Draft - University of Toronto T-Space · 71 runs of Chinook salmon, each named for the timing of...

Draft

Identifying hidden biocomplexity and genomic diversity in Chinook salmon, an imperiled species with a history of

anthropogenic influence

Journal: Canadian Journal of Fisheries and Aquatic Sciences

Manuscript ID cjfas-2019-0171.R1

Manuscript Type: Article

Date Submitted by the Author: 13-Aug-2019

Complete List of Authors: Meek, Mariah; Michigan State University, Department of Integrative Biology, and AgBio ResearchStephens, Molly; University of California MercedGoodbla, Alisha; University of California DavisMay, Bernie; University of California DavisBaerwald, Melinda; University of California Davis; California Department of Water Resources, Division of Environmental Services

Keyword: portfolio effect, RAD-seq, GENOMICS < General, POPULATION STRUCTURE < General, SALMON < Organisms

Is the invited manuscript for consideration in a Special

Issue? :Not applicable (regular submission)

https://mc06.manuscriptcentral.com/cjfas-pubs

Canadian Journal of Fisheries and Aquatic Sciences

Draft

1

IdentifyinghiddenbiocomplexityandgenomicdiversityinChinooksalmon,animperiled1specieswithahistoryofanthropogenicinfluence2MariahH.Meek1*,MollyR.Stephens2,AlishaGoodbla3,BernieMay3,MelindaR.Baerwald43

4

1.DepartmentofIntegrativeBiologyandAgBioResearch,MichiganStateUniversity,EastLansing,MI.Email:5

[email protected]

2.SchoolofNaturalScience,UniversityofCalifornia,Merced,Merced,CA.Email:[email protected]

3.DepartmentofAnimalScience,UniversityofCalifornia,Davis,Davis,CA.Email:AGoodbla:8

[email protected],BMay:[email protected]

4.CaliforniaDepartmentofWaterResources,DivisionofEnvironmentalServices,3500IndustrialBoulevard,10

WestSacramento,California95691,USA.Email:[email protected]

*correspondingauthor12

13

RunningHead:usinggenomicstoidentifybiocomplexity14

15

16

Page 1 of 51



Draft

2

ABSTRACT17

Biocomplexityisanimportantmechanismforpopulationresilienceinchanging18

environments.However,wearejustbeginningtounderstandhowtoidentify19

biocomplexitysowecanmanagespeciestopromoteresilienceandstability.Genomic20

techniquesareemergingasanimportantmethodforidentifyingbiocomplexity.Central21

Valley(CV)Chinooksalmonareanexampleofaspeciesatriskofextinctionifbetter22

methodsforidentifyingandprotectingbiocomplexityarenotemployed.Toaddressthis23

knowledgegap,weemployedrestriction-siteassociatedDNAsequencingtoconductthe24

firstgenomicstudyofallmajorpopulationsofCVChinooksalmon.Wefoundgreater25

populationstructureacrosstheCentralValleythanpreviouslydescribed.Additionally,we26

showevidencefordifferentiationandadaptationwithinmigratoryphenotypesdespite27

highlevelsofgeneflow.Toassistinmanagementpractices,wealsofindthatgenomicdata28

canvastlyimproveourabilitytoassignindividualstotheirnatalpopulations,evenasthey29

mixduringmigration.Theseresultsdemonstratehowgenomicstudycangreatlyimprove30

ourabilitytoidentifyandconservebiocomplexity.31

32

Page 2 of 51



Draft

3

INTRODUCTION33

Biocomplexityreflectsthelevelofintraspecificdiversitywithinandamongpopulations,34

andisimportantforthestabilityandresilienceofmetapopulations(Hilbornetal.2003).It35

isnecessarytoidentifythefactorsthatcontributetobiocomplexity,suchasgenetic36

diversity,forincreasedunderstandingoftherolesuchfactorsplayinspeciesresilienceand37

persistence.Additionally,thisisimportantfordevelopingsustainablemanagement38

practicesthatpromotelong-termstability(DedrickandBaskett2018).However,fine-scale39

informationongeneticdiversitycanbelackingforspeciesofconservationormanagement40

concern,compromisingtheabilityofresourcemanagerstomonitorandenhancegenetic41

variationessentialtothebiocomplexityofmetapopulations(Angelonietal.2012;Ouborg42

etal.2010).43

44

Pacificsalmonidsareanexcellentexampleoftheimportanceofbiocomplexityfor45

persistenceandharvestproduction.PacificsalmonidstocksinAlaskahavebeenshownto46

displayhighbiocomplexity,allowingforstabilityinannualreturns(Schindleretal.2010).47

