzeba M Rusinowski S Clifton‐Brown J McCal-mont J P Kiesel A hellip Mos M (2017) Heavy metal uptake
30 | CLIFTON‐BROWN ET AL
by novel miscanthus seed‐based hybrids cultivated in heavy metalcontaminated soil Civil and Environmental Engineering Reports26 121ndash132 httpsdoiorg101515ceer-2017-0040
Kuzovkina Y A amp Quigley M F (2005) Willows beyond wet-lands Uses of Salix L species for environmental projects WaterAir and Soil Pollution 162 183ndash204 httpsdoiorg101007s11270-005-6272-5
Lazarus W Headlee W L amp Zalesny R S (2015) Impacts of sup-plyshed‐level differences in productivity and land costs on the eco-nomics of hybrid poplar production in Minnesota USA BioEnergyResearch 8 231ndash248 httpsdoiorg101007s12155-014-9520-y
Lewandowski I (2015) Securing a sustainable biomass supply in agrowing bioeconomy Global Food Security 6 34ndash42 httpsdoiorg101016jgfs201510001
Lewandowski I (2016) The role of perennial biomass crops in agrowing bioeconomy In S Barth D Murphy‐Bokern O Kalin-ina G Taylor amp M Jones (Eds) Perennial biomass crops for aresource‐constrained world (pp 3ndash13) Cham SwitzerlandSpringer Switzerland AG
Li J L Meilan R Ma C Barish M amp Strauss S H (2008)Stability of herbicide resistance over 8 years of coppice in field‐grown genetically engineered poplars Western Journal of AppliedForestry 23 89ndash93
Li R Y amp Qu R D (2011) High throughput Agrobacterium‐medi-ated switchgrass transformation Biomass amp Bioenergy 35 1046ndash1054 httpsdoiorg101016jbiombioe201011025
Lindegaard K N amp Barker J H A (1997) Breeding willows forbiomass Aspects of Applied Biology 49 155ndash162
Linde‐Laursen I B (1993) Cytogenetic analysis of MiscanthusGiganteus an interspecific hybrid Hereditas 119 297ndash300httpsdoiorg101111j1601-5223199300297x
Liu Y Merrick P Zhang Z Z Ji C H Yang B amp Fei S Z(2018) Targeted mutagenesis in tetraploid switchgrass (Panicumvirgatum L) using CRISPRCas9 Plant Biotechnology Journal16 381ndash393
Liu L L Wu Y Q Wang Y W amp Samuels T (2012) A high‐density simple sequence repeat‐based genetic linkage map ofswitchgrass G3 (Bethesda Md) 2 357ndash370
Lovett A Suumlnnenberg G amp Dockerty T (2014) The availabilityof land for perennial energy crops in Great Britain GlobalChange Biology Bioenergy 6 99ndash107 httpsdoiorg101111gcbb12147
Lozano‐Juste J amp Cutler S R (2014) Plant genome engineering infull bloom Trends in Plant Science 19 284ndash287 httpsdoiorg101016jtplants201402014
Lu F Lipka A E Glaubitz J Elshire R Cherney J H CaslerM D hellip Costich D E (2013) Switchgrass Genomic DiversityPloidy and Evolution Novel Insights from a Network‐Based SNPDiscovery Protocol Plos Genetics 9 e1003215
Ludovisi R Tauro F Salvati R Khoury S Mugnozza G S ampHarfouche A (2017) UAV‐based thermal imaging for high‐throughput field phenotyping of black poplar response to droughtFrontiers in Plant Science 81681
Macalpine W Shield I amp Karp A (2010) Seed to near market vari-ety the BEGIN willow breeding pipeline 2003ndash2010 and beyond InBioten conference proceedings Birmingham pp 21ndash23
Macalpine W J Shield I F Trybush S O Hayes C M ampKarp A (2008) Overcoming barriers to crossing in willow (Salixspp) breeding Aspects of Applied Biology 90 173ndash180
Mahfouz M M (2017) The efficient tool CRISPR‐Cpf1 NaturePlants 3 17028
Malinowska M Donnison I S amp Robson P R (2017) Phenomicsanalysis of drought responses in Miscanthus collected from differ-ent geographical locations Global Change Biology Bioenergy 978ndash91
Maroder H L Prego I A Facciuto G R amp Maldonado S B(2000) Storage behaviour of Salix alba and Salix matsudanaseeds Annals of Botany 86 1017ndash1021 httpsdoiorg101006anbo20001265
Martinez‐Reyna J amp Vogel K (2002) Incompatibility systems inswitchgrass Crop Science 42 1800ndash1805 httpsdoiorg102135cropsci20021800
Maurya J P amp Bhalerao R P (2017) Photoperiod‐and tempera-ture‐mediated control of growth cessation and dormancy in treesA molecular perspective Annals of Botany 120 351ndash360httpsdoiorg101093aobmcx061
McCalmont J P Hastings A Mcnamara N P Richter G MRobson P Donnison I S amp Clifton‐Brown J (2017) Environ-mental costs and benefits of growing Miscanthus for bioenergy inthe UK Global Change Biology Bioenergy 9 489ndash507
McCracken A R amp Dawson W M (1997) Using mixtures of wil-low clones as a means of controlling rust disease Aspects ofApplied Biology 49 97ndash103
McKown A D Klaacutepště J Guy R D Geraldes A Porth I Hanne-mann JhellipDouglas C J (2014) Genome‐wide association implicatesnumerous genes underlying ecological trait variation in natural popula-tions ofPopulus trichocarpaNew Phytologist 203 535ndash553
Merrick P amp Fei S Z (2015) Plant regeneration and genetic transforma-tion in switchgrass ndash A review Journal of Integrative Agriculture 14483ndash493 httpsdoiorg101016S2095-3119(14)60921-7
Moreno‐Mateos M A Fernandez J P Rouet R Lane M AVejnar C E Mis E hellip Giraldez A J (2017) CRISPR‐Cpf1mediates efficient homology‐directed repair and temperature‐con-trolled genome editing Nature Communications 8 2024 httpsdoiorg101038s41467-017-01836-2
Mosseler A (1990) Hybrid performance and species crossabilityrelationships in willows (Salix) Canadian Journal of Botany 682329ndash2338
Muchero W Guo J DiFazio S P Chen J‐G Ranjan P Slavov GT hellip Tuskan G A (2015) High‐resolution genetic mapping ofallelic variants associated with cell wall chemistry in Populus BMCGenomics 16 24 httpsdoiorg101186s12864-015-1215-z
Neale D B amp Kremer A (2011) Forest tree genomics Growingresources and applications Nature Reviews Genetics 12 111ndash122 httpsdoiorg101038nrg2931
Nunn C Hastings A F S J Kalinina O Oumlzguumlven M SchuumlleH Tarakanov I G hellip Clifton‐Brown J C (2017) Environmen-tal influences on the growing season duration and ripening ofdiverse Miscanthus germplasm grown in six countries Frontiersin Plant Science 8 907 httpsdoiorg103389fpls201700907
Ogawa Y Honda M Kondo Y amp Hara‐Nishimura I (2016) Anefficient Agrobacterium‐mediated transformation method forswitchgrass genotypes using Type I callus Plant Biotechnology33 19ndash26
Ogawa Y Shirakawa M Koumoto Y Honda M Asami Y KondoY ampHara‐Nishimura I (2014) A simple and reliable multi‐gene trans-formation method for switchgrass Plant Cell Reports 33 1161ndash1172httpsdoiorg101007s00299-014-1605-8
CLIFTON‐BROWN ET AL | 31
Okada M Lanzatella C Saha M C Bouton J Wu R L amp Tobias CM (2010) Complete switchgrass genetic maps reveal subgenomecollinearity preferential pairing and multilocus interactions Genetics185 745ndash760 httpsdoiorg101534genetics110113910
Palomo‐Riacuteos E Macalpine W Shield I Amey J Karaoğlu C WestJ hellip Jones H D (2015) Efficient method for rapid multiplication ofclean and healthy willow clones via in vitro propagation with broadgenotype applicability Canadian Journal of Forest Research 451662ndash1667 httpsdoiorg101139cjfr-2015-0055
Pilate G Allona I Boerjan W Deacutejardin A Fladung M Gal-lardo F hellip Halpin C (2016) Lessons from 25 years of GM treefield trials in Europe and prospects for the future In C Vettori FGallardo H Haumlggman V Kazana F Migliacci G Pilateamp MFladung (Eds) Biosafety of forest transgenic trees (pp 67ndash100)Dordrecht Netherlands Springer Switzerland AG
Pinosio S Giacomello S Faivre-Rampant P Taylor G Jorge VLe Paslier M C amp Marroni F (2016) Characterization of thepoplar pan-genome by genome-wide identification of structuralvariation Molecular Biology and Evolution 33 2706ndash2719
Porth I Klaacutepště J Skyba O Friedmann M C Hannemann JEhlting J hellip Douglas C J (2013) Network analysis reveals therelationship among wood properties gene expression levels andgenotypes of natural Populus trichocarpa accessions New Phytol-ogist 200 727ndash742
Pugesgaard S Schelde K Larsen S U Laeligrke P E amp JoslashrgensenU (2015) Comparing annual and perennial crops for bioenergyproductionndashinfluence on nitrate leaching and energy balance Glo-bal Change Biology Bioenergy 7 1136ndash1149 httpsdoiorg101111gcbb12215
Quinn L D Allen D J amp Stewart J R (2010) Invasivenesspotential of Miscanthus sinensis Implications for bioenergy pro-duction in the United States Global Change Biology Bioenergy2 310ndash320 httpsdoiorg101111j1757-1707201001062x
Rae A M Robinson K M Street N R amp Taylor G (2004)Morphological and physiological traits influencing biomass pro-ductivity in short‐rotation coppice poplar Canadian Journal ofForest Research‐Revue Canadienne De Recherche Forestiere 341488ndash1498 httpsdoiorg101139x04-033
Rambaud C Arnoult S Bluteau A Mansard M C Blassiau Camp Brancourt‐Hulmel M (2013) Shoot organogenesis in threeMiscanthus species and evaluation for genetic uniformity usingAFLP analysis Plant Cell Tissue and Organ Culture (PCTOC)113 437ndash448 httpsdoiorg101007s11240-012-0284-9
Ramstein G P Evans J Kaeppler S M Mitchell R B Vogel K PBuell C R amp Casler M D (2016) Accuracy of genomic prediction inswitchgrass (Panicum virgatum L) improved by accounting for linkagedisequilibriumG3 (Bethesda Md) 6 1049ndash1062
Rayburn A L Crawford J Rayburn C M amp Juvik J A (2009)Genome size of three Miscanthus species Plant Molecular Biol-ogy Reporter 27 184 httpsdoiorg101007s11105-008-0070-3
Richardson J Isebrands J amp Ball J (2014) Ecology and physiol-ogy of poplars and willows In J Isebrands amp J Richardson(Eds) Poplars and willows Trees for society and the environ-ment (pp 92‐123) Boston MA and Rome CABI and FAO
Robison T L Rousseau R J amp Zhang J (2006) Biomass produc-tivity improvement for eastern cottonwood Biomass amp Bioenergy30 735ndash739 httpsdoiorg101016jbiombioe200601012
Robson P R Farrar K Gay A P Jensen E F Clifton‐Brown JC amp Donnison I S (2013) Variation in canopy duration in the
perennial biofuel crop Miscanthus reveals complex associationswith yield Journal of Experimental Botany 64 2373ndash2383
Robson P Jensen E Hawkins S White S R Kenobi K Clif-ton‐Brown J hellip Farrar K (2013) Accelerating the domestica-tion of a bioenergy crop Identifying and modelling morphologicaltargets for sustainable yield increase in Miscanthus Journal ofExperimental Botany 64 4143ndash4155
Sanderson M A Adler P R Boateng A A Casler M D ampSarath G (2006) Switchgrass as a biofuels feedstock in theUSA Canadian Journal of Plant Science 86 1315ndash1325httpsdoiorg104141P06-136
Scarlat N Dallemand J F Monforti‐Ferrario F amp Nita V(2015) The role of biomass and bioenergy in a future bioecon-omy Policies and facts Environmental Development 15 3ndash34httpsdoiorg101016jenvdev201503006
Serapiglia M J Gouker F E amp Smart L B (2014) Early selec-tion of novel triploid hybrids of shrub willow with improved bio-mass yield relative to diploids BMC Plant Biology 14 74httpsdoiorg1011861471-2229-14-74
Serba D Wu L Daverdin G Bahri B A Wang X Kilian Ahellip Devos K M (2013) Linkage maps of lowland and upland tet-raploid switchgrass ecotypes BioEnergy Research 6 953ndash965httpsdoiorg101007s12155-013-9315-6
Shakoor N Lee S amp Mockler T C (2017) High throughput phe-notyping to accelerate crop breeding and monitoring of diseases inthe field Current Opinion in Plant Biology 38 184ndash192 httpsdoiorg101016jpbi201705006
Shield I Macalpine W Hanley S amp Karp A (2015) Breedingwillow for short rotation coppice energy cropping In Industrialcrops Breeding for BioEnergy and Bioproducts (pp 67ndash80) NewYork NY Springer
Sjodin A Street N R Sandberg G Gustafsson P amp Jansson S (2009)The Populus Genome Integrative Explorer (PopGenIE) A new resourcefor exploring the Populus genomeNewPhytologist 182 1013ndash1025
Slavov G amp Davey C (2017) Integrated genomic prediction inbioenergy crops In Abstracts from the International Conferenceon Developing biomass crops for future climates pp 34 24ndash27September 2017 Oriel College Oxford wwwwatbioeu
Slavov G Davey C Bosch M Robson P Donnison I ampMackay I (2018a) Genomic index selection provides a pragmaticframework for setting and refining multi‐objective breeding targetsin Miscanthus Annals of Botany (in press)
Slavov G T DiFazio S P Martin J Schackwitz W MucheroW Rodgers‐Melnick E hellip Tuskan G A (2012) Genome rese-quencing reveals multiscale geographic structure and extensivelinkage disequilibrium in the forest tree Populus trichocarpa NewPhytologist 196 713ndash725
Slavov G Davey C Robson P Donnison I amp Mackay I(2018b) Domestication of Miscanthus through genomic indexselection In Plant and Animal Genome Conference 13 January2018 San Diego CA Retrieved from httpspagconfexcompagxxvimeetingappcgiPaper30622
Slavov G T Nipper R Robson P Farrar K Allison G GBosch M hellip Jensen E (2013) Genome‐wide association studiesand prediction of 17 traits related to phenology biomass and cellwall composition in the energy grass Miscanthus sinensis NewPhytologist 201 1227ndash1239
Slavov G Robson P Jensen E Hodgson E Farrar K AllisonG hellip Donnison I (2014) Contrasting geographic patterns of
32 | CLIFTON‐BROWN ET AL
genetic variation for molecular markers vs phenotypic traits in theenergy grass Miscanthus sinensis Global Change Biology Bioen-ergy 5 562ndash571
Ślusarkiewicz‐Jarzina A Ponitka A Cerazy‐Waliszewska J Woj-ciechowicz M K Sobańska K Jeżowski S amp Pniewski T(2017) Effective and simple in vitro regeneration system of Mis-canthus sinensis Mtimes giganteus and M sacchariflorus for plant-ing and biotechnology purposes Biomass and Bioenergy 107219ndash226 httpsdoiorg101016jbiombioe201710012
Smart L amp Cameron K (2012) Shrub willow In C Kole S Joshiamp D Shonnard (Eds) Handbook of bioenergy crop plants (pp687ndash708) Boca Raton FL Taylor and Francis Group
Stanton B J (2014) The domestication and conservation of Populusand Salix genetic resources (Chapter 4) In Isebrands J ampRichardson J (Eds) Poplars and willows Trees for society andthe environment (pp 124ndash200) Rome Italy CAB InternationalFood and Agricultural Organization of the United Nations
Stanton B J Neale D B amp Li S (2010) Populus breeding Fromthe classical to the genomic approach In S Jansson R Bha-lerao amp A T Groover (Eds) Genetics and genomics of popu-lus (pp 309ndash348) New York NY Springer
Stewart J R Toma Y Fernandez F G Nishiwaki A Yamada T ampBollero G (2009) The ecology and agronomy ofMiscanthus sinensisa species important to bioenergy crop development in its native range inJapan A reviewGlobal Change Biology Bioenergy 1 126ndash153
Stott K G (1992) Willows in the service of man Proceedings of the RoyalSociety of Edinburgh Section B Biological Sciences 98 169ndash182
Strauss S H Ma C Ault K amp Klocko A L (2016) Lessonsfrom two decades of field trials with genetically modified trees inthe USA Biology and regulatory compliance In C Vettori FGallardo H Haumlggman V Kazana F Migliacci G Pilate amp MFladung (Eds) Biosafety of forest transgenic trees (pp 101ndash124)Dordrecht Netherlands Springer
Suda Y amp Argus G W (1968) Chromosome numbers of some NorthAmerican Salix Brittonia 20 191ndash197 httpsdoiorg1023072805440
Tan B (2018) Genomic selection and genome‐wide association studies todissect quantitative traits in forest trees Umearing Sweden University
Tornqvist C Taylor M Jiang Y Evans J Buell C KaepplerS amp Casler M (2018) Quantitative trait locus mapping for flow-ering time in a lowland x upland switchgrass pseudo‐F2 popula-tion The Plant Genome 11 170093
Toacuteth G Hermann T Da Silva M amp Montanarella L (2016)Heavy metals in agricultural soils of the European Union withimplications for food safety Environment International 88 299ndash309 httpsdoiorg101016jenvint201512017
Tracy W Dawson J Moore V amp Fisch J (2016) Intellectual propertyrights and public plant breeding Recommendations and proceedings ofa conference on best practices for intellectual property protection of pub-lically developed plant germplasm (p 70) University of Wisconsin‐Madison Retrieved from httpshostcalswisceduagronomywp-contentuploadssites1620162005Proceedings-IPR-Finalpdf
Trybush S Jahodovaacute Š Macalpine W amp Karp A (2008) A geneticstudy of a Salix germplasm resource reveals new insights into rela-tionships among subgenera sections and species BioEnergyResearch 1 67ndash79 httpsdoiorg101007s12155-008-9007-9
Tsai C J (2013) Next‐generation sequencing for next‐generationbreeding and more New Phytologist 198 635ndash637 httpsdoiorg101111nph12245
Tuskan G A Difazio S Jansson S Bohlmann J Grigoriev I HellstenU hellip Rokhsar D (2006) The genome of black cottonwood Populustrichocarpa (Torr ampGray) Science 313 1596ndash1604
Tuskan G A amp Walsh M E (2001) Short‐rotation woody cropsystems atmospheric carbon dioxide and carbon management AUS case study The Forestry Chronicle 77 259ndash264 httpsdoiorg105558tfc77259-2
Van Den Broek R Treffers D‐J Meeusen M Van Wijk ANieuwlaar E amp Turkenburg W (2001) Green energy or organicfood A life‐cycle assessment comparing two uses of set‐asideland Journal of Industrial Ecology 5 65ndash87 httpsdoiorg101162108819801760049477
Van Der Schoot J Pospiskova M Vosman B amp Smulders M J M(2000) Development and characterization of microsatellite markers inblack poplar (Populus nigra L) Theoretical and Applied Genetics 101317ndash322 httpsdoiorg101007s001220051485
Van Der Weijde T Dolstra O Visser R G amp Trindade L M(2017a) Stability of cell wall composition and saccharificationefficiency in Miscanthus across diverse environments Frontiers inPlant Science 7 2004
Van der Weijde T Huxley L M Hawkins S Sembiring E HFarrar K Dolstra O hellip Trindade L M (2017) Impact ofdrought stress on growth and quality of miscanthus for biofuelproduction Global Change Biology Bioenergy 9 770ndash782
Van der Weijde T Kamei C L A Severing E I Torres A FGomez L D Dolstra O hellip Trindade L M (2017) Geneticcomplexity of miscanthus cell wall composition and biomass qual-ity for biofuels BMC Genomics 18 406
Van der Weijde T Kiesel A Iqbal Y Muylle H Dolstra O Visser RG FhellipTrindade LM (2017) Evaluation ofMiscanthus sinensis bio-mass quality as feedstock for conversion into different bioenergy prod-uctsGlobal Change Biology Bioenergy 9 176ndash190
Vanholme B Cesarino I Goeminne G Kim H Marroni F VanAcker R hellip Boerjan W (2013) Breeding with rare defectivealleles (BRDA) A natural Populus nigra HCT mutant with modi-fied lignin as a case study New Phytologist 198 765ndash776
Vogel K P Mitchell R B Casler M D amp Sarath G (2014)Registration of Liberty Switchgrass Journal of Plant Registra-tions 8 242ndash247 httpsdoiorg103198jpr2013120076crc
Wang X Yamada T Kong F‐J Abe Y Hoshino Y Sato Hhellip Yamada T (2011) Establishment of an efficient in vitro cul-ture and particle bombardment‐mediated transformation systems inMiscanthus sinensis Anderss a potential bioenergy crop GlobalChange Biology Bioenergy 3 322ndash332 httpsdoiorg101111j1757-1707201101090x
Watrud L S Lee E H Fairbrother A Burdick C Reichman JR Bollman M hellip Van de Water P K (2004) Evidence forlandscape‐level pollen‐mediated gene flow from genetically modi-fied creeping bentgrass with CP4 EPSPS as a marker Proceedingsof the National Academy of Sciences of the United States of Amer-ica 101 14533ndash14538
Weisgerber H (1993) Poplar breeding for the purpose of biomassproduction in short‐rotation periods in Germany ndash Problems andfirst findings Forestry Chronicle 69 727ndash729 httpsdoiorg105558tfc69727-6
Wheeler N C Steiner K C Schlarbaum S E amp Neale D B(2015) The evolution of forest genetics and tree improvementresearch in the United States Journal of Forestry 113 500ndash510httpsdoiorg105849jof14-120
CLIFTON‐BROWN ET AL | 33
Whittaker C Macalpine W J Yates N E amp Shield I (2016)Dry matter losses and methane emissions during wood chip stor-age The impact on full life cycle greenhouse gas savings of shortrotation coppice willow for heat BioEnergy Research 9 820ndash835 httpsdoiorg101007s12155-016-9728-0
Xi Q (2000) Investigation on the distribution and potential of giant grassesin China PhD thesis Goettingen Germany Cuvillier Verlag
Xi Q amp Jezowkski S (2004) Plant resources of Triarrhena andMiscanthus species in China and its meaning for Europe PlantBreeding and Seed Science 49 63ndash77
Xue S Kalinina O amp Lewandowski I (2015) Present and future optionsfor the improvement of Miscanthus propagation techniques Renewableand Sustainable Energy Reviews 49 1233ndash1246
Yin K Q Gao C X amp Qiu J L (2017) Progress and prospectsin plant genome editing Nature Plants 3 httpsdoiorg101038nplants2017107
YookM (2016)Genetic diversity and transcriptome analysis for salt toler-ance inMiscanthus PhD thesis Seoul National University Korea
Zaidi S S E A Mahfouz M M amp Mansoor S (2017) CRISPR‐Cpf1 A new tool for plant genome editing Trends in PlantScience 22 550ndash553 httpsdoiorg101016jtplants201705001
Zalesny R S Stanturf J A Gardiner E S Perdue J H YoungT M Coyle D R hellip Hass A (2016) Ecosystem services ofwoody crop production systems BioEnergy Research 9 465ndash491 httpsdoiorg101007s12155-016-9737-z
Zhang Q X Sun Y Hu H K Chen B Hong C T Guo H Phellip Zheng B S (2012) Micropropagation and plant regenerationfrom embryogenic callus of Miscanthus sinensis Vitro Cellular ampDevelopmental Biology‐Plant 48 50ndash57 httpsdoiorg101007s11627-011-9387-y
Zhang Y Zalapa J E Jakubowski A R Price D L AcharyaA Wei Y hellip Casler M D (2011) Post‐glacial evolution of
Panicum virgatum Centers of diversity and gene pools revealedby SSR markers and cpDNA sequences Genetica 139 933ndash948httpsdoiorg101007s10709-011-9597-6
Zhao H Wang B He J Yang J Pan L Sun D amp Peng J(2013) Genetic diversity and population structure of Miscanthussinensis germplasm in China PloS One 8 e75672
Zhou X H Jacobs T B Xue L J Harding S A amp Tsai C J(2015) Exploiting SNPs for biallelic CRISPR mutations in theoutcrossing woody perennial Populus reveals 4‐coumarate CoAligase specificity and redundancy New Phytologist 208 298ndash301
Zhou R Macaya‐Sanz D Rodgers‐Melnick E Carlson C HGouker F E Evans L M hellip DiFazio S P (2018) Characteri-zation of a large sex determination region in Salix purpurea L(Salicaceae) Molecular Genetics and Genomics 1ndash16 httpsdoiorg101007s00438-018-1473-y
Zsuffa L Mosseler A amp Raj Y (1984) Prospects for interspecifichybridization in willow for biomass production In K L Perttu(Ed) Ecology and management of forest biomass production sys-tems Report 15 (pp 261ndash281) Uppsala Sweden SwedishUniversity of Agricultural Sciences
How to cite this article Clifton‐Brown J HarfoucheA Casler MD et al Breeding progress andpreparedness for mass‐scale deployment of perenniallignocellulosic biomass crops switchgrassmiscanthus willow and poplar GCB Bioenergy201800000ndash000 httpsdoiorg101111gcbb12566
34 | CLIFTON‐BROWN ET AL
Graphical AbstractThe contents of this page will be used as part of the graphical abstract of html only
It will not be published as part of main article
Plant breeding links the research effort with commercial mass upscaling The authorsrsquo assessment of development status ofthe four species is shown (poplar having two one for short rotation coppice (SRC) poplar and one for the more traditionalshort rotation forestry (SRF)) Mass scale deployment needs developments outside the breeding arenas to drive breedingactivities more rapidly and extensively
TABLE
4(Contin
ued)
Breeding
techno
logy
step
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sWillow
Poplar
IT~1
20
US
525genotypesin
eastside
hybrid
programme
205westside
hybrid
programmeand~1
200
in
SoutheastPdelto
ides
program
8Chrom
osom
e
doublin
g
Arouteto
triploid
seeded
typesdoublin
gadiploid
parent
ofknow
n
breeding
valuefrom
recurrentselection
Needs
1ndash7to
help
identify
therightparental
lines
Doubled
plantsarenotorious
forrevertingto
diploid
US
~50plantstakenfrom
tetraploid
tooctaploid
which
arenow
infieldtrials
CN~1
0Msactriploids
DKSeveralattemptsin
mid
90s
FR5genotypesof
Msin
UK2Msin1Mflora
nd1Msac
US
~30
UKAttempted
butnot
routinelyused
US
Not
partof
theregular
breeding
becausethere
existnaturalpolyploids
NA
9DoubleHaploids
Arouteto
producing
homogeneous
progeny
andalso
forthe
introductio
nof
transgenes
orforgenome
editing
Needs
1ndash7to
help
identify
therightparental
lines
Fully
homozygousplantsare
weakandeasily
die
andmay
notflow
ersynchronously
US
Noneyetattempteddueto
poor
vigour
andviability
of
haploids
CN2genotypesof
Msin
andMflor
UKDihaploidsof
MsinMflor
and
Msacand2transgenic
Msinvia
anther
cultu
re
US
6createdandused
toidentify
miss‐identifying
paralogueloci
(causedby
recent
genome
duplicationin
miscanthus)Usedin
creatin
gthereferencegenomefor
MsinVeryweakplantsand
difficultto
retain
thelin
es
NA
NA
10Embryo
rescue
Anattractiv
etechniquefor
recovering
plantsfrom
sexual
crosses
The
majority
ofem
bryos
cannot
survivein
vivo
or
becomelong‐timedorm
ant
None
UKone3times
hybrid
UKEmbryosrescue
protocol
proved
robustat
8days
post‐pollin
ation
FRAnem
bryo
rescue
protocol
proved
robustandim
proved
hybridizationsuccessfor
Pdelto
ides
timesPtrichocarpa
crosses
11Po
llenstorage
Flexibility
tocross
interestingparents
with
outneed
of
flow
ering
synchronization
None
Nosuccessto
daten
otoriously
difficultwith
grassesincluding
sugarcane
Freshpollencommonly
used
incrossing
Pollenstoragein
some
interspecificcrosses
FRBothstored
(cryobank)
and
freshindividual
pollen
Note
AFL
Pam
plifiedfragmentlength
polymorphismCBDconventio
non
biological
diversity
CP
controlledpollinatio
nCpD
NAchloroplastDNAEWBP
Europeanwillow
breeding
programme
FSfull‐sib
GtimesE
genotype‐by‐environm
entGRIN
germ
plasm
resourcesinform
ationnetwork
GWRGreenWoodResourcesMflorMiscanthusflo
ridulusMsacMsacchariflo
rus
MsinMsinensisNAnotapplicableNRRINatural
Resources
ResearchInstitu
teOP
open
pollinatio
nRADseq
restrictionsite‐associatedDNA
sequencingSL
USw
edishUniversity
ofAgriculturalSciencesST
TSw
eTreeTech
UMDUniversity
ofMinnesota
Duluth
a ISO
Alpha‐2
lettercountrycodes
10 | CLIFTON‐BROWN ET AL
programmes mature The average rate of gain for biomassyield in long‐term switchgrass breeding programmes hasbeen 1ndash4 per year depending on ecotype populationand location of the breeding programme (Casler amp Vogel2014 Casler et al 2018) The hybrid derivative Libertyhas a biomass yield 43 higher than the better of its twoparents (Casler amp Vogel 2014 Vogel et al 2014) Thedevelopment of cold‐tolerant and late‐flowering lowland‐ecotype populations for the northern United States hasincreased biomass yields by 27 (Casler et al 2018)
Currently more than 20 recurrent selection populationsare being managed in the United States to select parentsfor improved yield yield resilience and compositional qual-ity of the biomass For the agronomic development andupscaling high seed multiplication rates need to be com-bined with lower seed dormancy to reduce both crop estab-lishment costs and risks Expresso is the first cultivar withsignificantly reduced seed dormancy which is the first steptowards development of domesticated populations (Casleret al 2015) Most phenotypic traits of interest to breeders
require a minimum of 2 years to be fully expressed whichresults in a breeding cycle that is at least two years Morecomplicated breeding programmes or traits that requiremore time to evaluate can extend the breeding cycle to 4ndash8 years per generation for example progeny testing forbiomass yield or field‐based selection for cold toleranceBreeding for a range of traits with such long cycles callsfor the development of molecular methods to reduce time-scales and improve breeding efficiency
Two association panels of switchgrass have been pheno-typically and genotypically characterized to identify quanti-tative trait loci (QTLs) that control important biomasstraits The northern panel consists of 60 populationsapproximately 65 from the upland ecotype The southernpanel consists of 48 populations approximately 65 fromthe lowland ecotype Numerous QTLs have been identifiedwithin the northern panel to date (Grabowski et al 2017)Both panels are the subject of additional studies focused onbiomass quality flowering and phenology and cold toler-ance Additionally numerous linkage maps have been
FIGURE 1 Cumulative minimum years needed for the conventional breeding cycle through the steps from wild germplasm to thecommercial hybrids in switchgrass miscanthus willow and poplar Information links between the steps are indicated by dotted arrows andhighlight the importance of long‐term informatics to maximize breeding gain
CLIFTON‐BROWN ET AL | 11
created by the pairwise crossing of individuals with diver-gent characteristics often to generate four‐way crosses thatare analysed as pseudo‐F2 crosses (Liu Wu Wang ampSamuels 2012 Okada et al 2010 Serba et al 2013Tornqvist et al 2018) Individual markers and QTLs iden-tified can be used to design marker‐assisted selection(MAS) programmes to accelerate breeding and increase itsefficiency Genomic prediction and selection (GS) holdseven more promise with the potential to double or triplethe rate of gain for biomass yield and other highly complexquantitative traits of switchgrass (Casler amp Ramstein 2018Ramstein et al 2016) The genome of switchgrass hasrecently been made public through the Joint Genome Insti-tute (httpsphytozomejgidoegov)
Transgenic approaches have been heavily relied upon togenerate unique genetic variants principally for traitsrelated to biomass quality (Merrick amp Fei 2015) Switch-grass is highly transformable using either Agrobacterium‐mediated transformation or biolistics bombardment butregeneration of plants is the bottleneck to these systemsTraditionally plants from the cultivar Alamo were the onlyregenerable genotypes but recent efforts have begun toidentify more genotypes from different populations that arecapable of both transformation and subsequent regeneration(King Bray Lafayette amp Parrott 2014 Li amp Qu 2011Ogawa et al 2014 Ogawa Honda Kondo amp Hara‐Nishi-mura 2016) Cell wall recalcitrance and improved sugarrelease are the most common targets for modification (Bis-wal et al 2018 Fu et al 2011) Transgenic approacheshave the potential to provide traits that cannot be bredusing natural genetic variability However they will stillrequire about 10ndash15 years and will cost $70ndash100 millionfor cultivar development and deployment (Harfouche Mei-lan amp Altman 2011) In addition there is commercialuncertainty due to the significant costs and unpredictabletimescales and outcomes of the regulatory approval processin the countries targeted for seed sales As seen in maizeone advantage of transgenic approaches is that they caneasily be incorporated into F1 hybrid cultivars (Casler2012 Casler et al 2012) but this does not decrease thetime required for cultivar development due to field evalua-tion and seed multiplication requirements
The potential impacts of unintentional gene flow andestablishment of non‐native transgene sequences in nativeprairie species via cross‐pollination are also major issues forthe environmental risk assessment These limit further thecommercialization of varieties made using these technolo-gies Although there is active research into switchgrass steril-ity mechanisms to curb unintended pollen‐mediated genetransfer it is likely that the first transgenic cultivars proposedfor release in the United States will be met with considerableopposition due to the potential for pollen flow to remainingwild prairie sites which account for lt1 of the original
prairie land area and are highly protected by various govern-mental and nongovernmental organizations (Casler et al2015) Evidence for landscape‐level pollen‐mediated geneflow from genetically modified Agrostis seed multiplicationfields (over a mountain range) to pollinate wild relatives(Watrud et al 2004) confirms the challenge of using trans-genic approaches Looking ahead genome editing technolo-gies hold considerable promise for creating targeted changesin phenotype (Burris Dlugosz Collins Stewart amp Lena-ghan 2016 Liu et al 2018) and at least in some jurisdic-tions it is likely that cultivars resulting from gene editingwill not need the same regulatory approval as GMOs (Jones2015a) However in July 2018 the European Court of Justice(ECJ) ruled that cultivars carrying mutations resulting fromgene editing should be regulated in the same way as GMOsThe ECJ ruled that such cultivars be distinguished from thosearising from untargeted mutation breeding which isexempted from regulation under Directive 200118EC
3 | MISCANTHUS
Miscanthus is indigenous to eastern Asia and Oceania whereit is traditionally used for forage thatching and papermaking(Xi 2000 Xi amp Jezowkski 2004) In the 1960s the highbiomass potential of a Japanese genotype introduced to Eur-ope by Danish nurseryman Aksel Olsen in 1935 was firstrecognized in Denmark (Linde‐Laursen 1993) Later thisaccession was characterized described and named asldquoM times giganteusrdquo (Greef amp Deuter 1993 Hodkinson ampRenvoize 2001) commonly abbreviated as Mxg It is a natu-rally occurring interspecies triploid hybrid between tetraploidM sacchariflorus (2n = 4x) and diploid M sinensis(2n = 2x) Despite its favourable agronomic characteristicsand ability to produce high yields in a wide range of environ-ments in Europe (Kalinina et al 2017) the risks of relianceon it as a single clone have been recognized Miscanthuslike switchgrass is outcrossing due to self‐incompatibility(Jiang et al 2017) Thus seeded hybrids are an option forcommercial breeding Miscanthus can also be vegetativelypropagated by rhizome or in vitro culture which allows thedevelopment of clones The breeding approaches are usuallybased on F1 crosses and recurrent selection cycles within thesynthetic populations There are several breeding pro-grammes that target improvement of miscanthus traitsincluding stress resilience targeted regional adaptation agro-nomic ldquoscalabilityrdquo through cheaper propagation fasterestablishment lower moisture and ash contents and greaterusable yield (Clifton‐Brown et al 2017)
Germplasm collections specifically to support breedingfor biomass started in the late 1980s and early 1990s inDenmark Germany and the United Kingdom (Clifton‐Brown Schwarz amp Hastings 2015) These collectionshave continued with successive expeditions from European
12 | CLIFTON‐BROWN ET AL
and US teams assembling diverse collections from a widegeographic range in eastern Asia including from ChinaJapan South Korea Russia and Taiwan (Hodkinson KlaasJones Prickett amp Barth 2015 Stewart et al 2009) Threekey miscanthus species for biomass production are M si-nensis M floridulus and M sacchariflorus M sinensis iswidely distributed throughout eastern Asia with an adap-tive range from the subtropics to southern Russia (Zhaoet al 2013) This species has small rhizomes and producesmany tightly packed shoots forming a ldquotuftrdquo M floridulushas a more southerly adaptive range with a rather similarmorphology to M sinensis but grows taller with thickerstems and is evergreen and less cold‐tolerant than the othermiscanthus species M sacchariflorus is the most northern‐adapted species ranging to 50 degN in eastern Russia (Clarket al 2016) Populations of diploid and tetraploid M sac-chariflorus are found in China (Xi 2000) and South Korea(Yook 2016) and eastern Russia but only tetraploids havebeen found in Japan (Clark Jin amp Petersen 2018)
Germplasm has been assembled from multiple collec-tions over the last century though some early collectionsare poorly documented This historical germplasm has beenused to initiate breeding programmes largely based on phe-notypic and genotypic characterization As many of theaccessions from these collections are ldquoorigin unknownrdquocrucial environmental envelope data are not available UK‐led expeditions started in 2006 and continued until 2011with European and Asian partners and have built up a com-prehensive collection of 1500 accessions from 500 sitesacross Eastern Asia including China Japan South Koreaand Taiwan These collections were guided using spatialclimatic data to identify variation in abiotic stress toleranceAccessions from these recent collections were planted fol-lowing quarantine in multilocation nursery trials at severallocations in Europe to examine trait expression in differentenvironments Based on the resulting phenotypic andmolecular marker data several studies (a) characterized pat-terns of population genetic structure (Slavov et al 20132014) (b) evaluated the statistical power of genomewideassociation studies (GWASs) and identified preliminarymarkerndashtrait associations (Slavov et al 2013 2014) and(c) assessed the potential of genomic prediction (Daveyet al 2017 Slavov et al 2014 2018b) Genomic indexselection in particular offers the possibility of exploringscenarios for different locations or industrial markets (Sla-vov et al 2018a 2018b)
Separately US‐led expeditions also collected about1500 accessions between 2010 and 2014 (Clark et al2014 2016 2018 2015) A comprehensive genetic analy-sis of the population structure has been produced by RAD-seq for M sinensis (Clark et al 2015 Van der WeijdeKamei et al 2017) and M sacchariflorus (Clark et al2018) Multilocation replicated field trials have also been
conducted on these materials in North America and inAsia GWAS has been conducted for both M sinensis anda subset of M sacchariflorus accessions (Clark et al2016) To date about 75 of these recent US‐led collec-tions are in nursery trials outside the United States Due tolengthy US quarantine procedures these are not yet avail-able for breeding in the United States However molecularanalyses have allowed us to identify and prioritize sets ofgenotypes that best encompass the genetic variation in eachspecies
While mostM sinensis accessions flower in northern Eur-ope very few M sacchariflorus accessions flower even inheated glasshouses For this reason the European pro-grammes in the United Kingdom the Netherlands and Francehave performed mainly M sinensis (intraspecies) hybridiza-tions (Table 4) Selected progeny become the parents of latergenerations (recurrent selection as in switchgrass) Seed setsof up to 400 seed per panicle occur inM sinensis In Aberyst-wyth and Illinois significant efforts to induce synchronousflowering in M sacchariflorus and M sinensis have beenmade because interspecies hybrids have proven higher yieldperformance and wide adaptability (Kalinina et al 2017) Ininterspecies pairwise crosses in glasshouses breathable bagsandor large crossing tubes or chambers in which two or morewhole plants fit are used for pollination control Seed sets arelower in bags than in the open air because bags restrict pollenmovement while increasing temperatures and reducinghumidity (Clifton‐Brown Senior amp Purdy 2018) About30 of attempted crosses produced 10 to 60 seeds per baggedpanicle The seed (thousand seed mass ranges from 05 to09 g) is threshed from the inflorescences and sown into mod-ular trays to produce plug plants which are then planted infield nurseries to identify key parental combinations
A breeding programme of this scale must serve theneeds of different environments accepting the commonpurpose is to optimize the interception of solar radiationAn ideal hybrid for a given environment combines adapta-tion to date of emergence with optimization of traits suchas height number of stems per plant flowering and senes-cence time to optimize solar interception to produce a highbiomass yield with low moisture content at harvest (Rob-son Farrar et al 2013 Robson Jensen et al 2013) By20132014 conventional breeding in Europe had producedintra‐ and interspecific fertile seeded hybrids When acohort (typically about 5) of outstanding crosses have beenidentified it is important to work on related upscaling mat-ters in parallel These are as follows
Assessment of the yield and critical traits in selectedhybrids using a network of field trials
Efficient cloning of the seed parents While in vitro andmacro‐cloning techniques are used some genotypes areamenable to neither technique
CLIFTON‐BROWN ET AL | 13
TABLE5
Status
ofmod
ernplantbreeding
techniqu
esin
four
leadingperennialbiom
asscrop
s(PBCs)
Breeding
techno
logy
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sW
illow
Pop
lar
MAS
Use
ofmarker
sequ
encesthat
correlatewith
atrait
allowingearly
prog
enyselectionor
rejectionlocatin
gkn
owngenesfor
useful
traits(suchas
height)from
other
speciesin
your
crop
Breeding
prog
ramme
relevant
biparental
crosses(Table
4)to
create
ldquomapping
popu
latio
nsrdquoand
QTLidentification
andestim
ationto
identifyrobu
stmarkers
alternatively
acomprehensive
panelof
unrelated
geno
typesfor
GWAS
Ineffectivewhere
traits
areaffected
bymany
geneswith
small
effects
US
Literally
dozens
ofstud
iesto
identify
SNPs
andmarkers
ofinterestNoattempts
atMASas
yetlargely
dueto
popu
latio
nspecificity
CNS
SRISS
Rmarkers
forM
sinMsacand
Mflor
FR2
Msin
mapping
popu
latio
ns(BFF
project)
NL2
Msin
mapping
popu
latio
ns(A
tienza
Ram
irezetal
2003
AtienzaSatovic
PetersenD
olstraamp
Martin
200
3Van
Der
Weijdeetal20
17a
Van
derW
eijde
Hux
ley
etal20
17
Van
derW
eijdeKam
ei
etal20
17V
ander
WeijdeKieseletal
2017
)SK
2GWASpanels(1
collectionof
Msin1
diploidbiparentalF 1
popu
latio
n)in
the
pipelin
eUK1
2families
tofind
QTLin
common
indifferentfam
ilies
ofthe
sameanddifferent
speciesandhy
brids
US
2largeGWAS
panels4
diploidbi‐
parentalF 1
popu
latio
ns
1F 2
popu
latio
n(add
ition
alin
pipelin
e)
UK16
families
for
QTLdiscov
ery
2crossesMASscreened
1po
tentialvariety
selected
US
9families
geno
typedby
GBS(8
areF 1o
neisF 2)a
numberof
prom
ising
QTLdevelopm
entof
markers
inprog
ress
FR(INRA)tested
inlargeFS
families
and
infactorialmating
design
low
efficiency
dueto
family
specificity
US
49families
GS
Metho
dto
accelerate
breeding
throug
hredu
cing
the
resourcesforcross
attemptsby
Aldquotraining
popu
latio
nrdquoof
individu
alsfrom
stages
abov
ethat
have
been
both
Risks
ofpo
orpredictio
nProg
eny
testingneedsto
becontinueddu
ring
the
timethetraining
setis
US
One
prog
rammeso
farmdash
USD
Ain
WisconsinThree
cycles
ofgeno
mic
selectioncompleted
CN~1
000
geno
types
NLprelim
inary
mod
elsforMsin
developed
UK~1
000
geno
types
Verysuitable
application
not
attempted
yet
Maybe
challeng
ingfor
interspecies
hybrids
FR(INRA)120
0geno
type
ofPop
ulus
nigraarebeingused
todevelop
intraspecificGS
(Con
tinues)
14 | CLIFTON‐BROWN ET AL
TABLE
5(Contin
ued)
Breeding
techno
logy
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sW
illow
Pop
lar
predictin
gthe
performance
ofprog
enyof
crosses
(and
inresearch
topredictbestparents
tousein
biparental
crosses)
geno
typedand
phenotyp
edto
developamod
elthattakesgeno
typic
datafrom
aldquocandidate
popu
latio
nrdquoof
untested
individu
als
andprod
uces
GEBVs(Jannink
Lorenzamp
Iwata
2010
)
beingph
enotyp
edand
geno
typed
Inthis
time
next‐generation
germ
plasm
andreadyto
beginthe
second
roun
dof
training
and
recalib
ratio
nof
geno
mic
predictio
nmod
els
US
Prelim
inary
mod
elsforMsac
and
Msin
developed
Work
isun
derw
ayto
deal
moreeffectivelywith
polyploidgenetics
calib
ratio
nsUS
125
0geno
types
arebeingused
todevelopGS
calib
ratio
ns
Traditio
nal
transgenesis
Efficient
introd
uctio
nof
ldquoforeign
rdquotraits
(possiblyfrom
other
genu
sspecies)
into
anelite
plant(ega
prov
enparent
ora
hybrid)that
needsa
simpletraitto
beim
prov
edValidate
cand
idategenesfrom
QTLstud
ies
MASand
know
ledg
eof
the
biolog
yof
thetrait
andsource
ofgenesto
confer
the
relevant
changesto
phenotyp
eWorking
transformation
protocol
IPissuescostof
regu
latory
approv
al
GMO
labelling
marketin
gissuestricky
tousetransgenes
inou
t‐breedersbecause
ofcomplexity
oftransformingandgene
flow
risks
US
Manyprog
ramsin
UnitedStates
are
creatin
gtransgenic
plantsusingAlamoas
asource
oftransformable
geno
typesManytraits
ofinterestNothing
commercial
yet
CNC
hang
shaBtg
ene
transformed
into
Msac
in20
04M
sin
transformed
with
MINAC2gene
and
Msactransform
edwith
Cry2A
ageneb
othwith
amarkerg
eneby
Agrob
acterium
JP1
Msintransgenic
forlow
temperature
tolerancewith
increased
expression
offructans
(unp
ublished)
NLImprov
ingprotocol
forM
sin
transformation
SK2
Msin
with
Arabido
psisand
Brachypod
ium
PhytochromeBgenes
(Hwang
Cho
etal
2014
Hwang
Lim
etal20
14)
UK2
Msin
and
1Mflorg
enotyp
es
UKRou
tine
transformationno
tyet
possibleResearchto
overcomerecalcitrance
ongo
ing
Currently
trying
differentspecies
andcond
ition
sPo
plar
transformationused
atpresent
US
Frequently
attempted
with
very
limitedsuccess
US
600transgenic
lines
have
been
characterized
mostly
performed
inthe
aspenhy
brids
reprod
uctiv
esterility
drou
ghttolerance
with
Kno
ckdo
wns
(Con
tinues)
CLIFTON‐BROWN ET AL | 15
TABLE
5(Contin
ued)
Breeding
techno
logy
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sW
illow
Pop
lar
variou
slytransformed
with
four
cellwall
genesiptand
uidA
genesusingtwo
selectionsystem
sby
biolisticsand
Agrob
acterium
Transform
edplantsare
beinganalysed
US
Prelim
inarywork
will
betakenforw
ardin
2018
ndash202
2in
CABBI
Genom
eediting
CRISPR
Refinem
entofexisting
traits
inusefulparents
orprom
isinghybrids
bygeneratingtargeted
mutations
ingenes
know
ntocontrolthe
traitof
interestF
irst
adouble‐stranded
breakismadeinthe
DNAwhich
isrepairedby
natural
DNArepair
machineryL
eads
tofram
eshiftSN
Por
can
usealdquorepair
templaterdquo
orcanbe
used
toinserta
transgene
intoaldquosafe
harbourlocusrdquoIt
couldbe
used
todeleterepressors(or
transcriptionfactors)
Identification
mapping
and
sequ
encing
oftarget
genes(from
DNA
sequ
ence)
avoiding
screening
outof
unintend
ededits
Manyregu
latory
authorities
have
not
decidedwhether
CRISPR
andother
geno
meediting
techno
logies
are
GMOsor
notIf
GMOsseecomment
onstage10
abov
eIf
noteditedcrop
swill
beregu
latedas
conv
entio
nalvarieties
CATechn
olog
yisstill
toonewand
switchg
rass
geno
meis
very
complexOther
labo
ratories
are
interestedbu
tno
tyet
mov
ingon
this
US
One
prog
rammeso
farmdash
USD
Ain
Albany
FRInitiated
in20
16(M
ISEDIT
project)
NLInitiatingin
2018
US
Initiatingin
2018
in
CABBI
UKNot
started
Not
possible
until
transformation
achieved
US
CRISPR
‐Cas9and
Cpf1have
been
successfully
developedin
Pop
ulus
andarehigh
lyefficientExamples
forlig
nin
biosyn
thesis
Note
BFF
BiomassFo
rtheFu
tureBtBacillus
thuringiensisCABBICenterforadvanced
bioenergyandbioproductsinnovatio
nCpf1
CRISPR
from
Prevotella
andFrancisella
1CRISPR
clusteredregularlyinterspaced
palin
drom
icrepeatsCRISPR
‐CasC
RISPR
‐associated
Cry2A
acrystaltoxins2A
asubfam
ilyproduced
byBtFS
full‐sibG
BS
genotyping
bysequencingG
EBVsgenomic‐estim
ated
breeding
valuesG
MOg
enetically
modi-
fied
organism
GS
genomicselection
GWAS
genomew
ideassociationstudy
INRAF
renchNationalInstituteforAgriculturalR
esearch
IPintellectualp
roperty
IPTisopentenyltransferase
ISSR
inter‐SSR
MflorMiscant-
husflo
ridulusMsacMsacchariflo
rusMsinMsinensisM
AS
marker‐a
ssistedselection
MISEDITm
iscanthusgene
editing
forseed‐propagatedtriploidsNACn
oapicalmeristemA
TAF1
2and
cup‐shaped
cotyledon2
‐lik
eQTLq
uantitativ
etraitlocusS
NP
singlenucleotid
epolymorphismS
SRsim
plesequence
repeatsUSD
AU
SDepartm
ento
fAgriculture
a ISO
Alpha‐2
lettercountrycodes
16 | CLIFTON‐BROWN ET AL
High seed production from field crossing trials con-ducted in locations where flowering in both seed andpollen parents is likely to happen synchronously
Scalable and adapted harvesting threshing and seed pro-cessing methods for producing high seed quality
The results of these parallel activities need to becombined to identify the upscaling pathway for eachhybrid if this cannot be achieved the hybrid will likelynot be commercially viable The UK‐led programme withpartners in Italy and Germany shows that seedbased mul-tiplication rates of 12000 are achievable several inter-specific hybrids (Clifton‐Brown et al 2017) Themultiplication rate of M sinensis is higher probably15000ndash10000 Conventional cloning from rhizome islimited to around 120 that is one ha could provide rhi-zomes for around 20 ha of new plantation