Thisisoftenreferredtoastheportfolioeffect,wheredifferencesamongindividual48

populationsinresponsetoenvironmentalconditionscandrivegreaterstabilityofthe49

wholepopulationcomplex,whencomparedtothestabilityofeachpopulationindividually50

(Hilbornetal.2003;Schindleretal.2010).Forexample,variabilityinannualreturnsof51

sockeyesalmon(Oncorhynchusnerka)intheBristolBayofAlaskaisovertwotimeslower52

thanreturnswouldbewithoutthepopulationandlifehistorydiversitypresentinthe53

system(Schindleretal.2010).Incontrast,demographicsynchronyhasincreasedoverpast54

decadesinChinooksalmon(O.tshawytscha)fromtheSnakeRiver,decreasingtheportfolio55

Page 3 of 51



Draft

4

effectandincreasingextinctionrisk(Mooreetal.2010).Thisincreaseinsynchronyhas56

beenattributedtohumanactivitiesthathavedecreasedgeneticdiversityandhomogenized57

habitats,suchashatcheryproductionanddamconstruction(Mooreetal.2010).58

59

Ourabilitytomonitorchangesinbiocomplexityislimitedbyourabilitytoidentify60

componentsofbiocomplexityandmanageforfinescaledifferencesinthesecomponents61

amongpopulations(Hilbornetal.2003;Mooreetal.2010).Advancesingenomic62

techniques,however,providethemeanstoaddressthisproblemthroughmuchfinerscale63

resolutionintheidentificationofbiocomplexity(e.g.Larsonetal.2017;Princeetal.2017).64

Byapplyinggenomicstoidentifyfinescalebiocomplexity,wecanthenmanagetoprotect65

allthediversityfoundwithinspecies,boostingtheportfolioeffectanddecreasing66

extinctionrisk.67

68

Inthisstudy,weapplygenomictoolstobetterunderstandthebiocomplexitycontained69

withinCentralValley(CV)Chinooksalmon.TheCentralValleyofCaliforniaishometofour70

runsofChinooksalmon,eachnamedforthetimingoftheirfreshwaterspawningmigration71

(Fall,Late-fall,Spring,andWinterruns).Theserunsaredifferentiatedbythetimingof72

majorlifehistorytransitions,andtheCentralValleyistheonlyareainthespeciesrange73

wherethesefourrunsco-occur(Williams2006).Giventhisdiversityinlifehistory,74

ChinooksalmonintheCVshouldhavearobustportfolio.However,CVChinookhave75

amongtheweakestportfoliointhespeciesrange(Griffithsetal.2014).76

77

Page 4 of 51



Draft

5

CVChinookwildpopulationabundanceshavebeeninseriousdeclineinrecentdecades78

(Yoshiyamaetal.1998),andseveralofthesepopulationsarepredictedtogoextinctin79

Californiain0-100yearsifnonewactionsaretaken(Moyleetal.2017).SacramentoRiver80

WinterrunandCentralValleySpringrunChinooksalmonarelistedundertheUnited81

StatesEndangeredSpeciesAct(ESA)asendangeredandthreatened,respectively(Federal82

Register1999,2005),whiletheFallandLate-fallrunsarefederalspeciesofconcern83

(Myersetal.1998).SpringandWinterrunsareeachdesignatedastheirownEvolutionary84

SignificantUnit(ESU)andgiventheprotectionof“species”undertheESA,whileFalland85

Late-fallaregroupedtogetherinasingleESU(Myersetal.1998).86

87

Apopulationcollapseintheearly2000sresultedintheclosingofthefisheryin2008and88

2009.Lossoflifehistoryandgeneticdiversityarecitedasmakingthestockssusceptibleto89

collapse(Lindleyetal.2009).Previousworkhasshownthatmuchofthebiocomplexityin90

theCentralValleyhasbeenlost(CarlsonandSatterthwaite2011).Thisincludesdecreased91

portfolioeffectsandincreasesinsynchronyamongpopulationsandresultingincreasesin92

varianceinpopulationabundances(CarlsonandSatterthwaite2011;Griffithsetal.2014).93