Multilocation field testing of wild and novel miscanthushybrids selected by breeding programmes in the Nether-lands and the United Kingdom was performed as part ofthe project Optimizing Miscanthus Biomass Production(OPTIMISC 2012ndash2016) These trials showed that com-mercial yields and biomass qualities (Kiesel et al 2017Van der Weijde et al 2017a Van der Weijde Kieselet al 2017) could be produced in a wide range of climatesand soil conditions from the temperate maritime climate ofwestern Wales to the continental climate of eastern Russiaand the Ukraine (Kalinina et al 2017) Extensive environ-mental measurements of soil and climate combined withgrowth monitoring are being used to understand abioticstresses (Nunn et al 2017 Van der Weijde Huxley et al2017) and develop genotype‐specific scenarios similar tothose reported earlier in Hastings et al (2009) Phenomicsexperiments on drought tolerance have been conducted onwild and improved germplasm (Malinowska Donnison ampRobson 2017 Van der Weijde Huxley et al 2017)Recently produced interspecific hybrids displaying excep-tional yield under drought (~30 greater than control Mxg)in field trials in Poland and Moldova are being furtherstudied in detail in the phenomics and genomics facility atAberystwyth to better understand genendashtrait associationswhich can be fed back into breeding
Intraspecific seeded hybrids of M sinensis produced inthe Netherlands and interspecific M sacchariflorus times M si-nensis hybrids produced by the UK‐led breeding programmehave entered yield testing in 2018 with the recently EU‐funded project ldquoGRowing Advanced industrial Crops on mar-ginal lands for biorEfineries (GRACE)rdquo (httpswwwgrace-bbieu) Substantial variation in biomass quality for sacchari-fication efficiency (glucose release as of dry matter) ashcontent and melting point has already been generated inintraspecific M sinensis hybrids (Van der Weijde Kiesel
et al 2017) across environments (Weijde Dolstra et al2017a) GRACE aims to establish more than 20 hectares ofnew inter‐ and intraspecific seeded hybrids across six Euro-pean countries This project is building the know‐how andagronomy needed to transition from small research plots tocommercial‐scale field sites and linking biomass productiondirectly to industrial applications The biomass produced byhybrids in different locations will be supplied to innovativeindustrial end‐users making a wide range of biobased prod-ucts both for chemicals and for energy In the United Statesmulti‐location yield were initiated in 2018 to evaluate new tri-ploid M times giganteus genotypes developed at Illinois Cur-rently infertile hybrids are favoured in the United Statesbecause this eliminates the risk of invasiveness from naturallydispersed viable seed The precautionary principle is appliedas fertile miscanthus has naturalized in several states (QuinnAllen amp Stewart 2010) North European multilocation fieldtrials in the EMI and OPTIMISC projects have shown thereis minimal risk of invasiveness even in years when fertileflowering hybrids produce viable seed Naturalized standshave not established here due perhaps to low dormancy pooroverwintering and low seedling competitive strength In addi-tion to breeding for nonshattering or sterile seeded hybridsQuinn et al (2010) suggest management strategies which canfurther minimize environmental opportunities to manage therisk of invasiveness
31 | Molecular breeding and biotechnology
In miscanthus new plant breeding techniques (Table 5)have focussed on developing molecular markers forbreeding in Europe the United States South Korea andJapan There are several publications on QTL mappingpopulations for key traits such as flowering (AtienzaRamirez amp Martin 2003) and compositional traits(Atienza Satovic Petersen Dolstra amp Martin 2003) Inthe United States and United Kingdom independent andinterconnected bi‐parental ldquomappingrdquo families have beenstudied (Dong et al 2018 Gifford Chae SwaminathanMoose amp Juvik 2015) alongside panels of diverse germ-plasm accessions for GWAS (Slavov et al 2013) Fur-ther developments calibrating GS with very large panelsof parents and cross progeny are underway (Daveyet al 2017) The recently completed first miscanthus ref-erence genome sequence is expected to improve the effi-ciency of MAS strategies and especially GWAS(httpsphytozomejgidoegovpzportalhtmlinfoalias=Org_Msinensis_er) For example without a referencegenome sequence Clark et al (2014) obtained 21207RADseq SNPs (single nucleotide polymorphisms) on apanel of 767 miscanthus genotypes (mostly M sinensis)but subsequent reanalysis of the RADseq data using the
CLIFTON‐BROWN ET AL | 17
new reference genome resulted in hundreds of thousandsof SNPs being called
Robust and effective in vitro regeneration systems havebeen developed for Miscanthus sinensis M times giganteusand M sacchariflorus (Dalton 2013 Guo et al 2013Hwang Cho et al 2014 Rambaud et al 2013 Ślusarkie-wicz‐Jarzina et al 2017 Wang et al 2011 Zhang et al2012) However there is still significant genotype speci-ficity and these methods need ldquoin‐houserdquo optimization anddevelopment to be used routinely They provide potentialroutes for rapid clonal propagation and also as a basis forgenetic transformation Stable transformation using bothbiolistics (Wang et al 2011) and Agrobacterium tumefa-ciens DNA delivery methods (Hwang Cho et al 2014Hwang Lim et al 2014) has been achieved in M sinen-sis The development of miscanthus transformation andgene editing to generate diplogametes for producing seed‐propagated triploid hybrids are performed as part of theFrench project MISEDIT (miscanthus gene editing forseed‐propagated triploids) There are no reports of genomeediting in any miscanthus species but new breeding inno-vations including genome editing are particularly relevantin this slow‐to‐breed nonfood bioenergy crop (Table 4)
4 | WILLOW
Willow (Salix spp) is a very diverse group of catkin‐bear-ing trees and shrubs Willow belongs to the family Sali-caceae which also includes the Populus genus There areapproximately 350 willow species (Argus 2007) foundmostly in temperate and arctic zones in the northern hemi-sphere A few are adapted to subtropical and tropicalzones The centre of diversity is believed to be in Asiawith over 200 species in China Around 120 species arefound in the former Soviet Union over 100 in NorthAmerica and around 65 species in Europe and one speciesis native to South America (Karp et al 2011) Willows aredioecious thus obligate outcrossers and highly heterozy-gous The haploid chromosome number is 19 (Hanley ampKarp 2014) Around 40 of species are polyploid (Sudaamp Argus 1968) ranging from triploids to the atypicaldodecaploid S maxxaliana with 2n=190 (Zsuffa et al1984)
Although almost exclusively native to the NorthernHemisphere willow has been grown around the globe formany thousands of years to support a wide range of appli-cations (Kuzovkina amp Quigley 2005 Stott 1992) How-ever it has been the focus of domestication for bioenergypurposes for only a relatively short period since the 1970sin North America and Europe For bioenergy breedershave focused their efforts on the shrub willows (subgenusVetix) because of their rapid juvenile growth rates as aresponse to coppicing on a 2‐ to 4‐year cycle that can be
accomplished using farm machinery rather than forestryequipment (Shield Macalpine Hanley amp Karp 2015Smart amp Cameron 2012)
Since shrub willow was not generally recognized as anagricultural crop until very recently there has been littlecommitment to building and maintaining germplasm reposi-tories of willow to support long‐term breeding One excep-tion is the United Kingdom where a large and well‐characterized Salix germplasm collection comprising over1500 accessions is held at Rothamsted Research (Stott1992 Trybush et al 2008) Originally initiated for use inbasketry in 1923 accessions have been added ever sinceIn the United States a germplasm collection of gt350accessions is located at Cornell University to support thebreeding programme there The UK and Cornell collectionshave a relatively small number of accessions in common(around 20) Taken together they represent much of thespecies diversity but only a small fraction of the overallgenetic diversity within the genus There are three activewillow breeding programmes in Europe RothamstedResearch (UK) Salixenergi Europa AB (SEE) and a pro-gramme at the University of Warmia and Mazury in Olsz-tyn (Poland) (abbreviations used in Table 1) There is oneactive US programme based at Cornell University Culti-vars are still being marketed by the European WillowBreeding Programme (EWBP) (UK) which was activelybreeding biomass varieties from 1996 to 2002 Cultivarsare protected by plant breedersrsquo rights (PBRs) in Europeand by plant patents in the United States The sharing ofgenetic resources in the willow community is generallyregulated by material transfer agreements (MTA) and tai-lored licensing agreements although the import of cuttingsinto North America is prohibited except under special quar-antine permit conditions
Efforts to augment breeding germplasm collection fromnature are continuing with phenotypic screening of wildgermplasm performed in field experiments with 177 S pur-purea genotypes in the United States (at sites in Genevaand Portland NY and Morgantown WV) that have beengenotyped using genotyping by sequencing (GBS) (Elshireet al 2011) In addition there are approximately 400accessions of S viminalis in Europe (near Pustnaumls Upp-sala Sweden and Woburn UK (Berlin et al 2014 Hal-lingbaumlck et al 2016) The S viminalis accessions wereinitially genotyped using 38 simple sequence repeats (SSR)markers to assess genetic diversity and screened with~1600 SNPs in genes of potential interest for phenologyand biomass traits Genetics and genomics combined withextensive phenotyping have substantially improved thegenetic basis of biomass‐related traits in willow and arenow being developed in targeted breeding via MAS Thisunderpinning work has been conducted on large specifi-cally developed biparental Salix mapping populations
18 | CLIFTON‐BROWN ET AL
(Hanley amp Karp 2014 Zhou et al 2018) as well asGWAS panels (Hallingbaumlck et al 2016)
Once promising parental combinations are identifiedcrosses are usually performed using fresh pollen frommaterial that has been subject to a phased removal fromcold storage (minus4degC) (Lindegaard amp Barker 1997 Macal-pine Shield Trybush Hayes amp Karp 2008 Mosseler1990) Pollen storage is useful in certain interspecific com-binations where flowering is not naturally synchronizedThis can be overcome by using pollen collection and stor-age protocol which involves extracting pollen using toluene(Kopp Maynard Niella Smart amp Abrahamson 2002)
The main breeding approach to improve willow yieldsrelies on species hybridization to capture hybrid vigour(Fabio et al 2017 Serapiglia Gouker amp Smart 2014) Inthe absence of genotypic models for heterosis breedershave extensively tested general and specific combiningability of parents to produce superior progeny The UKbreeding programmes (EWBP 1996ndash2002 and RothamstedResearch from 2003 on) have performed more than 1500exploratory cross‐pollinations The Cornell programme hassuccessfully completed about 550 crosses since 1998Investment into the characterization of genetic diversitycombined with progeny tests from exploratory crosses hasbeen used to produce hundreds of targeted intraspeciescrosses in the United Kingdom and United States respec-tively (see Table 1) To achieve long‐term gains beyond F1hybrids four intraspecific recurrent selection populationshave been created in the United Kingdom (for S dasycla-dos S viminalis and S miyabeana) and Cornell is pursu-ing recurrent selection of S purpurea Interspecifichybridizations with genotypes selected from the recurrentselection cycles are well advanced in willow with suchcrosses to date totalling 420 in the United Kingdom andover 100 in the United States
While species hybridization is common in Salix it isnot universal Of the crosses attempted about 50 hybri-dize and produce seed (Macalpine Shield amp Karp 2010)As the viability of seed from successful crosses is short (amatter of days at ambient temperatures) proper seed rear-ing and storage protocols are essential (Maroder PregoFacciuto amp Maldonado 2000)
Progeny from crosses are treated in different waysamong the breeding programmes at the seedling stage Inthe United States seedlings are planted into an irrigatedfield where plants are screened for two seasons beforebeing progressed to further field trials In the United King-dom seedlings are planted into trays of compost wherethey remain containerized in an irrigated nursery for theremainder of year one In the United Kingdom seedlingsare subject to two rounds of selection in the nursery yearThe first round takes place in September to select againstsusceptibility to rust infection (Melampsora spp) A second
round of selection in winter assesses tip damage from frostand giant willow aphid infestation In the United Stateswhere the rust pressure is lower screening for Melampsoraspp cannot be performed at the nursery stage Both pro-grammes monitor Melampsora spp pest susceptibilityyield and architecture over multiple years in field trialsSelected material is subject to two rounds of field trials fol-lowed by a final multilocation yield trial to identify vari-eties for commercialization
Promising selections (ie potential cultivars) need to beclonally propagated A rapid in vitro tissue culture propa-gation method has been developed (Palomo‐Riacuteos et al2015) This method can generate about 5000 viable trans-plantable clones from a single plant in just 24 weeks Anin vitro system can also accommodate early selection viamolecular or biochemical markers to increase selectionspeed Conventional breeding systems take 13 years viafour rounds of selection from crossing to selecting a variety(Figure 1) but this has the potential to be reduced to7 years if micropropagation and MAS selection are adopted(Hanley amp Karp 2014 Palomo‐Riacuteos et al 2015)
Willows are currently propagated commercially byplanting winter‐dormant stem cuttings in spring Commer-cial planting systems for willow use mechanical plantersthat cut and insert stem sections from whips into a well‐prepared soil One hectare of stock plants grown in specificmultiplication beds planted at 40000 plants per ha pro-duces planting material for 80 hectares of commercialshort‐rotation coppice willow annually (planted at 15000cuttings per hectare) (Whittaker et al 2016) When com-mercial plantations are established the industry standard isto plant intimate mixtures of ~5 diverse rust (Melampsoraspp)‐resistant varieties (McCracken amp Dawson 1997 VanDen Broek et al 2001)
The foundations for using new plant breeding tech-niques have been established with funding from both thepublic and the private sectors To establish QTL maps 16mapping populations from biparental crosses are understudy in the United Kingdom Nine are under study in theUnited States The average number of individuals in thesefamilies ranges from 150 to 947 (Hanley amp Karp 2014)GS is also being evaluated in S viminalis and preliminaryresults indicate that multiomic approaches combininggenomic and metabolomic data have great potential (Sla-vov amp Davey 2017) For both QTL and GS approachesthe field phenotyping demands are large as several thou-sand individuals need to be phenotyped for a wide rangeof traits These include the following dates of bud burstand growth succession stem height stem density wooddensity and disease resistance The greater the number ofindividuals the more precise the QTL marker maps andGS models are However the logistical and financial chal-lenges of phenotyping large numbers of individuals are
CLIFTON‐BROWN ET AL | 19
considerable because the willow crop is gt5 m tall in thesecond year There is tremendous potential to improve thethroughput of phenotyping using unmanned aerial systemswhich is being tested in the USDA National Institute ofFood and Agriculture (NIFA) Willow SkyCAP project atCornell Further investment in these approaches needs tobe sustained over many years fully realizes the potentialof a marker‐assisted selection programme for willow
To date despite considerable efforts in Europe and theUnited States to establish a routine transformation systemthere has not been a breakthrough in willow but attemptsare ongoing As some form of transformation is typically aprerequisite for genome editing techniques these have notyet been applied to willow
In Europe there are 53 short‐rotation coppice (SRC) bio-mass willow cultivars registered with the Community PlantVariety Office (CPVO) for PBRs of which ~25 are availablecommercially in the United Kingdom There are eightpatented cultivars commercially available in the UnitedStates In Sweden there are nine commercial cultivars regis-tered in Europe and two others which are unregistered(httpssalixenergiseplanting-material) Furthermore thereare about 20 precommercial hybrids in final yield trials inboth the United States and the United Kingdom It has beenestimated that it would take two years to produce the stockrequired to plant 50 ha commercially from the plant stock inthe final yield trials Breeding programmes have alreadydelivered rust‐resistant varieties and increases in yield to themarket The adoption of advanced breeding technologies willlikely lead to a step change in improving traits of interest
5 | POPLAR
Poplar a fast‐growing tree from the northern hemispherewith a small genome size has been adopted for commercialforestry and scientific purposes The genus Populus con-sists of about 29 species classified in six different sectionsPopulus (formerly Leuce) Tacamahaca Aigeiros AbasoTuranga and Leucoides (Eckenwalder 1996) The Populusspecies of most interest for breeding and testing in the Uni-ted States and Europe are P nigra P deltoides P maxi-mowiczii and P trichocarpa (Stanton 2014) Populusclones for biomass production are being developed byintra‐ and interspecies hybridization (DeWoody Trewin ampTaylor 2015 Richardson Isebrands amp Ball 2014 van derSchoot et al 2000) Recurrent selection approaches areused for gradual population improvement and to create eliteclonal lines for commercialization (Berguson McMahon ampRiemenschneider 2017 Neale amp Kremer 2011) Currentlypoplar breeding in the United States occurs in industrialand academic programmes located in the Southeast theMidwest and the Pacific Northwest These use six speciesand five interspecific taxa (Stanton 2014)
The southeastern programme historically focused onrecurrent selection of P deltoides from accessions made inthe lower Mississippi River alluvial plain (Robison Rous-seau amp Zhang 2006) More recently the genetic base hasbeen broadened to produce interspecific hybrids with resis-tance to the fungal infection Septoria musiva which causescankers
In the midwest of the United States populationimprovement efforts are focused on P deltoides selectionsfrom native provenances and hybrid crosses with acces-sions introduced from Europe Interspecific intercontinental(Europe and America) hybrid crosses between P nigra andP deltoides (P times canadensis) are behind many of theleading commercial hybrids which are the most advancedbreeding materials for many applications and regions InMinnesota previous breeding experience and efforts utiliz-ing P maximowiczii and P trichocarpa have been discon-tinued due to Septoria susceptibility and a lack of coldhardiness (Berguson et al 2017) Traits targeted forimprovement include yieldgrowth rate cold hardinessadventitious rooting resistance to Septoria and Melamp-sora leaf rust and stem form The Upper Midwest pro-gramme also carries out wide hybridizations within thesection Populus The P times wettsteinii (P trem-ula times P tremuloides) taxon is bred for gains in growthrate wood quality and resistance to the fungus Entoleucamammata which causes hypoxylon canker (David ampAnderson 2002)
In the Pacific Northwest GreenWood Resources Incleads poplar breeding that emphasizes interspecifichybrid improvement of P times generosa (P deltoides timesP trichocarpa and reciprocal) and P deltoides times P maxi-mowiczii taxa for coastal regions and the P times canadensistaxon for the drier continental regions Intraspecificimprovement of second‐generation breeding populations ofP deltoides P nigra P maximowiczii and P trichocarpaare also involved (Stanton et al 2010) The present focusof GreenWood Resourcesrsquo hybridization is bioenergy feed-stock improvement concentrating on coppice yield wood‐specific gravity and rate of sugar release
Industrial interest in poplar in the United States has his-torically come from the pulp and paper sector althoughveneer and dimensional lumber markets have been pursuedat times Currently the biomass market for liquid trans-portation fuels is being emphasized along with the use oftraditional and improved poplar genotypes for ecosystemservices such as phytoremediation (Tuskan amp Walsh 2001Zalesny et al 2016)
In Europe there are breeding programmes in FranceGermany Italy and Sweden These include the following(a) Alasia Franco Vivai (AFV) programme in northernItaly (b) the French programme led by the poplar Scien-tific Interest Group (GIS Peuplier) and carried out
20 | CLIFTON‐BROWN ET AL
collaboratively between the National Institute for Agricul-tural Research (INRA) the National Research Unit ofScience and Technology for Environment and Agriculture(IRSTEA) and the Forest Cellulose Wood Constructionand Furniture Technology Institute (FCBA) (c) the Ger-man programme at Northwest German Forest Research Sta-tion (NW‐FVA) at Hannoversch Muumlnden and (d) theSwedish programme at the Swedish University of Agricul-tural Sciences and SweTree Technologies AB (Table 1)
AFV leads an Italian poplar breeding programme usingextensive field‐grown germplasm collections of P albaP deltoides P nigra and P trichocarpa While interspeci-fic hybridization uses several taxa the focus is onP times canadensis The breeding programme addresses dis-ease resistance (Marssonina brunnea Melampsora larici‐populina Discosporium populeum and poplar mosaicvirus) growth rate and photoperiod adaptation AFV andGreenWood Resources collaborate in poplar improvementin Europe through the exchange of frozen pollen and seedfor reciprocal breeding projects Plantations in Poland andRomania are currently the focus of the collaboration
The ongoing French GIS Peuplier is developing a long‐term breeding programme based on intraspecific recurrentselection for the four parental species (P deltoides P tri-chocarpa P nigra and P maximowiczii) designed to betterbenefit from hybrid vigour demonstrated by the interspeci-fic crosses P canadensis P deltoides times P trichocarpaand P trichocarpa times P maximowiczii Current selectionpriorities are targeting adaptation to soil and climate condi-tions resistance and tolerance to the most economicallyimportant diseases and pests high volume production underSRC and traditional poplar cultivation regimes as well aswood quality of interest by different markets Currentlygenomic selection is under exploration to increase selectionaccuracy and selection intensity while maintaining geneticdiversity over generations
The German NW‐FVA programme is breeding intersec-tional AigeirosndashTacamahaca hybrids with a focus on resis-tance to Pollaccia elegans Xanthomonas populiDothichiza spp Marssonina brunnea and Melampsoraspp (Stanton 2014) Various cross combinations ofP maximowiczii P trichocarpa P nigra and P deltoideshave led to new cultivars suitable for deployment in vari-etal mixtures of five to ten genotypes of complementarystature high productivity and phenotypic stability (Weis-gerber 1993) The current priority is the selection of culti-vars for high‐yield short‐rotation biomass production Sixhundred P nigra genotypes are maintained in an ex situconservation programme An in situ P nigra conservationeffort involves an inventory of native stands which havebeen molecular fingerprinted for identity and diversity
The Swedish programme is concentrating on locallyadapted genotypes used for short‐rotation forestry (SRF)
because these meet the needs of the current pulping mar-kets Several field trials have shown that commercial poplarclones tested and deployed in Southern and Central Europeare not well adapted to photoperiods and low temperaturesin Sweden and in the Baltics Consequently SwedishUniversity of Agricultural Sciences and SweTree Technolo-gies AB started breeding in Sweden in 1990s to producepoplar clones better adapted to local climates and markets
51 | Molecular breeding technologies
Poplar genetic improvement cannot be rapidly achievedthrough traditional methods alone because of the longbreeding cycles outcrossing breeding systems and highheterozygosity Integrating modern genetic genomic andphenomics techniques with conventional breeding has thepotential to expedite poplar improvement
The genome of poplar has been sequenced (Tuskanet al 2006) It has an estimated genome size of485 plusmn 10 Mbp divided into 19 chromosomes This is smal-ler than other PBCs and makes poplar more amenable togenetic engineering (transgenesis) GS and genome editingPoplar has seen major investment in both the United Statesand Europe being the model system for woody perennialplant genetics and genomics research
52 | Targets for genetic modification
Traits targeted include wood properties (lignin content andcomposition) earlylate flowering male sterility to addressbiosafety regulation issues enhanced yield traits and herbi-cide tolerance These extensive transgenic experiments haveshown differences in recalcitrance to in vitro regenerationand genetic transformation in some of the most importantcommercial hybrid poplars (Alburquerque et al 2016)Further transgene expression stability is being studied Sofar China is the only country known to have commerciallyused transgenic insect‐resistant poplar A precommercialherbicide‐tolerant poplar was trialled for 8 years in the Uni-ted States (Li Meilan Ma Barish amp Strauss 2008) butcould not be released due to stringent environmental riskassessments required for regulatory approval This increasestranslation costs and delays reducing investor confidencefor commercial deployment (Harfouche et al 2011)
The first field trials of transgenic poplar were performedin France in 1987 (Fillatti Sellmer Mccown Haissig ampComai 1987) and in Belgium in 1988 (Deblock 1990)Although there have been a total of 28 research‐scale GMpoplar field trials approved in the European Union underCouncil Directive 90220EEC since October 1991 (inPoland Belgium Finland France Germany Spain Swe-den and in the United Kingdom (Pilate et al 2016) onlyauthorizations in Poland and Belgium are in place today In
CLIFTON‐BROWN ET AL | 21
the United States regulatory notifications and permits fornearly 20000 transgenic poplar trees derived from approxi-mately 600 different constructs have been issued since1995 by the USDAs Animal and Plant Health InspectionService (APHIS) (Strauss et al 2016)
53 | Genome editing CRISPR technologies
Clustered regularly interspaced palindromic repeats(CRISPR) and the CRISPR‐associated (CRISPR‐Cas)nucleases are a groundbreaking genome‐engineering toolthat complements classical plant breeding and transgenicmethods (Moreno‐Mateos et al 2017) Only two publishedstudies in poplar have applied the CRISPRCas9 technol-ogy One is in P tomentosa in which an endogenous phy-toene desaturase gene (PtoPDS) was successfully disruptedsite specifically in the first generation of transgenic plantsresulting in an albino and dwarf phenotype (Fan et al2015) The second was in P tremula times alba in whichhigh CRISPR‐Cas9 mutational efficiency was achieved forthree 4‐coumarateCoA ligase (4Cl) genes 4CL1 4CL2and 4CL5 associated with lignin and flavonoid biosynthe-sis (Zhou et al 2015) Due to its low cost precision andrapidness it is very probable that cultivars or clones pro-duced using CRISPR technology will be ready for market-ing in the near future (Yin et al 2017) Recently aCRISPR with a smaller associated endonuclease has beendiscovered from Prevotella and Francisella 1 (Cpf1) whichmay have advantages over Cas9 In addition there arereports of DNA‐free editing in plants using both CRISPRCpf1 and CRISPR Cas9 for example Ref (Kim et al2017 Mahfouz 2017 Zaidi et al 2017)
It remains unresolved whether plants modified by genomeediting will be regulated as genetically modified organisms(GMOs) by the relevant authorities in different countries(Lozano‐Juste amp Cutler 2014) Regulations to cover thesenew breeding techniques are still evolving but those coun-tries who have published specific guidance (including UnitedStates Argentina and Chile) are indicating that plants pos-sessing simple genome edits will not be regulated as conven-tional transgenesis (Jones 2015b) The first generation ofgenome‐edited crops will likely be phenocopy gene knock-outs that already exist to produce ldquonature identicalrdquo traits thatis traits that could also be derived by conventional breedingDespite this confidence in applying these new powerfulbreeding tools remains limited owing to the uncertain regula-tory environment in many parts of the world (Gao 2018)including the recent ECJ 2018 rulings mentioned earlier
54 | Genomics‐based breeding technologies
Poplar breeding programs are becoming well equipped withuseful genomics tools and resources that are critical to
explore genomewide variability and make use of the varia-tion for enhancing genetic gains Deep transcriptomesequencing resequencing of alternate genomes and GBStechnology for genomewide marker detection using next‐generation sequencing (NGS) are yielding valuable geno-mics tools GWAS with NGS‐based markers facilitatesmarker identification for MAS breeding by design and GS
GWAS approaches have provided a deeper understand-ing of genome function as well as allelic architectures ofcomplex traits (Huang et al 2010) and have been widelyimplemented in poplar for wood characteristics (Porthet al 2013) stomatal patterning carbon gain versus dis-ease resistance (McKown et al 2014) height and phenol-ogy (Evans et al 2014) cell wall chemistry (Mucheroet al 2015) growth and cell walls traits (Fahrenkroget al 2017) bark roughness (Bdeir et al 2017) andheight and diameter growth (Liu et al 2018) Using high‐throughput sequencing and genotyping platforms an enor-mous amount of SNP markers have been used to character-ize the linkage disequilibrium (LD) in poplar (eg Slavovet al 2012 discussed below)
The genetic architecture of photoperiodic traits in peren-nial trees is complex involving many loci However itshows high levels of conservation during evolution (Mau-rya amp Bhalerao 2017) These genomics tools can thereforebe used to address adaptation issues and fine‐tune themovement of elite lines into new environments For exam-ple poor timing of spring bud burst and autumn bud setcan result in frost damage resulting in yield losses (Ilstedt1996) These have been studied in P tremula genotypesalong a latitudinal cline in Sweden (~56ndash66oN) and haverevealed high nucleotide polymorphism in two nonsynony-mous SNPs within and around the phytochrome B2 locus(Ingvarsson Garcia Hall Luquez amp Jansson 2006 Ing-varsson Garcia Luquez Hall amp Jansson 2008) Rese-quencing 94 of these P tremula genotypes for GWASshowed that noncoding variation of a single genomicregion containing the PtFT2 gene described 65 ofobserved genetic variation in bud set along the latitudinalcline (Tan 2018)
Resequencing genomes is currently the most rapid andeffective method detecting genetic differences betweenvariants and for linking loci to complex and importantagronomical and biomass traits thus addressing breedingchallenges associated with long‐lived plants like poplars
To date whole genome resequencing initiatives havebeen launched for several poplar species and genotypes InPopulus LD studies based on genome resequencing sug-gested the feasibility of GWAS in undomesticated popula-tions (Slavov et al 2012) This plant population is beingused to inform breeding for bioenergy development Forexample the detection of reliable phenotypegenotype asso-ciations and molecular signatures of selection requires a
22 | CLIFTON‐BROWN ET AL
detailed knowledge about genomewide patterns of allelefrequency variation LD and recombination suggesting thatGWAS and GS in undomesticated populations may bemore feasible in Populus than previously assumed Slavovet al (2012) have resequenced 16 genomes of P tri-chocarpa and genotyped 120 trees from 10 subpopulationsusing 29213 SNPs (Geraldes et al 2013) The largest everSNP data set of genetic variations in poplar has recentlybeen released providing useful information for breedinghttpswwwbioenergycenterorgbescgwasindexcfmAlso deep sequencing of transcriptomes using RNA‐Seqhas been used for identification of functional genes andmolecular markers that is polymorphism markers andSSRs A multitissue and multiple experimental data set forP trichocarpa RNA‐Seq is publicly available (httpsjgidoegovdoe-jgi-plant-flagship-gene-atlas)
The availability of genomic information of DNA‐con-taining cell organelles (nucleus chloroplast and mitochon-dria) will also allow a holistic approach in poplarmolecular breeding in the near future (Kersten et al2016) Complete Populus genome sequences are availablefor nucleus (P trichocarpa section Tacamahaca) andchloroplasts (seven species and two clones from P trem-ula W52 and P tremula times P alba 717ndash1B4) A compara-tive approach revealed structural and functionalinformation broadening the knowledge base of PopuluscpDNA and stimulating future diagnostic marker develop-ment The availability of whole genome sequences of thesecellular compartments of P tremula holds promise forboosting marker‐assisted poplar breeding Other nucleargenome sequences from additional Populus species arenow available (eg P deltoides (httpsphytozomejgidoegovpz) and will become available in the forthcomingyears (eg P tremula and P tremuloidesmdashPopGenIE(Sjodin Street Sandberg Gustafsson amp Jansson 2009))Recently the characterization of the poplar pan‐genome bygenomewide identification of structural variation in threecrossable poplar species P nigra P deltoides and P tri-chocarpa revealed a deeper understanding of the role ofinter‐ and intraspecific structural variants in poplar pheno-type and may have important implications for breedingparticularly interspecific hybrids (Pinosio et al 2016)
GS has been proposed as an alternative to MAS in cropimprovement (Bernardo amp Yu 2007 Heffner Sorrells ampJannink 2009) GS is particularly well suited for specieswith long generation times for characteristics that displaymoderate‐to‐low heritability for traits that are expensive tomeasure and for selection of traits expressed late in the lifecycle as is the case for most traits of commercial value inforestry (Harfouche et al 2012) Current joint genomesequencing efforts to implement GS in poplar using geno-mic‐estimated breeding values for bioenergy conversiontraits from 49 P trichocarpa families and 20 full‐sib
progeny are taking place at the Oak Ridge National Labo-ratory and GreenWood Resources (Brian Stanton personalcommunication httpscbiornlgov) These data togetherwith the resequenced GWAS population data will be thebasis for developing GS algorithms Genomic breedingtools have been developed for the intraspecific programmetargeting yield resistance to Venturia shoot blightMelampsora leaf rust resistance to Cryptorhynchus lapathistem form wood‐specific gravity and wind firmness (Evanset al 2014 Guerra et al 2016) A newly developedldquobreeding with rare defective allelesrdquo (BRDA) technologyhas been developed to exploit natural variation of P nigraand identify defective variants of genes predicted by priortransgenic research to impact lignin properties Individualtrees carrying naturally defective alleles can then be incor-porated directly into breeding programs thereby bypassingthe need for transgenics (Vanholme et al 2013) Thisnovel breeding technology offers a reverse genetics com-plement to emerging GS for targeted improvement of quan-titative traits (Tsai 2013)
55 | Phenomics‐assisted breeding technology
Phenomics involves the characterization of phenomesmdashthefull set of phenotypes of given individual plants (HouleGovindaraju amp Omholt 2010) Traditional phenotypingtools which inefficiently measure a limited set of pheno-types have become a bottleneck in plant breeding studiesHigh‐throughput plant phenotyping facilities provide accu-rate screening of thousands of plant breeding lines clones orpopulations over time (Fu 2015) are critical for acceleratinggenomics‐based breeding Automated image collection andanalysis phenomics technologies allow accurate and nonde-structive measurements of a diversity of phenotypic traits inlarge breeding populations (Gegas Gay Camargo amp Doo-nan 2014 Goggin Lorence amp Topp 2015 Ludovisi et al2017 Shakoor Lee amp Mockler 2017) One important con-sideration is the identification of relevant and quantifiabletarget traits that are early diagnostic indicators of biomassyield Good progress has been made in elucidating theseunderpinning morpho‐physiological traits that are amenableto remote sensing in Populus (Harfouche Meilan amp Altman2014 Rae Robinson Street amp Taylor 2004) Morerecently Ludovisi et al (2017) developed a novel methodol-ogy for field phenomics of drought stress in a P nigra F2partially inbred population using thermal infrared imagesrecorded from an unmanned aerial vehicle‐based platform
Energy is the current main market for poplar biomass butthe market return provided is not sufficient to support pro-duction expansion even with added demand for environmen-tal and land management ldquoecosystem servicesrdquo such as thetreatment of effluent phytoremediation riparian buffer zonesand agro‐forestry plantings Aviation fuel is a significant
CLIFTON‐BROWN ET AL | 23
target market (Crawford et al 2016) To serve this marketand to reduce current carbon costs of production (Budsberget al 2016) key improvement traits in addition to yield(eg coppice regeneration pestdisease resistance water‐and nutrient‐use efficiencies) will be trace greenhouse gas(GHG) emissions (eg isoprene volatiles) site adaptabilityand biomass conversion efficiency Efforts are underway tohave national environmental protection agenciesrsquo approvalfor poplar hybrids qualifying for renewable energy credits
6 | REFLECTIONS ON THECOMMERCIALIZATIONCHALLENGE
The research and innovation activities reviewed in thispaper aim to advance the genetic improvement of speciesthat can provide feedstocks for bioenergy applicationsshould those markets eventually develop These marketsneed to generate sufficient revenue and adequately dis-tribute it to the actors along the value chain The work onall four crops shares one thing in common long‐termefforts to integrate fundamental knowledge into breedingand crop development along a research and development(RampD) pipeline The development of miscanthus led byAberystwyth University exemplifies the concerted researcheffort that has integrated the RampD activities from eight pro-jects over 14 years with background core research fundingalong an emerging innovation chain (Figure 2) This pro-gramme has produced a first range of conventionally bredseeded interspecies hybrids which are now in upscaling tri-als (Table 3) The application of molecular approaches(Table 5) with further conventional breeding (Table 4)offers the prospect of a second range of improved seededhybrids This example shows that research‐based support ofthe development of new crops or crop types requires along‐term commitment that goes beyond that normallyavailable from project‐based funding (Figure 1) Innovationin this sector requires continuous resourcing of conven-tional breeding operations and capability to minimize timeand investment losses caused by funding discontinuities
This challenge is increased further by the well‐knownmarket failure in the breeding of many agricultural cropspecies The UK Department for Environment Food andRural Affairs (Defra) examined the role of geneticimprovement in relation to nonmarket outcomes such asenvironmental protection and concluded that public invest-ment in breeding was required if profound market failure isto be addressed (Defra 2002) With the exception ofwidely grown hybrid crops such as maize and some high‐value horticultural crops royalties arising from plant breed-ersrsquo rights or other returns to breeders fail to adequatelycompensate for the full cost for research‐based plant breed-ing The result even for major crops such as wheat is sub‐
optimal investment and suboptimal returns for society Thismarket failure is especially acute for perennial crops devel-oped for improved sustainability rather than consumerappeal (Tracy et al 2016) Figure 3 illustrates the underly-ing challenge of capturing value for the breeding effortThe ldquovalley of deathrdquo that results from the low and delayedreturns to investment applies generally to the research‐to‐product innovation pipeline (Beard Ford Koutsky amp Spi-wak 2009) and certainly to most agricultural crop speciesHowever this schematic is particularly relevant to PBCsMost of the value for society from the improved breedingof these crops comes from changes in how agricultural landis used that is it depends on the increased production ofthese crops The value for society includes many ecosys-tems benefits the effects of a return to seminatural peren-nial crop cover that protects soils the increase in soilcarbon storage the protection of vulnerable land or the cul-tivation of polluted soils and the reductions in GHG emis-sions (Lewandowski 2016) By its very nature theproduction of biomass on agricultural land marginal forfood production challenges farm‐level profitability Thecosts of planting material and one‐time nature of cropestablishment are major early‐stage costs and therefore theopportunities for conventional royalty capture by breedersthat are manifold for annual crops are limited for PBCs(Hastings et al 2017) Public investment in developingPBCs for the nonfood biobased sector needs to providemore long‐term support for this critical foundation to a sus-tainable bioeconomy
7 | CONCLUSIONS
This paper provides an overview of research‐based plantbreeding in four leading PBCs For all four PBC generasignificant progress has been made in genetic improvementthrough collaboration between research scientists and thoseoperating ongoing breeding programmes Compared withthe main food crops most PBC breeding programmes dateback only a few decades (Table 1) This breeding efforthas thus co‐evolved with molecular biology and the result-ing ‐omics technologies that can support breeding Thedevelopment of all four PBCs has depended strongly onpublic investment in research and innovation The natureand driver of the investment varied In close associationwith public research organizations or universities all theseprogrammes started with germplasm collection and charac-terization which underpin the selection of parents forexploratory wide crosses for progeny testing (Figure 1Table 4)
Public support for switchgrass in North America wasexplicitly linked to plant breeding with 12 breeding pro-grammes supported in the United States and CanadaSwitchgrass breeding efforts to date using conventional
24 | CLIFTON‐BROWN ET AL
breeding have resulted in over 36 registered cultivars inthe United States (Table 1) with the development of dedi-cated biomass‐type cultivars coming within the past fewyears While ‐omics technologies have been incorporatedinto several of these breeding programmes they have notyet led to commercial deployment in either conventional orhybrid cultivars
Willow genetic improvement was led by the researchcommunity closely linked to plant breeding programmesWillow and poplar have the longest record of public invest-ment in genetic improvement that can be traced back to1920s in the United Kingdom and United States respec-tively Like switchgrass breeding programmes for willoware connected to public research efforts The United King-dom in partnership with the programme based at CornellUniversity remains the European leader in willowimprovement with a long‐term breeding effort closelylinked to supporting biological research at Rothamsted Inwillow F1 hybrids have produced impressive yield gainsover parental germplasm by capturing hybrid vigour Over30 willow clones are commercially available in the UnitedStates and Europe and a further ~90 are under precommer-cial testing (Table 1)
Compared with willow and poplar miscanthus is a rela-tive newcomer with all the current breeding programmesstarting in the 2000s Clonal M times giganteus propagated byrhizomes is expected to be replaced by more readily scal-able seeded hybrids from intra‐ (M sinensis) and inter‐(M sacchariflorus times M sinensis) species crosses with highseed multiplication rates (of gt2000) The first group ofhybrid cultivars is expected to be market‐ready around2022
Of the four genera used as PBCs Populus is the mostadvanced in terms of achievements in biological researchas a result of its use as a model for basic research of treesMuch of this biological research is not directly connectedto plant breeding Nevertheless reflecting the fact thatpoplar is widely grown as a single‐stem tree in SRF thereare about 60 commercially available clones and an addi-tional 80 clones in commercial pipelines (Table 1) Trans-genic poplar hybrids have moved beyond proof of conceptto commercial reality in China
Many PBC programmes have initiated long‐term con-ventional recurrent selection breeding cycles for populationimprovement which is a key process in increasing yieldthrough hybrid vigour As this approach requires manyyears most programmes are experimenting with molecularbreeding methods as these have the potential to accelerateprecision breeding For all four PBCs investments in basicgenetic and genomic resources including the developmentof mapping populations for QTLs and whole genomesequences are available to support long‐term advancesMore recently association genetics with panels of diversegermplasm are being used as training populations for GSmodels (Table 5) These efforts are benefitting from pub-licly available DNA sequences and whole genome assem-blies in crop databases Key to these accelerated breedingtechnologies are developments in novel phenomics tech-nologies to bridge the genotypephenotype gap In poplarnovel remote sensing field phenotyping is now beingdeployed to assist breeders These advances are being com-bined with in vitro and in planta modern molecular breed-ing techniques such as CRISPR (Table 5) CRISPRtechnology for genome editing has been proven in poplar
FIGURE 2 A schematic developmentpathway for miscanthus in the UnitedKingdom related to the investment in RampDprojects at Aberystwyth (top coloured areasfor projects in the three categories basicresearch breeding and commercialupscaling) leading to a projected croppingarea of 350000 ha by 2030 with clonal andsuccessive ranges of improved seed‐basedhybrids Purple represents theBiotechnology and Biological SciencesResearch Council (BBSRC) and brown theDepartment for Environment Food andRural Affairs (Defra) (UK Nationalfunding) blue bars represent EU fundingand green private sector funding (Terravestaand CERES) and GIANT‐LINK andMiscanthus Upscaling Technology (MUST)are publicndashprivate‐initiatives (PPI)
CLIFTON‐BROWN ET AL | 25
This technology is also being applied in switchgrass andmiscanthus (Table 5) but the future of CRISPR in com-mercial breeding for the European market is uncertain inthe light of recent ECJ 2018 rulings
There is integration of research and plant breeding itselfin all four PBCs Therefore estimating the ongoing costs ofmaintaining these breeding programmes is difficult Invest-ment in research also seeks wider benefits associated withtechnological advances in plant science rather than cultivardevelopment per se However in all cases the conventionalbreeding cycle shown in Figure 1 is the basic ldquoenginerdquo withmolecular technologies (‐omics) serving to accelerate thisengine The history of the development shows that the exis-tence of these breeding programmes is essential to gain ben-efits from the biological research Despite this it is thisessential step that is at most risk from reductions in invest-ment A conventional breeding programme typicallyrequires a breeder and several technicians who are supported
over the long term (20ndash30 years Figure 1) at costs of about05 to 10 million Euro per year (as of 2018) The analysisreported here shows that the time needed to perform onecycle of conventional breeding bringing germplasm fromthe wild to a commercial hybrid ranged from 11 years inswitchgrass to 26 years in poplar (Figure 1) In a maturecrop grown on over 100000 ha with effective cultivar pro-tection and a suitable business model this level of revenuecould come from royalties Until such levels are reachedPBCs lie in the innovation valley of death (Figure 3) andneed public support
Applying industrial ldquotechnology readiness levelsrdquo (TRL)originally developed for aerospace (Heacuteder 2017) to ourplant breeding efforts we estimate many promising hybridscultivars are at TRL levels of 3ndash4 In Table 3 experts ineach crop estimate that it would take 3 years from now toupscale planting material from leading cultivars in plot tri-als to 100 ha
FIGURE 3 A schematic relating some of the steps in the innovation chain from relatively basic crop science research through to thedeployment in commercial cropping systems and value chains The shape of the funnel above the expanding development and deploymentrepresents the availability of investment along the development chain from relatively basic research at the top to the upscaled deployment at thebottom Plant breeding links the research effort with the development of cropping systems The constriction represents the constrained fundingfor breeding that links conventional public research investment and the potential returns from commercial development The handover pointsbetween publicly funded work to develop the germplasm resources (often known as prebreeding) the breeding and the subsequent cropdevelopment are shown on the left The constriction point is aggravated by the lack academic rewards for this essential breeding activity Theoutcome is such that this innovation system is constrained by the precarious resourcing of plant breeding The authorsrsquo assessment ofdevelopment status of the four species is shown (poplar having two one for short‐rotation coppice (SRC) poplar and one for the more traditionalshort‐rotation forestry (SRF)) The four new perennial biomass crops (PBCs) are now in the critical phase of depending of plant breedingprogress without the income stream from a large crop production base
26 | CLIFTON‐BROWN ET AL
Taking the UK example mentioned in the introductionplanting rates of ~35000 ha per year from 2022 onwardsare needed to reach over 1 m ha by 2050 Ongoing workin the UK‐funded Miscanthus Upscaling Technology(MUST) project shows that ramping annual hybrid seedproduction from the current level of sufficient seed for10 ha in 2018 to 35000 ha would take about 5 yearsassuming no setbacks If current hybrids of any of the fourPBCs in the upscaling pipeline fail on any step for exam-ple lower than expected multiplication rates or unforeseenagronomic barriers then further selections from ongoingbreeding are needed to replace earlier candidates
In conclusion the breeding foundations have been laidwell for switchgrass miscanthus willow and poplar owingto public funding over the long time periods necessaryImproved cultivars or genotypes are available that could bescaled up over a few years if real sustained market oppor-tunities emerged in response to sustained favourable poli-cies and industrial market pull The potential contributionsof growing and using these PBCs for socioeconomic andenvironmental benefits are clear but how farmers andothers in commercial value chains are rewarded for mass‐scale deployment as is necessary is not obvious at presentTherefore mass‐scale deployment of these lignocellulosecrops needs developments outside the breeding arenas todrive breeding activities more rapidly and extensively
8 | UNCITED REFERENCE
Huang et al (2019)
ACKNOWLEDGEMENTS
UK The UK‐led miscanthus research and breeding wasmainly supported by the Biotechnology and BiologicalSciences Research Council (BBSRC) Department for Envi-ronment Food and Rural Affairs (Defra) the BBSRC CSPstrategic funding grant BBCSP17301 Innovate UKBBSRC ldquoMUSTrdquo BBN0161491 CERES Inc and Ter-ravesta Ltd through the GIANT‐LINK project (LK0863)Genomic selection and genomewide association study activi-ties were supported by BBSRC grant BBK01711X1 theBBSRC strategic programme grant on Energy Grasses ampBio‐refining BBSEW10963A01 The UK‐led willow RampDwork reported here was supported by BBSRC (BBSEC00005199 BBSEC00005201 BBG0162161 BBE0068331 BBG00580X1 and BBSEC000I0410) Defra(NF0424 NF0426) and the Department of Trade and Indus-try (DTI) (BW6005990000) IT The Brain Gain Program(Rientro dei cervelli) of the Italian Ministry of EducationUniversity and Research supports Antoine Harfouche USContributions by Gerald Tuskan to this manuscript were sup-ported by the Center for Bioenergy Innovation a US
Department of Energy Bioenergy Research Center supportedby the Office of Biological and Environmental Research inthe DOE Office of Science under contract number DE‐AC05‐00OR22725 Willow breeding efforts at CornellUniversity have been supported by grants from the USDepartment of Agriculture National Institute of Food andAgriculture Contributions by the University of Illinois weresupported primarily by the DOE Office of Science Office ofBiological and Environmental Research (BER) grant nosDE‐SC0006634 DE‐SC0012379 and DE‐SC0018420 (Cen-ter for Advanced Bioenergy and Bioproducts Innovation)and the Energy Biosciences Institute EU We would like tofurther acknowledge contributions from the EU projectsldquoOPTIMISCrdquo FP7‐289159 on miscanthus and ldquoWATBIOrdquoFP7‐311929 on poplar and miscanthus as well as ldquoGRACErdquoH2020‐EU326 Biobased Industries Joint Technology Ini-tiative (BBI‐JTI) Project ID 745012 on miscanthus
CONFLICT OF INTEREST
The authors declare that progress reported in this paperwhich includes input from industrial partners is not biasedby their business interests
ORCID
John Clifton‐Brown httporcidorg0000-0001-6477-5452Antoine Harfouche httporcidorg0000-0003-3998-0694Lawrence B Smart httporcidorg0000-0002-7812-7736Danny Awty‐Carroll httporcidorg0000-0001-5855-0775VasileBotnari httpsorcidorg0000-0002-0470-0384Lindsay V Clark httporcidorg0000-0002-3881-9252SalvatoreCosentino httporcidorg0000-0001-7076-8777Lin SHuang httporcidorg0000-0003-3079-326XJon P McCalmont httporcidorg0000-0002-5978-9574Paul Robson httporcidorg0000-0003-1841-3594AnatoliiSandu httporcidorg0000-0002-7900-448XDaniloScordia httporcidorg0000-0002-3822-788XRezaShafiei httporcidorg0000-0003-0891-0730Gail Taylor httporcidorg0000-0001-8470-6390IrisLewandowski httporcidorg0000-0002-0388-4521
REFERENCES
Alburquerque N Baldacci‐Cresp F Baucher M Casacuberta JM Collonnier C ElJaziri M hellip Burgos L (2016) New trans-formation technologies for trees In C Vettori F Gallardo H
CLIFTON‐BROWN ET AL | 27
Haumlggman V Kazana F Migliacci G Pilateamp M Fladung(Eds) Biosafety of forest transgenic trees (pp 31ndash66) DordrechtNetherlands Springer
Argus G W (2007) Salix (Salicaceae) distribution maps and a syn-opsis of their classification in North America north of MexicoHarvard Papers in Botany 12 335ndash368 httpsdoiorg1031001043-4534(2007)12[335SSDMAA]20CO2
Atienza S G Ramirez M C amp Martin A (2003) Mapping‐QTLscontrolling flowering date in Miscanthus sinensis Anderss CerealResearch Communications 31 265ndash271
Atienza S G Satovic Z Petersen K K Dolstra O amp Martin A(2003) Identification of QTLs influencing agronomic traits inMiscanthus sinensis Anderss I Total height flag‐leaf height andstem diameter Theoretical and Applied Genetics 107 123ndash129
Atienza S G Satovic Z Petersen K K Dolstra O amp Martin A(2003) Influencing combustion quality in Miscanthus sinensisAnderss Identification of QTLs for calcium phosphorus and sul-phur content Plant Breeding 122 141ndash145
Bdeir R Muchero W Yordanov Y Tuskan G A Busov V ampGailing O (2017) Quantitative trait locus mapping of Populusbark features and stem diameter BMC Plant Biology 17 224httpsdoiorg101186s12870-017-1166-4
Beard T R Ford G S Koutsky T M amp Spiwak L J (2009) AValley of Death in the innovation sequence An economic investi-gation Research Evaluation 18 343ndash356 httpsdoiorg103152095820209X481057
Berguson W McMahon B amp Riemenschneider D (2017) Addi-tive and non‐additive genetic variances for tree growth in severalhybrid poplar populations and implications regarding breedingstrategy Silvae Genetica 66(1) 33ndash39 httpsdoiorg101515sg-2017-0005
Berlin S Trybush S O Fogelqvist J Gyllenstrand N Halling-baumlck H R Aringhman I hellip Hanley S J (2014) Genetic diversitypopulation structure and phenotypic variation in European Salixviminalis L (Salicaceae) Tree Genetics amp Genomes 10 1595ndash1610 httpsdoiorg101007s11295-014-0782-5
Bernardo R amp Yu J (2007) Prospects for genomewide selectionfor quantitative traits in maize Crop Science 47 1082ndash1090httpsdoiorg102135cropsci2006110690
Biswal A K Atmodjo M A Li M Baxter H L Yoo C GPu Y hellip Mohnen D (2018) Sugar release and growth of bio-fuel crops are improved by downregulation of pectin biosynthesisNature Biotechnology 36 249ndash257
Bmu (2009) National biomass action plan for Germany (p 17) Bun-desministerium fuumlr Umwelt Naturschutz und ReaktorsicherheitBundesministerium fuumlr Ernaumlhrung (BMU) amp Landwirtschaft undVerbraucherschutz (BMELV) 11055 Berlin | Germany Retrievedfrom httpwwwbmeldeSharedDocsDownloadsENPublicationsBiomassActionPlanpdf__blob=publicationFile
Brummer E C (1999) Capturing heterosis in forage crop cultivardevelopment Crop Science 39 943ndash954 httpsdoiorg102135cropsci19990011183X003900040001x
Budsberg E Crawford J T Morgan H Chin W S Bura R ampGustafson R (2016) Hydrocarbon bio‐jet fuel from bioconver-sion of poplar biomass Life cycle assessment Biotechnology forBiofuels 9 170 httpsdoiorg101186s13068-016-0582-2
Burris K P Dlugosz E M Collins A G Stewart C N amp Lena-ghan S C (2016) Development of a rapid low‐cost protoplasttransfection system for switchgrass (Panicum virgatum L) Plant
Cell Reports 35 693ndash704 httpsdoiorg101007s00299-015-1913-7
Casler M (2012) Switchgrass breeding genetics and genomics InA Monti (Ed) Switchgrass (pp 29ndash54) London UK Springer
Casler M Mitchell R amp Vogel K (2012) Switchgrass In CKole C P Joshi amp D R Shonnard (Eds) Handbook of bioen-ergy crop plants Vol 2 New York NY Taylor amp Francis
Casler M D amp Ramstein G P (2018) Breeding for biomass yieldin switchgrass using surrogate measures of yield BioEnergyResearch 11 6ndash12
Casler M D Sosa S Hoffman L Mayton H Ernst C Adler PR hellip Bonos S A (2017) Biomass Yield of Switchgrass Culti-vars under High‐ versus Low‐Input Conditions Crop Science 57821ndash832 httpsdoiorg102135cropsci2016080698
Casler M D amp Vogel K P (2014) Selection for biomass yield inupland lowland and hybrid switchgrass Crop Science 54 626ndash636 httpsdoiorg102135cropsci2013040239
Casler M D Vogel K P amp Harrison M (2015) Switchgrassgermplasm resources Crop Science 55 2463ndash2478 httpsdoiorg102135cropsci2015020076
Casler M D Vogel K P Lee D K Mitchell R B Adler P RSulc R M hellip Moore K J (2018) 30 Years of progress towardincreasing biomass yield of switchgrass and big bluestem CropScience 58 1242
Clark L V Brummer J E Głowacka K Hall M C Heo K PengJ hellip Sacks E J (2014) A footprint of past climate change on thediversity and population structure of Miscanthus sinensis Annals ofBotany 114 97ndash107 httpsdoiorg101093aobmcu084
Clark L V Dzyubenko E Dzyubenko N Bagmet L SabitovA Chebukin P hellip Sacks E J (2016) Ecological characteristicsand in situ genetic associations for yield‐component traits of wildMiscanthus from eastern Russia Annals of Botany 118 941ndash955
Clark L V Jin X Petersen K K Anzoua K G Bagmet LChebukin P hellip Sacks E J(2018) Population structure of Mis-canthus sacchariflorus reveals two major polyploidization eventstetraploid‐mediated unidirectional introgression from diploid Msinensis and diversity centred around the Yellow Sea Annals ofBotany 20 1ndash18 httpsdoiorg101093aobmcy1161
Clark L V Stewart J R Nishiwaki A Toma Y Kjeldsen J BJoslashrgensen U hellip Sacks E J (2015) Genetic structure ofMiscanthus sinensis and Miscanthus sacchariflorus in Japan indi-cates a gradient of bidirectional but asymmetric introgressionJournal of Experimental Botany 66 4213ndash4225
Clifton‐Brown J Hastings A Mos M McCalmont J P Ashman CAwty‐Carroll DhellipCracroft‐EleyW (2017) Progress in upscalingMis-canthus biomass production for the European bio‐economy with seed‐based hybridsGlobal Change Biology Bioenergy 9 6ndash17
Clifton‐Brown J Schwarz K U amp Hastings A (2015) History ofthe development of Miscanthus as a bioenergy crop From smallbeginnings to potential realisation Biology and Environment‐Pro-ceedings of the Royal Irish Academy 115B 45ndash57
Clifton‐Brown J Senior H Purdy S Horsnell R Lankamp BMuumlennekhoff A-K hellip Bentley A R (2018) Investigating thepotential of novel non‐woven fabrics to increase pollination effi-ciency in plant breeding PloS One (Accepted)
Crawford J T Shan CW Budsberg E Morgan H Bura R ampGus-tafson R (2016) Hydrocarbon bio‐jet fuel from bioconversion ofpoplar biomass Techno‐economic assessment Biotechnology forBiofuels 9 141 httpsdoiorg101186s13068-016-0545-7
28 | CLIFTON‐BROWN ET AL
Dalton S (2013) Biotechnology of Miscanthus In S M Jain amp SD Gupta (Eds) Biotechnology of neglected and underutilizedcrops (pp 243ndash294) Dordrecht Netherlands Springer
Davey C L Robson P Hawkins S Farrar K Clifton‐Brown J CDonnison I S amp Slavov G T (2017) Genetic relationshipsbetween spring emergence canopy phenology and biomass yieldincrease the accuracy of genomic prediction in Miscanthus Journalof Experimental Botany 68 5093ndash5102 httpsdoiorg101093jxberx339
David X amp Anderson X (2002) Aspen and larch genetics cooper-ative annual report 13 St Paul MN Department of ForestResources University of Minnesota
Deblock M (1990) Factors influencing the tissue‐culture and theAgrobacterium‐Tumefaciens‐mediated transformation of hybridaspen and poplar clones Plant Physiology 93 1110ndash1116httpsdoiorg101104pp9331110
Defra (2002) The role of future public research investment in thegenetic improvement of UK‐grown crops (p 220) LondonRetrieved from sciencesearchdefragovukDefaultaspxMenu=MenuampModule=MoreampLocation=NoneampCompleted=220ampProjectID=10412
Dewoody J Trewin H amp Taylor G (2015) Genetic and morpho-logical differentiation in Populus nigra L Isolation by coloniza-tion or isolation by adaptation Molecular Ecology 24 2641ndash2655
Dong H Liu S Clark L V Sharma S Gifford J M Juvik JA hellip Sacks E J (2018) Genetic mapping of biomass yield inthree interconnected Miscanthus populations Global Change Biol-ogy Bioenergy 10 165ndash185
Dukes (2017) Digest of UK Energy Statistics (DUKES) (p 264) Lon-don UK Department for Business Energy amp Industrial StrategyRetrieved from wwwgovukgovernmentcollectionsdigest-of-uk-energy-statistics-dukes
Eckenwalder J (1996) Systematics and evolution of Populus In RStettler H Bradshaw J Heilman amp T Hinckley (Eds) Biologyof Populus and its implications for management and conservation(pp 7‐32) Ottawa ON National Research Council of Canada
Elshire R J Glaubitz J C Sun Q Poland J A Kawamoto KBuckler E S amp Mitchell S E (2011) A robust simple geno-typing‐by‐sequencing (GBS) approach for high diversity speciesPloS One 6 e19379
Evans H (2016) Bioenergy crops in the UK Case studies of suc-cessful whole farm integration (p 23) Loughborough UKEnergy Technologies Institute Retrieved from httpswwweticouklibrarybioenergy-crops-in-the-uk-case-studies-on-successful-whole-farm-integration-evidence-pack
Evans H (2017) Increasing UK biomass production through moreproductive use of land (p 7) Loughborough UK Energy Tech-nologies Institute Retrieved from httpswwweticouklibraryan-eti-perspective-increasing-uk-biomass-production-through-more-productive-use-of-land
Evans J Sanciangco M D Lau K H Crisovan E Barry KDaum C hellip Buell C R (2018) Extensive genetic diversity ispresent within North American switchgrass germplasm The PlantGenome 11 1ndash16 httpsdoiorg103835plantgenome2017060055
Evans L M Slavov G T Rodgers‐Melnick E Martin J RanjanP Muchero W hellip DiFazio S P (2014) Population genomicsof Populus trichocarpa identifies signatures of selection and
adaptive trait associations Nature Genetics 46 1089ndash1096httpsdoiorg101038ng3075
Fabio E S Volk T A Miller R O Serapiglia M J Gauch HG Van Rees K C J hellip Smart L B (2017) Genotypetimes envi-ronment interaction analysis of North American shrub willowyield trials confirms superior performance of triploid hybrids Glo-bal Change Biology Bioenergy 9 445ndash459 httpsdoiorg101111gcbb12344
Fahrenkrog A M Neves L G Resende M F R Vazquez A Ide los Campos G Dervinis C hellip Kirst M (2017) Genome‐wide association study reveals putative regulators of bioenergytraits in Populus deltoides New Phytologist 213 799ndash811
Fan D Liu T T Li C F Jiao B Li S Hou Y S amp Luo KM (2015) Efficient CRISPRCas9‐mediated targeted mutagenesisin Populus in the first generation Scientific Reports 5 12217
Feng X P Lourgant K Castric V Saumitou-Laprade P ZhengB S Jiang D M amp Brancourt-Hulmel M (2014) The discov-ery of natural accessions related to Miscanthus x giganteus usingchloroplast DNA Crop Science 54 1645ndash1655
Fillatti J J Sellmer J Mccown B Haissig B amp Comai L(1987) Agrobacterium-mediated transformation and regenerationof Populus Molecular amp General Genetics 206 192ndash199
Fu Y B (2015) Understanding crop genetic diversity under modernplant breeding Theoretical and Applied Genetics 128 2131ndash2142 httpsdoiorg101007s00122-015-2585-y
Fu C Mielenz J R Xiao X Ge Y Hamilton C Y RodriguezM hellip Wang Z‐Y (2011) Genetic manipulation of ligninreduces recalcitrance and improves ethanol production fromswitchgrass Proceedings of the National Academy of Sciences ofthe United States of America 108 3803ndash3808 httpsdoiorg101073pnas1100310108
Gao C (2018) The future of CRISPR technologies in agricultureNature Reviews Molecular Cell Biology 19(5) 275ndash276
Gegas V Gay A Camargo A amp Doonan J (2014) Challengesof Crop Phenomics in the Post-genomic Era In J Hancock (Ed)Phenomics (pp 142ndash171) Boca Raton FL CRC Press
Geraldes A DiFazio S P Slavov G T Ranjan P Muchero WHannemann J hellip Tuskan G A (2013) A 34K SNP genotypingarray for Populus trichocarpa Design application to the study ofnatural populations and transferability to other Populus speciesMolecular Ecology Resources 13 306ndash323
Gifford J M Chae W B Swaminathan K Moose S P amp JuvikJ A (2015) Mapping the genome of Miscanthus sinensis forQTL associated with biomass productivity Global Change Biol-ogy Bioenergy 7 797ndash810
Goggin F L Lorence A amp Topp C N (2015) Applying high‐throughput phenotyping to plant‐insect interactions Picturingmore resistant crops Current Opinion in Insect Science 9 69ndash76httpsdoiorg101016jcois201503002
Grabowski P P Evans J Daum C Deshpande S Barry K WKennedy M hellip Casler M D (2017) Genome‐wide associationswith flowering time in switchgrass using exome‐capture sequenc-ing data New Phytologist 213 154ndash169 httpsdoiorg101111nph14101
Greef J M amp Deuter M (1993) Syntaxonomy of Miscanthus xgiganteus GREEF et DEU Angewandte Botanik 67 87ndash90
Guerra F P Richards J H Fiehn O Famula R Stanton B JShuren R hellip Neale D B (2016) Analysis of the genetic varia-tion in growth ecophysiology and chemical and metabolomic
CLIFTON‐BROWN ET AL | 29
composition of wood of Populus trichocarpa provenances TreeGenetics amp Genomes 12 6 httpsdoiorg101007s11295-015-0965-8
Guo H P Shao R X Hong C T Hu H K Zheng B S ampZhang Q X (2013) Rapid in vitro propagation of bioenergy cropMiscanthus sacchariflorus Applied Mechanics and Materials 260181ndash186
Hallingbaumlck H R Fogelqvist J Powers S J Turrion‐Gomez JRossiter R Amey J hellip Roumlnnberg‐Waumlstljung A‐C (2016)Association mapping in Salix viminalis L (Salicaceae)ndashIdentifica-tion of candidate genes associated with growth and phenologyGlobal Change Biology Bioenergy 8 670ndash685
Hanley S J amp Karp A (2014) Genetic strategies for dissectingcomplex traits in biomass willows (Salix spp) Tree Physiology34 1167ndash1180 httpsdoiorg101093treephystpt089
Harfouche A Meilan R amp Altman A (2011) Tree genetic engi-neering and applications to sustainable forestry and biomass pro-duction Trends in Biotechnology 29 9ndash17 httpsdoiorg101016jtibtech201009003
Harfouche A Meilan R amp Altman A (2014) Molecular and phys-iological responses to abiotic stress in forest trees and their rele-vance to tree improvement Tree Physiology 34 1181ndash1198httpsdoiorg101093treephystpu012
Harfouche A Meilan R Kirst M Morgante M Boerjan WSabatti M amp Mugnozza G S (2012) Accelerating the domesti-cation of forest trees in a changing world Trends in PlantScience 17 64ndash72 httpsdoiorg101016jtplants201111005
Hastings A Clifton‐Brown J Wattenbach M Mitchell C PStampfl P amp Smith P (2009) Future energy potential of Mis-canthus in Europe Global Change Biology Bioenergy 1 180ndash196
Hastings A Mos M Yesufu J AMccalmont J Schwarz KShafei RhellipClifton-Brown J (2017) Economic and environmen-tal assessment of seed and rhizome propagated Miscanthus in theUK Frontiers in Plant Science 8 16 httpsdoiorg103389fpls201701058
Heacuteder M (2017) From NASA to EU The evolution of the TRLscale in Public Sector Innovation The Innovation Journal 22 1ndash23 httpsdoiorg103389fpls201701058
Heffner E L Sorrells M E amp Jannink J L (2009) Genomicselection for crop improvement Crop Science 49 1ndash12 httpsdoiorg102135cropsci2008080512
Hodkinson T R Klaas M Jones M B Prickett R amp Barth S(2015) Miscanthus A case study for the utilization of naturalgenetic variation Plant Genetic Resources‐Characterization andUtilization 13 219ndash237 httpsdoiorg101017S147926211400094X
Hodkinson T R amp Renvoize S (2001) Nomenclature of Miscant-hus xgiganteus (Poaceae) Kew Bulletin 56 759ndash760 httpsdoiorg1023074117709
Houle D Govindaraju D R amp Omholt S (2010) Phenomics Thenext challenge Nature Reviews Genetics 11 855ndash866 httpsdoiorg101038nrg2897
Huang X H Wei X H Sang T Zhao Q Feng Q Zhao Yhellip Han B (2010) Genome‐wide association studies of 14 agro-nomic traits in rice landraces Nature Genetics 42 961ndashU976
Hwang O‐J Cho M‐A Han Y‐J Kim Y‐M Lim S‐H KimD‐S hellip Kim J‐I (2014) Agrobacterium‐mediated genetic trans-formation of Miscanthus sinensis Plant Cell Tissue and OrganCulture (PCTOC) 117 51ndash63
Hwang O‐J Lim S‐H Han Y‐J Shin A‐Y Kim D‐S ampKim J‐I (2014) Phenotypic characterization of transgenic Mis-canthus sinensis plants overexpressing Arabidopsis phytochromeB International Journal of Photoenergy 2014501016
Ilstedt B (1996) Genetics and performance of Belgian poplar clonestested in Sweden International Journal of Forest Genetics 3183ndash195
Ingvarsson P K Garcia M V Hall D Luquez V amp Jansson S(2006) Clinal variation in phyB2 a candidate gene for day‐length‐induced growth cessation and bud set across a latitudinalgradient in European aspen (Populus tremula) Genetics 1721845ndash1853
Ingvarsson P K Garcia M V Luquez V Hall D amp Jansson S(2008) Nucleotide polymorphism and phenotypic associationswithin and around the phytochrome B2 locus in European aspen(Populus tremula Salicaceae) Genetics 178 2217ndash2226 httpsdoiorg101534genetics107082354
Jannink J‐L Lorenz A J amp Iwata H (2010) Genomic selectionin plant breeding From theory to practice Briefings in FunctionalGenomics 9 166ndash177 httpsdoiorg101093bfgpelq001
Jiang J Guan Y Mccormick S Juvik J Lubberstedt T amp FeiS‐Z (2017) Gametophytic self‐incompatibility is operative inMiscanthus sinensis (Poaceae) and is affected by pistil age CropScience 57 1948ndash1956
Jones H D (2015a) Future of breeding by genome editing is in thehands of regulators GM Crops amp Food 6 223ndash232
Jones H D (2015b) Regulatory uncertainty over genome editingNature Plants 1 14011
Kalinina O Nunn C Sanderson R Hastings A F S van der Weijde TOumlzguumlven MhellipClifton‐Brown J C (2017) ExtendingMiscanthus cul-tivation with novel germplasm at six contrasting sites Frontiers in PlantScience 8563 httpsdoiorg103389fpls201700563
Karp A Hanley S J Trybush S O Macalpine W Pei M ampShield I (2011) Genetic improvement of willow for bioenergyand biofuels free access Journal of Integrative Plant Biology 53151ndash165 httpsdoiorg101111j1744-7909201001015x
Kersten B Faivre Rampant P Mader M Le Paslier M‐C Bou-non R Berard A hellip Fladung M (2016) Genome sequences ofPopulus tremula chloroplast and mitochondrion Implications forholistic poplar breeding PloS One 11 e0147209 httpsdoiorg101371journalpone0147209
Kiesel A Nunn C Iqbal Y Van der Weijde T Wagner MOumlzguumlven M hellip Lewandowski I (2017) Site‐specific manage-ment of Miscanthus genotypes for combustion and anaerobicdigestion A comparison of energy yields Frontiers in PlantScience 8 347 httpsdoiorg103389fpls201700347
Kim H Kim S T Ryu J Kang B C Kim J S amp Kim S G(2017) CRISPRCpf1‐mediated DNA‐free plant genome editingNature Communications 8 14406 httpsdoiorg101038ncomms14406
King Z R Bray A L Lafayette P R amp Parrott W A (2014)Biolistic transformation of elite genotypes of switchgrass (Pan-icum virgatum L) Plant Cell Reports 33 313ndash322 httpsdoiorg101007s00299-013-1531-1
Kopp R F Maynard C A De Niella P R Smart L B amp Abra-hamson L P (2002) Collection and storage of pollen from Salix(Salicaceae) American Journal of Botany 89 248ndash252
Krzyżak J Pogrzeba M Rusinowski S Clifton‐Brown J McCal-mont J P Kiesel A hellip Mos M (2017) Heavy metal uptake
30 | CLIFTON‐BROWN ET AL
by novel miscanthus seed‐based hybrids cultivated in heavy metalcontaminated soil Civil and Environmental Engineering Reports26 121ndash132 httpsdoiorg101515ceer-2017-0040
Kuzovkina Y A amp Quigley M F (2005) Willows beyond wet-lands Uses of Salix L species for environmental projects WaterAir and Soil Pollution 162 183ndash204 httpsdoiorg101007s11270-005-6272-5
Lazarus W Headlee W L amp Zalesny R S (2015) Impacts of sup-plyshed‐level differences in productivity and land costs on the eco-nomics of hybrid poplar production in Minnesota USA BioEnergyResearch 8 231ndash248 httpsdoiorg101007s12155-014-9520-y
Lewandowski I (2015) Securing a sustainable biomass supply in agrowing bioeconomy Global Food Security 6 34ndash42 httpsdoiorg101016jgfs201510001
Lewandowski I (2016) The role of perennial biomass crops in agrowing bioeconomy In S Barth D Murphy‐Bokern O Kalin-ina G Taylor amp M Jones (Eds) Perennial biomass crops for aresource‐constrained world (pp 3ndash13) Cham SwitzerlandSpringer Switzerland AG
Li J L Meilan R Ma C Barish M amp Strauss S H (2008)Stability of herbicide resistance over 8 years of coppice in field‐grown genetically engineered poplars Western Journal of AppliedForestry 23 89ndash93
Li R Y amp Qu R D (2011) High throughput Agrobacterium‐medi-ated switchgrass transformation Biomass amp Bioenergy 35 1046ndash1054 httpsdoiorg101016jbiombioe201011025
Lindegaard K N amp Barker J H A (1997) Breeding willows forbiomass Aspects of Applied Biology 49 155ndash162
Linde‐Laursen I B (1993) Cytogenetic analysis of MiscanthusGiganteus an interspecific hybrid Hereditas 119 297ndash300httpsdoiorg101111j1601-5223199300297x
Liu Y Merrick P Zhang Z Z Ji C H Yang B amp Fei S Z(2018) Targeted mutagenesis in tetraploid switchgrass (Panicumvirgatum L) using CRISPRCas9 Plant Biotechnology Journal16 381ndash393
Liu L L Wu Y Q Wang Y W amp Samuels T (2012) A high‐density simple sequence repeat‐based genetic linkage map ofswitchgrass G3 (Bethesda Md) 2 357ndash370
Lovett A Suumlnnenberg G amp Dockerty T (2014) The availabilityof land for perennial energy crops in Great Britain GlobalChange Biology Bioenergy 6 99ndash107 httpsdoiorg101111gcbb12147
Lozano‐Juste J amp Cutler S R (2014) Plant genome engineering infull bloom Trends in Plant Science 19 284ndash287 httpsdoiorg101016jtplants201402014
Lu F Lipka A E Glaubitz J Elshire R Cherney J H CaslerM D hellip Costich D E (2013) Switchgrass Genomic DiversityPloidy and Evolution Novel Insights from a Network‐Based SNPDiscovery Protocol Plos Genetics 9 e1003215
Ludovisi R Tauro F Salvati R Khoury S Mugnozza G S ampHarfouche A (2017) UAV‐based thermal imaging for high‐throughput field phenotyping of black poplar response to droughtFrontiers in Plant Science 81681
Macalpine W Shield I amp Karp A (2010) Seed to near market vari-ety the BEGIN willow breeding pipeline 2003ndash2010 and beyond InBioten conference proceedings Birmingham pp 21ndash23
Macalpine W J Shield I F Trybush S O Hayes C M ampKarp A (2008) Overcoming barriers to crossing in willow (Salixspp) breeding Aspects of Applied Biology 90 173ndash180
Mahfouz M M (2017) The efficient tool CRISPR‐Cpf1 NaturePlants 3 17028
Malinowska M Donnison I S amp Robson P R (2017) Phenomicsanalysis of drought responses in Miscanthus collected from differ-ent geographical locations Global Change Biology Bioenergy 978ndash91
Maroder H L Prego I A Facciuto G R amp Maldonado S B(2000) Storage behaviour of Salix alba and Salix matsudanaseeds Annals of Botany 86 1017ndash1021 httpsdoiorg101006anbo20001265
Martinez‐Reyna J amp Vogel K (2002) Incompatibility systems inswitchgrass Crop Science 42 1800ndash1805 httpsdoiorg102135cropsci20021800
Maurya J P amp Bhalerao R P (2017) Photoperiod‐and tempera-ture‐mediated control of growth cessation and dormancy in treesA molecular perspective Annals of Botany 120 351ndash360httpsdoiorg101093aobmcx061
McCalmont J P Hastings A Mcnamara N P Richter G MRobson P Donnison I S amp Clifton‐Brown J (2017) Environ-mental costs and benefits of growing Miscanthus for bioenergy inthe UK Global Change Biology Bioenergy 9 489ndash507
McCracken A R amp Dawson W M (1997) Using mixtures of wil-low clones as a means of controlling rust disease Aspects ofApplied Biology 49 97ndash103
McKown A D Klaacutepště J Guy R D Geraldes A Porth I Hanne-mann JhellipDouglas C J (2014) Genome‐wide association implicatesnumerous genes underlying ecological trait variation in natural popula-tions ofPopulus trichocarpaNew Phytologist 203 535ndash553
Merrick P amp Fei S Z (2015) Plant regeneration and genetic transforma-tion in switchgrass ndash A review Journal of Integrative Agriculture 14483ndash493 httpsdoiorg101016S2095-3119(14)60921-7
Moreno‐Mateos M A Fernandez J P Rouet R Lane M AVejnar C E Mis E hellip Giraldez A J (2017) CRISPR‐Cpf1mediates efficient homology‐directed repair and temperature‐con-trolled genome editing Nature Communications 8 2024 httpsdoiorg101038s41467-017-01836-2
Mosseler A (1990) Hybrid performance and species crossabilityrelationships in willows (Salix) Canadian Journal of Botany 682329ndash2338
Muchero W Guo J DiFazio S P Chen J‐G Ranjan P Slavov GT hellip Tuskan G A (2015) High‐resolution genetic mapping ofallelic variants associated with cell wall chemistry in Populus BMCGenomics 16 24 httpsdoiorg101186s12864-015-1215-z
Neale D B amp Kremer A (2011) Forest tree genomics Growingresources and applications Nature Reviews Genetics 12 111ndash122 httpsdoiorg101038nrg2931
Nunn C Hastings A F S J Kalinina O Oumlzguumlven M SchuumlleH Tarakanov I G hellip Clifton‐Brown J C (2017) Environmen-tal influences on the growing season duration and ripening ofdiverse Miscanthus germplasm grown in six countries Frontiersin Plant Science 8 907 httpsdoiorg103389fpls201700907
Ogawa Y Honda M Kondo Y amp Hara‐Nishimura I (2016) Anefficient Agrobacterium‐mediated transformation method forswitchgrass genotypes using Type I callus Plant Biotechnology33 19ndash26
Ogawa Y Shirakawa M Koumoto Y Honda M Asami Y KondoY ampHara‐Nishimura I (2014) A simple and reliable multi‐gene trans-formation method for switchgrass Plant Cell Reports 33 1161ndash1172httpsdoiorg101007s00299-014-1605-8
CLIFTON‐BROWN ET AL | 31
Okada M Lanzatella C Saha M C Bouton J Wu R L amp Tobias CM (2010) Complete switchgrass genetic maps reveal subgenomecollinearity preferential pairing and multilocus interactions Genetics185 745ndash760 httpsdoiorg101534genetics110113910
Palomo‐Riacuteos E Macalpine W Shield I Amey J Karaoğlu C WestJ hellip Jones H D (2015) Efficient method for rapid multiplication ofclean and healthy willow clones via in vitro propagation with broadgenotype applicability Canadian Journal of Forest Research 451662ndash1667 httpsdoiorg101139cjfr-2015-0055
Pilate G Allona I Boerjan W Deacutejardin A Fladung M Gal-lardo F hellip Halpin C (2016) Lessons from 25 years of GM treefield trials in Europe and prospects for the future In C Vettori FGallardo H Haumlggman V Kazana F Migliacci G Pilateamp MFladung (Eds) Biosafety of forest transgenic trees (pp 67ndash100)Dordrecht Netherlands Springer Switzerland AG
Pinosio S Giacomello S Faivre-Rampant P Taylor G Jorge VLe Paslier M C amp Marroni F (2016) Characterization of thepoplar pan-genome by genome-wide identification of structuralvariation Molecular Biology and Evolution 33 2706ndash2719
Porth I Klaacutepště J Skyba O Friedmann M C Hannemann JEhlting J hellip Douglas C J (2013) Network analysis reveals therelationship among wood properties gene expression levels andgenotypes of natural Populus trichocarpa accessions New Phytol-ogist 200 727ndash742
Pugesgaard S Schelde K Larsen S U Laeligrke P E amp JoslashrgensenU (2015) Comparing annual and perennial crops for bioenergyproductionndashinfluence on nitrate leaching and energy balance Glo-bal Change Biology Bioenergy 7 1136ndash1149 httpsdoiorg101111gcbb12215
Quinn L D Allen D J amp Stewart J R (2010) Invasivenesspotential of Miscanthus sinensis Implications for bioenergy pro-duction in the United States Global Change Biology Bioenergy2 310ndash320 httpsdoiorg101111j1757-1707201001062x
Rae A M Robinson K M Street N R amp Taylor G (2004)Morphological and physiological traits influencing biomass pro-ductivity in short‐rotation coppice poplar Canadian Journal ofForest Research‐Revue Canadienne De Recherche Forestiere 341488ndash1498 httpsdoiorg101139x04-033
Rambaud C Arnoult S Bluteau A Mansard M C Blassiau Camp Brancourt‐Hulmel M (2013) Shoot organogenesis in threeMiscanthus species and evaluation for genetic uniformity usingAFLP analysis Plant Cell Tissue and Organ Culture (PCTOC)113 437ndash448 httpsdoiorg101007s11240-012-0284-9
Ramstein G P Evans J Kaeppler S M Mitchell R B Vogel K PBuell C R amp Casler M D (2016) Accuracy of genomic prediction inswitchgrass (Panicum virgatum L) improved by accounting for linkagedisequilibriumG3 (Bethesda Md) 6 1049ndash1062
Rayburn A L Crawford J Rayburn C M amp Juvik J A (2009)Genome size of three Miscanthus species Plant Molecular Biol-ogy Reporter 27 184 httpsdoiorg101007s11105-008-0070-3
Richardson J Isebrands J amp Ball J (2014) Ecology and physiol-ogy of poplars and willows In J Isebrands amp J Richardson(Eds) Poplars and willows Trees for society and the environ-ment (pp 92‐123) Boston MA and Rome CABI and FAO
Robison T L Rousseau R J amp Zhang J (2006) Biomass produc-tivity improvement for eastern cottonwood Biomass amp Bioenergy30 735ndash739 httpsdoiorg101016jbiombioe200601012
Robson P R Farrar K Gay A P Jensen E F Clifton‐Brown JC amp Donnison I S (2013) Variation in canopy duration in the
perennial biofuel crop Miscanthus reveals complex associationswith yield Journal of Experimental Botany 64 2373ndash2383
Robson P Jensen E Hawkins S White S R Kenobi K Clif-ton‐Brown J hellip Farrar K (2013) Accelerating the domestica-tion of a bioenergy crop Identifying and modelling morphologicaltargets for sustainable yield increase in Miscanthus Journal ofExperimental Botany 64 4143ndash4155
Sanderson M A Adler P R Boateng A A Casler M D ampSarath G (2006) Switchgrass as a biofuels feedstock in theUSA Canadian Journal of Plant Science 86 1315ndash1325httpsdoiorg104141P06-136
Scarlat N Dallemand J F Monforti‐Ferrario F amp Nita V(2015) The role of biomass and bioenergy in a future bioecon-omy Policies and facts Environmental Development 15 3ndash34httpsdoiorg101016jenvdev201503006
Serapiglia M J Gouker F E amp Smart L B (2014) Early selec-tion of novel triploid hybrids of shrub willow with improved bio-mass yield relative to diploids BMC Plant Biology 14 74httpsdoiorg1011861471-2229-14-74
Serba D Wu L Daverdin G Bahri B A Wang X Kilian Ahellip Devos K M (2013) Linkage maps of lowland and upland tet-raploid switchgrass ecotypes BioEnergy Research 6 953ndash965httpsdoiorg101007s12155-013-9315-6
Shakoor N Lee S amp Mockler T C (2017) High throughput phe-notyping to accelerate crop breeding and monitoring of diseases inthe field Current Opinion in Plant Biology 38 184ndash192 httpsdoiorg101016jpbi201705006
Shield I Macalpine W Hanley S amp Karp A (2015) Breedingwillow for short rotation coppice energy cropping In Industrialcrops Breeding for BioEnergy and Bioproducts (pp 67ndash80) NewYork NY Springer
Sjodin A Street N R Sandberg G Gustafsson P amp Jansson S (2009)The Populus Genome Integrative Explorer (PopGenIE) A new resourcefor exploring the Populus genomeNewPhytologist 182 1013ndash1025
Slavov G amp Davey C (2017) Integrated genomic prediction inbioenergy crops In Abstracts from the International Conferenceon Developing biomass crops for future climates pp 34 24ndash27September 2017 Oriel College Oxford wwwwatbioeu
Slavov G Davey C Bosch M Robson P Donnison I ampMackay I (2018a) Genomic index selection provides a pragmaticframework for setting and refining multi‐objective breeding targetsin Miscanthus Annals of Botany (in press)
Slavov G T DiFazio S P Martin J Schackwitz W MucheroW Rodgers‐Melnick E hellip Tuskan G A (2012) Genome rese-quencing reveals multiscale geographic structure and extensivelinkage disequilibrium in the forest tree Populus trichocarpa NewPhytologist 196 713ndash725
Slavov G Davey C Robson P Donnison I amp Mackay I(2018b) Domestication of Miscanthus through genomic indexselection In Plant and Animal Genome Conference 13 January2018 San Diego CA Retrieved from httpspagconfexcompagxxvimeetingappcgiPaper30622
Slavov G T Nipper R Robson P Farrar K Allison G GBosch M hellip Jensen E (2013) Genome‐wide association studiesand prediction of 17 traits related to phenology biomass and cellwall composition in the energy grass Miscanthus sinensis NewPhytologist 201 1227ndash1239
Slavov G Robson P Jensen E Hodgson E Farrar K AllisonG hellip Donnison I (2014) Contrasting geographic patterns of
32 | CLIFTON‐BROWN ET AL
genetic variation for molecular markers vs phenotypic traits in theenergy grass Miscanthus sinensis Global Change Biology Bioen-ergy 5 562ndash571
Ślusarkiewicz‐Jarzina A Ponitka A Cerazy‐Waliszewska J Woj-ciechowicz M K Sobańska K Jeżowski S amp Pniewski T(2017) Effective and simple in vitro regeneration system of Mis-canthus sinensis Mtimes giganteus and M sacchariflorus for plant-ing and biotechnology purposes Biomass and Bioenergy 107219ndash226 httpsdoiorg101016jbiombioe201710012
Smart L amp Cameron K (2012) Shrub willow In C Kole S Joshiamp D Shonnard (Eds) Handbook of bioenergy crop plants (pp687ndash708) Boca Raton FL Taylor and Francis Group
Stanton B J (2014) The domestication and conservation of Populusand Salix genetic resources (Chapter 4) In Isebrands J ampRichardson J (Eds) Poplars and willows Trees for society andthe environment (pp 124ndash200) Rome Italy CAB InternationalFood and Agricultural Organization of the United Nations
Stanton B J Neale D B amp Li S (2010) Populus breeding Fromthe classical to the genomic approach In S Jansson R Bha-lerao amp A T Groover (Eds) Genetics and genomics of popu-lus (pp 309ndash348) New York NY Springer
Stewart J R Toma Y Fernandez F G Nishiwaki A Yamada T ampBollero G (2009) The ecology and agronomy ofMiscanthus sinensisa species important to bioenergy crop development in its native range inJapan A reviewGlobal Change Biology Bioenergy 1 126ndash153
Stott K G (1992) Willows in the service of man Proceedings of the RoyalSociety of Edinburgh Section B Biological Sciences 98 169ndash182
Strauss S H Ma C Ault K amp Klocko A L (2016) Lessonsfrom two decades of field trials with genetically modified trees inthe USA Biology and regulatory compliance In C Vettori FGallardo H Haumlggman V Kazana F Migliacci G Pilate amp MFladung (Eds) Biosafety of forest transgenic trees (pp 101ndash124)Dordrecht Netherlands Springer
Suda Y amp Argus G W (1968) Chromosome numbers of some NorthAmerican Salix Brittonia 20 191ndash197 httpsdoiorg1023072805440
Tan B (2018) Genomic selection and genome‐wide association studies todissect quantitative traits in forest trees Umearing Sweden University
Tornqvist C Taylor M Jiang Y Evans J Buell C KaepplerS amp Casler M (2018) Quantitative trait locus mapping for flow-ering time in a lowland x upland switchgrass pseudo‐F2 popula-tion The Plant Genome 11 170093
Toacuteth G Hermann T Da Silva M amp Montanarella L (2016)Heavy metals in agricultural soils of the European Union withimplications for food safety Environment International 88 299ndash309 httpsdoiorg101016jenvint201512017
Tracy W Dawson J Moore V amp Fisch J (2016) Intellectual propertyrights and public plant breeding Recommendations and proceedings ofa conference on best practices for intellectual property protection of pub-lically developed plant germplasm (p 70) University of Wisconsin‐Madison Retrieved from httpshostcalswisceduagronomywp-contentuploadssites1620162005Proceedings-IPR-Finalpdf
Trybush S Jahodovaacute Š Macalpine W amp Karp A (2008) A geneticstudy of a Salix germplasm resource reveals new insights into rela-tionships among subgenera sections and species BioEnergyResearch 1 67ndash79 httpsdoiorg101007s12155-008-9007-9
Tsai C J (2013) Next‐generation sequencing for next‐generationbreeding and more New Phytologist 198 635ndash637 httpsdoiorg101111nph12245
Tuskan G A Difazio S Jansson S Bohlmann J Grigoriev I HellstenU hellip Rokhsar D (2006) The genome of black cottonwood Populustrichocarpa (Torr ampGray) Science 313 1596ndash1604
Tuskan G A amp Walsh M E (2001) Short‐rotation woody cropsystems atmospheric carbon dioxide and carbon management AUS case study The Forestry Chronicle 77 259ndash264 httpsdoiorg105558tfc77259-2
Van Den Broek R Treffers D‐J Meeusen M Van Wijk ANieuwlaar E amp Turkenburg W (2001) Green energy or organicfood A life‐cycle assessment comparing two uses of set‐asideland Journal of Industrial Ecology 5 65ndash87 httpsdoiorg101162108819801760049477
Van Der Schoot J Pospiskova M Vosman B amp Smulders M J M(2000) Development and characterization of microsatellite markers inblack poplar (Populus nigra L) Theoretical and Applied Genetics 101317ndash322 httpsdoiorg101007s001220051485
Van Der Weijde T Dolstra O Visser R G amp Trindade L M(2017a) Stability of cell wall composition and saccharificationefficiency in Miscanthus across diverse environments Frontiers inPlant Science 7 2004
Van der Weijde T Huxley L M Hawkins S Sembiring E HFarrar K Dolstra O hellip Trindade L M (2017) Impact ofdrought stress on growth and quality of miscanthus for biofuelproduction Global Change Biology Bioenergy 9 770ndash782
Van der Weijde T Kamei C L A Severing E I Torres A FGomez L D Dolstra O hellip Trindade L M (2017) Geneticcomplexity of miscanthus cell wall composition and biomass qual-ity for biofuels BMC Genomics 18 406
Van der Weijde T Kiesel A Iqbal Y Muylle H Dolstra O Visser RG FhellipTrindade LM (2017) Evaluation ofMiscanthus sinensis bio-mass quality as feedstock for conversion into different bioenergy prod-uctsGlobal Change Biology Bioenergy 9 176ndash190
Vanholme B Cesarino I Goeminne G Kim H Marroni F VanAcker R hellip Boerjan W (2013) Breeding with rare defectivealleles (BRDA) A natural Populus nigra HCT mutant with modi-fied lignin as a case study New Phytologist 198 765ndash776
Vogel K P Mitchell R B Casler M D amp Sarath G (2014)Registration of Liberty Switchgrass Journal of Plant Registra-tions 8 242ndash247 httpsdoiorg103198jpr2013120076crc
Wang X Yamada T Kong F‐J Abe Y Hoshino Y Sato Hhellip Yamada T (2011) Establishment of an efficient in vitro cul-ture and particle bombardment‐mediated transformation systems inMiscanthus sinensis Anderss a potential bioenergy crop GlobalChange Biology Bioenergy 3 322ndash332 httpsdoiorg101111j1757-1707201101090x
Watrud L S Lee E H Fairbrother A Burdick C Reichman JR Bollman M hellip Van de Water P K (2004) Evidence forlandscape‐level pollen‐mediated gene flow from genetically modi-fied creeping bentgrass with CP4 EPSPS as a marker Proceedingsof the National Academy of Sciences of the United States of Amer-ica 101 14533ndash14538
Weisgerber H (1993) Poplar breeding for the purpose of biomassproduction in short‐rotation periods in Germany ndash Problems andfirst findings Forestry Chronicle 69 727ndash729 httpsdoiorg105558tfc69727-6
Wheeler N C Steiner K C Schlarbaum S E amp Neale D B(2015) The evolution of forest genetics and tree improvementresearch in the United States Journal of Forestry 113 500ndash510httpsdoiorg105849jof14-120
CLIFTON‐BROWN ET AL | 33
Whittaker C Macalpine W J Yates N E amp Shield I (2016)Dry matter losses and methane emissions during wood chip stor-age The impact on full life cycle greenhouse gas savings of shortrotation coppice willow for heat BioEnergy Research 9 820ndash835 httpsdoiorg101007s12155-016-9728-0
Xi Q (2000) Investigation on the distribution and potential of giant grassesin China PhD thesis Goettingen Germany Cuvillier Verlag
Xi Q amp Jezowkski S (2004) Plant resources of Triarrhena andMiscanthus species in China and its meaning for Europe PlantBreeding and Seed Science 49 63ndash77
Xue S Kalinina O amp Lewandowski I (2015) Present and future optionsfor the improvement of Miscanthus propagation techniques Renewableand Sustainable Energy Reviews 49 1233ndash1246
Yin K Q Gao C X amp Qiu J L (2017) Progress and prospectsin plant genome editing Nature Plants 3 httpsdoiorg101038nplants2017107
YookM (2016)Genetic diversity and transcriptome analysis for salt toler-ance inMiscanthus PhD thesis Seoul National University Korea
Zaidi S S E A Mahfouz M M amp Mansoor S (2017) CRISPR‐Cpf1 A new tool for plant genome editing Trends in PlantScience 22 550ndash553 httpsdoiorg101016jtplants201705001
Zalesny R S Stanturf J A Gardiner E S Perdue J H YoungT M Coyle D R hellip Hass A (2016) Ecosystem services ofwoody crop production systems BioEnergy Research 9 465ndash491 httpsdoiorg101007s12155-016-9737-z
Zhang Q X Sun Y Hu H K Chen B Hong C T Guo H Phellip Zheng B S (2012) Micropropagation and plant regenerationfrom embryogenic callus of Miscanthus sinensis Vitro Cellular ampDevelopmental Biology‐Plant 48 50ndash57 httpsdoiorg101007s11627-011-9387-y