ThislossinbiocomplexityincludesgenetichomogenizationoftheFallrun(Williamsonand94

May2005)andintrogressionbetweenrunsduetohatcherypractices(CaliforniaHatchery95

ScientificReviewGroup2012).96

97

Inordertoprotectallthediversityfoundinaportfolioofstocksandpromotepopulation98

buffering,itisvitaltobeabletoaccuratelyassignindividualstotheirpopulationoforigin.99

Thiscanbeverydifficultwhenmultiplepopulationsinterminglealongtheirmigration100

Page 5 of 51



Draft

6

paths,asisthecasewithCVChinooksalmon.CVChinookrunscanco-occuronthe101

spawninggrounds,butevenmorecommonly,theyco-occuralongmigrationroutes,as102

juvenilesinrearinggrounds,andintheopenocean(Williams2006;2012).Forexample,103

allfourrunsofCVChinookusethelimitedfloodplainhabitatintheSanFranciscoEstuary104

tofeedandgrowbeforetheymoveouttotheoceanandtheycanoccurinthishabitatat105

overlappingtimes(Sommeretal.2001a;b).Currently,mucheffortisbeinginvestedin106

understandinghowdifferentpopulationsofjuvenilesusetheSanFranciscoBay-Delta107

system(Johnsonetal.2017).Becausetherearenomorphologicaldifferencesamongthe108

differentruns,thishasmadeaccuratelymonitoringandmanagingthedifferentpopulations109

veryerrorprone(Harveyetal.2014).110

111

Hatcherymanagementpracticescanalsodisruptthenaturalprocessesthatcreateand112

maintainpopulationdiversificationandbiocomplexity(HuberandCarlson2015;113

SatterthwaiteandCarlson2015).IntheCV,hatcherieswerecreatedtomitigateforthe114

negativeeffectsofthedamsonChinookpopulations,yetwenowknowhatcheriescanhave115

strongnegativeeffectsonwildpopulations(Arakietal.2007;butalsoseeHessetal.2012).116

TherearefivehatcheriesthroughouttheCVproducingFallrunpopulationsandoneeach117

fortheremainingruns.Hatcherypracticeshaveincludedtruckingandreleasingjuveniles118

downstreamofthehatcheryorintheSanFranciscoEstuarytodecreasejuvenilemortality,119

whichhindershomingduringspawningmigrations(CaliforniaHatcheryScientificReview120

Group2012).Additionally,theonlyCVhatcherytoproducebothspringandfallrun121

Chinooksalmon(theFeatherRiverHatchery)hascausedhybridizationbetweenthetwo122

runsintheirhatcherybreedingprogrambynotproperlyseparatingthebroodstockforthe123

Page 6 of 51



Draft

7

tworuns(CaliforniaHatcheryScientificReviewGroup2012).Thesepracticeshaveledto124

increasedstrayingamongpopulationsandincreasedsynchronyviademographiccoupling125

(HuberandCarlson2015;SatterthwaiteandCarlson2015).Thisinturnhasresultedinthe126

mixingofgenepoolsamongrunsandpopulationsthatwerepreviouslydistinct,and127

ultimatelycomplicatestheidentificationofmanagementunitsandassignmentof128

individualstothoseunits.129

130

PreviousgeneticstudiesofCVChinookpopulationshavereportedvaryinglevelsofgenetic131

distinctivenessamongtherunsandlocations(reviewedin:Lindleyetal.,2004;Williams,132