Zhang Y Zalapa J E Jakubowski A R Price D L AcharyaA Wei Y hellip Casler M D (2011) Post‐glacial evolution of
Panicum virgatum Centers of diversity and gene pools revealedby SSR markers and cpDNA sequences Genetica 139 933ndash948httpsdoiorg101007s10709-011-9597-6
Zhao H Wang B He J Yang J Pan L Sun D amp Peng J(2013) Genetic diversity and population structure of Miscanthussinensis germplasm in China PloS One 8 e75672
Zhou X H Jacobs T B Xue L J Harding S A amp Tsai C J(2015) Exploiting SNPs for biallelic CRISPR mutations in theoutcrossing woody perennial Populus reveals 4‐coumarate CoAligase specificity and redundancy New Phytologist 208 298ndash301
Zhou R Macaya‐Sanz D Rodgers‐Melnick E Carlson C HGouker F E Evans L M hellip DiFazio S P (2018) Characteri-zation of a large sex determination region in Salix purpurea L(Salicaceae) Molecular Genetics and Genomics 1ndash16 httpsdoiorg101007s00438-018-1473-y
Zsuffa L Mosseler A amp Raj Y (1984) Prospects for interspecifichybridization in willow for biomass production In K L Perttu(Ed) Ecology and management of forest biomass production sys-tems Report 15 (pp 261ndash281) Uppsala Sweden SwedishUniversity of Agricultural Sciences
How to cite this article Clifton‐Brown J HarfoucheA Casler MD et al Breeding progress andpreparedness for mass‐scale deployment of perenniallignocellulosic biomass crops switchgrassmiscanthus willow and poplar GCB Bioenergy201800000ndash000 httpsdoiorg101111gcbb12566
34 | CLIFTON‐BROWN ET AL
Graphical AbstractThe contents of this page will be used as part of the graphical abstract of html only
It will not be published as part of main article
Plant breeding links the research effort with commercial mass upscaling The authorsrsquo assessment of development status ofthe four species is shown (poplar having two one for short rotation coppice (SRC) poplar and one for the more traditionalshort rotation forestry (SRF)) Mass scale deployment needs developments outside the breeding arenas to drive breedingactivities more rapidly and extensively
programmes mature The average rate of gain for biomassyield in long‐term switchgrass breeding programmes hasbeen 1ndash4 per year depending on ecotype populationand location of the breeding programme (Casler amp Vogel2014 Casler et al 2018) The hybrid derivative Libertyhas a biomass yield 43 higher than the better of its twoparents (Casler amp Vogel 2014 Vogel et al 2014) Thedevelopment of cold‐tolerant and late‐flowering lowland‐ecotype populations for the northern United States hasincreased biomass yields by 27 (Casler et al 2018)
Currently more than 20 recurrent selection populationsare being managed in the United States to select parentsfor improved yield yield resilience and compositional qual-ity of the biomass For the agronomic development andupscaling high seed multiplication rates need to be com-bined with lower seed dormancy to reduce both crop estab-lishment costs and risks Expresso is the first cultivar withsignificantly reduced seed dormancy which is the first steptowards development of domesticated populations (Casleret al 2015) Most phenotypic traits of interest to breeders
require a minimum of 2 years to be fully expressed whichresults in a breeding cycle that is at least two years Morecomplicated breeding programmes or traits that requiremore time to evaluate can extend the breeding cycle to 4ndash8 years per generation for example progeny testing forbiomass yield or field‐based selection for cold toleranceBreeding for a range of traits with such long cycles callsfor the development of molecular methods to reduce time-scales and improve breeding efficiency
Two association panels of switchgrass have been pheno-typically and genotypically characterized to identify quanti-tative trait loci (QTLs) that control important biomasstraits The northern panel consists of 60 populationsapproximately 65 from the upland ecotype The southernpanel consists of 48 populations approximately 65 fromthe lowland ecotype Numerous QTLs have been identifiedwithin the northern panel to date (Grabowski et al 2017)Both panels are the subject of additional studies focused onbiomass quality flowering and phenology and cold toler-ance Additionally numerous linkage maps have been
FIGURE 1 Cumulative minimum years needed for the conventional breeding cycle through the steps from wild germplasm to thecommercial hybrids in switchgrass miscanthus willow and poplar Information links between the steps are indicated by dotted arrows andhighlight the importance of long‐term informatics to maximize breeding gain
CLIFTON‐BROWN ET AL | 11
created by the pairwise crossing of individuals with diver-gent characteristics often to generate four‐way crosses thatare analysed as pseudo‐F2 crosses (Liu Wu Wang ampSamuels 2012 Okada et al 2010 Serba et al 2013Tornqvist et al 2018) Individual markers and QTLs iden-tified can be used to design marker‐assisted selection(MAS) programmes to accelerate breeding and increase itsefficiency Genomic prediction and selection (GS) holdseven more promise with the potential to double or triplethe rate of gain for biomass yield and other highly complexquantitative traits of switchgrass (Casler amp Ramstein 2018Ramstein et al 2016) The genome of switchgrass hasrecently been made public through the Joint Genome Insti-tute (httpsphytozomejgidoegov)
Transgenic approaches have been heavily relied upon togenerate unique genetic variants principally for traitsrelated to biomass quality (Merrick amp Fei 2015) Switch-grass is highly transformable using either Agrobacterium‐mediated transformation or biolistics bombardment butregeneration of plants is the bottleneck to these systemsTraditionally plants from the cultivar Alamo were the onlyregenerable genotypes but recent efforts have begun toidentify more genotypes from different populations that arecapable of both transformation and subsequent regeneration(King Bray Lafayette amp Parrott 2014 Li amp Qu 2011Ogawa et al 2014 Ogawa Honda Kondo amp Hara‐Nishi-mura 2016) Cell wall recalcitrance and improved sugarrelease are the most common targets for modification (Bis-wal et al 2018 Fu et al 2011) Transgenic approacheshave the potential to provide traits that cannot be bredusing natural genetic variability However they will stillrequire about 10ndash15 years and will cost $70ndash100 millionfor cultivar development and deployment (Harfouche Mei-lan amp Altman 2011) In addition there is commercialuncertainty due to the significant costs and unpredictabletimescales and outcomes of the regulatory approval processin the countries targeted for seed sales As seen in maizeone advantage of transgenic approaches is that they caneasily be incorporated into F1 hybrid cultivars (Casler2012 Casler et al 2012) but this does not decrease thetime required for cultivar development due to field evalua-tion and seed multiplication requirements
The potential impacts of unintentional gene flow andestablishment of non‐native transgene sequences in nativeprairie species via cross‐pollination are also major issues forthe environmental risk assessment These limit further thecommercialization of varieties made using these technolo-gies Although there is active research into switchgrass steril-ity mechanisms to curb unintended pollen‐mediated genetransfer it is likely that the first transgenic cultivars proposedfor release in the United States will be met with considerableopposition due to the potential for pollen flow to remainingwild prairie sites which account for lt1 of the original
prairie land area and are highly protected by various govern-mental and nongovernmental organizations (Casler et al2015) Evidence for landscape‐level pollen‐mediated geneflow from genetically modified Agrostis seed multiplicationfields (over a mountain range) to pollinate wild relatives(Watrud et al 2004) confirms the challenge of using trans-genic approaches Looking ahead genome editing technolo-gies hold considerable promise for creating targeted changesin phenotype (Burris Dlugosz Collins Stewart amp Lena-ghan 2016 Liu et al 2018) and at least in some jurisdic-tions it is likely that cultivars resulting from gene editingwill not need the same regulatory approval as GMOs (Jones2015a) However in July 2018 the European Court of Justice(ECJ) ruled that cultivars carrying mutations resulting fromgene editing should be regulated in the same way as GMOsThe ECJ ruled that such cultivars be distinguished from thosearising from untargeted mutation breeding which isexempted from regulation under Directive 200118EC
3 | MISCANTHUS
Miscanthus is indigenous to eastern Asia and Oceania whereit is traditionally used for forage thatching and papermaking(Xi 2000 Xi amp Jezowkski 2004) In the 1960s the highbiomass potential of a Japanese genotype introduced to Eur-ope by Danish nurseryman Aksel Olsen in 1935 was firstrecognized in Denmark (Linde‐Laursen 1993) Later thisaccession was characterized described and named asldquoM times giganteusrdquo (Greef amp Deuter 1993 Hodkinson ampRenvoize 2001) commonly abbreviated as Mxg It is a natu-rally occurring interspecies triploid hybrid between tetraploidM sacchariflorus (2n = 4x) and diploid M sinensis(2n = 2x) Despite its favourable agronomic characteristicsand ability to produce high yields in a wide range of environ-ments in Europe (Kalinina et al 2017) the risks of relianceon it as a single clone have been recognized Miscanthuslike switchgrass is outcrossing due to self‐incompatibility(Jiang et al 2017) Thus seeded hybrids are an option forcommercial breeding Miscanthus can also be vegetativelypropagated by rhizome or in vitro culture which allows thedevelopment of clones The breeding approaches are usuallybased on F1 crosses and recurrent selection cycles within thesynthetic populations There are several breeding pro-grammes that target improvement of miscanthus traitsincluding stress resilience targeted regional adaptation agro-nomic ldquoscalabilityrdquo through cheaper propagation fasterestablishment lower moisture and ash contents and greaterusable yield (Clifton‐Brown et al 2017)
Germplasm collections specifically to support breedingfor biomass started in the late 1980s and early 1990s inDenmark Germany and the United Kingdom (Clifton‐Brown Schwarz amp Hastings 2015) These collectionshave continued with successive expeditions from European
12 | CLIFTON‐BROWN ET AL
and US teams assembling diverse collections from a widegeographic range in eastern Asia including from ChinaJapan South Korea Russia and Taiwan (Hodkinson KlaasJones Prickett amp Barth 2015 Stewart et al 2009) Threekey miscanthus species for biomass production are M si-nensis M floridulus and M sacchariflorus M sinensis iswidely distributed throughout eastern Asia with an adap-tive range from the subtropics to southern Russia (Zhaoet al 2013) This species has small rhizomes and producesmany tightly packed shoots forming a ldquotuftrdquo M floridulushas a more southerly adaptive range with a rather similarmorphology to M sinensis but grows taller with thickerstems and is evergreen and less cold‐tolerant than the othermiscanthus species M sacchariflorus is the most northern‐adapted species ranging to 50 degN in eastern Russia (Clarket al 2016) Populations of diploid and tetraploid M sac-chariflorus are found in China (Xi 2000) and South Korea(Yook 2016) and eastern Russia but only tetraploids havebeen found in Japan (Clark Jin amp Petersen 2018)
Germplasm has been assembled from multiple collec-tions over the last century though some early collectionsare poorly documented This historical germplasm has beenused to initiate breeding programmes largely based on phe-notypic and genotypic characterization As many of theaccessions from these collections are ldquoorigin unknownrdquocrucial environmental envelope data are not available UK‐led expeditions started in 2006 and continued until 2011with European and Asian partners and have built up a com-prehensive collection of 1500 accessions from 500 sitesacross Eastern Asia including China Japan South Koreaand Taiwan These collections were guided using spatialclimatic data to identify variation in abiotic stress toleranceAccessions from these recent collections were planted fol-lowing quarantine in multilocation nursery trials at severallocations in Europe to examine trait expression in differentenvironments Based on the resulting phenotypic andmolecular marker data several studies (a) characterized pat-terns of population genetic structure (Slavov et al 20132014) (b) evaluated the statistical power of genomewideassociation studies (GWASs) and identified preliminarymarkerndashtrait associations (Slavov et al 2013 2014) and(c) assessed the potential of genomic prediction (Daveyet al 2017 Slavov et al 2014 2018b) Genomic indexselection in particular offers the possibility of exploringscenarios for different locations or industrial markets (Sla-vov et al 2018a 2018b)
Separately US‐led expeditions also collected about1500 accessions between 2010 and 2014 (Clark et al2014 2016 2018 2015) A comprehensive genetic analy-sis of the population structure has been produced by RAD-seq for M sinensis (Clark et al 2015 Van der WeijdeKamei et al 2017) and M sacchariflorus (Clark et al2018) Multilocation replicated field trials have also been
conducted on these materials in North America and inAsia GWAS has been conducted for both M sinensis anda subset of M sacchariflorus accessions (Clark et al2016) To date about 75 of these recent US‐led collec-tions are in nursery trials outside the United States Due tolengthy US quarantine procedures these are not yet avail-able for breeding in the United States However molecularanalyses have allowed us to identify and prioritize sets ofgenotypes that best encompass the genetic variation in eachspecies
While mostM sinensis accessions flower in northern Eur-ope very few M sacchariflorus accessions flower even inheated glasshouses For this reason the European pro-grammes in the United Kingdom the Netherlands and Francehave performed mainly M sinensis (intraspecies) hybridiza-tions (Table 4) Selected progeny become the parents of latergenerations (recurrent selection as in switchgrass) Seed setsof up to 400 seed per panicle occur inM sinensis In Aberyst-wyth and Illinois significant efforts to induce synchronousflowering in M sacchariflorus and M sinensis have beenmade because interspecies hybrids have proven higher yieldperformance and wide adaptability (Kalinina et al 2017) Ininterspecies pairwise crosses in glasshouses breathable bagsandor large crossing tubes or chambers in which two or morewhole plants fit are used for pollination control Seed sets arelower in bags than in the open air because bags restrict pollenmovement while increasing temperatures and reducinghumidity (Clifton‐Brown Senior amp Purdy 2018) About30 of attempted crosses produced 10 to 60 seeds per baggedpanicle The seed (thousand seed mass ranges from 05 to09 g) is threshed from the inflorescences and sown into mod-ular trays to produce plug plants which are then planted infield nurseries to identify key parental combinations
A breeding programme of this scale must serve theneeds of different environments accepting the commonpurpose is to optimize the interception of solar radiationAn ideal hybrid for a given environment combines adapta-tion to date of emergence with optimization of traits suchas height number of stems per plant flowering and senes-cence time to optimize solar interception to produce a highbiomass yield with low moisture content at harvest (Rob-son Farrar et al 2013 Robson Jensen et al 2013) By20132014 conventional breeding in Europe had producedintra‐ and interspecific fertile seeded hybrids When acohort (typically about 5) of outstanding crosses have beenidentified it is important to work on related upscaling mat-ters in parallel These are as follows
Assessment of the yield and critical traits in selectedhybrids using a network of field trials
Efficient cloning of the seed parents While in vitro andmacro‐cloning techniques are used some genotypes areamenable to neither technique
CLIFTON‐BROWN ET AL | 13
TABLE5
Status
ofmod
ernplantbreeding
techniqu
esin
four
leadingperennialbiom
asscrop
s(PBCs)
Breeding
techno
logy
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sW
illow
Pop
lar
MAS
Use
ofmarker
sequ
encesthat
correlatewith
atrait
allowingearly
prog
enyselectionor
rejectionlocatin
gkn
owngenesfor
useful
traits(suchas
height)from
other
speciesin
your
crop
Breeding
prog
ramme
relevant
biparental
crosses(Table
4)to
create
ldquomapping
popu
latio
nsrdquoand
QTLidentification
andestim
ationto
identifyrobu
stmarkers
alternatively
acomprehensive
panelof
unrelated
geno
typesfor
GWAS
Ineffectivewhere
traits
areaffected
bymany
geneswith
small
effects
US
Literally
dozens
ofstud
iesto
identify
SNPs
andmarkers
ofinterestNoattempts
atMASas
yetlargely
dueto
popu
latio
nspecificity
CNS
SRISS
Rmarkers
forM
sinMsacand
Mflor
FR2
Msin
mapping
popu
latio
ns(BFF
project)
NL2
Msin
mapping
popu
latio
ns(A
tienza
Ram
irezetal
2003
AtienzaSatovic
PetersenD
olstraamp
Martin
200
3Van
Der
Weijdeetal20
17a
Van
derW
eijde
Hux
ley
etal20
17
Van
derW
eijdeKam
ei
etal20
17V
ander
WeijdeKieseletal
2017
)SK
2GWASpanels(1
collectionof
Msin1
diploidbiparentalF 1
popu
latio
n)in
the
pipelin
eUK1
2families
tofind
QTLin
common
indifferentfam
ilies
ofthe
sameanddifferent
speciesandhy
brids
US
2largeGWAS
panels4
diploidbi‐
parentalF 1
popu
latio
ns
1F 2
popu
latio
n(add
ition
alin
pipelin
e)
UK16
families
for
QTLdiscov
ery
2crossesMASscreened
1po
tentialvariety
selected
US
9families
geno
typedby
GBS(8
areF 1o
neisF 2)a
numberof
prom
ising
QTLdevelopm
entof
markers
inprog
ress
FR(INRA)tested
inlargeFS
families
and
infactorialmating
design
low
efficiency
dueto
family
specificity
US
49families
GS
Metho
dto
accelerate
breeding
throug
hredu
cing
the
resourcesforcross
attemptsby
Aldquotraining
popu
latio
nrdquoof
individu
alsfrom
stages
abov
ethat
have
been
both
Risks
ofpo
orpredictio
nProg
eny
testingneedsto
becontinueddu
ring
the
timethetraining
setis
US
One
prog
rammeso
farmdash
USD
Ain
WisconsinThree
cycles
ofgeno
mic
selectioncompleted
CN~1
000
geno
types
NLprelim
inary
mod
elsforMsin
developed
UK~1
000
geno
types
Verysuitable
application
not
attempted
yet
Maybe
challeng
ingfor
interspecies
hybrids
FR(INRA)120
0geno
type
ofPop
ulus
nigraarebeingused
todevelop
intraspecificGS
(Con
tinues)
14 | CLIFTON‐BROWN ET AL
TABLE
5(Contin
ued)
Breeding
techno
logy
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sW
illow
Pop
lar
predictin
gthe
performance
ofprog
enyof
crosses
(and
inresearch
topredictbestparents
tousein
biparental
crosses)
geno
typedand
phenotyp
edto
developamod
elthattakesgeno
typic
datafrom
aldquocandidate
popu
latio
nrdquoof
untested
individu
als
andprod
uces
GEBVs(Jannink
Lorenzamp
Iwata
2010
)
beingph
enotyp
edand
geno
typed
Inthis
time
next‐generation
germ
plasm
andreadyto
beginthe
second
roun
dof
training
and
recalib
ratio
nof
geno
mic
predictio
nmod
els
US
Prelim
inary
mod
elsforMsac
and
Msin
developed
Work
isun
derw
ayto
deal
moreeffectivelywith
polyploidgenetics
calib
ratio
nsUS
125
0geno
types
arebeingused
todevelopGS
calib
ratio
ns
Traditio
nal
transgenesis
Efficient
introd
uctio
nof
ldquoforeign
rdquotraits
(possiblyfrom
other
genu
sspecies)
into
anelite
plant(ega
prov
enparent
ora
hybrid)that
needsa
simpletraitto
beim
prov
edValidate
cand
idategenesfrom
QTLstud
ies
MASand
know
ledg
eof
the
biolog
yof
thetrait
andsource
ofgenesto
confer
the
relevant
changesto
phenotyp
eWorking
transformation
protocol
IPissuescostof
regu
latory
approv
al
GMO
labelling
marketin
gissuestricky
tousetransgenes
inou
t‐breedersbecause
ofcomplexity
oftransformingandgene
flow
risks
US
Manyprog
ramsin
UnitedStates
are
creatin
gtransgenic
plantsusingAlamoas
asource
oftransformable
geno
typesManytraits
ofinterestNothing
commercial
yet
CNC
hang
shaBtg
ene
transformed
into
Msac
in20
04M
sin
transformed
with
MINAC2gene
and
Msactransform
edwith
Cry2A
ageneb
othwith
amarkerg
eneby
Agrob
acterium
JP1
Msintransgenic
forlow
temperature
tolerancewith
increased
expression
offructans
(unp
ublished)
NLImprov
ingprotocol
forM
sin
transformation
SK2
Msin
with
Arabido
psisand
Brachypod
ium
PhytochromeBgenes
(Hwang
Cho
etal
2014
Hwang
Lim
etal20
14)
UK2
Msin
and
1Mflorg
enotyp
es
UKRou
tine
transformationno
tyet
possibleResearchto
overcomerecalcitrance
ongo
ing
Currently
trying
differentspecies
andcond
ition
sPo
plar
transformationused
atpresent
US
Frequently
attempted
with
very
limitedsuccess
US
600transgenic
lines
have
been
characterized
mostly
performed
inthe
aspenhy
brids
reprod
uctiv
esterility
drou
ghttolerance
with
Kno
ckdo
wns
(Con
tinues)
CLIFTON‐BROWN ET AL | 15
TABLE
5(Contin
ued)
Breeding
techno
logy
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sW
illow
Pop
lar
variou
slytransformed
with
four
cellwall
genesiptand
uidA
genesusingtwo
selectionsystem
sby
biolisticsand
Agrob
acterium
Transform
edplantsare
beinganalysed
US
Prelim
inarywork
will
betakenforw
ardin
2018
ndash202
2in
CABBI
Genom
eediting
CRISPR
Refinem
entofexisting
traits
inusefulparents
orprom
isinghybrids
bygeneratingtargeted
mutations
ingenes
know
ntocontrolthe
traitof
interestF
irst
adouble‐stranded
breakismadeinthe
DNAwhich
isrepairedby
natural
DNArepair
machineryL
eads
tofram
eshiftSN
Por
can
usealdquorepair
templaterdquo
orcanbe
used
toinserta
transgene
intoaldquosafe
harbourlocusrdquoIt
couldbe
used
todeleterepressors(or
transcriptionfactors)
Identification
mapping
and
sequ
encing
oftarget
genes(from
DNA
sequ
ence)
avoiding
screening
outof
unintend
ededits
Manyregu
latory
authorities
have
not
decidedwhether
CRISPR
andother
geno
meediting
techno
logies
are
GMOsor
notIf
GMOsseecomment
onstage10
abov
eIf
noteditedcrop
swill
beregu
latedas
conv
entio
nalvarieties
CATechn
olog
yisstill
toonewand
switchg
rass
geno
meis
very
complexOther
labo
ratories
are
interestedbu
tno
tyet
mov
ingon
this
US
One
prog
rammeso
farmdash
USD
Ain
Albany
FRInitiated
in20
16(M
ISEDIT
project)
NLInitiatingin
2018
US
Initiatingin
2018
in
CABBI
UKNot
started
Not
possible
until
transformation
achieved
US
CRISPR
‐Cas9and
Cpf1have
been
successfully
developedin
Pop
ulus
andarehigh
lyefficientExamples
forlig
nin
biosyn
thesis
Note
BFF
BiomassFo
rtheFu
tureBtBacillus
thuringiensisCABBICenterforadvanced
bioenergyandbioproductsinnovatio
nCpf1
CRISPR
from
Prevotella
andFrancisella
1CRISPR
clusteredregularlyinterspaced
palin
drom
icrepeatsCRISPR
‐CasC
RISPR
‐associated
Cry2A
acrystaltoxins2A
asubfam
ilyproduced
byBtFS
full‐sibG
BS
genotyping
bysequencingG
EBVsgenomic‐estim
ated
breeding
valuesG
MOg
enetically
modi-
fied
organism
GS
genomicselection
GWAS
genomew
ideassociationstudy
INRAF
renchNationalInstituteforAgriculturalR
esearch
IPintellectualp
roperty
IPTisopentenyltransferase
ISSR
inter‐SSR
MflorMiscant-
husflo
ridulusMsacMsacchariflo
rusMsinMsinensisM
AS
marker‐a
ssistedselection
MISEDITm
iscanthusgene
editing
forseed‐propagatedtriploidsNACn
oapicalmeristemA
TAF1
2and
cup‐shaped
cotyledon2
‐lik
eQTLq
uantitativ
etraitlocusS
NP
singlenucleotid
epolymorphismS
SRsim
plesequence
repeatsUSD
AU
SDepartm
ento
fAgriculture
a ISO
Alpha‐2
lettercountrycodes
16 | CLIFTON‐BROWN ET AL
High seed production from field crossing trials con-ducted in locations where flowering in both seed andpollen parents is likely to happen synchronously
Scalable and adapted harvesting threshing and seed pro-cessing methods for producing high seed quality
The results of these parallel activities need to becombined to identify the upscaling pathway for eachhybrid if this cannot be achieved the hybrid will likelynot be commercially viable The UK‐led programme withpartners in Italy and Germany shows that seedbased mul-tiplication rates of 12000 are achievable several inter-specific hybrids (Clifton‐Brown et al 2017) Themultiplication rate of M sinensis is higher probably15000ndash10000 Conventional cloning from rhizome islimited to around 120 that is one ha could provide rhi-zomes for around 20 ha of new plantation
Multilocation field testing of wild and novel miscanthushybrids selected by breeding programmes in the Nether-lands and the United Kingdom was performed as part ofthe project Optimizing Miscanthus Biomass Production(OPTIMISC 2012ndash2016) These trials showed that com-mercial yields and biomass qualities (Kiesel et al 2017Van der Weijde et al 2017a Van der Weijde Kieselet al 2017) could be produced in a wide range of climatesand soil conditions from the temperate maritime climate ofwestern Wales to the continental climate of eastern Russiaand the Ukraine (Kalinina et al 2017) Extensive environ-mental measurements of soil and climate combined withgrowth monitoring are being used to understand abioticstresses (Nunn et al 2017 Van der Weijde Huxley et al2017) and develop genotype‐specific scenarios similar tothose reported earlier in Hastings et al (2009) Phenomicsexperiments on drought tolerance have been conducted onwild and improved germplasm (Malinowska Donnison ampRobson 2017 Van der Weijde Huxley et al 2017)Recently produced interspecific hybrids displaying excep-tional yield under drought (~30 greater than control Mxg)in field trials in Poland and Moldova are being furtherstudied in detail in the phenomics and genomics facility atAberystwyth to better understand genendashtrait associationswhich can be fed back into breeding
Intraspecific seeded hybrids of M sinensis produced inthe Netherlands and interspecific M sacchariflorus times M si-nensis hybrids produced by the UK‐led breeding programmehave entered yield testing in 2018 with the recently EU‐funded project ldquoGRowing Advanced industrial Crops on mar-ginal lands for biorEfineries (GRACE)rdquo (httpswwwgrace-bbieu) Substantial variation in biomass quality for sacchari-fication efficiency (glucose release as of dry matter) ashcontent and melting point has already been generated inintraspecific M sinensis hybrids (Van der Weijde Kiesel
et al 2017) across environments (Weijde Dolstra et al2017a) GRACE aims to establish more than 20 hectares ofnew inter‐ and intraspecific seeded hybrids across six Euro-pean countries This project is building the know‐how andagronomy needed to transition from small research plots tocommercial‐scale field sites and linking biomass productiondirectly to industrial applications The biomass produced byhybrids in different locations will be supplied to innovativeindustrial end‐users making a wide range of biobased prod-ucts both for chemicals and for energy In the United Statesmulti‐location yield were initiated in 2018 to evaluate new tri-ploid M times giganteus genotypes developed at Illinois Cur-rently infertile hybrids are favoured in the United Statesbecause this eliminates the risk of invasiveness from naturallydispersed viable seed The precautionary principle is appliedas fertile miscanthus has naturalized in several states (QuinnAllen amp Stewart 2010) North European multilocation fieldtrials in the EMI and OPTIMISC projects have shown thereis minimal risk of invasiveness even in years when fertileflowering hybrids produce viable seed Naturalized standshave not established here due perhaps to low dormancy pooroverwintering and low seedling competitive strength In addi-tion to breeding for nonshattering or sterile seeded hybridsQuinn et al (2010) suggest management strategies which canfurther minimize environmental opportunities to manage therisk of invasiveness
31 | Molecular breeding and biotechnology
In miscanthus new plant breeding techniques (Table 5)have focussed on developing molecular markers forbreeding in Europe the United States South Korea andJapan There are several publications on QTL mappingpopulations for key traits such as flowering (AtienzaRamirez amp Martin 2003) and compositional traits(Atienza Satovic Petersen Dolstra amp Martin 2003) Inthe United States and United Kingdom independent andinterconnected bi‐parental ldquomappingrdquo families have beenstudied (Dong et al 2018 Gifford Chae SwaminathanMoose amp Juvik 2015) alongside panels of diverse germ-plasm accessions for GWAS (Slavov et al 2013) Fur-ther developments calibrating GS with very large panelsof parents and cross progeny are underway (Daveyet al 2017) The recently completed first miscanthus ref-erence genome sequence is expected to improve the effi-ciency of MAS strategies and especially GWAS(httpsphytozomejgidoegovpzportalhtmlinfoalias=Org_Msinensis_er) For example without a referencegenome sequence Clark et al (2014) obtained 21207RADseq SNPs (single nucleotide polymorphisms) on apanel of 767 miscanthus genotypes (mostly M sinensis)but subsequent reanalysis of the RADseq data using the
CLIFTON‐BROWN ET AL | 17
new reference genome resulted in hundreds of thousandsof SNPs being called
Robust and effective in vitro regeneration systems havebeen developed for Miscanthus sinensis M times giganteusand M sacchariflorus (Dalton 2013 Guo et al 2013Hwang Cho et al 2014 Rambaud et al 2013 Ślusarkie-wicz‐Jarzina et al 2017 Wang et al 2011 Zhang et al2012) However there is still significant genotype speci-ficity and these methods need ldquoin‐houserdquo optimization anddevelopment to be used routinely They provide potentialroutes for rapid clonal propagation and also as a basis forgenetic transformation Stable transformation using bothbiolistics (Wang et al 2011) and Agrobacterium tumefa-ciens DNA delivery methods (Hwang Cho et al 2014Hwang Lim et al 2014) has been achieved in M sinen-sis The development of miscanthus transformation andgene editing to generate diplogametes for producing seed‐propagated triploid hybrids are performed as part of theFrench project MISEDIT (miscanthus gene editing forseed‐propagated triploids) There are no reports of genomeediting in any miscanthus species but new breeding inno-vations including genome editing are particularly relevantin this slow‐to‐breed nonfood bioenergy crop (Table 4)
4 | WILLOW
Willow (Salix spp) is a very diverse group of catkin‐bear-ing trees and shrubs Willow belongs to the family Sali-caceae which also includes the Populus genus There areapproximately 350 willow species (Argus 2007) foundmostly in temperate and arctic zones in the northern hemi-sphere A few are adapted to subtropical and tropicalzones The centre of diversity is believed to be in Asiawith over 200 species in China Around 120 species arefound in the former Soviet Union over 100 in NorthAmerica and around 65 species in Europe and one speciesis native to South America (Karp et al 2011) Willows aredioecious thus obligate outcrossers and highly heterozy-gous The haploid chromosome number is 19 (Hanley ampKarp 2014) Around 40 of species are polyploid (Sudaamp Argus 1968) ranging from triploids to the atypicaldodecaploid S maxxaliana with 2n=190 (Zsuffa et al1984)
Although almost exclusively native to the NorthernHemisphere willow has been grown around the globe formany thousands of years to support a wide range of appli-cations (Kuzovkina amp Quigley 2005 Stott 1992) How-ever it has been the focus of domestication for bioenergypurposes for only a relatively short period since the 1970sin North America and Europe For bioenergy breedershave focused their efforts on the shrub willows (subgenusVetix) because of their rapid juvenile growth rates as aresponse to coppicing on a 2‐ to 4‐year cycle that can be
accomplished using farm machinery rather than forestryequipment (Shield Macalpine Hanley amp Karp 2015Smart amp Cameron 2012)
Since shrub willow was not generally recognized as anagricultural crop until very recently there has been littlecommitment to building and maintaining germplasm reposi-tories of willow to support long‐term breeding One excep-tion is the United Kingdom where a large and well‐characterized Salix germplasm collection comprising over1500 accessions is held at Rothamsted Research (Stott1992 Trybush et al 2008) Originally initiated for use inbasketry in 1923 accessions have been added ever sinceIn the United States a germplasm collection of gt350accessions is located at Cornell University to support thebreeding programme there The UK and Cornell collectionshave a relatively small number of accessions in common(around 20) Taken together they represent much of thespecies diversity but only a small fraction of the overallgenetic diversity within the genus There are three activewillow breeding programmes in Europe RothamstedResearch (UK) Salixenergi Europa AB (SEE) and a pro-gramme at the University of Warmia and Mazury in Olsz-tyn (Poland) (abbreviations used in Table 1) There is oneactive US programme based at Cornell University Culti-vars are still being marketed by the European WillowBreeding Programme (EWBP) (UK) which was activelybreeding biomass varieties from 1996 to 2002 Cultivarsare protected by plant breedersrsquo rights (PBRs) in Europeand by plant patents in the United States The sharing ofgenetic resources in the willow community is generallyregulated by material transfer agreements (MTA) and tai-lored licensing agreements although the import of cuttingsinto North America is prohibited except under special quar-antine permit conditions
Efforts to augment breeding germplasm collection fromnature are continuing with phenotypic screening of wildgermplasm performed in field experiments with 177 S pur-purea genotypes in the United States (at sites in Genevaand Portland NY and Morgantown WV) that have beengenotyped using genotyping by sequencing (GBS) (Elshireet al 2011) In addition there are approximately 400accessions of S viminalis in Europe (near Pustnaumls Upp-sala Sweden and Woburn UK (Berlin et al 2014 Hal-lingbaumlck et al 2016) The S viminalis accessions wereinitially genotyped using 38 simple sequence repeats (SSR)markers to assess genetic diversity and screened with~1600 SNPs in genes of potential interest for phenologyand biomass traits Genetics and genomics combined withextensive phenotyping have substantially improved thegenetic basis of biomass‐related traits in willow and arenow being developed in targeted breeding via MAS Thisunderpinning work has been conducted on large specifi-cally developed biparental Salix mapping populations
18 | CLIFTON‐BROWN ET AL
(Hanley amp Karp 2014 Zhou et al 2018) as well asGWAS panels (Hallingbaumlck et al 2016)
Once promising parental combinations are identifiedcrosses are usually performed using fresh pollen frommaterial that has been subject to a phased removal fromcold storage (minus4degC) (Lindegaard amp Barker 1997 Macal-pine Shield Trybush Hayes amp Karp 2008 Mosseler1990) Pollen storage is useful in certain interspecific com-binations where flowering is not naturally synchronizedThis can be overcome by using pollen collection and stor-age protocol which involves extracting pollen using toluene(Kopp Maynard Niella Smart amp Abrahamson 2002)
The main breeding approach to improve willow yieldsrelies on species hybridization to capture hybrid vigour(Fabio et al 2017 Serapiglia Gouker amp Smart 2014) Inthe absence of genotypic models for heterosis breedershave extensively tested general and specific combiningability of parents to produce superior progeny The UKbreeding programmes (EWBP 1996ndash2002 and RothamstedResearch from 2003 on) have performed more than 1500exploratory cross‐pollinations The Cornell programme hassuccessfully completed about 550 crosses since 1998Investment into the characterization of genetic diversitycombined with progeny tests from exploratory crosses hasbeen used to produce hundreds of targeted intraspeciescrosses in the United Kingdom and United States respec-tively (see Table 1) To achieve long‐term gains beyond F1hybrids four intraspecific recurrent selection populationshave been created in the United Kingdom (for S dasycla-dos S viminalis and S miyabeana) and Cornell is pursu-ing recurrent selection of S purpurea Interspecifichybridizations with genotypes selected from the recurrentselection cycles are well advanced in willow with suchcrosses to date totalling 420 in the United Kingdom andover 100 in the United States
While species hybridization is common in Salix it isnot universal Of the crosses attempted about 50 hybri-dize and produce seed (Macalpine Shield amp Karp 2010)As the viability of seed from successful crosses is short (amatter of days at ambient temperatures) proper seed rear-ing and storage protocols are essential (Maroder PregoFacciuto amp Maldonado 2000)
Progeny from crosses are treated in different waysamong the breeding programmes at the seedling stage Inthe United States seedlings are planted into an irrigatedfield where plants are screened for two seasons beforebeing progressed to further field trials In the United King-dom seedlings are planted into trays of compost wherethey remain containerized in an irrigated nursery for theremainder of year one In the United Kingdom seedlingsare subject to two rounds of selection in the nursery yearThe first round takes place in September to select againstsusceptibility to rust infection (Melampsora spp) A second
round of selection in winter assesses tip damage from frostand giant willow aphid infestation In the United Stateswhere the rust pressure is lower screening for Melampsoraspp cannot be performed at the nursery stage Both pro-grammes monitor Melampsora spp pest susceptibilityyield and architecture over multiple years in field trialsSelected material is subject to two rounds of field trials fol-lowed by a final multilocation yield trial to identify vari-eties for commercialization
Promising selections (ie potential cultivars) need to beclonally propagated A rapid in vitro tissue culture propa-gation method has been developed (Palomo‐Riacuteos et al2015) This method can generate about 5000 viable trans-plantable clones from a single plant in just 24 weeks Anin vitro system can also accommodate early selection viamolecular or biochemical markers to increase selectionspeed Conventional breeding systems take 13 years viafour rounds of selection from crossing to selecting a variety(Figure 1) but this has the potential to be reduced to7 years if micropropagation and MAS selection are adopted(Hanley amp Karp 2014 Palomo‐Riacuteos et al 2015)
Willows are currently propagated commercially byplanting winter‐dormant stem cuttings in spring Commer-cial planting systems for willow use mechanical plantersthat cut and insert stem sections from whips into a well‐prepared soil One hectare of stock plants grown in specificmultiplication beds planted at 40000 plants per ha pro-duces planting material for 80 hectares of commercialshort‐rotation coppice willow annually (planted at 15000cuttings per hectare) (Whittaker et al 2016) When com-mercial plantations are established the industry standard isto plant intimate mixtures of ~5 diverse rust (Melampsoraspp)‐resistant varieties (McCracken amp Dawson 1997 VanDen Broek et al 2001)
The foundations for using new plant breeding tech-niques have been established with funding from both thepublic and the private sectors To establish QTL maps 16mapping populations from biparental crosses are understudy in the United Kingdom Nine are under study in theUnited States The average number of individuals in thesefamilies ranges from 150 to 947 (Hanley amp Karp 2014)GS is also being evaluated in S viminalis and preliminaryresults indicate that multiomic approaches combininggenomic and metabolomic data have great potential (Sla-vov amp Davey 2017) For both QTL and GS approachesthe field phenotyping demands are large as several thou-sand individuals need to be phenotyped for a wide rangeof traits These include the following dates of bud burstand growth succession stem height stem density wooddensity and disease resistance The greater the number ofindividuals the more precise the QTL marker maps andGS models are However the logistical and financial chal-lenges of phenotyping large numbers of individuals are
CLIFTON‐BROWN ET AL | 19
considerable because the willow crop is gt5 m tall in thesecond year There is tremendous potential to improve thethroughput of phenotyping using unmanned aerial systemswhich is being tested in the USDA National Institute ofFood and Agriculture (NIFA) Willow SkyCAP project atCornell Further investment in these approaches needs tobe sustained over many years fully realizes the potentialof a marker‐assisted selection programme for willow
To date despite considerable efforts in Europe and theUnited States to establish a routine transformation systemthere has not been a breakthrough in willow but attemptsare ongoing As some form of transformation is typically aprerequisite for genome editing techniques these have notyet been applied to willow
In Europe there are 53 short‐rotation coppice (SRC) bio-mass willow cultivars registered with the Community PlantVariety Office (CPVO) for PBRs of which ~25 are availablecommercially in the United Kingdom There are eightpatented cultivars commercially available in the UnitedStates In Sweden there are nine commercial cultivars regis-tered in Europe and two others which are unregistered(httpssalixenergiseplanting-material) Furthermore thereare about 20 precommercial hybrids in final yield trials inboth the United States and the United Kingdom It has beenestimated that it would take two years to produce the stockrequired to plant 50 ha commercially from the plant stock inthe final yield trials Breeding programmes have alreadydelivered rust‐resistant varieties and increases in yield to themarket The adoption of advanced breeding technologies willlikely lead to a step change in improving traits of interest
5 | POPLAR
Poplar a fast‐growing tree from the northern hemispherewith a small genome size has been adopted for commercialforestry and scientific purposes The genus Populus con-sists of about 29 species classified in six different sectionsPopulus (formerly Leuce) Tacamahaca Aigeiros AbasoTuranga and Leucoides (Eckenwalder 1996) The Populusspecies of most interest for breeding and testing in the Uni-ted States and Europe are P nigra P deltoides P maxi-mowiczii and P trichocarpa (Stanton 2014) Populusclones for biomass production are being developed byintra‐ and interspecies hybridization (DeWoody Trewin ampTaylor 2015 Richardson Isebrands amp Ball 2014 van derSchoot et al 2000) Recurrent selection approaches areused for gradual population improvement and to create eliteclonal lines for commercialization (Berguson McMahon ampRiemenschneider 2017 Neale amp Kremer 2011) Currentlypoplar breeding in the United States occurs in industrialand academic programmes located in the Southeast theMidwest and the Pacific Northwest These use six speciesand five interspecific taxa (Stanton 2014)
The southeastern programme historically focused onrecurrent selection of P deltoides from accessions made inthe lower Mississippi River alluvial plain (Robison Rous-seau amp Zhang 2006) More recently the genetic base hasbeen broadened to produce interspecific hybrids with resis-tance to the fungal infection Septoria musiva which causescankers
In the midwest of the United States populationimprovement efforts are focused on P deltoides selectionsfrom native provenances and hybrid crosses with acces-sions introduced from Europe Interspecific intercontinental(Europe and America) hybrid crosses between P nigra andP deltoides (P times canadensis) are behind many of theleading commercial hybrids which are the most advancedbreeding materials for many applications and regions InMinnesota previous breeding experience and efforts utiliz-ing P maximowiczii and P trichocarpa have been discon-tinued due to Septoria susceptibility and a lack of coldhardiness (Berguson et al 2017) Traits targeted forimprovement include yieldgrowth rate cold hardinessadventitious rooting resistance to Septoria and Melamp-sora leaf rust and stem form The Upper Midwest pro-gramme also carries out wide hybridizations within thesection Populus The P times wettsteinii (P trem-ula times P tremuloides) taxon is bred for gains in growthrate wood quality and resistance to the fungus Entoleucamammata which causes hypoxylon canker (David ampAnderson 2002)
In the Pacific Northwest GreenWood Resources Incleads poplar breeding that emphasizes interspecifichybrid improvement of P times generosa (P deltoides timesP trichocarpa and reciprocal) and P deltoides times P maxi-mowiczii taxa for coastal regions and the P times canadensistaxon for the drier continental regions Intraspecificimprovement of second‐generation breeding populations ofP deltoides P nigra P maximowiczii and P trichocarpaare also involved (Stanton et al 2010) The present focusof GreenWood Resourcesrsquo hybridization is bioenergy feed-stock improvement concentrating on coppice yield wood‐specific gravity and rate of sugar release
Industrial interest in poplar in the United States has his-torically come from the pulp and paper sector althoughveneer and dimensional lumber markets have been pursuedat times Currently the biomass market for liquid trans-portation fuels is being emphasized along with the use oftraditional and improved poplar genotypes for ecosystemservices such as phytoremediation (Tuskan amp Walsh 2001Zalesny et al 2016)
In Europe there are breeding programmes in FranceGermany Italy and Sweden These include the following(a) Alasia Franco Vivai (AFV) programme in northernItaly (b) the French programme led by the poplar Scien-tific Interest Group (GIS Peuplier) and carried out
20 | CLIFTON‐BROWN ET AL
collaboratively between the National Institute for Agricul-tural Research (INRA) the National Research Unit ofScience and Technology for Environment and Agriculture(IRSTEA) and the Forest Cellulose Wood Constructionand Furniture Technology Institute (FCBA) (c) the Ger-man programme at Northwest German Forest Research Sta-tion (NW‐FVA) at Hannoversch Muumlnden and (d) theSwedish programme at the Swedish University of Agricul-tural Sciences and SweTree Technologies AB (Table 1)
AFV leads an Italian poplar breeding programme usingextensive field‐grown germplasm collections of P albaP deltoides P nigra and P trichocarpa While interspeci-fic hybridization uses several taxa the focus is onP times canadensis The breeding programme addresses dis-ease resistance (Marssonina brunnea Melampsora larici‐populina Discosporium populeum and poplar mosaicvirus) growth rate and photoperiod adaptation AFV andGreenWood Resources collaborate in poplar improvementin Europe through the exchange of frozen pollen and seedfor reciprocal breeding projects Plantations in Poland andRomania are currently the focus of the collaboration
The ongoing French GIS Peuplier is developing a long‐term breeding programme based on intraspecific recurrentselection for the four parental species (P deltoides P tri-chocarpa P nigra and P maximowiczii) designed to betterbenefit from hybrid vigour demonstrated by the interspeci-fic crosses P canadensis P deltoides times P trichocarpaand P trichocarpa times P maximowiczii Current selectionpriorities are targeting adaptation to soil and climate condi-tions resistance and tolerance to the most economicallyimportant diseases and pests high volume production underSRC and traditional poplar cultivation regimes as well aswood quality of interest by different markets Currentlygenomic selection is under exploration to increase selectionaccuracy and selection intensity while maintaining geneticdiversity over generations