2006).Studieshaverangedfromthosethatreportdivergenceamongallfourruntypes133

(Banksetal.2000),tostudiesthatfindevidenceforandagainstintrogressionbetweenthe134

runs(Banksetal.2000;Garzaetal.2008;O’Malleyetal.2013;Clementoetal.2014).These135

conflictinggeneticreportsmakeitdifficulttodesignmanagementactionstoprotect136

biocomplexityinCVChinookpopulations.Allofthesegeneticstudiesusedlimitedmarker137

sets(range:10-95markers).Studiesthattakeadvantageofadvancednext-generation138

sequencingtechniques,samplingagreaterproportionofthegenome,areneededtoclarify139

therelationshipsamongCVChinooksalmonandelucidatehowgenomicinformationcan140

assistinconservingbiocomplexityinthisspecies.141

142

Wetakearestriction-siteassociatedDNAsequencing(RAD-seq,Bairdetal.2008;Etteret143

al.2011)approachtoconductthefirstcomprehensivegenome-widegeneticstudyofallthe144

majorpopulationsofChinooksalmonfoundintheCentralValley.Ofprimaryimportanceis145

identifyingtheuniquegeneticdiversitycontainedwithinandamongtheCVChinook146

Page 7 of 51



Draft

8

salmonpopulationcomplex.Weaimedtodothisbyinvestigatingthepatternsof147

diversificationshownacrosstheCVChinooksalmongenome.Specifically,weaskthe148

followingquestions:149

i. WhatisthepopulationstructureofCVChinooksalmonandhowis150

diversitypartitionedamongmigratoryphenotypes?151

ii. AretheresignalsoffinescalepopulationstructurewithinFalland152

Springruns,despitehighlevelsofhumanmediatedgeneflow?153

iii. Canweassignindividualstotheirmigratoryphenotypes(akarun154

type)andpopulationsusinggenomicdata?155

Informationresultingfromthesequestionswillbegreatlybeneficialforidentifyingthe156

biocomplexityfoundwithinCVChinooksalmon,andforenablingresourcemanagersto157

developstrategiesforprotectinggeneticvariation.158

159

MATERIALSANDMETHODS160

161

Samplecollection162

163

WeobtainedfintissuesamplesofadultChinooksalmonfromtheCaliforniaDept.ofFish164

andWildlifeAnadromousResourcesTissueArchive.Sampleswereoriginallycollected165

duringspawningmigrationsandcomefromallmajorpopulationswithinthefourChinook166

salmonruns(Fall,Late-fall,Spring,Winter)(Table1,Figure1).Weanalyzed28-32167

individualsperpopulation.Previousworkhasshownthissamplesizetobemorethan168

adequateforcapturinggeneticvariationwithgenomicdata(Nazarenoetal.2017).169

Page 8 of 51



Draft

9

Populationswererepresentedbysamplescollectedfromtwoormoreyears,tocapture170

temporalvariation.However,previousworkhasshowntemporalvariationwithin171

populationsofChinooksalmontoberelativelysmallcomparedtovariationamong172

populations(Banksetal.2000;Beachametal.2006;Narumetal.2008).173

174

Molecularbiology175

176

WeusedthemolecularmethodsoutlinedinMeeketal.(2016).Briefly,weextracted177

genomicDNAandconstructedRADlibrariesusingtheSbfIrestrictionenzymefollowingthe178

protocolfromMilleretal.(2012).Eachsamplewasligatedwithauniquecustomsixbase179

pairbarcodeandwemultiplexed30–47individualsperlibrary.Wesequencedthelibraries180

as100basepairsingle-endreadsonanIlluminaHiSeq2000(VincentJ.CoatesGenomics181

SequencingLaboratory,Berkeley,CA),runningasinglelibraryperlane.182

183

SNPgenotyping184

185

WealignedthesequencesfromeachindividualtotheRADlociinMeeketal.(2016)using186

Bowtie(Langmeadetal.2009)andperformedgenotypingandqualityfilteringfollowing187

themethodsdetailedinMeeketal.(2016).Insummary,wetrimmedthereadsfromthe3’188

endto92bpandeliminatedthosewitha>20%probabilityofsequencingerrorbasedon189

PHREDscores.Wealsoeliminatedthosethathadoneormoreambiguousbasecalls.We190

usedthegenotypingmethodofLewetal.(2015).Aftergenotyping,weremovedindividuals191

thatweregenotypedatfewerthanthelowerconfidenceintervalofgenotypedlociper192

Page 9 of 51



Draft

10

individual(<6,621loci).Wethenremovedlocithatweregenotypedatfewerthan70%of193

theremainingindividuals,followedbyremovingindividualsthatweregenotypedatfewer194

than70%ofremainingloci.195

196

AlignmenttoChinooklinkagemap197

WealignedourRADlocitotheintegratedChinooklinkagemap,usingtheconsensusfemale198

map(McKinneyetal.2016).Weusedblast+(Camachoetal.2009)withdefaultsettingsto199

makethealignments.Wefurtherfilteredtheblast+outputinR(RDevelopmentCoreTeam200

2005),onlyretainingalignmentswherethelengthwas≥78basepairs,1orfewer201

mismatches,percentidentitywas≥95,andtherewerenogaps.202

203

PopulationStructureandGeneticDiversityAnalyses204

205

Weanalyzedthedatasetforpopulationstructureusingtwomethods.First,weusedthe206

programSTRUCTURE(Pritchardetal.2000)toidentifygeneticallydistinctpopulationsin207

ourdataset.Weranfiveiterationsofeachmodelwith3-8clusters,withaburn-inof20,000208

stepsfollowedby750,000steps.WethenemployedtheprogramsCLUMPP(Jakobssonand209

Rosenberg2007),ingreedymode,andStructureHarvester(EarlandVonHoldt2011)to210

averageoverreplicates.WevisualizedourplotsusingcustomcodeinR(RDevelopment211

CoreTeam2005),availableuponrequest.212

213

Tofurtherexplorepopulationstructure,weconductedadiscriminantanalysisofprincipal214

components(DAPC),asimplementedintheRpackageadegenet(JombartandAhmed215

Page 10 of 51



Draft

11

2011).ThismethodallowsanalysisofgroupclusteringsimilartoBayesianmethods,such216

asSTRUCTURE(Pritchardetal.2000),withouttheassumptionsofHardy-Weinbergand217

linkageequilibria.DAPChasbeenshowntobebetteratfindingfinescalestructurethanthe218

programSTRUCTURE(Jombartetal.2010;Benestanetal.2015).WerantheDAPConceto219

optimizethenumberofprincipalcomponents(PCs)thatwereretained,testingupto25220

clustersandretaining250PCs.Inthisrun,thealphascoreshowedthatretaining11-14PCs221

providestheoptimalnumberwithoutoverfittingthedata.InthefinalDAPCanalysis,we222

testeduptotenclusters,retainingalldiscriminantfunctionsandtheoptimalnumberof223

principalcomponentsforeachnumberofclusters(K).ResultsoftheDAPCanalysis224

identifiedseveralindividualsthatappearedtobemislabeled,misidentifiedinthefield,or225

werestrays,astheyhadhighprobabilitiesofclusteringwithadifferentrun.Weremoved226

anyindividualwhosemembershipprobabilitytoanalternatepopulationwasgreaterthan227

0.85(Individualsremoved:1F_BUT,2F_MKH,2L_USR,and1S_DER--seelocation228

abbreviationsinTable1).WethencalculatedWeirandCockerham’sunbiasedestimatorof229

FST(WeirandCockerham1984)asimplementedinGENODIVE(MeirmansandVan230

Tienderen2004),using999permutationsanddeterminingsignificancewithaBenjamini231

andYekutieli(2001)FalseDiscoveryRate-correctedvalue(p<0.015),asperNarum(2006).232