The German NW‐FVA programme is breeding intersec-tional AigeirosndashTacamahaca hybrids with a focus on resis-tance to Pollaccia elegans Xanthomonas populiDothichiza spp Marssonina brunnea and Melampsoraspp (Stanton 2014) Various cross combinations ofP maximowiczii P trichocarpa P nigra and P deltoideshave led to new cultivars suitable for deployment in vari-etal mixtures of five to ten genotypes of complementarystature high productivity and phenotypic stability (Weis-gerber 1993) The current priority is the selection of culti-vars for high‐yield short‐rotation biomass production Sixhundred P nigra genotypes are maintained in an ex situconservation programme An in situ P nigra conservationeffort involves an inventory of native stands which havebeen molecular fingerprinted for identity and diversity
The Swedish programme is concentrating on locallyadapted genotypes used for short‐rotation forestry (SRF)
because these meet the needs of the current pulping mar-kets Several field trials have shown that commercial poplarclones tested and deployed in Southern and Central Europeare not well adapted to photoperiods and low temperaturesin Sweden and in the Baltics Consequently SwedishUniversity of Agricultural Sciences and SweTree Technolo-gies AB started breeding in Sweden in 1990s to producepoplar clones better adapted to local climates and markets
51 | Molecular breeding technologies
Poplar genetic improvement cannot be rapidly achievedthrough traditional methods alone because of the longbreeding cycles outcrossing breeding systems and highheterozygosity Integrating modern genetic genomic andphenomics techniques with conventional breeding has thepotential to expedite poplar improvement
The genome of poplar has been sequenced (Tuskanet al 2006) It has an estimated genome size of485 plusmn 10 Mbp divided into 19 chromosomes This is smal-ler than other PBCs and makes poplar more amenable togenetic engineering (transgenesis) GS and genome editingPoplar has seen major investment in both the United Statesand Europe being the model system for woody perennialplant genetics and genomics research
52 | Targets for genetic modification
Traits targeted include wood properties (lignin content andcomposition) earlylate flowering male sterility to addressbiosafety regulation issues enhanced yield traits and herbi-cide tolerance These extensive transgenic experiments haveshown differences in recalcitrance to in vitro regenerationand genetic transformation in some of the most importantcommercial hybrid poplars (Alburquerque et al 2016)Further transgene expression stability is being studied Sofar China is the only country known to have commerciallyused transgenic insect‐resistant poplar A precommercialherbicide‐tolerant poplar was trialled for 8 years in the Uni-ted States (Li Meilan Ma Barish amp Strauss 2008) butcould not be released due to stringent environmental riskassessments required for regulatory approval This increasestranslation costs and delays reducing investor confidencefor commercial deployment (Harfouche et al 2011)
The first field trials of transgenic poplar were performedin France in 1987 (Fillatti Sellmer Mccown Haissig ampComai 1987) and in Belgium in 1988 (Deblock 1990)Although there have been a total of 28 research‐scale GMpoplar field trials approved in the European Union underCouncil Directive 90220EEC since October 1991 (inPoland Belgium Finland France Germany Spain Swe-den and in the United Kingdom (Pilate et al 2016) onlyauthorizations in Poland and Belgium are in place today In
CLIFTON‐BROWN ET AL | 21
the United States regulatory notifications and permits fornearly 20000 transgenic poplar trees derived from approxi-mately 600 different constructs have been issued since1995 by the USDAs Animal and Plant Health InspectionService (APHIS) (Strauss et al 2016)
53 | Genome editing CRISPR technologies
Clustered regularly interspaced palindromic repeats(CRISPR) and the CRISPR‐associated (CRISPR‐Cas)nucleases are a groundbreaking genome‐engineering toolthat complements classical plant breeding and transgenicmethods (Moreno‐Mateos et al 2017) Only two publishedstudies in poplar have applied the CRISPRCas9 technol-ogy One is in P tomentosa in which an endogenous phy-toene desaturase gene (PtoPDS) was successfully disruptedsite specifically in the first generation of transgenic plantsresulting in an albino and dwarf phenotype (Fan et al2015) The second was in P tremula times alba in whichhigh CRISPR‐Cas9 mutational efficiency was achieved forthree 4‐coumarateCoA ligase (4Cl) genes 4CL1 4CL2and 4CL5 associated with lignin and flavonoid biosynthe-sis (Zhou et al 2015) Due to its low cost precision andrapidness it is very probable that cultivars or clones pro-duced using CRISPR technology will be ready for market-ing in the near future (Yin et al 2017) Recently aCRISPR with a smaller associated endonuclease has beendiscovered from Prevotella and Francisella 1 (Cpf1) whichmay have advantages over Cas9 In addition there arereports of DNA‐free editing in plants using both CRISPRCpf1 and CRISPR Cas9 for example Ref (Kim et al2017 Mahfouz 2017 Zaidi et al 2017)
It remains unresolved whether plants modified by genomeediting will be regulated as genetically modified organisms(GMOs) by the relevant authorities in different countries(Lozano‐Juste amp Cutler 2014) Regulations to cover thesenew breeding techniques are still evolving but those coun-tries who have published specific guidance (including UnitedStates Argentina and Chile) are indicating that plants pos-sessing simple genome edits will not be regulated as conven-tional transgenesis (Jones 2015b) The first generation ofgenome‐edited crops will likely be phenocopy gene knock-outs that already exist to produce ldquonature identicalrdquo traits thatis traits that could also be derived by conventional breedingDespite this confidence in applying these new powerfulbreeding tools remains limited owing to the uncertain regula-tory environment in many parts of the world (Gao 2018)including the recent ECJ 2018 rulings mentioned earlier
54 | Genomics‐based breeding technologies
Poplar breeding programs are becoming well equipped withuseful genomics tools and resources that are critical to
explore genomewide variability and make use of the varia-tion for enhancing genetic gains Deep transcriptomesequencing resequencing of alternate genomes and GBStechnology for genomewide marker detection using next‐generation sequencing (NGS) are yielding valuable geno-mics tools GWAS with NGS‐based markers facilitatesmarker identification for MAS breeding by design and GS
GWAS approaches have provided a deeper understand-ing of genome function as well as allelic architectures ofcomplex traits (Huang et al 2010) and have been widelyimplemented in poplar for wood characteristics (Porthet al 2013) stomatal patterning carbon gain versus dis-ease resistance (McKown et al 2014) height and phenol-ogy (Evans et al 2014) cell wall chemistry (Mucheroet al 2015) growth and cell walls traits (Fahrenkroget al 2017) bark roughness (Bdeir et al 2017) andheight and diameter growth (Liu et al 2018) Using high‐throughput sequencing and genotyping platforms an enor-mous amount of SNP markers have been used to character-ize the linkage disequilibrium (LD) in poplar (eg Slavovet al 2012 discussed below)
The genetic architecture of photoperiodic traits in peren-nial trees is complex involving many loci However itshows high levels of conservation during evolution (Mau-rya amp Bhalerao 2017) These genomics tools can thereforebe used to address adaptation issues and fine‐tune themovement of elite lines into new environments For exam-ple poor timing of spring bud burst and autumn bud setcan result in frost damage resulting in yield losses (Ilstedt1996) These have been studied in P tremula genotypesalong a latitudinal cline in Sweden (~56ndash66oN) and haverevealed high nucleotide polymorphism in two nonsynony-mous SNPs within and around the phytochrome B2 locus(Ingvarsson Garcia Hall Luquez amp Jansson 2006 Ing-varsson Garcia Luquez Hall amp Jansson 2008) Rese-quencing 94 of these P tremula genotypes for GWASshowed that noncoding variation of a single genomicregion containing the PtFT2 gene described 65 ofobserved genetic variation in bud set along the latitudinalcline (Tan 2018)
Resequencing genomes is currently the most rapid andeffective method detecting genetic differences betweenvariants and for linking loci to complex and importantagronomical and biomass traits thus addressing breedingchallenges associated with long‐lived plants like poplars
To date whole genome resequencing initiatives havebeen launched for several poplar species and genotypes InPopulus LD studies based on genome resequencing sug-gested the feasibility of GWAS in undomesticated popula-tions (Slavov et al 2012) This plant population is beingused to inform breeding for bioenergy development Forexample the detection of reliable phenotypegenotype asso-ciations and molecular signatures of selection requires a
22 | CLIFTON‐BROWN ET AL
detailed knowledge about genomewide patterns of allelefrequency variation LD and recombination suggesting thatGWAS and GS in undomesticated populations may bemore feasible in Populus than previously assumed Slavovet al (2012) have resequenced 16 genomes of P tri-chocarpa and genotyped 120 trees from 10 subpopulationsusing 29213 SNPs (Geraldes et al 2013) The largest everSNP data set of genetic variations in poplar has recentlybeen released providing useful information for breedinghttpswwwbioenergycenterorgbescgwasindexcfmAlso deep sequencing of transcriptomes using RNA‐Seqhas been used for identification of functional genes andmolecular markers that is polymorphism markers andSSRs A multitissue and multiple experimental data set forP trichocarpa RNA‐Seq is publicly available (httpsjgidoegovdoe-jgi-plant-flagship-gene-atlas)
The availability of genomic information of DNA‐con-taining cell organelles (nucleus chloroplast and mitochon-dria) will also allow a holistic approach in poplarmolecular breeding in the near future (Kersten et al2016) Complete Populus genome sequences are availablefor nucleus (P trichocarpa section Tacamahaca) andchloroplasts (seven species and two clones from P trem-ula W52 and P tremula times P alba 717ndash1B4) A compara-tive approach revealed structural and functionalinformation broadening the knowledge base of PopuluscpDNA and stimulating future diagnostic marker develop-ment The availability of whole genome sequences of thesecellular compartments of P tremula holds promise forboosting marker‐assisted poplar breeding Other nucleargenome sequences from additional Populus species arenow available (eg P deltoides (httpsphytozomejgidoegovpz) and will become available in the forthcomingyears (eg P tremula and P tremuloidesmdashPopGenIE(Sjodin Street Sandberg Gustafsson amp Jansson 2009))Recently the characterization of the poplar pan‐genome bygenomewide identification of structural variation in threecrossable poplar species P nigra P deltoides and P tri-chocarpa revealed a deeper understanding of the role ofinter‐ and intraspecific structural variants in poplar pheno-type and may have important implications for breedingparticularly interspecific hybrids (Pinosio et al 2016)
GS has been proposed as an alternative to MAS in cropimprovement (Bernardo amp Yu 2007 Heffner Sorrells ampJannink 2009) GS is particularly well suited for specieswith long generation times for characteristics that displaymoderate‐to‐low heritability for traits that are expensive tomeasure and for selection of traits expressed late in the lifecycle as is the case for most traits of commercial value inforestry (Harfouche et al 2012) Current joint genomesequencing efforts to implement GS in poplar using geno-mic‐estimated breeding values for bioenergy conversiontraits from 49 P trichocarpa families and 20 full‐sib
progeny are taking place at the Oak Ridge National Labo-ratory and GreenWood Resources (Brian Stanton personalcommunication httpscbiornlgov) These data togetherwith the resequenced GWAS population data will be thebasis for developing GS algorithms Genomic breedingtools have been developed for the intraspecific programmetargeting yield resistance to Venturia shoot blightMelampsora leaf rust resistance to Cryptorhynchus lapathistem form wood‐specific gravity and wind firmness (Evanset al 2014 Guerra et al 2016) A newly developedldquobreeding with rare defective allelesrdquo (BRDA) technologyhas been developed to exploit natural variation of P nigraand identify defective variants of genes predicted by priortransgenic research to impact lignin properties Individualtrees carrying naturally defective alleles can then be incor-porated directly into breeding programs thereby bypassingthe need for transgenics (Vanholme et al 2013) Thisnovel breeding technology offers a reverse genetics com-plement to emerging GS for targeted improvement of quan-titative traits (Tsai 2013)
55 | Phenomics‐assisted breeding technology
Phenomics involves the characterization of phenomesmdashthefull set of phenotypes of given individual plants (HouleGovindaraju amp Omholt 2010) Traditional phenotypingtools which inefficiently measure a limited set of pheno-types have become a bottleneck in plant breeding studiesHigh‐throughput plant phenotyping facilities provide accu-rate screening of thousands of plant breeding lines clones orpopulations over time (Fu 2015) are critical for acceleratinggenomics‐based breeding Automated image collection andanalysis phenomics technologies allow accurate and nonde-structive measurements of a diversity of phenotypic traits inlarge breeding populations (Gegas Gay Camargo amp Doo-nan 2014 Goggin Lorence amp Topp 2015 Ludovisi et al2017 Shakoor Lee amp Mockler 2017) One important con-sideration is the identification of relevant and quantifiabletarget traits that are early diagnostic indicators of biomassyield Good progress has been made in elucidating theseunderpinning morpho‐physiological traits that are amenableto remote sensing in Populus (Harfouche Meilan amp Altman2014 Rae Robinson Street amp Taylor 2004) Morerecently Ludovisi et al (2017) developed a novel methodol-ogy for field phenomics of drought stress in a P nigra F2partially inbred population using thermal infrared imagesrecorded from an unmanned aerial vehicle‐based platform
Energy is the current main market for poplar biomass butthe market return provided is not sufficient to support pro-duction expansion even with added demand for environmen-tal and land management ldquoecosystem servicesrdquo such as thetreatment of effluent phytoremediation riparian buffer zonesand agro‐forestry plantings Aviation fuel is a significant
CLIFTON‐BROWN ET AL | 23
target market (Crawford et al 2016) To serve this marketand to reduce current carbon costs of production (Budsberget al 2016) key improvement traits in addition to yield(eg coppice regeneration pestdisease resistance water‐and nutrient‐use efficiencies) will be trace greenhouse gas(GHG) emissions (eg isoprene volatiles) site adaptabilityand biomass conversion efficiency Efforts are underway tohave national environmental protection agenciesrsquo approvalfor poplar hybrids qualifying for renewable energy credits
6 | REFLECTIONS ON THECOMMERCIALIZATIONCHALLENGE
The research and innovation activities reviewed in thispaper aim to advance the genetic improvement of speciesthat can provide feedstocks for bioenergy applicationsshould those markets eventually develop These marketsneed to generate sufficient revenue and adequately dis-tribute it to the actors along the value chain The work onall four crops shares one thing in common long‐termefforts to integrate fundamental knowledge into breedingand crop development along a research and development(RampD) pipeline The development of miscanthus led byAberystwyth University exemplifies the concerted researcheffort that has integrated the RampD activities from eight pro-jects over 14 years with background core research fundingalong an emerging innovation chain (Figure 2) This pro-gramme has produced a first range of conventionally bredseeded interspecies hybrids which are now in upscaling tri-als (Table 3) The application of molecular approaches(Table 5) with further conventional breeding (Table 4)offers the prospect of a second range of improved seededhybrids This example shows that research‐based support ofthe development of new crops or crop types requires along‐term commitment that goes beyond that normallyavailable from project‐based funding (Figure 1) Innovationin this sector requires continuous resourcing of conven-tional breeding operations and capability to minimize timeand investment losses caused by funding discontinuities
This challenge is increased further by the well‐knownmarket failure in the breeding of many agricultural cropspecies The UK Department for Environment Food andRural Affairs (Defra) examined the role of geneticimprovement in relation to nonmarket outcomes such asenvironmental protection and concluded that public invest-ment in breeding was required if profound market failure isto be addressed (Defra 2002) With the exception ofwidely grown hybrid crops such as maize and some high‐value horticultural crops royalties arising from plant breed-ersrsquo rights or other returns to breeders fail to adequatelycompensate for the full cost for research‐based plant breed-ing The result even for major crops such as wheat is sub‐
optimal investment and suboptimal returns for society Thismarket failure is especially acute for perennial crops devel-oped for improved sustainability rather than consumerappeal (Tracy et al 2016) Figure 3 illustrates the underly-ing challenge of capturing value for the breeding effortThe ldquovalley of deathrdquo that results from the low and delayedreturns to investment applies generally to the research‐to‐product innovation pipeline (Beard Ford Koutsky amp Spi-wak 2009) and certainly to most agricultural crop speciesHowever this schematic is particularly relevant to PBCsMost of the value for society from the improved breedingof these crops comes from changes in how agricultural landis used that is it depends on the increased production ofthese crops The value for society includes many ecosys-tems benefits the effects of a return to seminatural peren-nial crop cover that protects soils the increase in soilcarbon storage the protection of vulnerable land or the cul-tivation of polluted soils and the reductions in GHG emis-sions (Lewandowski 2016) By its very nature theproduction of biomass on agricultural land marginal forfood production challenges farm‐level profitability Thecosts of planting material and one‐time nature of cropestablishment are major early‐stage costs and therefore theopportunities for conventional royalty capture by breedersthat are manifold for annual crops are limited for PBCs(Hastings et al 2017) Public investment in developingPBCs for the nonfood biobased sector needs to providemore long‐term support for this critical foundation to a sus-tainable bioeconomy
7 | CONCLUSIONS
This paper provides an overview of research‐based plantbreeding in four leading PBCs For all four PBC generasignificant progress has been made in genetic improvementthrough collaboration between research scientists and thoseoperating ongoing breeding programmes Compared withthe main food crops most PBC breeding programmes dateback only a few decades (Table 1) This breeding efforthas thus co‐evolved with molecular biology and the result-ing ‐omics technologies that can support breeding Thedevelopment of all four PBCs has depended strongly onpublic investment in research and innovation The natureand driver of the investment varied In close associationwith public research organizations or universities all theseprogrammes started with germplasm collection and charac-terization which underpin the selection of parents forexploratory wide crosses for progeny testing (Figure 1Table 4)
Public support for switchgrass in North America wasexplicitly linked to plant breeding with 12 breeding pro-grammes supported in the United States and CanadaSwitchgrass breeding efforts to date using conventional
24 | CLIFTON‐BROWN ET AL
breeding have resulted in over 36 registered cultivars inthe United States (Table 1) with the development of dedi-cated biomass‐type cultivars coming within the past fewyears While ‐omics technologies have been incorporatedinto several of these breeding programmes they have notyet led to commercial deployment in either conventional orhybrid cultivars
Willow genetic improvement was led by the researchcommunity closely linked to plant breeding programmesWillow and poplar have the longest record of public invest-ment in genetic improvement that can be traced back to1920s in the United Kingdom and United States respec-tively Like switchgrass breeding programmes for willoware connected to public research efforts The United King-dom in partnership with the programme based at CornellUniversity remains the European leader in willowimprovement with a long‐term breeding effort closelylinked to supporting biological research at Rothamsted Inwillow F1 hybrids have produced impressive yield gainsover parental germplasm by capturing hybrid vigour Over30 willow clones are commercially available in the UnitedStates and Europe and a further ~90 are under precommer-cial testing (Table 1)
Compared with willow and poplar miscanthus is a rela-tive newcomer with all the current breeding programmesstarting in the 2000s Clonal M times giganteus propagated byrhizomes is expected to be replaced by more readily scal-able seeded hybrids from intra‐ (M sinensis) and inter‐(M sacchariflorus times M sinensis) species crosses with highseed multiplication rates (of gt2000) The first group ofhybrid cultivars is expected to be market‐ready around2022
Of the four genera used as PBCs Populus is the mostadvanced in terms of achievements in biological researchas a result of its use as a model for basic research of treesMuch of this biological research is not directly connectedto plant breeding Nevertheless reflecting the fact thatpoplar is widely grown as a single‐stem tree in SRF thereare about 60 commercially available clones and an addi-tional 80 clones in commercial pipelines (Table 1) Trans-genic poplar hybrids have moved beyond proof of conceptto commercial reality in China
Many PBC programmes have initiated long‐term con-ventional recurrent selection breeding cycles for populationimprovement which is a key process in increasing yieldthrough hybrid vigour As this approach requires manyyears most programmes are experimenting with molecularbreeding methods as these have the potential to accelerateprecision breeding For all four PBCs investments in basicgenetic and genomic resources including the developmentof mapping populations for QTLs and whole genomesequences are available to support long‐term advancesMore recently association genetics with panels of diversegermplasm are being used as training populations for GSmodels (Table 5) These efforts are benefitting from pub-licly available DNA sequences and whole genome assem-blies in crop databases Key to these accelerated breedingtechnologies are developments in novel phenomics tech-nologies to bridge the genotypephenotype gap In poplarnovel remote sensing field phenotyping is now beingdeployed to assist breeders These advances are being com-bined with in vitro and in planta modern molecular breed-ing techniques such as CRISPR (Table 5) CRISPRtechnology for genome editing has been proven in poplar
FIGURE 2 A schematic developmentpathway for miscanthus in the UnitedKingdom related to the investment in RampDprojects at Aberystwyth (top coloured areasfor projects in the three categories basicresearch breeding and commercialupscaling) leading to a projected croppingarea of 350000 ha by 2030 with clonal andsuccessive ranges of improved seed‐basedhybrids Purple represents theBiotechnology and Biological SciencesResearch Council (BBSRC) and brown theDepartment for Environment Food andRural Affairs (Defra) (UK Nationalfunding) blue bars represent EU fundingand green private sector funding (Terravestaand CERES) and GIANT‐LINK andMiscanthus Upscaling Technology (MUST)are publicndashprivate‐initiatives (PPI)
CLIFTON‐BROWN ET AL | 25
This technology is also being applied in switchgrass andmiscanthus (Table 5) but the future of CRISPR in com-mercial breeding for the European market is uncertain inthe light of recent ECJ 2018 rulings
There is integration of research and plant breeding itselfin all four PBCs Therefore estimating the ongoing costs ofmaintaining these breeding programmes is difficult Invest-ment in research also seeks wider benefits associated withtechnological advances in plant science rather than cultivardevelopment per se However in all cases the conventionalbreeding cycle shown in Figure 1 is the basic ldquoenginerdquo withmolecular technologies (‐omics) serving to accelerate thisengine The history of the development shows that the exis-tence of these breeding programmes is essential to gain ben-efits from the biological research Despite this it is thisessential step that is at most risk from reductions in invest-ment A conventional breeding programme typicallyrequires a breeder and several technicians who are supported
over the long term (20ndash30 years Figure 1) at costs of about05 to 10 million Euro per year (as of 2018) The analysisreported here shows that the time needed to perform onecycle of conventional breeding bringing germplasm fromthe wild to a commercial hybrid ranged from 11 years inswitchgrass to 26 years in poplar (Figure 1) In a maturecrop grown on over 100000 ha with effective cultivar pro-tection and a suitable business model this level of revenuecould come from royalties Until such levels are reachedPBCs lie in the innovation valley of death (Figure 3) andneed public support
Applying industrial ldquotechnology readiness levelsrdquo (TRL)originally developed for aerospace (Heacuteder 2017) to ourplant breeding efforts we estimate many promising hybridscultivars are at TRL levels of 3ndash4 In Table 3 experts ineach crop estimate that it would take 3 years from now toupscale planting material from leading cultivars in plot tri-als to 100 ha
FIGURE 3 A schematic relating some of the steps in the innovation chain from relatively basic crop science research through to thedeployment in commercial cropping systems and value chains The shape of the funnel above the expanding development and deploymentrepresents the availability of investment along the development chain from relatively basic research at the top to the upscaled deployment at thebottom Plant breeding links the research effort with the development of cropping systems The constriction represents the constrained fundingfor breeding that links conventional public research investment and the potential returns from commercial development The handover pointsbetween publicly funded work to develop the germplasm resources (often known as prebreeding) the breeding and the subsequent cropdevelopment are shown on the left The constriction point is aggravated by the lack academic rewards for this essential breeding activity Theoutcome is such that this innovation system is constrained by the precarious resourcing of plant breeding The authorsrsquo assessment ofdevelopment status of the four species is shown (poplar having two one for short‐rotation coppice (SRC) poplar and one for the more traditionalshort‐rotation forestry (SRF)) The four new perennial biomass crops (PBCs) are now in the critical phase of depending of plant breedingprogress without the income stream from a large crop production base
26 | CLIFTON‐BROWN ET AL
Taking the UK example mentioned in the introductionplanting rates of ~35000 ha per year from 2022 onwardsare needed to reach over 1 m ha by 2050 Ongoing workin the UK‐funded Miscanthus Upscaling Technology(MUST) project shows that ramping annual hybrid seedproduction from the current level of sufficient seed for10 ha in 2018 to 35000 ha would take about 5 yearsassuming no setbacks If current hybrids of any of the fourPBCs in the upscaling pipeline fail on any step for exam-ple lower than expected multiplication rates or unforeseenagronomic barriers then further selections from ongoingbreeding are needed to replace earlier candidates
In conclusion the breeding foundations have been laidwell for switchgrass miscanthus willow and poplar owingto public funding over the long time periods necessaryImproved cultivars or genotypes are available that could bescaled up over a few years if real sustained market oppor-tunities emerged in response to sustained favourable poli-cies and industrial market pull The potential contributionsof growing and using these PBCs for socioeconomic andenvironmental benefits are clear but how farmers andothers in commercial value chains are rewarded for mass‐scale deployment as is necessary is not obvious at presentTherefore mass‐scale deployment of these lignocellulosecrops needs developments outside the breeding arenas todrive breeding activities more rapidly and extensively
8 | UNCITED REFERENCE
Huang et al (2019)
ACKNOWLEDGEMENTS
UK The UK‐led miscanthus research and breeding wasmainly supported by the Biotechnology and BiologicalSciences Research Council (BBSRC) Department for Envi-ronment Food and Rural Affairs (Defra) the BBSRC CSPstrategic funding grant BBCSP17301 Innovate UKBBSRC ldquoMUSTrdquo BBN0161491 CERES Inc and Ter-ravesta Ltd through the GIANT‐LINK project (LK0863)Genomic selection and genomewide association study activi-ties were supported by BBSRC grant BBK01711X1 theBBSRC strategic programme grant on Energy Grasses ampBio‐refining BBSEW10963A01 The UK‐led willow RampDwork reported here was supported by BBSRC (BBSEC00005199 BBSEC00005201 BBG0162161 BBE0068331 BBG00580X1 and BBSEC000I0410) Defra(NF0424 NF0426) and the Department of Trade and Indus-try (DTI) (BW6005990000) IT The Brain Gain Program(Rientro dei cervelli) of the Italian Ministry of EducationUniversity and Research supports Antoine Harfouche USContributions by Gerald Tuskan to this manuscript were sup-ported by the Center for Bioenergy Innovation a US
Department of Energy Bioenergy Research Center supportedby the Office of Biological and Environmental Research inthe DOE Office of Science under contract number DE‐AC05‐00OR22725 Willow breeding efforts at CornellUniversity have been supported by grants from the USDepartment of Agriculture National Institute of Food andAgriculture Contributions by the University of Illinois weresupported primarily by the DOE Office of Science Office ofBiological and Environmental Research (BER) grant nosDE‐SC0006634 DE‐SC0012379 and DE‐SC0018420 (Cen-ter for Advanced Bioenergy and Bioproducts Innovation)and the Energy Biosciences Institute EU We would like tofurther acknowledge contributions from the EU projectsldquoOPTIMISCrdquo FP7‐289159 on miscanthus and ldquoWATBIOrdquoFP7‐311929 on poplar and miscanthus as well as ldquoGRACErdquoH2020‐EU326 Biobased Industries Joint Technology Ini-tiative (BBI‐JTI) Project ID 745012 on miscanthus
CONFLICT OF INTEREST
The authors declare that progress reported in this paperwhich includes input from industrial partners is not biasedby their business interests
ORCID
John Clifton‐Brown httporcidorg0000-0001-6477-5452Antoine Harfouche httporcidorg0000-0003-3998-0694Lawrence B Smart httporcidorg0000-0002-7812-7736Danny Awty‐Carroll httporcidorg0000-0001-5855-0775VasileBotnari httpsorcidorg0000-0002-0470-0384Lindsay V Clark httporcidorg0000-0002-3881-9252SalvatoreCosentino httporcidorg0000-0001-7076-8777Lin SHuang httporcidorg0000-0003-3079-326XJon P McCalmont httporcidorg0000-0002-5978-9574Paul Robson httporcidorg0000-0003-1841-3594AnatoliiSandu httporcidorg0000-0002-7900-448XDaniloScordia httporcidorg0000-0002-3822-788XRezaShafiei httporcidorg0000-0003-0891-0730Gail Taylor httporcidorg0000-0001-8470-6390IrisLewandowski httporcidorg0000-0002-0388-4521
REFERENCES
Alburquerque N Baldacci‐Cresp F Baucher M Casacuberta JM Collonnier C ElJaziri M hellip Burgos L (2016) New trans-formation technologies for trees In C Vettori F Gallardo H
CLIFTON‐BROWN ET AL | 27
Haumlggman V Kazana F Migliacci G Pilateamp M Fladung(Eds) Biosafety of forest transgenic trees (pp 31ndash66) DordrechtNetherlands Springer
Argus G W (2007) Salix (Salicaceae) distribution maps and a syn-opsis of their classification in North America north of MexicoHarvard Papers in Botany 12 335ndash368 httpsdoiorg1031001043-4534(2007)12[335SSDMAA]20CO2
Atienza S G Ramirez M C amp Martin A (2003) Mapping‐QTLscontrolling flowering date in Miscanthus sinensis Anderss CerealResearch Communications 31 265ndash271
Atienza S G Satovic Z Petersen K K Dolstra O amp Martin A(2003) Identification of QTLs influencing agronomic traits inMiscanthus sinensis Anderss I Total height flag‐leaf height andstem diameter Theoretical and Applied Genetics 107 123ndash129
Atienza S G Satovic Z Petersen K K Dolstra O amp Martin A(2003) Influencing combustion quality in Miscanthus sinensisAnderss Identification of QTLs for calcium phosphorus and sul-phur content Plant Breeding 122 141ndash145
Bdeir R Muchero W Yordanov Y Tuskan G A Busov V ampGailing O (2017) Quantitative trait locus mapping of Populusbark features and stem diameter BMC Plant Biology 17 224httpsdoiorg101186s12870-017-1166-4
Beard T R Ford G S Koutsky T M amp Spiwak L J (2009) AValley of Death in the innovation sequence An economic investi-gation Research Evaluation 18 343ndash356 httpsdoiorg103152095820209X481057
Berguson W McMahon B amp Riemenschneider D (2017) Addi-tive and non‐additive genetic variances for tree growth in severalhybrid poplar populations and implications regarding breedingstrategy Silvae Genetica 66(1) 33ndash39 httpsdoiorg101515sg-2017-0005
Berlin S Trybush S O Fogelqvist J Gyllenstrand N Halling-baumlck H R Aringhman I hellip Hanley S J (2014) Genetic diversitypopulation structure and phenotypic variation in European Salixviminalis L (Salicaceae) Tree Genetics amp Genomes 10 1595ndash1610 httpsdoiorg101007s11295-014-0782-5
Bernardo R amp Yu J (2007) Prospects for genomewide selectionfor quantitative traits in maize Crop Science 47 1082ndash1090httpsdoiorg102135cropsci2006110690
Biswal A K Atmodjo M A Li M Baxter H L Yoo C GPu Y hellip Mohnen D (2018) Sugar release and growth of bio-fuel crops are improved by downregulation of pectin biosynthesisNature Biotechnology 36 249ndash257
Bmu (2009) National biomass action plan for Germany (p 17) Bun-desministerium fuumlr Umwelt Naturschutz und ReaktorsicherheitBundesministerium fuumlr Ernaumlhrung (BMU) amp Landwirtschaft undVerbraucherschutz (BMELV) 11055 Berlin | Germany Retrievedfrom httpwwwbmeldeSharedDocsDownloadsENPublicationsBiomassActionPlanpdf__blob=publicationFile
Brummer E C (1999) Capturing heterosis in forage crop cultivardevelopment Crop Science 39 943ndash954 httpsdoiorg102135cropsci19990011183X003900040001x
Budsberg E Crawford J T Morgan H Chin W S Bura R ampGustafson R (2016) Hydrocarbon bio‐jet fuel from bioconver-sion of poplar biomass Life cycle assessment Biotechnology forBiofuels 9 170 httpsdoiorg101186s13068-016-0582-2
Burris K P Dlugosz E M Collins A G Stewart C N amp Lena-ghan S C (2016) Development of a rapid low‐cost protoplasttransfection system for switchgrass (Panicum virgatum L) Plant
Cell Reports 35 693ndash704 httpsdoiorg101007s00299-015-1913-7
Casler M (2012) Switchgrass breeding genetics and genomics InA Monti (Ed) Switchgrass (pp 29ndash54) London UK Springer
Casler M Mitchell R amp Vogel K (2012) Switchgrass In CKole C P Joshi amp D R Shonnard (Eds) Handbook of bioen-ergy crop plants Vol 2 New York NY Taylor amp Francis
Casler M D amp Ramstein G P (2018) Breeding for biomass yieldin switchgrass using surrogate measures of yield BioEnergyResearch 11 6ndash12
Casler M D Sosa S Hoffman L Mayton H Ernst C Adler PR hellip Bonos S A (2017) Biomass Yield of Switchgrass Culti-vars under High‐ versus Low‐Input Conditions Crop Science 57821ndash832 httpsdoiorg102135cropsci2016080698
Casler M D amp Vogel K P (2014) Selection for biomass yield inupland lowland and hybrid switchgrass Crop Science 54 626ndash636 httpsdoiorg102135cropsci2013040239
Casler M D Vogel K P amp Harrison M (2015) Switchgrassgermplasm resources Crop Science 55 2463ndash2478 httpsdoiorg102135cropsci2015020076
Casler M D Vogel K P Lee D K Mitchell R B Adler P RSulc R M hellip Moore K J (2018) 30 Years of progress towardincreasing biomass yield of switchgrass and big bluestem CropScience 58 1242
Clark L V Brummer J E Głowacka K Hall M C Heo K PengJ hellip Sacks E J (2014) A footprint of past climate change on thediversity and population structure of Miscanthus sinensis Annals ofBotany 114 97ndash107 httpsdoiorg101093aobmcu084
Clark L V Dzyubenko E Dzyubenko N Bagmet L SabitovA Chebukin P hellip Sacks E J (2016) Ecological characteristicsand in situ genetic associations for yield‐component traits of wildMiscanthus from eastern Russia Annals of Botany 118 941ndash955
Clark L V Jin X Petersen K K Anzoua K G Bagmet LChebukin P hellip Sacks E J(2018) Population structure of Mis-canthus sacchariflorus reveals two major polyploidization eventstetraploid‐mediated unidirectional introgression from diploid Msinensis and diversity centred around the Yellow Sea Annals ofBotany 20 1ndash18 httpsdoiorg101093aobmcy1161
Clark L V Stewart J R Nishiwaki A Toma Y Kjeldsen J BJoslashrgensen U hellip Sacks E J (2015) Genetic structure ofMiscanthus sinensis and Miscanthus sacchariflorus in Japan indi-cates a gradient of bidirectional but asymmetric introgressionJournal of Experimental Botany 66 4213ndash4225
Clifton‐Brown J Hastings A Mos M McCalmont J P Ashman CAwty‐Carroll DhellipCracroft‐EleyW (2017) Progress in upscalingMis-canthus biomass production for the European bio‐economy with seed‐based hybridsGlobal Change Biology Bioenergy 9 6ndash17
Clifton‐Brown J Schwarz K U amp Hastings A (2015) History ofthe development of Miscanthus as a bioenergy crop From smallbeginnings to potential realisation Biology and Environment‐Pro-ceedings of the Royal Irish Academy 115B 45ndash57
Clifton‐Brown J Senior H Purdy S Horsnell R Lankamp BMuumlennekhoff A-K hellip Bentley A R (2018) Investigating thepotential of novel non‐woven fabrics to increase pollination effi-ciency in plant breeding PloS One (Accepted)
Crawford J T Shan CW Budsberg E Morgan H Bura R ampGus-tafson R (2016) Hydrocarbon bio‐jet fuel from bioconversion ofpoplar biomass Techno‐economic assessment Biotechnology forBiofuels 9 141 httpsdoiorg101186s13068-016-0545-7
28 | CLIFTON‐BROWN ET AL
Dalton S (2013) Biotechnology of Miscanthus In S M Jain amp SD Gupta (Eds) Biotechnology of neglected and underutilizedcrops (pp 243ndash294) Dordrecht Netherlands Springer
Davey C L Robson P Hawkins S Farrar K Clifton‐Brown J CDonnison I S amp Slavov G T (2017) Genetic relationshipsbetween spring emergence canopy phenology and biomass yieldincrease the accuracy of genomic prediction in Miscanthus Journalof Experimental Botany 68 5093ndash5102 httpsdoiorg101093jxberx339
David X amp Anderson X (2002) Aspen and larch genetics cooper-ative annual report 13 St Paul MN Department of ForestResources University of Minnesota
Deblock M (1990) Factors influencing the tissue‐culture and theAgrobacterium‐Tumefaciens‐mediated transformation of hybridaspen and poplar clones Plant Physiology 93 1110ndash1116httpsdoiorg101104pp9331110
Defra (2002) The role of future public research investment in thegenetic improvement of UK‐grown crops (p 220) LondonRetrieved from sciencesearchdefragovukDefaultaspxMenu=MenuampModule=MoreampLocation=NoneampCompleted=220ampProjectID=10412
Dewoody J Trewin H amp Taylor G (2015) Genetic and morpho-logical differentiation in Populus nigra L Isolation by coloniza-tion or isolation by adaptation Molecular Ecology 24 2641ndash2655
Dong H Liu S Clark L V Sharma S Gifford J M Juvik JA hellip Sacks E J (2018) Genetic mapping of biomass yield inthree interconnected Miscanthus populations Global Change Biol-ogy Bioenergy 10 165ndash185
Dukes (2017) Digest of UK Energy Statistics (DUKES) (p 264) Lon-don UK Department for Business Energy amp Industrial StrategyRetrieved from wwwgovukgovernmentcollectionsdigest-of-uk-energy-statistics-dukes
Eckenwalder J (1996) Systematics and evolution of Populus In RStettler H Bradshaw J Heilman amp T Hinckley (Eds) Biologyof Populus and its implications for management and conservation(pp 7‐32) Ottawa ON National Research Council of Canada
Elshire R J Glaubitz J C Sun Q Poland J A Kawamoto KBuckler E S amp Mitchell S E (2011) A robust simple geno-typing‐by‐sequencing (GBS) approach for high diversity speciesPloS One 6 e19379
Evans H (2016) Bioenergy crops in the UK Case studies of suc-cessful whole farm integration (p 23) Loughborough UKEnergy Technologies Institute Retrieved from httpswwweticouklibrarybioenergy-crops-in-the-uk-case-studies-on-successful-whole-farm-integration-evidence-pack
Evans H (2017) Increasing UK biomass production through moreproductive use of land (p 7) Loughborough UK Energy Tech-nologies Institute Retrieved from httpswwweticouklibraryan-eti-perspective-increasing-uk-biomass-production-through-more-productive-use-of-land
Evans J Sanciangco M D Lau K H Crisovan E Barry KDaum C hellip Buell C R (2018) Extensive genetic diversity ispresent within North American switchgrass germplasm The PlantGenome 11 1ndash16 httpsdoiorg103835plantgenome2017060055
Evans L M Slavov G T Rodgers‐Melnick E Martin J RanjanP Muchero W hellip DiFazio S P (2014) Population genomicsof Populus trichocarpa identifies signatures of selection and
adaptive trait associations Nature Genetics 46 1089ndash1096httpsdoiorg101038ng3075
Fabio E S Volk T A Miller R O Serapiglia M J Gauch HG Van Rees K C J hellip Smart L B (2017) Genotypetimes envi-ronment interaction analysis of North American shrub willowyield trials confirms superior performance of triploid hybrids Glo-bal Change Biology Bioenergy 9 445ndash459 httpsdoiorg101111gcbb12344
Fahrenkrog A M Neves L G Resende M F R Vazquez A Ide los Campos G Dervinis C hellip Kirst M (2017) Genome‐wide association study reveals putative regulators of bioenergytraits in Populus deltoides New Phytologist 213 799ndash811
Fan D Liu T T Li C F Jiao B Li S Hou Y S amp Luo KM (2015) Efficient CRISPRCas9‐mediated targeted mutagenesisin Populus in the first generation Scientific Reports 5 12217
Feng X P Lourgant K Castric V Saumitou-Laprade P ZhengB S Jiang D M amp Brancourt-Hulmel M (2014) The discov-ery of natural accessions related to Miscanthus x giganteus usingchloroplast DNA Crop Science 54 1645ndash1655
Fillatti J J Sellmer J Mccown B Haissig B amp Comai L(1987) Agrobacterium-mediated transformation and regenerationof Populus Molecular amp General Genetics 206 192ndash199
Fu Y B (2015) Understanding crop genetic diversity under modernplant breeding Theoretical and Applied Genetics 128 2131ndash2142 httpsdoiorg101007s00122-015-2585-y
Fu C Mielenz J R Xiao X Ge Y Hamilton C Y RodriguezM hellip Wang Z‐Y (2011) Genetic manipulation of ligninreduces recalcitrance and improves ethanol production fromswitchgrass Proceedings of the National Academy of Sciences ofthe United States of America 108 3803ndash3808 httpsdoiorg101073pnas1100310108
Gao C (2018) The future of CRISPR technologies in agricultureNature Reviews Molecular Cell Biology 19(5) 275ndash276
Gegas V Gay A Camargo A amp Doonan J (2014) Challengesof Crop Phenomics in the Post-genomic Era In J Hancock (Ed)Phenomics (pp 142ndash171) Boca Raton FL CRC Press
Geraldes A DiFazio S P Slavov G T Ranjan P Muchero WHannemann J hellip Tuskan G A (2013) A 34K SNP genotypingarray for Populus trichocarpa Design application to the study ofnatural populations and transferability to other Populus speciesMolecular Ecology Resources 13 306ndash323
Gifford J M Chae W B Swaminathan K Moose S P amp JuvikJ A (2015) Mapping the genome of Miscanthus sinensis forQTL associated with biomass productivity Global Change Biol-ogy Bioenergy 7 797ndash810
Goggin F L Lorence A amp Topp C N (2015) Applying high‐throughput phenotyping to plant‐insect interactions Picturingmore resistant crops Current Opinion in Insect Science 9 69ndash76httpsdoiorg101016jcois201503002
Grabowski P P Evans J Daum C Deshpande S Barry K WKennedy M hellip Casler M D (2017) Genome‐wide associationswith flowering time in switchgrass using exome‐capture sequenc-ing data New Phytologist 213 154ndash169 httpsdoiorg101111nph14101
Greef J M amp Deuter M (1993) Syntaxonomy of Miscanthus xgiganteus GREEF et DEU Angewandte Botanik 67 87ndash90
Guerra F P Richards J H Fiehn O Famula R Stanton B JShuren R hellip Neale D B (2016) Analysis of the genetic varia-tion in growth ecophysiology and chemical and metabolomic
CLIFTON‐BROWN ET AL | 29
composition of wood of Populus trichocarpa provenances TreeGenetics amp Genomes 12 6 httpsdoiorg101007s11295-015-0965-8
Guo H P Shao R X Hong C T Hu H K Zheng B S ampZhang Q X (2013) Rapid in vitro propagation of bioenergy cropMiscanthus sacchariflorus Applied Mechanics and Materials 260181ndash186
Hallingbaumlck H R Fogelqvist J Powers S J Turrion‐Gomez JRossiter R Amey J hellip Roumlnnberg‐Waumlstljung A‐C (2016)Association mapping in Salix viminalis L (Salicaceae)ndashIdentifica-tion of candidate genes associated with growth and phenologyGlobal Change Biology Bioenergy 8 670ndash685
Hanley S J amp Karp A (2014) Genetic strategies for dissectingcomplex traits in biomass willows (Salix spp) Tree Physiology34 1167ndash1180 httpsdoiorg101093treephystpt089
Harfouche A Meilan R amp Altman A (2011) Tree genetic engi-neering and applications to sustainable forestry and biomass pro-duction Trends in Biotechnology 29 9ndash17 httpsdoiorg101016jtibtech201009003
Harfouche A Meilan R amp Altman A (2014) Molecular and phys-iological responses to abiotic stress in forest trees and their rele-vance to tree improvement Tree Physiology 34 1181ndash1198httpsdoiorg101093treephystpu012
Harfouche A Meilan R Kirst M Morgante M Boerjan WSabatti M amp Mugnozza G S (2012) Accelerating the domesti-cation of forest trees in a changing world Trends in PlantScience 17 64ndash72 httpsdoiorg101016jtplants201111005
Hastings A Clifton‐Brown J Wattenbach M Mitchell C PStampfl P amp Smith P (2009) Future energy potential of Mis-canthus in Europe Global Change Biology Bioenergy 1 180ndash196
Hastings A Mos M Yesufu J AMccalmont J Schwarz KShafei RhellipClifton-Brown J (2017) Economic and environmen-tal assessment of seed and rhizome propagated Miscanthus in theUK Frontiers in Plant Science 8 16 httpsdoiorg103389fpls201701058
Heacuteder M (2017) From NASA to EU The evolution of the TRLscale in Public Sector Innovation The Innovation Journal 22 1ndash23 httpsdoiorg103389fpls201701058
Heffner E L Sorrells M E amp Jannink J L (2009) Genomicselection for crop improvement Crop Science 49 1ndash12 httpsdoiorg102135cropsci2008080512
Hodkinson T R Klaas M Jones M B Prickett R amp Barth S(2015) Miscanthus A case study for the utilization of naturalgenetic variation Plant Genetic Resources‐Characterization andUtilization 13 219ndash237 httpsdoiorg101017S147926211400094X
Hodkinson T R amp Renvoize S (2001) Nomenclature of Miscant-hus xgiganteus (Poaceae) Kew Bulletin 56 759ndash760 httpsdoiorg1023074117709
Houle D Govindaraju D R amp Omholt S (2010) Phenomics Thenext challenge Nature Reviews Genetics 11 855ndash866 httpsdoiorg101038nrg2897
Huang X H Wei X H Sang T Zhao Q Feng Q Zhao Yhellip Han B (2010) Genome‐wide association studies of 14 agro-nomic traits in rice landraces Nature Genetics 42 961ndashU976
Hwang O‐J Cho M‐A Han Y‐J Kim Y‐M Lim S‐H KimD‐S hellip Kim J‐I (2014) Agrobacterium‐mediated genetic trans-formation of Miscanthus sinensis Plant Cell Tissue and OrganCulture (PCTOC) 117 51ndash63