Wealsocalculatedallelicrichness,observedandexpectedheterozygosity(Nei1987),and233

inbreedingcoefficientsusingGENODIVE.Weconductedananalysisofmolecularvariance234

(AMOVA),usingtheinfiniteallelesmodeland1000permutations,toinvestigatethe235

partitioningofgeneticvariationacrosspopulationsandrunsinGENODIVE.Wecalculated236

effectivepopulationsize(Ne)withNeEstimator(Doetal.2014)usingthefollowing237

parameters:thelinkagedisequilibriummodelwithrandommating,amiminumallele238

Page 11 of 51



Draft

12

frequencycutoffof0.02,andcalculatingtheparametric95%confidenceintervalforeach239

estimate.Weremovedlocifromthedatasetthatwerefoundtobepotentialoutliers(see240

methodsbelow)andonlyincludedthoseSNPsthatcouldbelocatedonthelinkagemap.We241

incorporatedchromosomeinformationinthemodelsolinkagedisequilibriumcalculations242

wereonlymadeinpairwisecomparisonsbetweenlociondifferentchromosomes(the“LD243

locuspairingacrosschromosomes”optioninNeEstimator).244

245

WedeterminedpresenceofisolationbydistancebytestingforcorrelationbetweenFST246

(transformedtoFST/(1-FST))andriverdistanceusingapartialManteltestinRusingthe247

veganpackage(Oksanenetal.2017),accountingforruntypeinadissimilaritymatrixand248

usingruntypeasthestrata.Wecalculatedriverdistancebymeasuringtheriverdistance249

betweenrivermouthsinGoogleEarth.Wealsotestedforcorrelationsbetweenriver250

distanceandFSTwithinFallandSpringrunsusingaManteltest.251

252

AlignmenttotheGREB1Llocus253

WealignedourRADlocitotheGREB1Lscaffoldshowntobeassociatedwithpremature254

andmatureruntiminginChinookandsteelheadsalmoninotherpartsoftherange(Prince255

et.al2017;Thompsonetal.2018).Weusedblast+(Camachoetal.2009)usingthedefault256

settingstomakethealignmentsandthenlookedforalignmentsintheregionsidentifiedby257

Princeet.al(2017)andThompsonetal.(2019).Usingthesamemethods,wealsoaligned258

ourlocitotheGREB1LtoROCK1regionidentifiedbyNarumetal.(2018)asbeingrelated259

tomigrationphenotypes.260

261

Page 12 of 51



Draft

13

AssignmentTesting262

WeevaluatedtheabilityoftheSNPdatasettoassignindividualstothedifferentlyidentified263

groups.Wefirstmadeadatasetofonlylociwithlessthan10%missingdata.Wethen264

conductedleave-one-outtestsinGENODIVE(settings:0.005inplaceofzerofrequencies,265

likelihoodratioteststatistic,alphalevelof0.001,and500permutations)toassessthe266

dataset’sabilitytocorrectlyassignindividualstotheirpopulationoforigin.Wetested267

severalconfigurationsofuniquepopulations:1.Eachtributaryorhatcheryasaunique268

population,2.AllFallrunindividualsincludedinonepopulation,and3.CombiningSpring269

runfromMillandDeerCreeksinonepopulation,butleavingtheotherSpringrun270

populationsasindividualpopulations.Thisallowedustoevaluatewhattributariesand271

hatcheriescouldbeuniquelyidentifiedwiththisdataset.272

273

IdentifyingCandidateLociUnderSelection274

FallrunChinooksalmonistheonlyrunthatstilloccursinboththeSacramentoRiverbasin275

andtheSanJoaquinRiverbasin.PreviousworkhasshownthatFallrunaregenetically276

indistinguishableacrosstheirrangeintheCentralValley(Banksetal.2000;Williamson277

andMay2005).WewantedtoexplorethisfindingfurtherusingoursetofdiscoveredSNPs.278

ToseeifthereishiddengeneticdifferentiationamongFallrunpopulations,weconducted279

anoutliertesttoidentifylocithatshowsignalsofbeingunderselection.Ourgoalwasto280

comparethepopulationstructurefoundusingtheentireSNPdatasetwiththepopulation281

structurefoundamongFallrunusingjusttheFallrunoutlierSNPs.282

283

Page 13 of 51



Draft

14

Toidentifyoutliers,weemployedtheFDIST2approachimplementedinArlequinv3.5284

(Excoffieretal.2009;ExcoffierandLischer2010).Afterrunning20,000coalescent285

simulations,locithatwereoutsidethe99%quantilewereidentifiedasbeingcandidatesfor286

selection.Wethencreatedadatasetofjustthelocithatwereidentifiedascandidatesfor287

positiveselection.Weevaluatedpopulationclusteringbasedonthesetoflociunder288

positiveselectionusingthesameDAPCmethodsasdescribedabove.Wealignedthe289

outlierstotheSalmosalar(Davidsonetal.2010;ICSASG_v2:Genebankaccession290

GCA_000233375.4)andO.mykissgenomes(GenBankAssemblyAccession291

#GCA_002163495.1),usingBowtie2(LangmeadandSalzberg2012)anddefaultsettings,to292

investigatefunctionalannotationsoftheoutliers.TheS.salargenomeprovidedthemost293

annotations,soweproceededwiththisgenomeforannotatingoutliersfortheSpringand294

Fallruns.Weusedthefollowingalignmentcriteriaforannotation:alignmentlength>75295

basepairs,percentidentity>90%,e-value>0.0001.296

297

Tofurtherinvestigatethevalidityofouroutlierresultsandthepotentialforfalsepositives,298

wecreatedanulldatasetofFallrunindividuals,randomlyreorganizingindividualsinto299

populations.Werandomlyassignedindividuals(usingtherandomnumbergenerator300

functionrunifinR)tothesamenumberofpopulationswiththesamenumberof301

individualsperpopulationastherealdataset.Wethenconductedthesameanalysesfor302

detectingoutliersasdescribedabove.Giventhatthereisnobiologicalmeaningbehindour303

randomlycreatedpopulations,weexpectedthatthereshouldbenooutliersdetected.If304

outliersweredetected,itwouldindicatethattheoutlierdetectionmethodispronetoa305

highnumberoffalsepositivesandlesscertaintycanbeappliedtotheresults.306

Page 14 of 51



Draft

15

307

WeconductedthesameoutlieranalysesontheSpringrunpopulations,toseeifthereis308

anysignalofselectiondrivingdifferentiationamongthedifferentSpringrunpopulations,309

despiteallextantpopulationsbeinglocatedintheSacramentoRiverbasin.310

311

RESULTS312

313

SequencingandSNPgenotyping314

Weobtainedanaverageof248,595,544readspersequencinglibrary(range:136,417,117-315

460,215,599).Thenumberofuniquereadsperindividualrangedfrom15,822-316

24,137,958,withanaverageof2,985,730perindividual.Themeannumberofgenotyped317

lociforeachindividualwas18,041(CI:6,621–22,969).Weremovedsixindividualsthat318

weretypedatfewerthanthelowerconfidenceinterval(<6,621loci).Weusedthefinalset319

of11,783SNPloci,asdescribedinMeeketal.(2016)forgenotyping.ThisSNPdataset320

excludedlociwithaminorallelefrequencyof<0.01andobservedheterozygosity>0.55in321

ordertoremovepotentialparalogs.Table1showsthenumberofindividualsper322

populationthatremainedafteralsofilteringoutindividualsthatwerenotgenotypedinat323

least70%ofthe11,783SNPloci(12-30individualsperpopulation).Themeanreaddepth324

perlocuswas22.4(95%CI:22.3–22.6).Wealigned6,666ofourlocitotheconsensus325

Chinooksalmonlinkagemap.326

327

GeneticDiversityandPopulationStructure328

Page 15 of 51



Draft

16

Meanexpectedheterozygocity(He)acrossallpopulationswas0.24(S.D.:0.08),withWinter329

runhavingthelowestHeof0.19andallothergroupsbeingbetween0.24-0.25(Table1).330

Theinbreedingcoefficientforallpopulationswasslightlynegative(range:-0.072---0.03),331