Hwang O‐J Lim S‐H Han Y‐J Shin A‐Y Kim D‐S ampKim J‐I (2014) Phenotypic characterization of transgenic Mis-canthus sinensis plants overexpressing Arabidopsis phytochromeB International Journal of Photoenergy 2014501016
Ilstedt B (1996) Genetics and performance of Belgian poplar clonestested in Sweden International Journal of Forest Genetics 3183ndash195
Ingvarsson P K Garcia M V Hall D Luquez V amp Jansson S(2006) Clinal variation in phyB2 a candidate gene for day‐length‐induced growth cessation and bud set across a latitudinalgradient in European aspen (Populus tremula) Genetics 1721845ndash1853
Ingvarsson P K Garcia M V Luquez V Hall D amp Jansson S(2008) Nucleotide polymorphism and phenotypic associationswithin and around the phytochrome B2 locus in European aspen(Populus tremula Salicaceae) Genetics 178 2217ndash2226 httpsdoiorg101534genetics107082354
Jannink J‐L Lorenz A J amp Iwata H (2010) Genomic selectionin plant breeding From theory to practice Briefings in FunctionalGenomics 9 166ndash177 httpsdoiorg101093bfgpelq001
Jiang J Guan Y Mccormick S Juvik J Lubberstedt T amp FeiS‐Z (2017) Gametophytic self‐incompatibility is operative inMiscanthus sinensis (Poaceae) and is affected by pistil age CropScience 57 1948ndash1956
Jones H D (2015a) Future of breeding by genome editing is in thehands of regulators GM Crops amp Food 6 223ndash232
Jones H D (2015b) Regulatory uncertainty over genome editingNature Plants 1 14011
Kalinina O Nunn C Sanderson R Hastings A F S van der Weijde TOumlzguumlven MhellipClifton‐Brown J C (2017) ExtendingMiscanthus cul-tivation with novel germplasm at six contrasting sites Frontiers in PlantScience 8563 httpsdoiorg103389fpls201700563
Karp A Hanley S J Trybush S O Macalpine W Pei M ampShield I (2011) Genetic improvement of willow for bioenergyand biofuels free access Journal of Integrative Plant Biology 53151ndash165 httpsdoiorg101111j1744-7909201001015x
Kersten B Faivre Rampant P Mader M Le Paslier M‐C Bou-non R Berard A hellip Fladung M (2016) Genome sequences ofPopulus tremula chloroplast and mitochondrion Implications forholistic poplar breeding PloS One 11 e0147209 httpsdoiorg101371journalpone0147209
Kiesel A Nunn C Iqbal Y Van der Weijde T Wagner MOumlzguumlven M hellip Lewandowski I (2017) Site‐specific manage-ment of Miscanthus genotypes for combustion and anaerobicdigestion A comparison of energy yields Frontiers in PlantScience 8 347 httpsdoiorg103389fpls201700347
Kim H Kim S T Ryu J Kang B C Kim J S amp Kim S G(2017) CRISPRCpf1‐mediated DNA‐free plant genome editingNature Communications 8 14406 httpsdoiorg101038ncomms14406
King Z R Bray A L Lafayette P R amp Parrott W A (2014)Biolistic transformation of elite genotypes of switchgrass (Pan-icum virgatum L) Plant Cell Reports 33 313ndash322 httpsdoiorg101007s00299-013-1531-1
Kopp R F Maynard C A De Niella P R Smart L B amp Abra-hamson L P (2002) Collection and storage of pollen from Salix(Salicaceae) American Journal of Botany 89 248ndash252
Krzyżak J Pogrzeba M Rusinowski S Clifton‐Brown J McCal-mont J P Kiesel A hellip Mos M (2017) Heavy metal uptake
30 | CLIFTON‐BROWN ET AL
by novel miscanthus seed‐based hybrids cultivated in heavy metalcontaminated soil Civil and Environmental Engineering Reports26 121ndash132 httpsdoiorg101515ceer-2017-0040
Kuzovkina Y A amp Quigley M F (2005) Willows beyond wet-lands Uses of Salix L species for environmental projects WaterAir and Soil Pollution 162 183ndash204 httpsdoiorg101007s11270-005-6272-5
Lazarus W Headlee W L amp Zalesny R S (2015) Impacts of sup-plyshed‐level differences in productivity and land costs on the eco-nomics of hybrid poplar production in Minnesota USA BioEnergyResearch 8 231ndash248 httpsdoiorg101007s12155-014-9520-y
Lewandowski I (2015) Securing a sustainable biomass supply in agrowing bioeconomy Global Food Security 6 34ndash42 httpsdoiorg101016jgfs201510001
Lewandowski I (2016) The role of perennial biomass crops in agrowing bioeconomy In S Barth D Murphy‐Bokern O Kalin-ina G Taylor amp M Jones (Eds) Perennial biomass crops for aresource‐constrained world (pp 3ndash13) Cham SwitzerlandSpringer Switzerland AG
Li J L Meilan R Ma C Barish M amp Strauss S H (2008)Stability of herbicide resistance over 8 years of coppice in field‐grown genetically engineered poplars Western Journal of AppliedForestry 23 89ndash93
Li R Y amp Qu R D (2011) High throughput Agrobacterium‐medi-ated switchgrass transformation Biomass amp Bioenergy 35 1046ndash1054 httpsdoiorg101016jbiombioe201011025
Lindegaard K N amp Barker J H A (1997) Breeding willows forbiomass Aspects of Applied Biology 49 155ndash162
Linde‐Laursen I B (1993) Cytogenetic analysis of MiscanthusGiganteus an interspecific hybrid Hereditas 119 297ndash300httpsdoiorg101111j1601-5223199300297x
Liu Y Merrick P Zhang Z Z Ji C H Yang B amp Fei S Z(2018) Targeted mutagenesis in tetraploid switchgrass (Panicumvirgatum L) using CRISPRCas9 Plant Biotechnology Journal16 381ndash393
Liu L L Wu Y Q Wang Y W amp Samuels T (2012) A high‐density simple sequence repeat‐based genetic linkage map ofswitchgrass G3 (Bethesda Md) 2 357ndash370
Lovett A Suumlnnenberg G amp Dockerty T (2014) The availabilityof land for perennial energy crops in Great Britain GlobalChange Biology Bioenergy 6 99ndash107 httpsdoiorg101111gcbb12147
Lozano‐Juste J amp Cutler S R (2014) Plant genome engineering infull bloom Trends in Plant Science 19 284ndash287 httpsdoiorg101016jtplants201402014
Lu F Lipka A E Glaubitz J Elshire R Cherney J H CaslerM D hellip Costich D E (2013) Switchgrass Genomic DiversityPloidy and Evolution Novel Insights from a Network‐Based SNPDiscovery Protocol Plos Genetics 9 e1003215
Ludovisi R Tauro F Salvati R Khoury S Mugnozza G S ampHarfouche A (2017) UAV‐based thermal imaging for high‐throughput field phenotyping of black poplar response to droughtFrontiers in Plant Science 81681
Macalpine W Shield I amp Karp A (2010) Seed to near market vari-ety the BEGIN willow breeding pipeline 2003ndash2010 and beyond InBioten conference proceedings Birmingham pp 21ndash23
Macalpine W J Shield I F Trybush S O Hayes C M ampKarp A (2008) Overcoming barriers to crossing in willow (Salixspp) breeding Aspects of Applied Biology 90 173ndash180
Mahfouz M M (2017) The efficient tool CRISPR‐Cpf1 NaturePlants 3 17028
Malinowska M Donnison I S amp Robson P R (2017) Phenomicsanalysis of drought responses in Miscanthus collected from differ-ent geographical locations Global Change Biology Bioenergy 978ndash91
Maroder H L Prego I A Facciuto G R amp Maldonado S B(2000) Storage behaviour of Salix alba and Salix matsudanaseeds Annals of Botany 86 1017ndash1021 httpsdoiorg101006anbo20001265
Martinez‐Reyna J amp Vogel K (2002) Incompatibility systems inswitchgrass Crop Science 42 1800ndash1805 httpsdoiorg102135cropsci20021800
Maurya J P amp Bhalerao R P (2017) Photoperiod‐and tempera-ture‐mediated control of growth cessation and dormancy in treesA molecular perspective Annals of Botany 120 351ndash360httpsdoiorg101093aobmcx061
McCalmont J P Hastings A Mcnamara N P Richter G MRobson P Donnison I S amp Clifton‐Brown J (2017) Environ-mental costs and benefits of growing Miscanthus for bioenergy inthe UK Global Change Biology Bioenergy 9 489ndash507
McCracken A R amp Dawson W M (1997) Using mixtures of wil-low clones as a means of controlling rust disease Aspects ofApplied Biology 49 97ndash103
McKown A D Klaacutepště J Guy R D Geraldes A Porth I Hanne-mann JhellipDouglas C J (2014) Genome‐wide association implicatesnumerous genes underlying ecological trait variation in natural popula-tions ofPopulus trichocarpaNew Phytologist 203 535ndash553
Merrick P amp Fei S Z (2015) Plant regeneration and genetic transforma-tion in switchgrass ndash A review Journal of Integrative Agriculture 14483ndash493 httpsdoiorg101016S2095-3119(14)60921-7
Moreno‐Mateos M A Fernandez J P Rouet R Lane M AVejnar C E Mis E hellip Giraldez A J (2017) CRISPR‐Cpf1mediates efficient homology‐directed repair and temperature‐con-trolled genome editing Nature Communications 8 2024 httpsdoiorg101038s41467-017-01836-2
Mosseler A (1990) Hybrid performance and species crossabilityrelationships in willows (Salix) Canadian Journal of Botany 682329ndash2338
Muchero W Guo J DiFazio S P Chen J‐G Ranjan P Slavov GT hellip Tuskan G A (2015) High‐resolution genetic mapping ofallelic variants associated with cell wall chemistry in Populus BMCGenomics 16 24 httpsdoiorg101186s12864-015-1215-z
Neale D B amp Kremer A (2011) Forest tree genomics Growingresources and applications Nature Reviews Genetics 12 111ndash122 httpsdoiorg101038nrg2931
Nunn C Hastings A F S J Kalinina O Oumlzguumlven M SchuumlleH Tarakanov I G hellip Clifton‐Brown J C (2017) Environmen-tal influences on the growing season duration and ripening ofdiverse Miscanthus germplasm grown in six countries Frontiersin Plant Science 8 907 httpsdoiorg103389fpls201700907
Ogawa Y Honda M Kondo Y amp Hara‐Nishimura I (2016) Anefficient Agrobacterium‐mediated transformation method forswitchgrass genotypes using Type I callus Plant Biotechnology33 19ndash26
Ogawa Y Shirakawa M Koumoto Y Honda M Asami Y KondoY ampHara‐Nishimura I (2014) A simple and reliable multi‐gene trans-formation method for switchgrass Plant Cell Reports 33 1161ndash1172httpsdoiorg101007s00299-014-1605-8
CLIFTON‐BROWN ET AL | 31
Okada M Lanzatella C Saha M C Bouton J Wu R L amp Tobias CM (2010) Complete switchgrass genetic maps reveal subgenomecollinearity preferential pairing and multilocus interactions Genetics185 745ndash760 httpsdoiorg101534genetics110113910
Palomo‐Riacuteos E Macalpine W Shield I Amey J Karaoğlu C WestJ hellip Jones H D (2015) Efficient method for rapid multiplication ofclean and healthy willow clones via in vitro propagation with broadgenotype applicability Canadian Journal of Forest Research 451662ndash1667 httpsdoiorg101139cjfr-2015-0055
Pilate G Allona I Boerjan W Deacutejardin A Fladung M Gal-lardo F hellip Halpin C (2016) Lessons from 25 years of GM treefield trials in Europe and prospects for the future In C Vettori FGallardo H Haumlggman V Kazana F Migliacci G Pilateamp MFladung (Eds) Biosafety of forest transgenic trees (pp 67ndash100)Dordrecht Netherlands Springer Switzerland AG
Pinosio S Giacomello S Faivre-Rampant P Taylor G Jorge VLe Paslier M C amp Marroni F (2016) Characterization of thepoplar pan-genome by genome-wide identification of structuralvariation Molecular Biology and Evolution 33 2706ndash2719
Porth I Klaacutepště J Skyba O Friedmann M C Hannemann JEhlting J hellip Douglas C J (2013) Network analysis reveals therelationship among wood properties gene expression levels andgenotypes of natural Populus trichocarpa accessions New Phytol-ogist 200 727ndash742
Pugesgaard S Schelde K Larsen S U Laeligrke P E amp JoslashrgensenU (2015) Comparing annual and perennial crops for bioenergyproductionndashinfluence on nitrate leaching and energy balance Glo-bal Change Biology Bioenergy 7 1136ndash1149 httpsdoiorg101111gcbb12215
Quinn L D Allen D J amp Stewart J R (2010) Invasivenesspotential of Miscanthus sinensis Implications for bioenergy pro-duction in the United States Global Change Biology Bioenergy2 310ndash320 httpsdoiorg101111j1757-1707201001062x
Rae A M Robinson K M Street N R amp Taylor G (2004)Morphological and physiological traits influencing biomass pro-ductivity in short‐rotation coppice poplar Canadian Journal ofForest Research‐Revue Canadienne De Recherche Forestiere 341488ndash1498 httpsdoiorg101139x04-033
Rambaud C Arnoult S Bluteau A Mansard M C Blassiau Camp Brancourt‐Hulmel M (2013) Shoot organogenesis in threeMiscanthus species and evaluation for genetic uniformity usingAFLP analysis Plant Cell Tissue and Organ Culture (PCTOC)113 437ndash448 httpsdoiorg101007s11240-012-0284-9
Ramstein G P Evans J Kaeppler S M Mitchell R B Vogel K PBuell C R amp Casler M D (2016) Accuracy of genomic prediction inswitchgrass (Panicum virgatum L) improved by accounting for linkagedisequilibriumG3 (Bethesda Md) 6 1049ndash1062
Rayburn A L Crawford J Rayburn C M amp Juvik J A (2009)Genome size of three Miscanthus species Plant Molecular Biol-ogy Reporter 27 184 httpsdoiorg101007s11105-008-0070-3
Richardson J Isebrands J amp Ball J (2014) Ecology and physiol-ogy of poplars and willows In J Isebrands amp J Richardson(Eds) Poplars and willows Trees for society and the environ-ment (pp 92‐123) Boston MA and Rome CABI and FAO
Robison T L Rousseau R J amp Zhang J (2006) Biomass produc-tivity improvement for eastern cottonwood Biomass amp Bioenergy30 735ndash739 httpsdoiorg101016jbiombioe200601012
Robson P R Farrar K Gay A P Jensen E F Clifton‐Brown JC amp Donnison I S (2013) Variation in canopy duration in the
perennial biofuel crop Miscanthus reveals complex associationswith yield Journal of Experimental Botany 64 2373ndash2383
Robson P Jensen E Hawkins S White S R Kenobi K Clif-ton‐Brown J hellip Farrar K (2013) Accelerating the domestica-tion of a bioenergy crop Identifying and modelling morphologicaltargets for sustainable yield increase in Miscanthus Journal ofExperimental Botany 64 4143ndash4155
Sanderson M A Adler P R Boateng A A Casler M D ampSarath G (2006) Switchgrass as a biofuels feedstock in theUSA Canadian Journal of Plant Science 86 1315ndash1325httpsdoiorg104141P06-136
Scarlat N Dallemand J F Monforti‐Ferrario F amp Nita V(2015) The role of biomass and bioenergy in a future bioecon-omy Policies and facts Environmental Development 15 3ndash34httpsdoiorg101016jenvdev201503006
Serapiglia M J Gouker F E amp Smart L B (2014) Early selec-tion of novel triploid hybrids of shrub willow with improved bio-mass yield relative to diploids BMC Plant Biology 14 74httpsdoiorg1011861471-2229-14-74
Serba D Wu L Daverdin G Bahri B A Wang X Kilian Ahellip Devos K M (2013) Linkage maps of lowland and upland tet-raploid switchgrass ecotypes BioEnergy Research 6 953ndash965httpsdoiorg101007s12155-013-9315-6
Shakoor N Lee S amp Mockler T C (2017) High throughput phe-notyping to accelerate crop breeding and monitoring of diseases inthe field Current Opinion in Plant Biology 38 184ndash192 httpsdoiorg101016jpbi201705006
Shield I Macalpine W Hanley S amp Karp A (2015) Breedingwillow for short rotation coppice energy cropping In Industrialcrops Breeding for BioEnergy and Bioproducts (pp 67ndash80) NewYork NY Springer
Sjodin A Street N R Sandberg G Gustafsson P amp Jansson S (2009)The Populus Genome Integrative Explorer (PopGenIE) A new resourcefor exploring the Populus genomeNewPhytologist 182 1013ndash1025
Slavov G amp Davey C (2017) Integrated genomic prediction inbioenergy crops In Abstracts from the International Conferenceon Developing biomass crops for future climates pp 34 24ndash27September 2017 Oriel College Oxford wwwwatbioeu
Slavov G Davey C Bosch M Robson P Donnison I ampMackay I (2018a) Genomic index selection provides a pragmaticframework for setting and refining multi‐objective breeding targetsin Miscanthus Annals of Botany (in press)
Slavov G T DiFazio S P Martin J Schackwitz W MucheroW Rodgers‐Melnick E hellip Tuskan G A (2012) Genome rese-quencing reveals multiscale geographic structure and extensivelinkage disequilibrium in the forest tree Populus trichocarpa NewPhytologist 196 713ndash725
Slavov G Davey C Robson P Donnison I amp Mackay I(2018b) Domestication of Miscanthus through genomic indexselection In Plant and Animal Genome Conference 13 January2018 San Diego CA Retrieved from httpspagconfexcompagxxvimeetingappcgiPaper30622
Slavov G T Nipper R Robson P Farrar K Allison G GBosch M hellip Jensen E (2013) Genome‐wide association studiesand prediction of 17 traits related to phenology biomass and cellwall composition in the energy grass Miscanthus sinensis NewPhytologist 201 1227ndash1239
Slavov G Robson P Jensen E Hodgson E Farrar K AllisonG hellip Donnison I (2014) Contrasting geographic patterns of
32 | CLIFTON‐BROWN ET AL
genetic variation for molecular markers vs phenotypic traits in theenergy grass Miscanthus sinensis Global Change Biology Bioen-ergy 5 562ndash571
Ślusarkiewicz‐Jarzina A Ponitka A Cerazy‐Waliszewska J Woj-ciechowicz M K Sobańska K Jeżowski S amp Pniewski T(2017) Effective and simple in vitro regeneration system of Mis-canthus sinensis Mtimes giganteus and M sacchariflorus for plant-ing and biotechnology purposes Biomass and Bioenergy 107219ndash226 httpsdoiorg101016jbiombioe201710012
Smart L amp Cameron K (2012) Shrub willow In C Kole S Joshiamp D Shonnard (Eds) Handbook of bioenergy crop plants (pp687ndash708) Boca Raton FL Taylor and Francis Group
Stanton B J (2014) The domestication and conservation of Populusand Salix genetic resources (Chapter 4) In Isebrands J ampRichardson J (Eds) Poplars and willows Trees for society andthe environment (pp 124ndash200) Rome Italy CAB InternationalFood and Agricultural Organization of the United Nations
Stanton B J Neale D B amp Li S (2010) Populus breeding Fromthe classical to the genomic approach In S Jansson R Bha-lerao amp A T Groover (Eds) Genetics and genomics of popu-lus (pp 309ndash348) New York NY Springer
Stewart J R Toma Y Fernandez F G Nishiwaki A Yamada T ampBollero G (2009) The ecology and agronomy ofMiscanthus sinensisa species important to bioenergy crop development in its native range inJapan A reviewGlobal Change Biology Bioenergy 1 126ndash153
Stott K G (1992) Willows in the service of man Proceedings of the RoyalSociety of Edinburgh Section B Biological Sciences 98 169ndash182
Strauss S H Ma C Ault K amp Klocko A L (2016) Lessonsfrom two decades of field trials with genetically modified trees inthe USA Biology and regulatory compliance In C Vettori FGallardo H Haumlggman V Kazana F Migliacci G Pilate amp MFladung (Eds) Biosafety of forest transgenic trees (pp 101ndash124)Dordrecht Netherlands Springer
Suda Y amp Argus G W (1968) Chromosome numbers of some NorthAmerican Salix Brittonia 20 191ndash197 httpsdoiorg1023072805440
Tan B (2018) Genomic selection and genome‐wide association studies todissect quantitative traits in forest trees Umearing Sweden University
Tornqvist C Taylor M Jiang Y Evans J Buell C KaepplerS amp Casler M (2018) Quantitative trait locus mapping for flow-ering time in a lowland x upland switchgrass pseudo‐F2 popula-tion The Plant Genome 11 170093
Toacuteth G Hermann T Da Silva M amp Montanarella L (2016)Heavy metals in agricultural soils of the European Union withimplications for food safety Environment International 88 299ndash309 httpsdoiorg101016jenvint201512017
Tracy W Dawson J Moore V amp Fisch J (2016) Intellectual propertyrights and public plant breeding Recommendations and proceedings ofa conference on best practices for intellectual property protection of pub-lically developed plant germplasm (p 70) University of Wisconsin‐Madison Retrieved from httpshostcalswisceduagronomywp-contentuploadssites1620162005Proceedings-IPR-Finalpdf
Trybush S Jahodovaacute Š Macalpine W amp Karp A (2008) A geneticstudy of a Salix germplasm resource reveals new insights into rela-tionships among subgenera sections and species BioEnergyResearch 1 67ndash79 httpsdoiorg101007s12155-008-9007-9
Tsai C J (2013) Next‐generation sequencing for next‐generationbreeding and more New Phytologist 198 635ndash637 httpsdoiorg101111nph12245
Tuskan G A Difazio S Jansson S Bohlmann J Grigoriev I HellstenU hellip Rokhsar D (2006) The genome of black cottonwood Populustrichocarpa (Torr ampGray) Science 313 1596ndash1604
Tuskan G A amp Walsh M E (2001) Short‐rotation woody cropsystems atmospheric carbon dioxide and carbon management AUS case study The Forestry Chronicle 77 259ndash264 httpsdoiorg105558tfc77259-2
Van Den Broek R Treffers D‐J Meeusen M Van Wijk ANieuwlaar E amp Turkenburg W (2001) Green energy or organicfood A life‐cycle assessment comparing two uses of set‐asideland Journal of Industrial Ecology 5 65ndash87 httpsdoiorg101162108819801760049477
Van Der Schoot J Pospiskova M Vosman B amp Smulders M J M(2000) Development and characterization of microsatellite markers inblack poplar (Populus nigra L) Theoretical and Applied Genetics 101317ndash322 httpsdoiorg101007s001220051485
Van Der Weijde T Dolstra O Visser R G amp Trindade L M(2017a) Stability of cell wall composition and saccharificationefficiency in Miscanthus across diverse environments Frontiers inPlant Science 7 2004
Van der Weijde T Huxley L M Hawkins S Sembiring E HFarrar K Dolstra O hellip Trindade L M (2017) Impact ofdrought stress on growth and quality of miscanthus for biofuelproduction Global Change Biology Bioenergy 9 770ndash782
Van der Weijde T Kamei C L A Severing E I Torres A FGomez L D Dolstra O hellip Trindade L M (2017) Geneticcomplexity of miscanthus cell wall composition and biomass qual-ity for biofuels BMC Genomics 18 406
Van der Weijde T Kiesel A Iqbal Y Muylle H Dolstra O Visser RG FhellipTrindade LM (2017) Evaluation ofMiscanthus sinensis bio-mass quality as feedstock for conversion into different bioenergy prod-uctsGlobal Change Biology Bioenergy 9 176ndash190
Vanholme B Cesarino I Goeminne G Kim H Marroni F VanAcker R hellip Boerjan W (2013) Breeding with rare defectivealleles (BRDA) A natural Populus nigra HCT mutant with modi-fied lignin as a case study New Phytologist 198 765ndash776
Vogel K P Mitchell R B Casler M D amp Sarath G (2014)Registration of Liberty Switchgrass Journal of Plant Registra-tions 8 242ndash247 httpsdoiorg103198jpr2013120076crc
Wang X Yamada T Kong F‐J Abe Y Hoshino Y Sato Hhellip Yamada T (2011) Establishment of an efficient in vitro cul-ture and particle bombardment‐mediated transformation systems inMiscanthus sinensis Anderss a potential bioenergy crop GlobalChange Biology Bioenergy 3 322ndash332 httpsdoiorg101111j1757-1707201101090x
Watrud L S Lee E H Fairbrother A Burdick C Reichman JR Bollman M hellip Van de Water P K (2004) Evidence forlandscape‐level pollen‐mediated gene flow from genetically modi-fied creeping bentgrass with CP4 EPSPS as a marker Proceedingsof the National Academy of Sciences of the United States of Amer-ica 101 14533ndash14538
Weisgerber H (1993) Poplar breeding for the purpose of biomassproduction in short‐rotation periods in Germany ndash Problems andfirst findings Forestry Chronicle 69 727ndash729 httpsdoiorg105558tfc69727-6
Wheeler N C Steiner K C Schlarbaum S E amp Neale D B(2015) The evolution of forest genetics and tree improvementresearch in the United States Journal of Forestry 113 500ndash510httpsdoiorg105849jof14-120
CLIFTON‐BROWN ET AL | 33
Whittaker C Macalpine W J Yates N E amp Shield I (2016)Dry matter losses and methane emissions during wood chip stor-age The impact on full life cycle greenhouse gas savings of shortrotation coppice willow for heat BioEnergy Research 9 820ndash835 httpsdoiorg101007s12155-016-9728-0
Xi Q (2000) Investigation on the distribution and potential of giant grassesin China PhD thesis Goettingen Germany Cuvillier Verlag
Xi Q amp Jezowkski S (2004) Plant resources of Triarrhena andMiscanthus species in China and its meaning for Europe PlantBreeding and Seed Science 49 63ndash77
Xue S Kalinina O amp Lewandowski I (2015) Present and future optionsfor the improvement of Miscanthus propagation techniques Renewableand Sustainable Energy Reviews 49 1233ndash1246
Yin K Q Gao C X amp Qiu J L (2017) Progress and prospectsin plant genome editing Nature Plants 3 httpsdoiorg101038nplants2017107
YookM (2016)Genetic diversity and transcriptome analysis for salt toler-ance inMiscanthus PhD thesis Seoul National University Korea
Zaidi S S E A Mahfouz M M amp Mansoor S (2017) CRISPR‐Cpf1 A new tool for plant genome editing Trends in PlantScience 22 550ndash553 httpsdoiorg101016jtplants201705001
Zalesny R S Stanturf J A Gardiner E S Perdue J H YoungT M Coyle D R hellip Hass A (2016) Ecosystem services ofwoody crop production systems BioEnergy Research 9 465ndash491 httpsdoiorg101007s12155-016-9737-z
Zhang Q X Sun Y Hu H K Chen B Hong C T Guo H Phellip Zheng B S (2012) Micropropagation and plant regenerationfrom embryogenic callus of Miscanthus sinensis Vitro Cellular ampDevelopmental Biology‐Plant 48 50ndash57 httpsdoiorg101007s11627-011-9387-y
Zhang Y Zalapa J E Jakubowski A R Price D L AcharyaA Wei Y hellip Casler M D (2011) Post‐glacial evolution of
Panicum virgatum Centers of diversity and gene pools revealedby SSR markers and cpDNA sequences Genetica 139 933ndash948httpsdoiorg101007s10709-011-9597-6
Zhao H Wang B He J Yang J Pan L Sun D amp Peng J(2013) Genetic diversity and population structure of Miscanthussinensis germplasm in China PloS One 8 e75672
Zhou X H Jacobs T B Xue L J Harding S A amp Tsai C J(2015) Exploiting SNPs for biallelic CRISPR mutations in theoutcrossing woody perennial Populus reveals 4‐coumarate CoAligase specificity and redundancy New Phytologist 208 298ndash301
Zhou R Macaya‐Sanz D Rodgers‐Melnick E Carlson C HGouker F E Evans L M hellip DiFazio S P (2018) Characteri-zation of a large sex determination region in Salix purpurea L(Salicaceae) Molecular Genetics and Genomics 1ndash16 httpsdoiorg101007s00438-018-1473-y
Zsuffa L Mosseler A amp Raj Y (1984) Prospects for interspecifichybridization in willow for biomass production In K L Perttu(Ed) Ecology and management of forest biomass production sys-tems Report 15 (pp 261ndash281) Uppsala Sweden SwedishUniversity of Agricultural Sciences
How to cite this article Clifton‐Brown J HarfoucheA Casler MD et al Breeding progress andpreparedness for mass‐scale deployment of perenniallignocellulosic biomass crops switchgrassmiscanthus willow and poplar GCB Bioenergy201800000ndash000 httpsdoiorg101111gcbb12566
34 | CLIFTON‐BROWN ET AL
Graphical AbstractThe contents of this page will be used as part of the graphical abstract of html only
It will not be published as part of main article
Plant breeding links the research effort with commercial mass upscaling The authorsrsquo assessment of development status ofthe four species is shown (poplar having two one for short rotation coppice (SRC) poplar and one for the more traditionalshort rotation forestry (SRF)) Mass scale deployment needs developments outside the breeding arenas to drive breedingactivities more rapidly and extensively
created by the pairwise crossing of individuals with diver-gent characteristics often to generate four‐way crosses thatare analysed as pseudo‐F2 crosses (Liu Wu Wang ampSamuels 2012 Okada et al 2010 Serba et al 2013Tornqvist et al 2018) Individual markers and QTLs iden-tified can be used to design marker‐assisted selection(MAS) programmes to accelerate breeding and increase itsefficiency Genomic prediction and selection (GS) holdseven more promise with the potential to double or triplethe rate of gain for biomass yield and other highly complexquantitative traits of switchgrass (Casler amp Ramstein 2018Ramstein et al 2016) The genome of switchgrass hasrecently been made public through the Joint Genome Insti-tute (httpsphytozomejgidoegov)
Transgenic approaches have been heavily relied upon togenerate unique genetic variants principally for traitsrelated to biomass quality (Merrick amp Fei 2015) Switch-grass is highly transformable using either Agrobacterium‐mediated transformation or biolistics bombardment butregeneration of plants is the bottleneck to these systemsTraditionally plants from the cultivar Alamo were the onlyregenerable genotypes but recent efforts have begun toidentify more genotypes from different populations that arecapable of both transformation and subsequent regeneration(King Bray Lafayette amp Parrott 2014 Li amp Qu 2011Ogawa et al 2014 Ogawa Honda Kondo amp Hara‐Nishi-mura 2016) Cell wall recalcitrance and improved sugarrelease are the most common targets for modification (Bis-wal et al 2018 Fu et al 2011) Transgenic approacheshave the potential to provide traits that cannot be bredusing natural genetic variability However they will stillrequire about 10ndash15 years and will cost $70ndash100 millionfor cultivar development and deployment (Harfouche Mei-lan amp Altman 2011) In addition there is commercialuncertainty due to the significant costs and unpredictabletimescales and outcomes of the regulatory approval processin the countries targeted for seed sales As seen in maizeone advantage of transgenic approaches is that they caneasily be incorporated into F1 hybrid cultivars (Casler2012 Casler et al 2012) but this does not decrease thetime required for cultivar development due to field evalua-tion and seed multiplication requirements
The potential impacts of unintentional gene flow andestablishment of non‐native transgene sequences in nativeprairie species via cross‐pollination are also major issues forthe environmental risk assessment These limit further thecommercialization of varieties made using these technolo-gies Although there is active research into switchgrass steril-ity mechanisms to curb unintended pollen‐mediated genetransfer it is likely that the first transgenic cultivars proposedfor release in the United States will be met with considerableopposition due to the potential for pollen flow to remainingwild prairie sites which account for lt1 of the original
prairie land area and are highly protected by various govern-mental and nongovernmental organizations (Casler et al2015) Evidence for landscape‐level pollen‐mediated geneflow from genetically modified Agrostis seed multiplicationfields (over a mountain range) to pollinate wild relatives(Watrud et al 2004) confirms the challenge of using trans-genic approaches Looking ahead genome editing technolo-gies hold considerable promise for creating targeted changesin phenotype (Burris Dlugosz Collins Stewart amp Lena-ghan 2016 Liu et al 2018) and at least in some jurisdic-tions it is likely that cultivars resulting from gene editingwill not need the same regulatory approval as GMOs (Jones2015a) However in July 2018 the European Court of Justice(ECJ) ruled that cultivars carrying mutations resulting fromgene editing should be regulated in the same way as GMOsThe ECJ ruled that such cultivars be distinguished from thosearising from untargeted mutation breeding which isexempted from regulation under Directive 200118EC
3 | MISCANTHUS
Miscanthus is indigenous to eastern Asia and Oceania whereit is traditionally used for forage thatching and papermaking(Xi 2000 Xi amp Jezowkski 2004) In the 1960s the highbiomass potential of a Japanese genotype introduced to Eur-ope by Danish nurseryman Aksel Olsen in 1935 was firstrecognized in Denmark (Linde‐Laursen 1993) Later thisaccession was characterized described and named asldquoM times giganteusrdquo (Greef amp Deuter 1993 Hodkinson ampRenvoize 2001) commonly abbreviated as Mxg It is a natu-rally occurring interspecies triploid hybrid between tetraploidM sacchariflorus (2n = 4x) and diploid M sinensis(2n = 2x) Despite its favourable agronomic characteristicsand ability to produce high yields in a wide range of environ-ments in Europe (Kalinina et al 2017) the risks of relianceon it as a single clone have been recognized Miscanthuslike switchgrass is outcrossing due to self‐incompatibility(Jiang et al 2017) Thus seeded hybrids are an option forcommercial breeding Miscanthus can also be vegetativelypropagated by rhizome or in vitro culture which allows thedevelopment of clones The breeding approaches are usuallybased on F1 crosses and recurrent selection cycles within thesynthetic populations There are several breeding pro-grammes that target improvement of miscanthus traitsincluding stress resilience targeted regional adaptation agro-nomic ldquoscalabilityrdquo through cheaper propagation fasterestablishment lower moisture and ash contents and greaterusable yield (Clifton‐Brown et al 2017)
Germplasm collections specifically to support breedingfor biomass started in the late 1980s and early 1990s inDenmark Germany and the United Kingdom (Clifton‐Brown Schwarz amp Hastings 2015) These collectionshave continued with successive expeditions from European
12 | CLIFTON‐BROWN ET AL
and US teams assembling diverse collections from a widegeographic range in eastern Asia including from ChinaJapan South Korea Russia and Taiwan (Hodkinson KlaasJones Prickett amp Barth 2015 Stewart et al 2009) Threekey miscanthus species for biomass production are M si-nensis M floridulus and M sacchariflorus M sinensis iswidely distributed throughout eastern Asia with an adap-tive range from the subtropics to southern Russia (Zhaoet al 2013) This species has small rhizomes and producesmany tightly packed shoots forming a ldquotuftrdquo M floridulushas a more southerly adaptive range with a rather similarmorphology to M sinensis but grows taller with thickerstems and is evergreen and less cold‐tolerant than the othermiscanthus species M sacchariflorus is the most northern‐adapted species ranging to 50 degN in eastern Russia (Clarket al 2016) Populations of diploid and tetraploid M sac-chariflorus are found in China (Xi 2000) and South Korea(Yook 2016) and eastern Russia but only tetraploids havebeen found in Japan (Clark Jin amp Petersen 2018)
Germplasm has been assembled from multiple collec-tions over the last century though some early collectionsare poorly documented This historical germplasm has beenused to initiate breeding programmes largely based on phe-notypic and genotypic characterization As many of theaccessions from these collections are ldquoorigin unknownrdquocrucial environmental envelope data are not available UK‐led expeditions started in 2006 and continued until 2011with European and Asian partners and have built up a com-prehensive collection of 1500 accessions from 500 sitesacross Eastern Asia including China Japan South Koreaand Taiwan These collections were guided using spatialclimatic data to identify variation in abiotic stress toleranceAccessions from these recent collections were planted fol-lowing quarantine in multilocation nursery trials at severallocations in Europe to examine trait expression in differentenvironments Based on the resulting phenotypic andmolecular marker data several studies (a) characterized pat-terns of population genetic structure (Slavov et al 20132014) (b) evaluated the statistical power of genomewideassociation studies (GWASs) and identified preliminarymarkerndashtrait associations (Slavov et al 2013 2014) and(c) assessed the potential of genomic prediction (Daveyet al 2017 Slavov et al 2014 2018b) Genomic indexselection in particular offers the possibility of exploringscenarios for different locations or industrial markets (Sla-vov et al 2018a 2018b)
Separately US‐led expeditions also collected about1500 accessions between 2010 and 2014 (Clark et al2014 2016 2018 2015) A comprehensive genetic analy-sis of the population structure has been produced by RAD-seq for M sinensis (Clark et al 2015 Van der WeijdeKamei et al 2017) and M sacchariflorus (Clark et al2018) Multilocation replicated field trials have also been
conducted on these materials in North America and inAsia GWAS has been conducted for both M sinensis anda subset of M sacchariflorus accessions (Clark et al2016) To date about 75 of these recent US‐led collec-tions are in nursery trials outside the United States Due tolengthy US quarantine procedures these are not yet avail-able for breeding in the United States However molecularanalyses have allowed us to identify and prioritize sets ofgenotypes that best encompass the genetic variation in eachspecies
While mostM sinensis accessions flower in northern Eur-ope very few M sacchariflorus accessions flower even inheated glasshouses For this reason the European pro-grammes in the United Kingdom the Netherlands and Francehave performed mainly M sinensis (intraspecies) hybridiza-tions (Table 4) Selected progeny become the parents of latergenerations (recurrent selection as in switchgrass) Seed setsof up to 400 seed per panicle occur inM sinensis In Aberyst-wyth and Illinois significant efforts to induce synchronousflowering in M sacchariflorus and M sinensis have beenmade because interspecies hybrids have proven higher yieldperformance and wide adaptability (Kalinina et al 2017) Ininterspecies pairwise crosses in glasshouses breathable bagsandor large crossing tubes or chambers in which two or morewhole plants fit are used for pollination control Seed sets arelower in bags than in the open air because bags restrict pollenmovement while increasing temperatures and reducinghumidity (Clifton‐Brown Senior amp Purdy 2018) About30 of attempted crosses produced 10 to 60 seeds per baggedpanicle The seed (thousand seed mass ranges from 05 to09 g) is threshed from the inflorescences and sown into mod-ular trays to produce plug plants which are then planted infield nurseries to identify key parental combinations
A breeding programme of this scale must serve theneeds of different environments accepting the commonpurpose is to optimize the interception of solar radiationAn ideal hybrid for a given environment combines adapta-tion to date of emergence with optimization of traits suchas height number of stems per plant flowering and senes-cence time to optimize solar interception to produce a highbiomass yield with low moisture content at harvest (Rob-son Farrar et al 2013 Robson Jensen et al 2013) By20132014 conventional breeding in Europe had producedintra‐ and interspecific fertile seeded hybrids When acohort (typically about 5) of outstanding crosses have beenidentified it is important to work on related upscaling mat-ters in parallel These are as follows
Assessment of the yield and critical traits in selectedhybrids using a network of field trials
Efficient cloning of the seed parents While in vitro andmacro‐cloning techniques are used some genotypes areamenable to neither technique
CLIFTON‐BROWN ET AL | 13
TABLE5
Status
ofmod
ernplantbreeding
techniqu
esin
four
leadingperennialbiom
asscrop
s(PBCs)
Breeding
techno
logy
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sW
illow
Pop
lar
MAS
Use
ofmarker
sequ
encesthat
correlatewith
atrait
allowingearly
prog
enyselectionor
rejectionlocatin
gkn
owngenesfor
useful
traits(suchas
height)from
other
speciesin
your
crop
Breeding
prog
ramme
relevant
biparental
crosses(Table
4)to
create
ldquomapping
popu
latio
nsrdquoand
QTLidentification
andestim
ationto
identifyrobu
stmarkers
alternatively
acomprehensive
panelof
unrelated
geno
typesfor
GWAS
Ineffectivewhere
traits
areaffected
bymany
geneswith
small
effects
US
Literally
dozens
ofstud
iesto
identify
SNPs
andmarkers
ofinterestNoattempts
atMASas
yetlargely
dueto
popu
latio
nspecificity
CNS
SRISS
Rmarkers
forM
sinMsacand
Mflor
FR2
Msin
mapping
popu
latio
ns(BFF
project)
NL2
Msin
mapping
popu
latio
ns(A
tienza
Ram
irezetal
2003
AtienzaSatovic
PetersenD
olstraamp
Martin
200
3Van
Der
Weijdeetal20
17a
Van
derW
eijde
Hux
ley
etal20
17
Van
derW
eijdeKam
ei
etal20
17V
ander
WeijdeKieseletal
2017
)SK
2GWASpanels(1
collectionof
Msin1
diploidbiparentalF 1
popu
latio
n)in
the
pipelin
eUK1
2families
tofind
QTLin
common
indifferentfam
ilies
ofthe
sameanddifferent
speciesandhy
brids
US
2largeGWAS
panels4
diploidbi‐
parentalF 1
popu
latio
ns
1F 2
popu
latio
n(add
ition
alin
pipelin
e)
UK16
families
for
QTLdiscov
ery
2crossesMASscreened
1po
tentialvariety
selected
US
9families
geno
typedby
GBS(8
areF 1o
neisF 2)a
numberof
prom
ising
QTLdevelopm
entof
markers
inprog
ress
FR(INRA)tested
inlargeFS
families
and
infactorialmating
design
low
efficiency
dueto
family
specificity
US
49families
GS
Metho
dto
accelerate
breeding
throug
hredu
cing
the
resourcesforcross
attemptsby
Aldquotraining
popu
latio
nrdquoof
individu
alsfrom
stages
abov
ethat
have
been
both
Risks
ofpo
orpredictio
nProg
eny
testingneedsto
becontinueddu
ring
the
timethetraining
setis
US
One
prog
rammeso
farmdash
USD
Ain
WisconsinThree
cycles
ofgeno
mic
selectioncompleted
CN~1
000
geno
types
NLprelim
inary
mod
elsforMsin
developed
UK~1
000
geno
types
Verysuitable
application
not
attempted
yet
Maybe
challeng
ingfor
interspecies
hybrids
FR(INRA)120
0geno
type
ofPop
ulus
nigraarebeingused
todevelop
intraspecificGS
(Con
tinues)
14 | CLIFTON‐BROWN ET AL
TABLE
5(Contin
ued)
Breeding
techno
logy
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sW
illow
Pop
lar
predictin
gthe
performance
ofprog
enyof
crosses
(and
inresearch
topredictbestparents
tousein
biparental
crosses)
geno
typedand
phenotyp
edto
developamod
elthattakesgeno
typic
datafrom
aldquocandidate
popu
latio
nrdquoof
untested
individu
als
andprod
uces
GEBVs(Jannink
Lorenzamp
Iwata
2010
)
beingph
enotyp
edand
geno
typed
Inthis
time
next‐generation
germ
plasm
andreadyto
beginthe
second
roun
dof
training
and
recalib
ratio
nof
geno
mic
predictio
nmod
els
US
Prelim
inary
mod
elsforMsac
and
Msin
developed
Work
isun
derw
ayto
deal
moreeffectivelywith
polyploidgenetics
calib
ratio
nsUS
125
0geno
types
arebeingused
todevelopGS
calib
ratio
ns
Traditio
nal
transgenesis
Efficient
introd
uctio
nof
ldquoforeign
rdquotraits
(possiblyfrom
other
genu
sspecies)
into
anelite
plant(ega
prov
enparent
ora
hybrid)that
needsa
simpletraitto
beim
prov
edValidate
cand
idategenesfrom
QTLstud
ies
MASand
know
ledg
eof
the
biolog
yof
thetrait
andsource
ofgenesto
confer
the
relevant
changesto
phenotyp
eWorking
transformation
protocol
IPissuescostof
regu
latory
approv
al
GMO
labelling
marketin
gissuestricky
tousetransgenes
inou
t‐breedersbecause
ofcomplexity
oftransformingandgene
flow
risks
US
Manyprog
ramsin
UnitedStates
are
creatin
gtransgenic
plantsusingAlamoas
asource
oftransformable
geno
typesManytraits
ofinterestNothing
commercial
yet
CNC
hang
shaBtg
ene
transformed
into
Msac
in20
04M
sin
transformed
with
MINAC2gene
and
Msactransform
edwith
Cry2A
ageneb
othwith
amarkerg
eneby
Agrob
acterium
JP1
Msintransgenic
forlow
temperature
tolerancewith
increased
expression
offructans
(unp
ublished)
NLImprov
ingprotocol
forM
sin
transformation
SK2
Msin
with
Arabido
psisand
Brachypod
ium
PhytochromeBgenes
(Hwang
Cho
etal
2014
Hwang
Lim
etal20
14)
UK2
Msin
and
1Mflorg
enotyp
es
UKRou
tine
transformationno
tyet
possibleResearchto
overcomerecalcitrance
ongo
ing
Currently
trying
differentspecies
andcond
ition
sPo
plar
transformationused
atpresent
US
Frequently
attempted
with
very
limitedsuccess
US
600transgenic
lines
have
been
characterized
mostly
performed
inthe
aspenhy
brids
reprod
uctiv
esterility
drou
ghttolerance
with
Kno
ckdo
wns
(Con
tinues)
CLIFTON‐BROWN ET AL | 15
TABLE
5(Contin
ued)
Breeding
techno
logy
Use
Prerequisite
steps
Lim
itatio
nsSw
itchg
rass
Miscanthu
sW
illow
Pop
lar
variou
slytransformed
with
four
cellwall
genesiptand
uidA
genesusingtwo
selectionsystem
sby
biolisticsand
Agrob
acterium
Transform
edplantsare
beinganalysed
US
Prelim
inarywork
will
betakenforw
ardin
2018
ndash202
2in
CABBI
Genom
eediting
CRISPR
Refinem
entofexisting
traits
inusefulparents
orprom
isinghybrids
bygeneratingtargeted
mutations
ingenes
know
ntocontrolthe
traitof
interestF
irst
adouble‐stranded
breakismadeinthe
DNAwhich
isrepairedby
natural
DNArepair
machineryL
eads
tofram
eshiftSN
Por
can
usealdquorepair
templaterdquo
orcanbe
used
toinserta
transgene
intoaldquosafe
harbourlocusrdquoIt
couldbe
used
todeleterepressors(or
transcriptionfactors)
Identification
mapping
and
sequ
encing
oftarget
genes(from
DNA
sequ
ence)
avoiding
screening
outof
unintend
ededits
Manyregu
latory
authorities
have
not
decidedwhether
CRISPR
andother
geno
meediting
techno
logies
are
GMOsor
notIf
GMOsseecomment
onstage10
abov
eIf
noteditedcrop
swill
beregu
latedas
conv
entio
nalvarieties
CATechn
olog
yisstill
toonewand
switchg
rass
geno
meis
very
complexOther
labo
ratories
are
interestedbu
tno
tyet
mov
ingon
this
US
One
prog
rammeso
farmdash
USD
Ain
Albany
FRInitiated
in20
16(M
ISEDIT
project)
NLInitiatingin
2018
US
Initiatingin
2018
in
CABBI
UKNot
started
Not
possible
until
transformation
achieved
US
CRISPR
‐Cas9and
Cpf1have
been
successfully
developedin
Pop
ulus
andarehigh
lyefficientExamples
forlig
nin
biosyn
thesis
Note
BFF
BiomassFo
rtheFu
tureBtBacillus
thuringiensisCABBICenterforadvanced
bioenergyandbioproductsinnovatio
nCpf1
CRISPR
from
Prevotella
andFrancisella
1CRISPR
clusteredregularlyinterspaced
palin
drom
icrepeatsCRISPR
‐CasC
RISPR
‐associated
Cry2A
acrystaltoxins2A
asubfam
ilyproduced
byBtFS
full‐sibG
BS
genotyping
bysequencingG
EBVsgenomic‐estim
ated
breeding
valuesG
MOg
enetically
modi-
fied
organism
GS
genomicselection
GWAS
genomew
ideassociationstudy
INRAF
renchNationalInstituteforAgriculturalR
esearch
IPintellectualp
roperty
IPTisopentenyltransferase
ISSR
inter‐SSR
MflorMiscant-
husflo
ridulusMsacMsacchariflo
rusMsinMsinensisM
AS
marker‐a
ssistedselection
MISEDITm
iscanthusgene
editing
forseed‐propagatedtriploidsNACn
oapicalmeristemA
TAF1
2and
cup‐shaped
cotyledon2
‐lik
eQTLq
uantitativ
etraitlocusS
NP
singlenucleotid
epolymorphismS
SRsim
plesequence
repeatsUSD
AU
SDepartm
ento
fAgriculture
a ISO
Alpha‐2
lettercountrycodes
16 | CLIFTON‐BROWN ET AL
High seed production from field crossing trials con-ducted in locations where flowering in both seed andpollen parents is likely to happen synchronously
Scalable and adapted harvesting threshing and seed pro-cessing methods for producing high seed quality
The results of these parallel activities need to becombined to identify the upscaling pathway for eachhybrid if this cannot be achieved the hybrid will likelynot be commercially viable The UK‐led programme withpartners in Italy and Germany shows that seedbased mul-tiplication rates of 12000 are achievable several inter-specific hybrids (Clifton‐Brown et al 2017) Themultiplication rate of M sinensis is higher probably15000ndash10000 Conventional cloning from rhizome islimited to around 120 that is one ha could provide rhi-zomes for around 20 ha of new plantation
Multilocation field testing of wild and novel miscanthushybrids selected by breeding programmes in the Nether-lands and the United Kingdom was performed as part ofthe project Optimizing Miscanthus Biomass Production(OPTIMISC 2012ndash2016) These trials showed that com-mercial yields and biomass qualities (Kiesel et al 2017Van der Weijde et al 2017a Van der Weijde Kieselet al 2017) could be produced in a wide range of climatesand soil conditions from the temperate maritime climate ofwestern Wales to the continental climate of eastern Russiaand the Ukraine (Kalinina et al 2017) Extensive environ-mental measurements of soil and climate combined withgrowth monitoring are being used to understand abioticstresses (Nunn et al 2017 Van der Weijde Huxley et al2017) and develop genotype‐specific scenarios similar tothose reported earlier in Hastings et al (2009) Phenomicsexperiments on drought tolerance have been conducted onwild and improved germplasm (Malinowska Donnison ampRobson 2017 Van der Weijde Huxley et al 2017)Recently produced interspecific hybrids displaying excep-tional yield under drought (~30 greater than control Mxg)in field trials in Poland and Moldova are being furtherstudied in detail in the phenomics and genomics facility atAberystwyth to better understand genendashtrait associationswhich can be fed back into breeding
Intraspecific seeded hybrids of M sinensis produced inthe Netherlands and interspecific M sacchariflorus times M si-nensis hybrids produced by the UK‐led breeding programmehave entered yield testing in 2018 with the recently EU‐funded project ldquoGRowing Advanced industrial Crops on mar-ginal lands for biorEfineries (GRACE)rdquo (httpswwwgrace-bbieu) Substantial variation in biomass quality for sacchari-fication efficiency (glucose release as of dry matter) ashcontent and melting point has already been generated inintraspecific M sinensis hybrids (Van der Weijde Kiesel
et al 2017) across environments (Weijde Dolstra et al2017a) GRACE aims to establish more than 20 hectares ofnew inter‐ and intraspecific seeded hybrids across six Euro-pean countries This project is building the know‐how andagronomy needed to transition from small research plots tocommercial‐scale field sites and linking biomass productiondirectly to industrial applications The biomass produced byhybrids in different locations will be supplied to innovativeindustrial end‐users making a wide range of biobased prod-ucts both for chemicals and for energy In the United Statesmulti‐location yield were initiated in 2018 to evaluate new tri-ploid M times giganteus genotypes developed at Illinois Cur-rently infertile hybrids are favoured in the United Statesbecause this eliminates the risk of invasiveness from naturallydispersed viable seed The precautionary principle is appliedas fertile miscanthus has naturalized in several states (QuinnAllen amp Stewart 2010) North European multilocation fieldtrials in the EMI and OPTIMISC projects have shown thereis minimal risk of invasiveness even in years when fertileflowering hybrids produce viable seed Naturalized standshave not established here due perhaps to low dormancy pooroverwintering and low seedling competitive strength In addi-tion to breeding for nonshattering or sterile seeded hybridsQuinn et al (2010) suggest management strategies which canfurther minimize environmental opportunities to manage therisk of invasiveness
31 | Molecular breeding and biotechnology
In miscanthus new plant breeding techniques (Table 5)have focussed on developing molecular markers forbreeding in Europe the United States South Korea andJapan There are several publications on QTL mappingpopulations for key traits such as flowering (AtienzaRamirez amp Martin 2003) and compositional traits(Atienza Satovic Petersen Dolstra amp Martin 2003) Inthe United States and United Kingdom independent andinterconnected bi‐parental ldquomappingrdquo families have beenstudied (Dong et al 2018 Gifford Chae SwaminathanMoose amp Juvik 2015) alongside panels of diverse germ-plasm accessions for GWAS (Slavov et al 2013) Fur-ther developments calibrating GS with very large panelsof parents and cross progeny are underway (Daveyet al 2017) The recently completed first miscanthus ref-erence genome sequence is expected to improve the effi-ciency of MAS strategies and especially GWAS(httpsphytozomejgidoegovpzportalhtmlinfoalias=Org_Msinensis_er) For example without a referencegenome sequence Clark et al (2014) obtained 21207RADseq SNPs (single nucleotide polymorphisms) on apanel of 767 miscanthus genotypes (mostly M sinensis)but subsequent reanalysis of the RADseq data using the
CLIFTON‐BROWN ET AL | 17
new reference genome resulted in hundreds of thousandsof SNPs being called
Robust and effective in vitro regeneration systems havebeen developed for Miscanthus sinensis M times giganteusand M sacchariflorus (Dalton 2013 Guo et al 2013Hwang Cho et al 2014 Rambaud et al 2013 Ślusarkie-wicz‐Jarzina et al 2017 Wang et al 2011 Zhang et al2012) However there is still significant genotype speci-ficity and these methods need ldquoin‐houserdquo optimization anddevelopment to be used routinely They provide potentialroutes for rapid clonal propagation and also as a basis forgenetic transformation Stable transformation using bothbiolistics (Wang et al 2011) and Agrobacterium tumefa-ciens DNA delivery methods (Hwang Cho et al 2014Hwang Lim et al 2014) has been achieved in M sinen-sis The development of miscanthus transformation andgene editing to generate diplogametes for producing seed‐propagated triploid hybrids are performed as part of theFrench project MISEDIT (miscanthus gene editing forseed‐propagated triploids) There are no reports of genomeediting in any miscanthus species but new breeding inno-vations including genome editing are particularly relevantin this slow‐to‐breed nonfood bioenergy crop (Table 4)
4 | WILLOW
Willow (Salix spp) is a very diverse group of catkin‐bear-ing trees and shrubs Willow belongs to the family Sali-caceae which also includes the Populus genus There areapproximately 350 willow species (Argus 2007) foundmostly in temperate and arctic zones in the northern hemi-sphere A few are adapted to subtropical and tropicalzones The centre of diversity is believed to be in Asiawith over 200 species in China Around 120 species arefound in the former Soviet Union over 100 in NorthAmerica and around 65 species in Europe and one speciesis native to South America (Karp et al 2011) Willows aredioecious thus obligate outcrossers and highly heterozy-gous The haploid chromosome number is 19 (Hanley ampKarp 2014) Around 40 of species are polyploid (Sudaamp Argus 1968) ranging from triploids to the atypicaldodecaploid S maxxaliana with 2n=190 (Zsuffa et al1984)
Although almost exclusively native to the NorthernHemisphere willow has been grown around the globe formany thousands of years to support a wide range of appli-cations (Kuzovkina amp Quigley 2005 Stott 1992) How-ever it has been the focus of domestication for bioenergypurposes for only a relatively short period since the 1970sin North America and Europe For bioenergy breedershave focused their efforts on the shrub willows (subgenusVetix) because of their rapid juvenile growth rates as aresponse to coppicing on a 2‐ to 4‐year cycle that can be
accomplished using farm machinery rather than forestryequipment (Shield Macalpine Hanley amp Karp 2015Smart amp Cameron 2012)
Since shrub willow was not generally recognized as anagricultural crop until very recently there has been littlecommitment to building and maintaining germplasm reposi-tories of willow to support long‐term breeding One excep-tion is the United Kingdom where a large and well‐characterized Salix germplasm collection comprising over1500 accessions is held at Rothamsted Research (Stott1992 Trybush et al 2008) Originally initiated for use inbasketry in 1923 accessions have been added ever sinceIn the United States a germplasm collection of gt350accessions is located at Cornell University to support thebreeding programme there The UK and Cornell collectionshave a relatively small number of accessions in common(around 20) Taken together they represent much of thespecies diversity but only a small fraction of the overallgenetic diversity within the genus There are three activewillow breeding programmes in Europe RothamstedResearch (UK) Salixenergi Europa AB (SEE) and a pro-gramme at the University of Warmia and Mazury in Olsz-tyn (Poland) (abbreviations used in Table 1) There is oneactive US programme based at Cornell University Culti-vars are still being marketed by the European WillowBreeding Programme (EWBP) (UK) which was activelybreeding biomass varieties from 1996 to 2002 Cultivarsare protected by plant breedersrsquo rights (PBRs) in Europeand by plant patents in the United States The sharing ofgenetic resources in the willow community is generallyregulated by material transfer agreements (MTA) and tai-lored licensing agreements although the import of cuttingsinto North America is prohibited except under special quar-antine permit conditions
Efforts to augment breeding germplasm collection fromnature are continuing with phenotypic screening of wildgermplasm performed in field experiments with 177 S pur-purea genotypes in the United States (at sites in Genevaand Portland NY and Morgantown WV) that have beengenotyped using genotyping by sequencing (GBS) (Elshireet al 2011) In addition there are approximately 400accessions of S viminalis in Europe (near Pustnaumls Upp-sala Sweden and Woburn UK (Berlin et al 2014 Hal-lingbaumlck et al 2016) The S viminalis accessions wereinitially genotyped using 38 simple sequence repeats (SSR)markers to assess genetic diversity and screened with~1600 SNPs in genes of potential interest for phenologyand biomass traits Genetics and genomics combined withextensive phenotyping have substantially improved thegenetic basis of biomass‐related traits in willow and arenow being developed in targeted breeding via MAS Thisunderpinning work has been conducted on large specifi-cally developed biparental Salix mapping populations
18 | CLIFTON‐BROWN ET AL
(Hanley amp Karp 2014 Zhou et al 2018) as well asGWAS panels (Hallingbaumlck et al 2016)
Once promising parental combinations are identifiedcrosses are usually performed using fresh pollen frommaterial that has been subject to a phased removal fromcold storage (minus4degC) (Lindegaard amp Barker 1997 Macal-pine Shield Trybush Hayes amp Karp 2008 Mosseler1990) Pollen storage is useful in certain interspecific com-binations where flowering is not naturally synchronizedThis can be overcome by using pollen collection and stor-age protocol which involves extracting pollen using toluene(Kopp Maynard Niella Smart amp Abrahamson 2002)
The main breeding approach to improve willow yieldsrelies on species hybridization to capture hybrid vigour(Fabio et al 2017 Serapiglia Gouker amp Smart 2014) Inthe absence of genotypic models for heterosis breedershave extensively tested general and specific combiningability of parents to produce superior progeny The UKbreeding programmes (EWBP 1996ndash2002 and RothamstedResearch from 2003 on) have performed more than 1500exploratory cross‐pollinations The Cornell programme hassuccessfully completed about 550 crosses since 1998Investment into the characterization of genetic diversitycombined with progeny tests from exploratory crosses hasbeen used to produce hundreds of targeted intraspeciescrosses in the United Kingdom and United States respec-tively (see Table 1) To achieve long‐term gains beyond F1hybrids four intraspecific recurrent selection populationshave been created in the United Kingdom (for S dasycla-dos S viminalis and S miyabeana) and Cornell is pursu-ing recurrent selection of S purpurea Interspecifichybridizations with genotypes selected from the recurrentselection cycles are well advanced in willow with suchcrosses to date totalling 420 in the United Kingdom andover 100 in the United States
While species hybridization is common in Salix it isnot universal Of the crosses attempted about 50 hybri-dize and produce seed (Macalpine Shield amp Karp 2010)As the viability of seed from successful crosses is short (amatter of days at ambient temperatures) proper seed rear-ing and storage protocols are essential (Maroder PregoFacciuto amp Maldonado 2000)
Progeny from crosses are treated in different waysamong the breeding programmes at the seedling stage Inthe United States seedlings are planted into an irrigatedfield where plants are screened for two seasons beforebeing progressed to further field trials In the United King-dom seedlings are planted into trays of compost wherethey remain containerized in an irrigated nursery for theremainder of year one In the United Kingdom seedlingsare subject to two rounds of selection in the nursery yearThe first round takes place in September to select againstsusceptibility to rust infection (Melampsora spp) A second
round of selection in winter assesses tip damage from frostand giant willow aphid infestation In the United Stateswhere the rust pressure is lower screening for Melampsoraspp cannot be performed at the nursery stage Both pro-grammes monitor Melampsora spp pest susceptibilityyield and architecture over multiple years in field trialsSelected material is subject to two rounds of field trials fol-lowed by a final multilocation yield trial to identify vari-eties for commercialization
Promising selections (ie potential cultivars) need to beclonally propagated A rapid in vitro tissue culture propa-gation method has been developed (Palomo‐Riacuteos et al2015) This method can generate about 5000 viable trans-plantable clones from a single plant in just 24 weeks Anin vitro system can also accommodate early selection viamolecular or biochemical markers to increase selectionspeed Conventional breeding systems take 13 years viafour rounds of selection from crossing to selecting a variety(Figure 1) but this has the potential to be reduced to7 years if micropropagation and MAS selection are adopted(Hanley amp Karp 2014 Palomo‐Riacuteos et al 2015)
Willows are currently propagated commercially byplanting winter‐dormant stem cuttings in spring Commer-cial planting systems for willow use mechanical plantersthat cut and insert stem sections from whips into a well‐prepared soil One hectare of stock plants grown in specificmultiplication beds planted at 40000 plants per ha pro-duces planting material for 80 hectares of commercialshort‐rotation coppice willow annually (planted at 15000cuttings per hectare) (Whittaker et al 2016) When com-mercial plantations are established the industry standard isto plant intimate mixtures of ~5 diverse rust (Melampsoraspp)‐resistant varieties (McCracken amp Dawson 1997 VanDen Broek et al 2001)
The foundations for using new plant breeding tech-niques have been established with funding from both thepublic and the private sectors To establish QTL maps 16mapping populations from biparental crosses are understudy in the United Kingdom Nine are under study in theUnited States The average number of individuals in thesefamilies ranges from 150 to 947 (Hanley amp Karp 2014)GS is also being evaluated in S viminalis and preliminaryresults indicate that multiomic approaches combininggenomic and metabolomic data have great potential (Sla-vov amp Davey 2017) For both QTL and GS approachesthe field phenotyping demands are large as several thou-sand individuals need to be phenotyped for a wide rangeof traits These include the following dates of bud burstand growth succession stem height stem density wooddensity and disease resistance The greater the number ofindividuals the more precise the QTL marker maps andGS models are However the logistical and financial chal-lenges of phenotyping large numbers of individuals are
CLIFTON‐BROWN ET AL | 19
considerable because the willow crop is gt5 m tall in thesecond year There is tremendous potential to improve thethroughput of phenotyping using unmanned aerial systemswhich is being tested in the USDA National Institute ofFood and Agriculture (NIFA) Willow SkyCAP project atCornell Further investment in these approaches needs tobe sustained over many years fully realizes the potentialof a marker‐assisted selection programme for willow
To date despite considerable efforts in Europe and theUnited States to establish a routine transformation systemthere has not been a breakthrough in willow but attemptsare ongoing As some form of transformation is typically aprerequisite for genome editing techniques these have notyet been applied to willow
In Europe there are 53 short‐rotation coppice (SRC) bio-mass willow cultivars registered with the Community PlantVariety Office (CPVO) for PBRs of which ~25 are availablecommercially in the United Kingdom There are eightpatented cultivars commercially available in the UnitedStates In Sweden there are nine commercial cultivars regis-tered in Europe and two others which are unregistered(httpssalixenergiseplanting-material) Furthermore thereare about 20 precommercial hybrids in final yield trials inboth the United States and the United Kingdom It has beenestimated that it would take two years to produce the stockrequired to plant 50 ha commercially from the plant stock inthe final yield trials Breeding programmes have alreadydelivered rust‐resistant varieties and increases in yield to themarket The adoption of advanced breeding technologies willlikely lead to a step change in improving traits of interest
5 | POPLAR
Poplar a fast‐growing tree from the northern hemispherewith a small genome size has been adopted for commercialforestry and scientific purposes The genus Populus con-sists of about 29 species classified in six different sectionsPopulus (formerly Leuce) Tacamahaca Aigeiros AbasoTuranga and Leucoides (Eckenwalder 1996) The Populusspecies of most interest for breeding and testing in the Uni-ted States and Europe are P nigra P deltoides P maxi-mowiczii and P trichocarpa (Stanton 2014) Populusclones for biomass production are being developed byintra‐ and interspecies hybridization (DeWoody Trewin ampTaylor 2015 Richardson Isebrands amp Ball 2014 van derSchoot et al 2000) Recurrent selection approaches areused for gradual population improvement and to create eliteclonal lines for commercialization (Berguson McMahon ampRiemenschneider 2017 Neale amp Kremer 2011) Currentlypoplar breeding in the United States occurs in industrialand academic programmes located in the Southeast theMidwest and the Pacific Northwest These use six speciesand five interspecific taxa (Stanton 2014)
The southeastern programme historically focused onrecurrent selection of P deltoides from accessions made inthe lower Mississippi River alluvial plain (Robison Rous-seau amp Zhang 2006) More recently the genetic base hasbeen broadened to produce interspecific hybrids with resis-tance to the fungal infection Septoria musiva which causescankers
In the midwest of the United States populationimprovement efforts are focused on P deltoides selectionsfrom native provenances and hybrid crosses with acces-sions introduced from Europe Interspecific intercontinental(Europe and America) hybrid crosses between P nigra andP deltoides (P times canadensis) are behind many of theleading commercial hybrids which are the most advancedbreeding materials for many applications and regions InMinnesota previous breeding experience and efforts utiliz-ing P maximowiczii and P trichocarpa have been discon-tinued due to Septoria susceptibility and a lack of coldhardiness (Berguson et al 2017) Traits targeted forimprovement include yieldgrowth rate cold hardinessadventitious rooting resistance to Septoria and Melamp-sora leaf rust and stem form The Upper Midwest pro-gramme also carries out wide hybridizations within thesection Populus The P times wettsteinii (P trem-ula times P tremuloides) taxon is bred for gains in growthrate wood quality and resistance to the fungus Entoleucamammata which causes hypoxylon canker (David ampAnderson 2002)
In the Pacific Northwest GreenWood Resources Incleads poplar breeding that emphasizes interspecifichybrid improvement of P times generosa (P deltoides timesP trichocarpa and reciprocal) and P deltoides times P maxi-mowiczii taxa for coastal regions and the P times canadensistaxon for the drier continental regions Intraspecificimprovement of second‐generation breeding populations ofP deltoides P nigra P maximowiczii and P trichocarpaare also involved (Stanton et al 2010) The present focusof GreenWood Resourcesrsquo hybridization is bioenergy feed-stock improvement concentrating on coppice yield wood‐specific gravity and rate of sugar release
Industrial interest in poplar in the United States has his-torically come from the pulp and paper sector althoughveneer and dimensional lumber markets have been pursuedat times Currently the biomass market for liquid trans-portation fuels is being emphasized along with the use oftraditional and improved poplar genotypes for ecosystemservices such as phytoremediation (Tuskan amp Walsh 2001Zalesny et al 2016)
In Europe there are breeding programmes in FranceGermany Italy and Sweden These include the following(a) Alasia Franco Vivai (AFV) programme in northernItaly (b) the French programme led by the poplar Scien-tific Interest Group (GIS Peuplier) and carried out
20 | CLIFTON‐BROWN ET AL
collaboratively between the National Institute for Agricul-tural Research (INRA) the National Research Unit ofScience and Technology for Environment and Agriculture(IRSTEA) and the Forest Cellulose Wood Constructionand Furniture Technology Institute (FCBA) (c) the Ger-man programme at Northwest German Forest Research Sta-tion (NW‐FVA) at Hannoversch Muumlnden and (d) theSwedish programme at the Swedish University of Agricul-tural Sciences and SweTree Technologies AB (Table 1)
AFV leads an Italian poplar breeding programme usingextensive field‐grown germplasm collections of P albaP deltoides P nigra and P trichocarpa While interspeci-fic hybridization uses several taxa the focus is onP times canadensis The breeding programme addresses dis-ease resistance (Marssonina brunnea Melampsora larici‐populina Discosporium populeum and poplar mosaicvirus) growth rate and photoperiod adaptation AFV andGreenWood Resources collaborate in poplar improvementin Europe through the exchange of frozen pollen and seedfor reciprocal breeding projects Plantations in Poland andRomania are currently the focus of the collaboration
The ongoing French GIS Peuplier is developing a long‐term breeding programme based on intraspecific recurrentselection for the four parental species (P deltoides P tri-chocarpa P nigra and P maximowiczii) designed to betterbenefit from hybrid vigour demonstrated by the interspeci-fic crosses P canadensis P deltoides times P trichocarpaand P trichocarpa times P maximowiczii Current selectionpriorities are targeting adaptation to soil and climate condi-tions resistance and tolerance to the most economicallyimportant diseases and pests high volume production underSRC and traditional poplar cultivation regimes as well aswood quality of interest by different markets Currentlygenomic selection is under exploration to increase selectionaccuracy and selection intensity while maintaining geneticdiversity over generations
The German NW‐FVA programme is breeding intersec-tional AigeirosndashTacamahaca hybrids with a focus on resis-tance to Pollaccia elegans Xanthomonas populiDothichiza spp Marssonina brunnea and Melampsoraspp (Stanton 2014) Various cross combinations ofP maximowiczii P trichocarpa P nigra and P deltoideshave led to new cultivars suitable for deployment in vari-etal mixtures of five to ten genotypes of complementarystature high productivity and phenotypic stability (Weis-gerber 1993) The current priority is the selection of culti-vars for high‐yield short‐rotation biomass production Sixhundred P nigra genotypes are maintained in an ex situconservation programme An in situ P nigra conservationeffort involves an inventory of native stands which havebeen molecular fingerprinted for identity and diversity
The Swedish programme is concentrating on locallyadapted genotypes used for short‐rotation forestry (SRF)
because these meet the needs of the current pulping mar-kets Several field trials have shown that commercial poplarclones tested and deployed in Southern and Central Europeare not well adapted to photoperiods and low temperaturesin Sweden and in the Baltics Consequently SwedishUniversity of Agricultural Sciences and SweTree Technolo-gies AB started breeding in Sweden in 1990s to producepoplar clones better adapted to local climates and markets
51 | Molecular breeding technologies
Poplar genetic improvement cannot be rapidly achievedthrough traditional methods alone because of the longbreeding cycles outcrossing breeding systems and highheterozygosity Integrating modern genetic genomic andphenomics techniques with conventional breeding has thepotential to expedite poplar improvement
The genome of poplar has been sequenced (Tuskanet al 2006) It has an estimated genome size of485 plusmn 10 Mbp divided into 19 chromosomes This is smal-ler than other PBCs and makes poplar more amenable togenetic engineering (transgenesis) GS and genome editingPoplar has seen major investment in both the United Statesand Europe being the model system for woody perennialplant genetics and genomics research
52 | Targets for genetic modification
Traits targeted include wood properties (lignin content andcomposition) earlylate flowering male sterility to addressbiosafety regulation issues enhanced yield traits and herbi-cide tolerance These extensive transgenic experiments haveshown differences in recalcitrance to in vitro regenerationand genetic transformation in some of the most importantcommercial hybrid poplars (Alburquerque et al 2016)Further transgene expression stability is being studied Sofar China is the only country known to have commerciallyused transgenic insect‐resistant poplar A precommercialherbicide‐tolerant poplar was trialled for 8 years in the Uni-ted States (Li Meilan Ma Barish amp Strauss 2008) butcould not be released due to stringent environmental riskassessments required for regulatory approval This increasestranslation costs and delays reducing investor confidencefor commercial deployment (Harfouche et al 2011)
The first field trials of transgenic poplar were performedin France in 1987 (Fillatti Sellmer Mccown Haissig ampComai 1987) and in Belgium in 1988 (Deblock 1990)Although there have been a total of 28 research‐scale GMpoplar field trials approved in the European Union underCouncil Directive 90220EEC since October 1991 (inPoland Belgium Finland France Germany Spain Swe-den and in the United Kingdom (Pilate et al 2016) onlyauthorizations in Poland and Belgium are in place today In
CLIFTON‐BROWN ET AL | 21
the United States regulatory notifications and permits fornearly 20000 transgenic poplar trees derived from approxi-mately 600 different constructs have been issued since1995 by the USDAs Animal and Plant Health InspectionService (APHIS) (Strauss et al 2016)
53 | Genome editing CRISPR technologies
Clustered regularly interspaced palindromic repeats(CRISPR) and the CRISPR‐associated (CRISPR‐Cas)nucleases are a groundbreaking genome‐engineering toolthat complements classical plant breeding and transgenicmethods (Moreno‐Mateos et al 2017) Only two publishedstudies in poplar have applied the CRISPRCas9 technol-ogy One is in P tomentosa in which an endogenous phy-toene desaturase gene (PtoPDS) was successfully disruptedsite specifically in the first generation of transgenic plantsresulting in an albino and dwarf phenotype (Fan et al2015) The second was in P tremula times alba in whichhigh CRISPR‐Cas9 mutational efficiency was achieved forthree 4‐coumarateCoA ligase (4Cl) genes 4CL1 4CL2and 4CL5 associated with lignin and flavonoid biosynthe-sis (Zhou et al 2015) Due to its low cost precision andrapidness it is very probable that cultivars or clones pro-duced using CRISPR technology will be ready for market-ing in the near future (Yin et al 2017) Recently aCRISPR with a smaller associated endonuclease has beendiscovered from Prevotella and Francisella 1 (Cpf1) whichmay have advantages over Cas9 In addition there arereports of DNA‐free editing in plants using both CRISPRCpf1 and CRISPR Cas9 for example Ref (Kim et al2017 Mahfouz 2017 Zaidi et al 2017)
It remains unresolved whether plants modified by genomeediting will be regulated as genetically modified organisms(GMOs) by the relevant authorities in different countries(Lozano‐Juste amp Cutler 2014) Regulations to cover thesenew breeding techniques are still evolving but those coun-tries who have published specific guidance (including UnitedStates Argentina and Chile) are indicating that plants pos-sessing simple genome edits will not be regulated as conven-tional transgenesis (Jones 2015b) The first generation ofgenome‐edited crops will likely be phenocopy gene knock-outs that already exist to produce ldquonature identicalrdquo traits thatis traits that could also be derived by conventional breedingDespite this confidence in applying these new powerfulbreeding tools remains limited owing to the uncertain regula-tory environment in many parts of the world (Gao 2018)including the recent ECJ 2018 rulings mentioned earlier
54 | Genomics‐based breeding technologies
Poplar breeding programs are becoming well equipped withuseful genomics tools and resources that are critical to
explore genomewide variability and make use of the varia-tion for enhancing genetic gains Deep transcriptomesequencing resequencing of alternate genomes and GBStechnology for genomewide marker detection using next‐generation sequencing (NGS) are yielding valuable geno-mics tools GWAS with NGS‐based markers facilitatesmarker identification for MAS breeding by design and GS
GWAS approaches have provided a deeper understand-ing of genome function as well as allelic architectures ofcomplex traits (Huang et al 2010) and have been widelyimplemented in poplar for wood characteristics (Porthet al 2013) stomatal patterning carbon gain versus dis-ease resistance (McKown et al 2014) height and phenol-ogy (Evans et al 2014) cell wall chemistry (Mucheroet al 2015) growth and cell walls traits (Fahrenkroget al 2017) bark roughness (Bdeir et al 2017) andheight and diameter growth (Liu et al 2018) Using high‐throughput sequencing and genotyping platforms an enor-mous amount of SNP markers have been used to character-ize the linkage disequilibrium (LD) in poplar (eg Slavovet al 2012 discussed below)
The genetic architecture of photoperiodic traits in peren-nial trees is complex involving many loci However itshows high levels of conservation during evolution (Mau-rya amp Bhalerao 2017) These genomics tools can thereforebe used to address adaptation issues and fine‐tune themovement of elite lines into new environments For exam-ple poor timing of spring bud burst and autumn bud setcan result in frost damage resulting in yield losses (Ilstedt1996) These have been studied in P tremula genotypesalong a latitudinal cline in Sweden (~56ndash66oN) and haverevealed high nucleotide polymorphism in two nonsynony-mous SNPs within and around the phytochrome B2 locus(Ingvarsson Garcia Hall Luquez amp Jansson 2006 Ing-varsson Garcia Luquez Hall amp Jansson 2008) Rese-quencing 94 of these P tremula genotypes for GWASshowed that noncoding variation of a single genomicregion containing the PtFT2 gene described 65 ofobserved genetic variation in bud set along the latitudinalcline (Tan 2018)
Resequencing genomes is currently the most rapid andeffective method detecting genetic differences betweenvariants and for linking loci to complex and importantagronomical and biomass traits thus addressing breedingchallenges associated with long‐lived plants like poplars
To date whole genome resequencing initiatives havebeen launched for several poplar species and genotypes InPopulus LD studies based on genome resequencing sug-gested the feasibility of GWAS in undomesticated popula-tions (Slavov et al 2012) This plant population is beingused to inform breeding for bioenergy development Forexample the detection of reliable phenotypegenotype asso-ciations and molecular signatures of selection requires a
22 | CLIFTON‐BROWN ET AL
detailed knowledge about genomewide patterns of allelefrequency variation LD and recombination suggesting thatGWAS and GS in undomesticated populations may bemore feasible in Populus than previously assumed Slavovet al (2012) have resequenced 16 genomes of P tri-chocarpa and genotyped 120 trees from 10 subpopulationsusing 29213 SNPs (Geraldes et al 2013) The largest everSNP data set of genetic variations in poplar has recentlybeen released providing useful information for breedinghttpswwwbioenergycenterorgbescgwasindexcfmAlso deep sequencing of transcriptomes using RNA‐Seqhas been used for identification of functional genes andmolecular markers that is polymorphism markers andSSRs A multitissue and multiple experimental data set forP trichocarpa RNA‐Seq is publicly available (httpsjgidoegovdoe-jgi-plant-flagship-gene-atlas)
The availability of genomic information of DNA‐con-taining cell organelles (nucleus chloroplast and mitochon-dria) will also allow a holistic approach in poplarmolecular breeding in the near future (Kersten et al2016) Complete Populus genome sequences are availablefor nucleus (P trichocarpa section Tacamahaca) andchloroplasts (seven species and two clones from P trem-ula W52 and P tremula times P alba 717ndash1B4) A compara-tive approach revealed structural and functionalinformation broadening the knowledge base of PopuluscpDNA and stimulating future diagnostic marker develop-ment The availability of whole genome sequences of thesecellular compartments of P tremula holds promise forboosting marker‐assisted poplar breeding Other nucleargenome sequences from additional Populus species arenow available (eg P deltoides (httpsphytozomejgidoegovpz) and will become available in the forthcomingyears (eg P tremula and P tremuloidesmdashPopGenIE(Sjodin Street Sandberg Gustafsson amp Jansson 2009))Recently the characterization of the poplar pan‐genome bygenomewide identification of structural variation in threecrossable poplar species P nigra P deltoides and P tri-chocarpa revealed a deeper understanding of the role ofinter‐ and intraspecific structural variants in poplar pheno-type and may have important implications for breedingparticularly interspecific hybrids (Pinosio et al 2016)
GS has been proposed as an alternative to MAS in cropimprovement (Bernardo amp Yu 2007 Heffner Sorrells ampJannink 2009) GS is particularly well suited for specieswith long generation times for characteristics that displaymoderate‐to‐low heritability for traits that are expensive tomeasure and for selection of traits expressed late in the lifecycle as is the case for most traits of commercial value inforestry (Harfouche et al 2012) Current joint genomesequencing efforts to implement GS in poplar using geno-mic‐estimated breeding values for bioenergy conversiontraits from 49 P trichocarpa families and 20 full‐sib
progeny are taking place at the Oak Ridge National Labo-ratory and GreenWood Resources (Brian Stanton personalcommunication httpscbiornlgov) These data togetherwith the resequenced GWAS population data will be thebasis for developing GS algorithms Genomic breedingtools have been developed for the intraspecific programmetargeting yield resistance to Venturia shoot blightMelampsora leaf rust resistance to Cryptorhynchus lapathistem form wood‐specific gravity and wind firmness (Evanset al 2014 Guerra et al 2016) A newly developedldquobreeding with rare defective allelesrdquo (BRDA) technologyhas been developed to exploit natural variation of P nigraand identify defective variants of genes predicted by priortransgenic research to impact lignin properties Individualtrees carrying naturally defective alleles can then be incor-porated directly into breeding programs thereby bypassingthe need for transgenics (Vanholme et al 2013) Thisnovel breeding technology offers a reverse genetics com-plement to emerging GS for targeted improvement of quan-titative traits (Tsai 2013)
55 | Phenomics‐assisted breeding technology
Phenomics involves the characterization of phenomesmdashthefull set of phenotypes of given individual plants (HouleGovindaraju amp Omholt 2010) Traditional phenotypingtools which inefficiently measure a limited set of pheno-types have become a bottleneck in plant breeding studiesHigh‐throughput plant phenotyping facilities provide accu-rate screening of thousands of plant breeding lines clones orpopulations over time (Fu 2015) are critical for acceleratinggenomics‐based breeding Automated image collection andanalysis phenomics technologies allow accurate and nonde-structive measurements of a diversity of phenotypic traits inlarge breeding populations (Gegas Gay Camargo amp Doo-nan 2014 Goggin Lorence amp Topp 2015 Ludovisi et al2017 Shakoor Lee amp Mockler 2017) One important con-sideration is the identification of relevant and quantifiabletarget traits that are early diagnostic indicators of biomassyield Good progress has been made in elucidating theseunderpinning morpho‐physiological traits that are amenableto remote sensing in Populus (Harfouche Meilan amp Altman2014 Rae Robinson Street amp Taylor 2004) Morerecently Ludovisi et al (2017) developed a novel methodol-ogy for field phenomics of drought stress in a P nigra F2partially inbred population using thermal infrared imagesrecorded from an unmanned aerial vehicle‐based platform
Energy is the current main market for poplar biomass butthe market return provided is not sufficient to support pro-duction expansion even with added demand for environmen-tal and land management ldquoecosystem servicesrdquo such as thetreatment of effluent phytoremediation riparian buffer zonesand agro‐forestry plantings Aviation fuel is a significant
CLIFTON‐BROWN ET AL | 23
target market (Crawford et al 2016) To serve this marketand to reduce current carbon costs of production (Budsberget al 2016) key improvement traits in addition to yield(eg coppice regeneration pestdisease resistance water‐and nutrient‐use efficiencies) will be trace greenhouse gas(GHG) emissions (eg isoprene volatiles) site adaptabilityand biomass conversion efficiency Efforts are underway tohave national environmental protection agenciesrsquo approvalfor poplar hybrids qualifying for renewable energy credits
6 | REFLECTIONS ON THECOMMERCIALIZATIONCHALLENGE
The research and innovation activities reviewed in thispaper aim to advance the genetic improvement of speciesthat can provide feedstocks for bioenergy applicationsshould those markets eventually develop These marketsneed to generate sufficient revenue and adequately dis-tribute it to the actors along the value chain The work onall four crops shares one thing in common long‐termefforts to integrate fundamental knowledge into breedingand crop development along a research and development(RampD) pipeline The development of miscanthus led byAberystwyth University exemplifies the concerted researcheffort that has integrated the RampD activities from eight pro-jects over 14 years with background core research fundingalong an emerging innovation chain (Figure 2) This pro-gramme has produced a first range of conventionally bredseeded interspecies hybrids which are now in upscaling tri-als (Table 3) The application of molecular approaches(Table 5) with further conventional breeding (Table 4)offers the prospect of a second range of improved seededhybrids This example shows that research‐based support ofthe development of new crops or crop types requires along‐term commitment that goes beyond that normallyavailable from project‐based funding (Figure 1) Innovationin this sector requires continuous resourcing of conven-tional breeding operations and capability to minimize timeand investment losses caused by funding discontinuities
This challenge is increased further by the well‐knownmarket failure in the breeding of many agricultural cropspecies The UK Department for Environment Food andRural Affairs (Defra) examined the role of geneticimprovement in relation to nonmarket outcomes such asenvironmental protection and concluded that public invest-ment in breeding was required if profound market failure isto be addressed (Defra 2002) With the exception ofwidely grown hybrid crops such as maize and some high‐value horticultural crops royalties arising from plant breed-ersrsquo rights or other returns to breeders fail to adequatelycompensate for the full cost for research‐based plant breed-ing The result even for major crops such as wheat is sub‐
optimal investment and suboptimal returns for society Thismarket failure is especially acute for perennial crops devel-oped for improved sustainability rather than consumerappeal (Tracy et al 2016) Figure 3 illustrates the underly-ing challenge of capturing value for the breeding effortThe ldquovalley of deathrdquo that results from the low and delayedreturns to investment applies generally to the research‐to‐product innovation pipeline (Beard Ford Koutsky amp Spi-wak 2009) and certainly to most agricultural crop speciesHowever this schematic is particularly relevant to PBCsMost of the value for society from the improved breedingof these crops comes from changes in how agricultural landis used that is it depends on the increased production ofthese crops The value for society includes many ecosys-tems benefits the effects of a return to seminatural peren-nial crop cover that protects soils the increase in soilcarbon storage the protection of vulnerable land or the cul-tivation of polluted soils and the reductions in GHG emis-sions (Lewandowski 2016) By its very nature theproduction of biomass on agricultural land marginal forfood production challenges farm‐level profitability Thecosts of planting material and one‐time nature of cropestablishment are major early‐stage costs and therefore theopportunities for conventional royalty capture by breedersthat are manifold for annual crops are limited for PBCs(Hastings et al 2017) Public investment in developingPBCs for the nonfood biobased sector needs to providemore long‐term support for this critical foundation to a sus-tainable bioeconomy
7 | CONCLUSIONS
This paper provides an overview of research‐based plantbreeding in four leading PBCs For all four PBC generasignificant progress has been made in genetic improvementthrough collaboration between research scientists and thoseoperating ongoing breeding programmes Compared withthe main food crops most PBC breeding programmes dateback only a few decades (Table 1) This breeding efforthas thus co‐evolved with molecular biology and the result-ing ‐omics technologies that can support breeding Thedevelopment of all four PBCs has depended strongly onpublic investment in research and innovation The natureand driver of the investment varied In close associationwith public research organizations or universities all theseprogrammes started with germplasm collection and charac-terization which underpin the selection of parents forexploratory wide crosses for progeny testing (Figure 1Table 4)
Public support for switchgrass in North America wasexplicitly linked to plant breeding with 12 breeding pro-grammes supported in the United States and CanadaSwitchgrass breeding efforts to date using conventional
24 | CLIFTON‐BROWN ET AL
breeding have resulted in over 36 registered cultivars inthe United States (Table 1) with the development of dedi-cated biomass‐type cultivars coming within the past fewyears While ‐omics technologies have been incorporatedinto several of these breeding programmes they have notyet led to commercial deployment in either conventional orhybrid cultivars
Willow genetic improvement was led by the researchcommunity closely linked to plant breeding programmesWillow and poplar have the longest record of public invest-ment in genetic improvement that can be traced back to1920s in the United Kingdom and United States respec-tively Like switchgrass breeding programmes for willoware connected to public research efforts The United King-dom in partnership with the programme based at CornellUniversity remains the European leader in willowimprovement with a long‐term breeding effort closelylinked to supporting biological research at Rothamsted Inwillow F1 hybrids have produced impressive yield gainsover parental germplasm by capturing hybrid vigour Over30 willow clones are commercially available in the UnitedStates and Europe and a further ~90 are under precommer-cial testing (Table 1)
Compared with willow and poplar miscanthus is a rela-tive newcomer with all the current breeding programmesstarting in the 2000s Clonal M times giganteus propagated byrhizomes is expected to be replaced by more readily scal-able seeded hybrids from intra‐ (M sinensis) and inter‐(M sacchariflorus times M sinensis) species crosses with highseed multiplication rates (of gt2000) The first group ofhybrid cultivars is expected to be market‐ready around2022
Of the four genera used as PBCs Populus is the mostadvanced in terms of achievements in biological researchas a result of its use as a model for basic research of treesMuch of this biological research is not directly connectedto plant breeding Nevertheless reflecting the fact thatpoplar is widely grown as a single‐stem tree in SRF thereare about 60 commercially available clones and an addi-tional 80 clones in commercial pipelines (Table 1) Trans-genic poplar hybrids have moved beyond proof of conceptto commercial reality in China
Many PBC programmes have initiated long‐term con-ventional recurrent selection breeding cycles for populationimprovement which is a key process in increasing yieldthrough hybrid vigour As this approach requires manyyears most programmes are experimenting with molecularbreeding methods as these have the potential to accelerateprecision breeding For all four PBCs investments in basicgenetic and genomic resources including the developmentof mapping populations for QTLs and whole genomesequences are available to support long‐term advancesMore recently association genetics with panels of diversegermplasm are being used as training populations for GSmodels (Table 5) These efforts are benefitting from pub-licly available DNA sequences and whole genome assem-blies in crop databases Key to these accelerated breedingtechnologies are developments in novel phenomics tech-nologies to bridge the genotypephenotype gap In poplarnovel remote sensing field phenotyping is now beingdeployed to assist breeders These advances are being com-bined with in vitro and in planta modern molecular breed-ing techniques such as CRISPR (Table 5) CRISPRtechnology for genome editing has been proven in poplar
FIGURE 2 A schematic developmentpathway for miscanthus in the UnitedKingdom related to the investment in RampDprojects at Aberystwyth (top coloured areasfor projects in the three categories basicresearch breeding and commercialupscaling) leading to a projected croppingarea of 350000 ha by 2030 with clonal andsuccessive ranges of improved seed‐basedhybrids Purple represents theBiotechnology and Biological SciencesResearch Council (BBSRC) and brown theDepartment for Environment Food andRural Affairs (Defra) (UK Nationalfunding) blue bars represent EU fundingand green private sector funding (Terravestaand CERES) and GIANT‐LINK andMiscanthus Upscaling Technology (MUST)are publicndashprivate‐initiatives (PPI)
CLIFTON‐BROWN ET AL | 25
This technology is also being applied in switchgrass andmiscanthus (Table 5) but the future of CRISPR in com-mercial breeding for the European market is uncertain inthe light of recent ECJ 2018 rulings
There is integration of research and plant breeding itselfin all four PBCs Therefore estimating the ongoing costs ofmaintaining these breeding programmes is difficult Invest-ment in research also seeks wider benefits associated withtechnological advances in plant science rather than cultivardevelopment per se However in all cases the conventionalbreeding cycle shown in Figure 1 is the basic ldquoenginerdquo withmolecular technologies (‐omics) serving to accelerate thisengine The history of the development shows that the exis-tence of these breeding programmes is essential to gain ben-efits from the biological research Despite this it is thisessential step that is at most risk from reductions in invest-ment A conventional breeding programme typicallyrequires a breeder and several technicians who are supported
over the long term (20ndash30 years Figure 1) at costs of about05 to 10 million Euro per year (as of 2018) The analysisreported here shows that the time needed to perform onecycle of conventional breeding bringing germplasm fromthe wild to a commercial hybrid ranged from 11 years inswitchgrass to 26 years in poplar (Figure 1) In a maturecrop grown on over 100000 ha with effective cultivar pro-tection and a suitable business model this level of revenuecould come from royalties Until such levels are reachedPBCs lie in the innovation valley of death (Figure 3) andneed public support
Applying industrial ldquotechnology readiness levelsrdquo (TRL)originally developed for aerospace (Heacuteder 2017) to ourplant breeding efforts we estimate many promising hybridscultivars are at TRL levels of 3ndash4 In Table 3 experts ineach crop estimate that it would take 3 years from now toupscale planting material from leading cultivars in plot tri-als to 100 ha
FIGURE 3 A schematic relating some of the steps in the innovation chain from relatively basic crop science research through to thedeployment in commercial cropping systems and value chains The shape of the funnel above the expanding development and deploymentrepresents the availability of investment along the development chain from relatively basic research at the top to the upscaled deployment at thebottom Plant breeding links the research effort with the development of cropping systems The constriction represents the constrained fundingfor breeding that links conventional public research investment and the potential returns from commercial development The handover pointsbetween publicly funded work to develop the germplasm resources (often known as prebreeding) the breeding and the subsequent cropdevelopment are shown on the left The constriction point is aggravated by the lack academic rewards for this essential breeding activity Theoutcome is such that this innovation system is constrained by the precarious resourcing of plant breeding The authorsrsquo assessment ofdevelopment status of the four species is shown (poplar having two one for short‐rotation coppice (SRC) poplar and one for the more traditionalshort‐rotation forestry (SRF)) The four new perennial biomass crops (PBCs) are now in the critical phase of depending of plant breedingprogress without the income stream from a large crop production base
26 | CLIFTON‐BROWN ET AL
Taking the UK example mentioned in the introductionplanting rates of ~35000 ha per year from 2022 onwardsare needed to reach over 1 m ha by 2050 Ongoing workin the UK‐funded Miscanthus Upscaling Technology(MUST) project shows that ramping annual hybrid seedproduction from the current level of sufficient seed for10 ha in 2018 to 35000 ha would take about 5 yearsassuming no setbacks If current hybrids of any of the fourPBCs in the upscaling pipeline fail on any step for exam-ple lower than expected multiplication rates or unforeseenagronomic barriers then further selections from ongoingbreeding are needed to replace earlier candidates
In conclusion the breeding foundations have been laidwell for switchgrass miscanthus willow and poplar owingto public funding over the long time periods necessaryImproved cultivars or genotypes are available that could bescaled up over a few years if real sustained market oppor-tunities emerged in response to sustained favourable poli-cies and industrial market pull The potential contributionsof growing and using these PBCs for socioeconomic andenvironmental benefits are clear but how farmers andothers in commercial value chains are rewarded for mass‐scale deployment as is necessary is not obvious at presentTherefore mass‐scale deployment of these lignocellulosecrops needs developments outside the breeding arenas todrive breeding activities more rapidly and extensively
8 | UNCITED REFERENCE
Huang et al (2019)
ACKNOWLEDGEMENTS
UK The UK‐led miscanthus research and breeding wasmainly supported by the Biotechnology and BiologicalSciences Research Council (BBSRC) Department for Envi-ronment Food and Rural Affairs (Defra) the BBSRC CSPstrategic funding grant BBCSP17301 Innovate UKBBSRC ldquoMUSTrdquo BBN0161491 CERES Inc and Ter-ravesta Ltd through the GIANT‐LINK project (LK0863)Genomic selection and genomewide association study activi-ties were supported by BBSRC grant BBK01711X1 theBBSRC strategic programme grant on Energy Grasses ampBio‐refining BBSEW10963A01 The UK‐led willow RampDwork reported here was supported by BBSRC (BBSEC00005199 BBSEC00005201 BBG0162161 BBE0068331 BBG00580X1 and BBSEC000I0410) Defra(NF0424 NF0426) and the Department of Trade and Indus-try (DTI) (BW6005990000) IT The Brain Gain Program(Rientro dei cervelli) of the Italian Ministry of EducationUniversity and Research supports Antoine Harfouche USContributions by Gerald Tuskan to this manuscript were sup-ported by the Center for Bioenergy Innovation a US
Department of Energy Bioenergy Research Center supportedby the Office of Biological and Environmental Research inthe DOE Office of Science under contract number DE‐AC05‐00OR22725 Willow breeding efforts at CornellUniversity have been supported by grants from the USDepartment of Agriculture National Institute of Food andAgriculture Contributions by the University of Illinois weresupported primarily by the DOE Office of Science Office ofBiological and Environmental Research (BER) grant nosDE‐SC0006634 DE‐SC0012379 and DE‐SC0018420 (Cen-ter for Advanced Bioenergy and Bioproducts Innovation)and the Energy Biosciences Institute EU We would like tofurther acknowledge contributions from the EU projectsldquoOPTIMISCrdquo FP7‐289159 on miscanthus and ldquoWATBIOrdquoFP7‐311929 on poplar and miscanthus as well as ldquoGRACErdquoH2020‐EU326 Biobased Industries Joint Technology Ini-tiative (BBI‐JTI) Project ID 745012 on miscanthus
CONFLICT OF INTEREST
The authors declare that progress reported in this paperwhich includes input from industrial partners is not biasedby their business interests
ORCID
John Clifton‐Brown httporcidorg0000-0001-6477-5452Antoine Harfouche httporcidorg0000-0003-3998-0694Lawrence B Smart httporcidorg0000-0002-7812-7736Danny Awty‐Carroll httporcidorg0000-0001-5855-0775VasileBotnari httpsorcidorg0000-0002-0470-0384Lindsay V Clark httporcidorg0000-0002-3881-9252SalvatoreCosentino httporcidorg0000-0001-7076-8777Lin SHuang httporcidorg0000-0003-3079-326XJon P McCalmont httporcidorg0000-0002-5978-9574Paul Robson httporcidorg0000-0003-1841-3594AnatoliiSandu httporcidorg0000-0002-7900-448XDaniloScordia httporcidorg0000-0002-3822-788XRezaShafiei httporcidorg0000-0003-0891-0730Gail Taylor httporcidorg0000-0001-8470-6390IrisLewandowski httporcidorg0000-0002-0388-4521
REFERENCES
Alburquerque N Baldacci‐Cresp F Baucher M Casacuberta JM Collonnier C ElJaziri M hellip Burgos L (2016) New trans-formation technologies for trees In C Vettori F Gallardo H
CLIFTON‐BROWN ET AL | 27
Haumlggman V Kazana F Migliacci G Pilateamp M Fladung(Eds) Biosafety of forest transgenic trees (pp 31ndash66) DordrechtNetherlands Springer
Argus G W (2007) Salix (Salicaceae) distribution maps and a syn-opsis of their classification in North America north of MexicoHarvard Papers in Botany 12 335ndash368 httpsdoiorg1031001043-4534(2007)12[335SSDMAA]20CO2
Atienza S G Ramirez M C amp Martin A (2003) Mapping‐QTLscontrolling flowering date in Miscanthus sinensis Anderss CerealResearch Communications 31 265ndash271
Atienza S G Satovic Z Petersen K K Dolstra O amp Martin A(2003) Identification of QTLs influencing agronomic traits inMiscanthus sinensis Anderss I Total height flag‐leaf height andstem diameter Theoretical and Applied Genetics 107 123ndash129
Atienza S G Satovic Z Petersen K K Dolstra O amp Martin A(2003) Influencing combustion quality in Miscanthus sinensisAnderss Identification of QTLs for calcium phosphorus and sul-phur content Plant Breeding 122 141ndash145
Bdeir R Muchero W Yordanov Y Tuskan G A Busov V ampGailing O (2017) Quantitative trait locus mapping of Populusbark features and stem diameter BMC Plant Biology 17 224httpsdoiorg101186s12870-017-1166-4
Beard T R Ford G S Koutsky T M amp Spiwak L J (2009) AValley of Death in the innovation sequence An economic investi-gation Research Evaluation 18 343ndash356 httpsdoiorg103152095820209X481057
Berguson W McMahon B amp Riemenschneider D (2017) Addi-tive and non‐additive genetic variances for tree growth in severalhybrid poplar populations and implications regarding breedingstrategy Silvae Genetica 66(1) 33ndash39 httpsdoiorg101515sg-2017-0005
Berlin S Trybush S O Fogelqvist J Gyllenstrand N Halling-baumlck H R Aringhman I hellip Hanley S J (2014) Genetic diversitypopulation structure and phenotypic variation in European Salixviminalis L (Salicaceae) Tree Genetics amp Genomes 10 1595ndash1610 httpsdoiorg101007s11295-014-0782-5
Bernardo R amp Yu J (2007) Prospects for genomewide selectionfor quantitative traits in maize Crop Science 47 1082ndash1090httpsdoiorg102135cropsci2006110690
Biswal A K Atmodjo M A Li M Baxter H L Yoo C GPu Y hellip Mohnen D (2018) Sugar release and growth of bio-fuel crops are improved by downregulation of pectin biosynthesisNature Biotechnology 36 249ndash257
Bmu (2009) National biomass action plan for Germany (p 17) Bun-desministerium fuumlr Umwelt Naturschutz und ReaktorsicherheitBundesministerium fuumlr Ernaumlhrung (BMU) amp Landwirtschaft undVerbraucherschutz (BMELV) 11055 Berlin | Germany Retrievedfrom httpwwwbmeldeSharedDocsDownloadsENPublicationsBiomassActionPlanpdf__blob=publicationFile
Brummer E C (1999) Capturing heterosis in forage crop cultivardevelopment Crop Science 39 943ndash954 httpsdoiorg102135cropsci19990011183X003900040001x
Budsberg E Crawford J T Morgan H Chin W S Bura R ampGustafson R (2016) Hydrocarbon bio‐jet fuel from bioconver-sion of poplar biomass Life cycle assessment Biotechnology forBiofuels 9 170 httpsdoiorg101186s13068-016-0582-2
Burris K P Dlugosz E M Collins A G Stewart C N amp Lena-ghan S C (2016) Development of a rapid low‐cost protoplasttransfection system for switchgrass (Panicum virgatum L) Plant
Cell Reports 35 693ndash704 httpsdoiorg101007s00299-015-1913-7
Casler M (2012) Switchgrass breeding genetics and genomics InA Monti (Ed) Switchgrass (pp 29ndash54) London UK Springer
Casler M Mitchell R amp Vogel K (2012) Switchgrass In CKole C P Joshi amp D R Shonnard (Eds) Handbook of bioen-ergy crop plants Vol 2 New York NY Taylor amp Francis
Casler M D amp Ramstein G P (2018) Breeding for biomass yieldin switchgrass using surrogate measures of yield BioEnergyResearch 11 6ndash12
Casler M D Sosa S Hoffman L Mayton H Ernst C Adler PR hellip Bonos S A (2017) Biomass Yield of Switchgrass Culti-vars under High‐ versus Low‐Input Conditions Crop Science 57821ndash832 httpsdoiorg102135cropsci2016080698
Casler M D amp Vogel K P (2014) Selection for biomass yield inupland lowland and hybrid switchgrass Crop Science 54 626ndash636 httpsdoiorg102135cropsci2013040239
Casler M D Vogel K P amp Harrison M (2015) Switchgrassgermplasm resources Crop Science 55 2463ndash2478 httpsdoiorg102135cropsci2015020076
Casler M D Vogel K P Lee D K Mitchell R B Adler P RSulc R M hellip Moore K J (2018) 30 Years of progress towardincreasing biomass yield of switchgrass and big bluestem CropScience 58 1242
Clark L V Brummer J E Głowacka K Hall M C Heo K PengJ hellip Sacks E J (2014) A footprint of past climate change on thediversity and population structure of Miscanthus sinensis Annals ofBotany 114 97ndash107 httpsdoiorg101093aobmcu084
Clark L V Dzyubenko E Dzyubenko N Bagmet L SabitovA Chebukin P hellip Sacks E J (2016) Ecological characteristicsand in situ genetic associations for yield‐component traits of wildMiscanthus from eastern Russia Annals of Botany 118 941ndash955
Clark L V Jin X Petersen K K Anzoua K G Bagmet LChebukin P hellip Sacks E J(2018) Population structure of Mis-canthus sacchariflorus reveals two major polyploidization eventstetraploid‐mediated unidirectional introgression from diploid Msinensis and diversity centred around the Yellow Sea Annals ofBotany 20 1ndash18 httpsdoiorg101093aobmcy1161
Clark L V Stewart J R Nishiwaki A Toma Y Kjeldsen J BJoslashrgensen U hellip Sacks E J (2015) Genetic structure ofMiscanthus sinensis and Miscanthus sacchariflorus in Japan indi-cates a gradient of bidirectional but asymmetric introgressionJournal of Experimental Botany 66 4213ndash4225
Clifton‐Brown J Hastings A Mos M McCalmont J P Ashman CAwty‐Carroll DhellipCracroft‐EleyW (2017) Progress in upscalingMis-canthus biomass production for the European bio‐economy with seed‐based hybridsGlobal Change Biology Bioenergy 9 6ndash17
Clifton‐Brown J Schwarz K U amp Hastings A (2015) History ofthe development of Miscanthus as a bioenergy crop From smallbeginnings to potential realisation Biology and Environment‐Pro-ceedings of the Royal Irish Academy 115B 45ndash57
Clifton‐Brown J Senior H Purdy S Horsnell R Lankamp BMuumlennekhoff A-K hellip Bentley A R (2018) Investigating thepotential of novel non‐woven fabrics to increase pollination effi-ciency in plant breeding PloS One (Accepted)
Crawford J T Shan CW Budsberg E Morgan H Bura R ampGus-tafson R (2016) Hydrocarbon bio‐jet fuel from bioconversion ofpoplar biomass Techno‐economic assessment Biotechnology forBiofuels 9 141 httpsdoiorg101186s13068-016-0545-7
28 | CLIFTON‐BROWN ET AL
Dalton S (2013) Biotechnology of Miscanthus In S M Jain amp SD Gupta (Eds) Biotechnology of neglected and underutilizedcrops (pp 243ndash294) Dordrecht Netherlands Springer
Davey C L Robson P Hawkins S Farrar K Clifton‐Brown J CDonnison I S amp Slavov G T (2017) Genetic relationshipsbetween spring emergence canopy phenology and biomass yieldincrease the accuracy of genomic prediction in Miscanthus Journalof Experimental Botany 68 5093ndash5102 httpsdoiorg101093jxberx339
David X amp Anderson X (2002) Aspen and larch genetics cooper-ative annual report 13 St Paul MN Department of ForestResources University of Minnesota
Deblock M (1990) Factors influencing the tissue‐culture and theAgrobacterium‐Tumefaciens‐mediated transformation of hybridaspen and poplar clones Plant Physiology 93 1110ndash1116httpsdoiorg101104pp9331110
Defra (2002) The role of future public research investment in thegenetic improvement of UK‐grown crops (p 220) LondonRetrieved from sciencesearchdefragovukDefaultaspxMenu=MenuampModule=MoreampLocation=NoneampCompleted=220ampProjectID=10412
Dewoody J Trewin H amp Taylor G (2015) Genetic and morpho-logical differentiation in Populus nigra L Isolation by coloniza-tion or isolation by adaptation Molecular Ecology 24 2641ndash2655
Dong H Liu S Clark L V Sharma S Gifford J M Juvik JA hellip Sacks E J (2018) Genetic mapping of biomass yield inthree interconnected Miscanthus populations Global Change Biol-ogy Bioenergy 10 165ndash185
Dukes (2017) Digest of UK Energy Statistics (DUKES) (p 264) Lon-don UK Department for Business Energy amp Industrial StrategyRetrieved from wwwgovukgovernmentcollectionsdigest-of-uk-energy-statistics-dukes
Eckenwalder J (1996) Systematics and evolution of Populus In RStettler H Bradshaw J Heilman amp T Hinckley (Eds) Biologyof Populus and its implications for management and conservation(pp 7‐32) Ottawa ON National Research Council of Canada
Elshire R J Glaubitz J C Sun Q Poland J A Kawamoto KBuckler E S amp Mitchell S E (2011) A robust simple geno-typing‐by‐sequencing (GBS) approach for high diversity speciesPloS One 6 e19379
Evans H (2016) Bioenergy crops in the UK Case studies of suc-cessful whole farm integration (p 23) Loughborough UKEnergy Technologies Institute Retrieved from httpswwweticouklibrarybioenergy-crops-in-the-uk-case-studies-on-successful-whole-farm-integration-evidence-pack
Evans H (2017) Increasing UK biomass production through moreproductive use of land (p 7) Loughborough UK Energy Tech-nologies Institute Retrieved from httpswwweticouklibraryan-eti-perspective-increasing-uk-biomass-production-through-more-productive-use-of-land
Evans J Sanciangco M D Lau K H Crisovan E Barry KDaum C hellip Buell C R (2018) Extensive genetic diversity ispresent within North American switchgrass germplasm The PlantGenome 11 1ndash16 httpsdoiorg103835plantgenome2017060055
Evans L M Slavov G T Rodgers‐Melnick E Martin J RanjanP Muchero W hellip DiFazio S P (2014) Population genomicsof Populus trichocarpa identifies signatures of selection and
adaptive trait associations Nature Genetics 46 1089ndash1096httpsdoiorg101038ng3075
Fabio E S Volk T A Miller R O Serapiglia M J Gauch HG Van Rees K C J hellip Smart L B (2017) Genotypetimes envi-ronment interaction analysis of North American shrub willowyield trials confirms superior performance of triploid hybrids Glo-bal Change Biology Bioenergy 9 445ndash459 httpsdoiorg101111gcbb12344
Fahrenkrog A M Neves L G Resende M F R Vazquez A Ide los Campos G Dervinis C hellip Kirst M (2017) Genome‐wide association study reveals putative regulators of bioenergytraits in Populus deltoides New Phytologist 213 799ndash811
Fan D Liu T T Li C F Jiao B Li S Hou Y S amp Luo KM (2015) Efficient CRISPRCas9‐mediated targeted mutagenesisin Populus in the first generation Scientific Reports 5 12217
Feng X P Lourgant K Castric V Saumitou-Laprade P ZhengB S Jiang D M amp Brancourt-Hulmel M (2014) The discov-ery of natural accessions related to Miscanthus x giganteus usingchloroplast DNA Crop Science 54 1645ndash1655
Fillatti J J Sellmer J Mccown B Haissig B amp Comai L(1987) Agrobacterium-mediated transformation and regenerationof Populus Molecular amp General Genetics 206 192ndash199
Fu Y B (2015) Understanding crop genetic diversity under modernplant breeding Theoretical and Applied Genetics 128 2131ndash2142 httpsdoiorg101007s00122-015-2585-y
Fu C Mielenz J R Xiao X Ge Y Hamilton C Y RodriguezM hellip Wang Z‐Y (2011) Genetic manipulation of ligninreduces recalcitrance and improves ethanol production fromswitchgrass Proceedings of the National Academy of Sciences ofthe United States of America 108 3803ndash3808 httpsdoiorg101073pnas1100310108
Gao C (2018) The future of CRISPR technologies in agricultureNature Reviews Molecular Cell Biology 19(5) 275ndash276
Gegas V Gay A Camargo A amp Doonan J (2014) Challengesof Crop Phenomics in the Post-genomic Era In J Hancock (Ed)Phenomics (pp 142ndash171) Boca Raton FL CRC Press
Geraldes A DiFazio S P Slavov G T Ranjan P Muchero WHannemann J hellip Tuskan G A (2013) A 34K SNP genotypingarray for Populus trichocarpa Design application to the study ofnatural populations and transferability to other Populus speciesMolecular Ecology Resources 13 306ndash323
Gifford J M Chae W B Swaminathan K Moose S P amp JuvikJ A (2015) Mapping the genome of Miscanthus sinensis forQTL associated with biomass productivity Global Change Biol-ogy Bioenergy 7 797ndash810
Goggin F L Lorence A amp Topp C N (2015) Applying high‐throughput phenotyping to plant‐insect interactions Picturingmore resistant crops Current Opinion in Insect Science 9 69ndash76httpsdoiorg101016jcois201503002
Grabowski P P Evans J Daum C Deshpande S Barry K WKennedy M hellip Casler M D (2017) Genome‐wide associationswith flowering time in switchgrass using exome‐capture sequenc-ing data New Phytologist 213 154ndash169 httpsdoiorg101111nph14101
Greef J M amp Deuter M (1993) Syntaxonomy of Miscanthus xgiganteus GREEF et DEU Angewandte Botanik 67 87ndash90
Guerra F P Richards J H Fiehn O Famula R Stanton B JShuren R hellip Neale D B (2016) Analysis of the genetic varia-tion in growth ecophysiology and chemical and metabolomic
CLIFTON‐BROWN ET AL | 29
composition of wood of Populus trichocarpa provenances TreeGenetics amp Genomes 12 6 httpsdoiorg101007s11295-015-0965-8
Guo H P Shao R X Hong C T Hu H K Zheng B S ampZhang Q X (2013) Rapid in vitro propagation of bioenergy cropMiscanthus sacchariflorus Applied Mechanics and Materials 260181ndash186
Hallingbaumlck H R Fogelqvist J Powers S J Turrion‐Gomez JRossiter R Amey J hellip Roumlnnberg‐Waumlstljung A‐C (2016)Association mapping in Salix viminalis L (Salicaceae)ndashIdentifica-tion of candidate genes associated with growth and phenologyGlobal Change Biology Bioenergy 8 670ndash685
Hanley S J amp Karp A (2014) Genetic strategies for dissectingcomplex traits in biomass willows (Salix spp) Tree Physiology34 1167ndash1180 httpsdoiorg101093treephystpt089
Harfouche A Meilan R amp Altman A (2011) Tree genetic engi-neering and applications to sustainable forestry and biomass pro-duction Trends in Biotechnology 29 9ndash17 httpsdoiorg101016jtibtech201009003
Harfouche A Meilan R amp Altman A (2014) Molecular and phys-iological responses to abiotic stress in forest trees and their rele-vance to tree improvement Tree Physiology 34 1181ndash1198httpsdoiorg101093treephystpu012
Harfouche A Meilan R Kirst M Morgante M Boerjan WSabatti M amp Mugnozza G S (2012) Accelerating the domesti-cation of forest trees in a changing world Trends in PlantScience 17 64ndash72 httpsdoiorg101016jtplants201111005
Hastings A Clifton‐Brown J Wattenbach M Mitchell C PStampfl P amp Smith P (2009) Future energy potential of Mis-canthus in Europe Global Change Biology Bioenergy 1 180ndash196
Hastings A Mos M Yesufu J AMccalmont J Schwarz KShafei RhellipClifton-Brown J (2017) Economic and environmen-tal assessment of seed and rhizome propagated Miscanthus in theUK Frontiers in Plant Science 8 16 httpsdoiorg103389fpls201701058
Heacuteder M (2017) From NASA to EU The evolution of the TRLscale in Public Sector Innovation The Innovation Journal 22 1ndash23 httpsdoiorg103389fpls201701058
Heffner E L Sorrells M E amp Jannink J L (2009) Genomicselection for crop improvement Crop Science 49 1ndash12 httpsdoiorg102135cropsci2008080512
Hodkinson T R Klaas M Jones M B Prickett R amp Barth S(2015) Miscanthus A case study for the utilization of naturalgenetic variation Plant Genetic Resources‐Characterization andUtilization 13 219ndash237 httpsdoiorg101017S147926211400094X
Hodkinson T R amp Renvoize S (2001) Nomenclature of Miscant-hus xgiganteus (Poaceae) Kew Bulletin 56 759ndash760 httpsdoiorg1023074117709
Houle D Govindaraju D R amp Omholt S (2010) Phenomics Thenext challenge Nature Reviews Genetics 11 855ndash866 httpsdoiorg101038nrg2897
Huang X H Wei X H Sang T Zhao Q Feng Q Zhao Yhellip Han B (2010) Genome‐wide association studies of 14 agro-nomic traits in rice landraces Nature Genetics 42 961ndashU976
Hwang O‐J Cho M‐A Han Y‐J Kim Y‐M Lim S‐H KimD‐S hellip Kim J‐I (2014) Agrobacterium‐mediated genetic trans-formation of Miscanthus sinensis Plant Cell Tissue and OrganCulture (PCTOC) 117 51ndash63
Hwang O‐J Lim S‐H Han Y‐J Shin A‐Y Kim D‐S ampKim J‐I (2014) Phenotypic characterization of transgenic Mis-canthus sinensis plants overexpressing Arabidopsis phytochromeB International Journal of Photoenergy 2014501016
Ilstedt B (1996) Genetics and performance of Belgian poplar clonestested in Sweden International Journal of Forest Genetics 3183ndash195
Ingvarsson P K Garcia M V Hall D Luquez V amp Jansson S(2006) Clinal variation in phyB2 a candidate gene for day‐length‐induced growth cessation and bud set across a latitudinalgradient in European aspen (Populus tremula) Genetics 1721845ndash1853
Ingvarsson P K Garcia M V Luquez V Hall D amp Jansson S(2008) Nucleotide polymorphism and phenotypic associationswithin and around the phytochrome B2 locus in European aspen(Populus tremula Salicaceae) Genetics 178 2217ndash2226 httpsdoiorg101534genetics107082354
Jannink J‐L Lorenz A J amp Iwata H (2010) Genomic selectionin plant breeding From theory to practice Briefings in FunctionalGenomics 9 166ndash177 httpsdoiorg101093bfgpelq001
Jiang J Guan Y Mccormick S Juvik J Lubberstedt T amp FeiS‐Z (2017) Gametophytic self‐incompatibility is operative inMiscanthus sinensis (Poaceae) and is affected by pistil age CropScience 57 1948ndash1956
Jones H D (2015a) Future of breeding by genome editing is in thehands of regulators GM Crops amp Food 6 223ndash232
Jones H D (2015b) Regulatory uncertainty over genome editingNature Plants 1 14011
Kalinina O Nunn C Sanderson R Hastings A F S van der Weijde TOumlzguumlven MhellipClifton‐Brown J C (2017) ExtendingMiscanthus cul-tivation with novel germplasm at six contrasting sites Frontiers in PlantScience 8563 httpsdoiorg103389fpls201700563
Karp A Hanley S J Trybush S O Macalpine W Pei M ampShield I (2011) Genetic improvement of willow for bioenergyand biofuels free access Journal of Integrative Plant Biology 53151ndash165 httpsdoiorg101111j1744-7909201001015x
Kersten B Faivre Rampant P Mader M Le Paslier M‐C Bou-non R Berard A hellip Fladung M (2016) Genome sequences ofPopulus tremula chloroplast and mitochondrion Implications forholistic poplar breeding PloS One 11 e0147209 httpsdoiorg101371journalpone0147209
Kiesel A Nunn C Iqbal Y Van der Weijde T Wagner MOumlzguumlven M hellip Lewandowski I (2017) Site‐specific manage-ment of Miscanthus genotypes for combustion and anaerobicdigestion A comparison of energy yields Frontiers in PlantScience 8 347 httpsdoiorg103389fpls201700347
Kim H Kim S T Ryu J Kang B C Kim J S amp Kim S G(2017) CRISPRCpf1‐mediated DNA‐free plant genome editingNature Communications 8 14406 httpsdoiorg101038ncomms14406
King Z R Bray A L Lafayette P R amp Parrott W A (2014)Biolistic transformation of elite genotypes of switchgrass (Pan-icum virgatum L) Plant Cell Reports 33 313ndash322 httpsdoiorg101007s00299-013-1531-1
Kopp R F Maynard C A De Niella P R Smart L B amp Abra-hamson L P (2002) Collection and storage of pollen from Salix(Salicaceae) American Journal of Botany 89 248ndash252
Krzyżak J Pogrzeba M Rusinowski S Clifton‐Brown J McCal-mont J P Kiesel A hellip Mos M (2017) Heavy metal uptake
30 | CLIFTON‐BROWN ET AL
by novel miscanthus seed‐based hybrids cultivated in heavy metalcontaminated soil Civil and Environmental Engineering Reports26 121ndash132 httpsdoiorg101515ceer-2017-0040
Kuzovkina Y A amp Quigley M F (2005) Willows beyond wet-lands Uses of Salix L species for environmental projects WaterAir and Soil Pollution 162 183ndash204 httpsdoiorg101007s11270-005-6272-5
Lazarus W Headlee W L amp Zalesny R S (2015) Impacts of sup-plyshed‐level differences in productivity and land costs on the eco-nomics of hybrid poplar production in Minnesota USA BioEnergyResearch 8 231ndash248 httpsdoiorg101007s12155-014-9520-y
Lewandowski I (2015) Securing a sustainable biomass supply in agrowing bioeconomy Global Food Security 6 34ndash42 httpsdoiorg101016jgfs201510001
Lewandowski I (2016) The role of perennial biomass crops in agrowing bioeconomy In S Barth D Murphy‐Bokern O Kalin-ina G Taylor amp M Jones (Eds) Perennial biomass crops for aresource‐constrained world (pp 3ndash13) Cham SwitzerlandSpringer Switzerland AG
Li J L Meilan R Ma C Barish M amp Strauss S H (2008)Stability of herbicide resistance over 8 years of coppice in field‐grown genetically engineered poplars Western Journal of AppliedForestry 23 89ndash93
Li R Y amp Qu R D (2011) High throughput Agrobacterium‐medi-ated switchgrass transformation Biomass amp Bioenergy 35 1046ndash1054 httpsdoiorg101016jbiombioe201011025
Lindegaard K N amp Barker J H A (1997) Breeding willows forbiomass Aspects of Applied Biology 49 155ndash162
Linde‐Laursen I B (1993) Cytogenetic analysis of MiscanthusGiganteus an interspecific hybrid Hereditas 119 297ndash300httpsdoiorg101111j1601-5223199300297x
Liu Y Merrick P Zhang Z Z Ji C H Yang B amp Fei S Z(2018) Targeted mutagenesis in tetraploid switchgrass (Panicumvirgatum L) using CRISPRCas9 Plant Biotechnology Journal16 381ndash393
Liu L L Wu Y Q Wang Y W amp Samuels T (2012) A high‐density simple sequence repeat‐based genetic linkage map ofswitchgrass G3 (Bethesda Md) 2 357ndash370
Lovett A Suumlnnenberg G amp Dockerty T (2014) The availabilityof land for perennial energy crops in Great Britain GlobalChange Biology Bioenergy 6 99ndash107 httpsdoiorg101111gcbb12147
Lozano‐Juste J amp Cutler S R (2014) Plant genome engineering infull bloom Trends in Plant Science 19 284ndash287 httpsdoiorg101016jtplants201402014
Lu F Lipka A E Glaubitz J Elshire R Cherney J H CaslerM D hellip Costich D E (2013) Switchgrass Genomic DiversityPloidy and Evolution Novel Insights from a Network‐Based SNPDiscovery Protocol Plos Genetics 9 e1003215
Ludovisi R Tauro F Salvati R Khoury S Mugnozza G S ampHarfouche A (2017) UAV‐based thermal imaging for high‐throughput field phenotyping of black poplar response to droughtFrontiers in Plant Science 81681
Macalpine W Shield I amp Karp A (2010) Seed to near market vari-ety the BEGIN willow breeding pipeline 2003ndash2010 and beyond InBioten conference proceedings Birmingham pp 21ndash23
Macalpine W J Shield I F Trybush S O Hayes C M ampKarp A (2008) Overcoming barriers to crossing in willow (Salixspp) breeding Aspects of Applied Biology 90 173ndash180
Mahfouz M M (2017) The efficient tool CRISPR‐Cpf1 NaturePlants 3 17028
Malinowska M Donnison I S amp Robson P R (2017) Phenomicsanalysis of drought responses in Miscanthus collected from differ-ent geographical locations Global Change Biology Bioenergy 978ndash91
Maroder H L Prego I A Facciuto G R amp Maldonado S B(2000) Storage behaviour of Salix alba and Salix matsudanaseeds Annals of Botany 86 1017ndash1021 httpsdoiorg101006anbo20001265
Martinez‐Reyna J amp Vogel K (2002) Incompatibility systems inswitchgrass Crop Science 42 1800ndash1805 httpsdoiorg102135cropsci20021800
Maurya J P amp Bhalerao R P (2017) Photoperiod‐and tempera-ture‐mediated control of growth cessation and dormancy in treesA molecular perspective Annals of Botany 120 351ndash360httpsdoiorg101093aobmcx061
McCalmont J P Hastings A Mcnamara N P Richter G MRobson P Donnison I S amp Clifton‐Brown J (2017) Environ-mental costs and benefits of growing Miscanthus for bioenergy inthe UK Global Change Biology Bioenergy 9 489ndash507
McCracken A R amp Dawson W M (1997) Using mixtures of wil-low clones as a means of controlling rust disease Aspects ofApplied Biology 49 97ndash103
McKown A D Klaacutepště J Guy R D Geraldes A Porth I Hanne-mann JhellipDouglas C J (2014) Genome‐wide association implicatesnumerous genes underlying ecological trait variation in natural popula-tions ofPopulus trichocarpaNew Phytologist 203 535ndash553
Merrick P amp Fei S Z (2015) Plant regeneration and genetic transforma-tion in switchgrass ndash A review Journal of Integrative Agriculture 14483ndash493 httpsdoiorg101016S2095-3119(14)60921-7
Moreno‐Mateos M A Fernandez J P Rouet R Lane M AVejnar C E Mis E hellip Giraldez A J (2017) CRISPR‐Cpf1mediates efficient homology‐directed repair and temperature‐con-trolled genome editing Nature Communications 8 2024 httpsdoiorg101038s41467-017-01836-2
Mosseler A (1990) Hybrid performance and species crossabilityrelationships in willows (Salix) Canadian Journal of Botany 682329ndash2338
Muchero W Guo J DiFazio S P Chen J‐G Ranjan P Slavov GT hellip Tuskan G A (2015) High‐resolution genetic mapping ofallelic variants associated with cell wall chemistry in Populus BMCGenomics 16 24 httpsdoiorg101186s12864-015-1215-z
Neale D B amp Kremer A (2011) Forest tree genomics Growingresources and applications Nature Reviews Genetics 12 111ndash122 httpsdoiorg101038nrg2931
Nunn C Hastings A F S J Kalinina O Oumlzguumlven M SchuumlleH Tarakanov I G hellip Clifton‐Brown J C (2017) Environmen-tal influences on the growing season duration and ripening ofdiverse Miscanthus germplasm grown in six countries Frontiersin Plant Science 8 907 httpsdoiorg103389fpls201700907
Ogawa Y Honda M Kondo Y amp Hara‐Nishimura I (2016) Anefficient Agrobacterium‐mediated transformation method forswitchgrass genotypes using Type I callus Plant Biotechnology33 19ndash26
Ogawa Y Shirakawa M Koumoto Y Honda M Asami Y KondoY ampHara‐Nishimura I (2014) A simple and reliable multi‐gene trans-formation method for switchgrass Plant Cell Reports 33 1161ndash1172httpsdoiorg101007s00299-014-1605-8
CLIFTON‐BROWN ET AL | 31
Okada M Lanzatella C Saha M C Bouton J Wu R L amp Tobias CM (2010) Complete switchgrass genetic maps reveal subgenomecollinearity preferential pairing and multilocus interactions Genetics185 745ndash760 httpsdoiorg101534genetics110113910
Palomo‐Riacuteos E Macalpine W Shield I Amey J Karaoğlu C WestJ hellip Jones H D (2015) Efficient method for rapid multiplication ofclean and healthy willow clones via in vitro propagation with broadgenotype applicability Canadian Journal of Forest Research 451662ndash1667 httpsdoiorg101139cjfr-2015-0055
Pilate G Allona I Boerjan W Deacutejardin A Fladung M Gal-lardo F hellip Halpin C (2016) Lessons from 25 years of GM treefield trials in Europe and prospects for the future In C Vettori FGallardo H Haumlggman V Kazana F Migliacci G Pilateamp MFladung (Eds) Biosafety of forest transgenic trees (pp 67ndash100)Dordrecht Netherlands Springer Switzerland AG
Pinosio S Giacomello S Faivre-Rampant P Taylor G Jorge VLe Paslier M C amp Marroni F (2016) Characterization of thepoplar pan-genome by genome-wide identification of structuralvariation Molecular Biology and Evolution 33 2706ndash2719
Porth I Klaacutepště J Skyba O Friedmann M C Hannemann JEhlting J hellip Douglas C J (2013) Network analysis reveals therelationship among wood properties gene expression levels andgenotypes of natural Populus trichocarpa accessions New Phytol-ogist 200 727ndash742
Pugesgaard S Schelde K Larsen S U Laeligrke P E amp JoslashrgensenU (2015) Comparing annual and perennial crops for bioenergyproductionndashinfluence on nitrate leaching and energy balance Glo-bal Change Biology Bioenergy 7 1136ndash1149 httpsdoiorg101111gcbb12215
Quinn L D Allen D J amp Stewart J R (2010) Invasivenesspotential of Miscanthus sinensis Implications for bioenergy pro-duction in the United States Global Change Biology Bioenergy2 310ndash320 httpsdoiorg101111j1757-1707201001062x
Rae A M Robinson K M Street N R amp Taylor G (2004)Morphological and physiological traits influencing biomass pro-ductivity in short‐rotation coppice poplar Canadian Journal ofForest Research‐Revue Canadienne De Recherche Forestiere 341488ndash1498 httpsdoiorg101139x04-033
Rambaud C Arnoult S Bluteau A Mansard M C Blassiau Camp Brancourt‐Hulmel M (2013) Shoot organogenesis in threeMiscanthus species and evaluation for genetic uniformity usingAFLP analysis Plant Cell Tissue and Organ Culture (PCTOC)113 437ndash448 httpsdoiorg101007s11240-012-0284-9
Ramstein G P Evans J Kaeppler S M Mitchell R B Vogel K PBuell C R amp Casler M D (2016) Accuracy of genomic prediction inswitchgrass (Panicum virgatum L) improved by accounting for linkagedisequilibriumG3 (Bethesda Md) 6 1049ndash1062
Rayburn A L Crawford J Rayburn C M amp Juvik J A (2009)Genome size of three Miscanthus species Plant Molecular Biol-ogy Reporter 27 184 httpsdoiorg101007s11105-008-0070-3
Richardson J Isebrands J amp Ball J (2014) Ecology and physiol-ogy of poplars and willows In J Isebrands amp J Richardson(Eds) Poplars and willows Trees for society and the environ-ment (pp 92‐123) Boston MA and Rome CABI and FAO
Robison T L Rousseau R J amp Zhang J (2006) Biomass produc-tivity improvement for eastern cottonwood Biomass amp Bioenergy30 735ndash739 httpsdoiorg101016jbiombioe200601012
Robson P R Farrar K Gay A P Jensen E F Clifton‐Brown JC amp Donnison I S (2013) Variation in canopy duration in the
perennial biofuel crop Miscanthus reveals complex associationswith yield Journal of Experimental Botany 64 2373ndash2383
Robson P Jensen E Hawkins S White S R Kenobi K Clif-ton‐Brown J hellip Farrar K (2013) Accelerating the domestica-tion of a bioenergy crop Identifying and modelling morphologicaltargets for sustainable yield increase in Miscanthus Journal ofExperimental Botany 64 4143ndash4155
Sanderson M A Adler P R Boateng A A Casler M D ampSarath G (2006) Switchgrass as a biofuels feedstock in theUSA Canadian Journal of Plant Science 86 1315ndash1325httpsdoiorg104141P06-136
Scarlat N Dallemand J F Monforti‐Ferrario F amp Nita V(2015) The role of biomass and bioenergy in a future bioecon-omy Policies and facts Environmental Development 15 3ndash34httpsdoiorg101016jenvdev201503006
Serapiglia M J Gouker F E amp Smart L B (2014) Early selec-tion of novel triploid hybrids of shrub willow with improved bio-mass yield relative to diploids BMC Plant Biology 14 74httpsdoiorg1011861471-2229-14-74
Serba D Wu L Daverdin G Bahri B A Wang X Kilian Ahellip Devos K M (2013) Linkage maps of lowland and upland tet-raploid switchgrass ecotypes BioEnergy Research 6 953ndash965httpsdoiorg101007s12155-013-9315-6
Shakoor N Lee S amp Mockler T C (2017) High throughput phe-notyping to accelerate crop breeding and monitoring of diseases inthe field Current Opinion in Plant Biology 38 184ndash192 httpsdoiorg101016jpbi201705006
Shield I Macalpine W Hanley S amp Karp A (2015) Breedingwillow for short rotation coppice energy cropping In Industrialcrops Breeding for BioEnergy and Bioproducts (pp 67ndash80) NewYork NY Springer
Sjodin A Street N R Sandberg G Gustafsson P amp Jansson S (2009)The Populus Genome Integrative Explorer (PopGenIE) A new resourcefor exploring the Populus genomeNewPhytologist 182 1013ndash1025
Slavov G amp Davey C (2017) Integrated genomic prediction inbioenergy crops In Abstracts from the International Conferenceon Developing biomass crops for future climates pp 34 24ndash27September 2017 Oriel College Oxford wwwwatbioeu
Slavov G Davey C Bosch M Robson P Donnison I ampMackay I (2018a) Genomic index selection provides a pragmaticframework for setting and refining multi‐objective breeding targetsin Miscanthus Annals of Botany (in press)
Slavov G T DiFazio S P Martin J Schackwitz W MucheroW Rodgers‐Melnick E hellip Tuskan G A (2012) Genome rese-quencing reveals multiscale geographic structure and extensivelinkage disequilibrium in the forest tree Populus trichocarpa NewPhytologist 196 713ndash725
Slavov G Davey C Robson P Donnison I amp Mackay I(2018b) Domestication of Miscanthus through genomic indexselection In Plant and Animal Genome Conference 13 January2018 San Diego CA Retrieved from httpspagconfexcompagxxvimeetingappcgiPaper30622
Slavov G T Nipper R Robson P Farrar K Allison G GBosch M hellip Jensen E (2013) Genome‐wide association studiesand prediction of 17 traits related to phenology biomass and cellwall composition in the energy grass Miscanthus sinensis NewPhytologist 201 1227ndash1239
Slavov G Robson P Jensen E Hodgson E Farrar K AllisonG hellip Donnison I (2014) Contrasting geographic patterns of
32 | CLIFTON‐BROWN ET AL
genetic variation for molecular markers vs phenotypic traits in theenergy grass Miscanthus sinensis Global Change Biology Bioen-ergy 5 562ndash571
Ślusarkiewicz‐Jarzina A Ponitka A Cerazy‐Waliszewska J Woj-ciechowicz M K Sobańska K Jeżowski S amp Pniewski T(2017) Effective and simple in vitro regeneration system of Mis-canthus sinensis Mtimes giganteus and M sacchariflorus for plant-ing and biotechnology purposes Biomass and Bioenergy 107219ndash226 httpsdoiorg101016jbiombioe201710012
Smart L amp Cameron K (2012) Shrub willow In C Kole S Joshiamp D Shonnard (Eds) Handbook of bioenergy crop plants (pp687ndash708) Boca Raton FL Taylor and Francis Group
Stanton B J (2014) The domestication and conservation of Populusand Salix genetic resources (Chapter 4) In Isebrands J ampRichardson J (Eds) Poplars and willows Trees for society andthe environment (pp 124ndash200) Rome Italy CAB InternationalFood and Agricultural Organization of the United Nations
Stanton B J Neale D B amp Li S (2010) Populus breeding Fromthe classical to the genomic approach In S Jansson R Bha-lerao amp A T Groover (Eds) Genetics and genomics of popu-lus (pp 309ndash348) New York NY Springer
Stewart J R Toma Y Fernandez F G Nishiwaki A Yamada T ampBollero G (2009) The ecology and agronomy ofMiscanthus sinensisa species important to bioenergy crop development in its native range inJapan A reviewGlobal Change Biology Bioenergy 1 126ndash153
Stott K G (1992) Willows in the service of man Proceedings of the RoyalSociety of Edinburgh Section B Biological Sciences 98 169ndash182
Strauss S H Ma C Ault K amp Klocko A L (2016) Lessonsfrom two decades of field trials with genetically modified trees inthe USA Biology and regulatory compliance In C Vettori FGallardo H Haumlggman V Kazana F Migliacci G Pilate amp MFladung (Eds) Biosafety of forest transgenic trees (pp 101ndash124)Dordrecht Netherlands Springer
Suda Y amp Argus G W (1968) Chromosome numbers of some NorthAmerican Salix Brittonia 20 191ndash197 httpsdoiorg1023072805440
Tan B (2018) Genomic selection and genome‐wide association studies todissect quantitative traits in forest trees Umearing Sweden University
Tornqvist C Taylor M Jiang Y Evans J Buell C KaepplerS amp Casler M (2018) Quantitative trait locus mapping for flow-ering time in a lowland x upland switchgrass pseudo‐F2 popula-tion The Plant Genome 11 170093
Toacuteth G Hermann T Da Silva M amp Montanarella L (2016)Heavy metals in agricultural soils of the European Union withimplications for food safety Environment International 88 299ndash309 httpsdoiorg101016jenvint201512017
Tracy W Dawson J Moore V amp Fisch J (2016) Intellectual propertyrights and public plant breeding Recommendations and proceedings ofa conference on best practices for intellectual property protection of pub-lically developed plant germplasm (p 70) University of Wisconsin‐Madison Retrieved from httpshostcalswisceduagronomywp-contentuploadssites1620162005Proceedings-IPR-Finalpdf
Trybush S Jahodovaacute Š Macalpine W amp Karp A (2008) A geneticstudy of a Salix germplasm resource reveals new insights into rela-tionships among subgenera sections and species BioEnergyResearch 1 67ndash79 httpsdoiorg101007s12155-008-9007-9
Tsai C J (2013) Next‐generation sequencing for next‐generationbreeding and more New Phytologist 198 635ndash637 httpsdoiorg101111nph12245
Tuskan G A Difazio S Jansson S Bohlmann J Grigoriev I HellstenU hellip Rokhsar D (2006) The genome of black cottonwood Populustrichocarpa (Torr ampGray) Science 313 1596ndash1604
Tuskan G A amp Walsh M E (2001) Short‐rotation woody cropsystems atmospheric carbon dioxide and carbon management AUS case study The Forestry Chronicle 77 259ndash264 httpsdoiorg105558tfc77259-2
Van Den Broek R Treffers D‐J Meeusen M Van Wijk ANieuwlaar E amp Turkenburg W (2001) Green energy or organicfood A life‐cycle assessment comparing two uses of set‐asideland Journal of Industrial Ecology 5 65ndash87 httpsdoiorg101162108819801760049477
Van Der Schoot J Pospiskova M Vosman B amp Smulders M J M(2000) Development and characterization of microsatellite markers inblack poplar (Populus nigra L) Theoretical and Applied Genetics 101317ndash322 httpsdoiorg101007s001220051485
Van Der Weijde T Dolstra O Visser R G amp Trindade L M(2017a) Stability of cell wall composition and saccharificationefficiency in Miscanthus across diverse environments Frontiers inPlant Science 7 2004
Van der Weijde T Huxley L M Hawkins S Sembiring E HFarrar K Dolstra O hellip Trindade L M (2017) Impact ofdrought stress on growth and quality of miscanthus for biofuelproduction Global Change Biology Bioenergy 9 770ndash782
Van der Weijde T Kamei C L A Severing E I Torres A FGomez L D Dolstra O hellip Trindade L M (2017) Geneticcomplexity of miscanthus cell wall composition and biomass qual-ity for biofuels BMC Genomics 18 406
Van der Weijde T Kiesel A Iqbal Y Muylle H Dolstra O Visser RG FhellipTrindade LM (2017) Evaluation ofMiscanthus sinensis bio-mass quality as feedstock for conversion into different bioenergy prod-uctsGlobal Change Biology Bioenergy 9 176ndash190
Vanholme B Cesarino I Goeminne G Kim H Marroni F VanAcker R hellip Boerjan W (2013) Breeding with rare defectivealleles (BRDA) A natural Populus nigra HCT mutant with modi-fied lignin as a case study New Phytologist 198 765ndash776
Vogel K P Mitchell R B Casler M D amp Sarath G (2014)Registration of Liberty Switchgrass Journal of Plant Registra-tions 8 242ndash247 httpsdoiorg103198jpr2013120076crc
Wang X Yamada T Kong F‐J Abe Y Hoshino Y Sato Hhellip Yamada T (2011) Establishment of an efficient in vitro cul-ture and particle bombardment‐mediated transformation systems inMiscanthus sinensis Anderss a potential bioenergy crop GlobalChange Biology Bioenergy 3 322ndash332 httpsdoiorg101111j1757-1707201101090x
Watrud L S Lee E H Fairbrother A Burdick C Reichman JR Bollman M hellip Van de Water P K (2004) Evidence forlandscape‐level pollen‐mediated gene flow from genetically modi-fied creeping bentgrass with CP4 EPSPS as a marker Proceedingsof the National Academy of Sciences of the United States of Amer-ica 101 14533ndash14538
Weisgerber H (1993) Poplar breeding for the purpose of biomassproduction in short‐rotation periods in Germany ndash Problems andfirst findings Forestry Chronicle 69 727ndash729 httpsdoiorg105558tfc69727-6
Wheeler N C Steiner K C Schlarbaum S E amp Neale D B(2015) The evolution of forest genetics and tree improvementresearch in the United States Journal of Forestry 113 500ndash510httpsdoiorg105849jof14-120
CLIFTON‐BROWN ET AL | 33
Whittaker C Macalpine W J Yates N E amp Shield I (2016)Dry matter losses and methane emissions during wood chip stor-age The impact on full life cycle greenhouse gas savings of shortrotation coppice willow for heat BioEnergy Research 9 820ndash835 httpsdoiorg101007s12155-016-9728-0
Xi Q (2000) Investigation on the distribution and potential of giant grassesin China PhD thesis Goettingen Germany Cuvillier Verlag
Xi Q amp Jezowkski S (2004) Plant resources of Triarrhena andMiscanthus species in China and its meaning for Europe PlantBreeding and Seed Science 49 63ndash77
Xue S Kalinina O amp Lewandowski I (2015) Present and future optionsfor the improvement of Miscanthus propagation techniques Renewableand Sustainable Energy Reviews 49 1233ndash1246
Yin K Q Gao C X amp Qiu J L (2017) Progress and prospectsin plant genome editing Nature Plants 3 httpsdoiorg101038nplants2017107
YookM (2016)Genetic diversity and transcriptome analysis for salt toler-ance inMiscanthus PhD thesis Seoul National University Korea
Zaidi S S E A Mahfouz M M amp Mansoor S (2017) CRISPR‐Cpf1 A new tool for plant genome editing Trends in PlantScience 22 550ndash553 httpsdoiorg101016jtplants201705001
Zalesny R S Stanturf J A Gardiner E S Perdue J H YoungT M Coyle D R hellip Hass A (2016) Ecosystem services ofwoody crop production systems BioEnergy Research 9 465ndash491 httpsdoiorg101007s12155-016-9737-z
Zhang Q X Sun Y Hu H K Chen B Hong C T Guo H Phellip Zheng B S (2012) Micropropagation and plant regenerationfrom embryogenic callus of Miscanthus sinensis Vitro Cellular ampDevelopmental Biology‐Plant 48 50ndash57 httpsdoiorg101007s11627-011-9387-y
Zhang Y Zalapa J E Jakubowski A R Price D L AcharyaA Wei Y hellip Casler M D (2011) Post‐glacial evolution of
Panicum virgatum Centers of diversity and gene pools revealedby SSR markers and cpDNA sequences Genetica 139 933ndash948httpsdoiorg101007s10709-011-9597-6
Zhao H Wang B He J Yang J Pan L Sun D amp Peng J(2013) Genetic diversity and population structure of Miscanthussinensis germplasm in China PloS One 8 e75672
Zhou X H Jacobs T B Xue L J Harding S A amp Tsai C J(2015) Exploiting SNPs for biallelic CRISPR mutations in theoutcrossing woody perennial Populus reveals 4‐coumarate CoAligase specificity and redundancy New Phytologist 208 298ndash301
Zhou R Macaya‐Sanz D Rodgers‐Melnick E Carlson C HGouker F E Evans L M hellip DiFazio S P (2018) Characteri-zation of a large sex determination region in Salix purpurea L(Salicaceae) Molecular Genetics and Genomics 1ndash16 httpsdoiorg101007s00438-018-1473-y
Zsuffa L Mosseler A amp Raj Y (1984) Prospects for interspecifichybridization in willow for biomass production In K L Perttu(Ed) Ecology and management of forest biomass production sys-tems Report 15 (pp 261ndash281) Uppsala Sweden SwedishUniversity of Agricultural Sciences
How to cite this article Clifton‐Brown J HarfoucheA Casler MD et al Breeding progress andpreparedness for mass‐scale deployment of perenniallignocellulosic biomass crops switchgrassmiscanthus willow and poplar GCB Bioenergy201800000ndash000 httpsdoiorg101111gcbb12566
34 | CLIFTON‐BROWN ET AL
Graphical AbstractThe contents of this page will be used as part of the graphical abstract of html only
It will not be published as part of main article
Plant breeding links the research effort with commercial mass upscaling The authorsrsquo assessment of development status ofthe four species is shown (poplar having two one for short rotation coppice (SRC) poplar and one for the more traditionalshort rotation forestry (SRF)) Mass scale deployment needs developments outside the breeding arenas to drive breedingactivities more rapidly and extensively