withall95%confidenceintervalsbelow0.332

333

TheAMOVAanalysisshowedthehighestamountofbetweengroupgeneticvariationis334

partitionedamongruns,followedbyamongpopulationsnestedwithinruns(Table2).335

STRUCTUREanalysesshowedthatpopulationstructuringwith6clusters(K=6)hadboth336

thehighestlikelihoodandthehighestdeltaK(FigureS1,TableS1,Evanno,Regnaut,&337

Goudet,2005),butthelikelihoodsplateauedbetweenK=3andK=7.The6-clustermodel338

groupedindividualsbyruntype,withsomesubstructurewithinSpringrunpopulations339

(FigureS1).SpringruninButteCreekclustereduniquely,whereasMillandDeerCreek340

clusteredtogether.SpringrunfromtheFeatherRiverHatcheryshowedadmixturebetween341

auniquegeneticlineageandtheFallruncluster.AllofFallrundisplayedprimary342

membershipinthesameuniquecluster.FallrunfromtheFeatherRiverHatcheryshowed343

someadmixturewiththeuniqueFeatherRiverHatcherySpringruncluster.Thereisalsoa344

signatureofauniqueLate-fallrunlineage,howeveritisadmixedwiththeFallruncluster.345

TheDAPCanalysisshowedsimilarresults,howevereachrunandSpringrunpopulation346

showedgreaterdistinctionandlessindividualadmixturethanintheSTRUCTUREanalysis347

(Figure2andFiguresS2-S3).IntheDAPCanalysisforK=6,Late-fallclustereduniquely,as348

didtheFeatherRiverHatcherySpringrun.Additionally,intheK=7model,Fallruninthe349

ColemanHatcherydisplayeduniqueclustermembership.350

351

Page 16 of 51



Draft

17

Wefoundsignificantdifferentiationwithinandamongthedifferentruns,basedonFst352

values(range:0-0.161,Table3).ThehighestvalueswerebetweenWinterrunandallother353

populations(range:0.135-0.161).WithinSpring-run,Fstvaluesrangedfrom0.004-0.033,354

withallvaluesbeingsignificant.FstwithinFallrunrangedfrom0-0.005.Fallrunfromthe355

ColemanFishHatcherywassignificantlydifferentfromallotherFallrunpopulations,as356

wasFallrunfromtheFeatherRiverHatchery.FallrunfromMillCreekwasalso357

significantlydifferentfromtheNimbusHatcherypopulationandtheStanislausandMerced358

Rivers.359

360

Usingadatasetof4689locithatwereidentifiedneutralmarkersandcouldbeplacedon361

theChinooklinkagemap,wecalculatedNeforthesixgroupingsidentifiedintheDAPC362

analysis.TheconfidenceintervalfortheNeestimateincludedinfinityforallgroups,with363

theexceptionofSpringrunfromMill/DeerCreeksandWinterrun.TheNeestimatefor364

SpringrunfromMill/DeerCreekswas591.1(CI:558.3-628.0)andforWinterrunwas365

376.4(CI:352.6-403.5).EstimatesofinfinityforNearelikelyduetosamplingerror366

outweighingtheeffectsofgeneticdrift(i.e.samplesizetoosmalltocapturetheeffectsof367

geneticdriftinlargepopulations,seeWaplesandDo(2010)forfurtherdiscussionofthis368

phenomenon).369

370

Therewassignificantisolationbydistancewhenallrunswereusedinthemodelandrunis371

accountedforinthepartialManteltest(P=0.007).Isolationbydistancewasnotsignificant372

whenlimitedtoonlySpringrun(P=0.125),butwashighlysignificantwithinFallrun373

populations(P=0.007,FigureS4).374

Page 17 of 51



Draft

18

375

AlignmenttoGREB1Lloci376

WefoundoneSNPinourfinaldatasetthatalignedwiththeruntimingassociatedGREB1L377

locus.Thislocus(SNP#R008612)alignedtoposition595079-595161ofthescaffold378

79929edescribedinPrinceet.al(2017).Thegenotypesatthislocuswerenearlyperfectly379

associatedwithpre-mature(SpringandWinterrun)andmature(FallandLate-fall)run380

timings(Figure3).Interestingly,allSpringrunindividualswerehomozygousforoneallele,381

withtheexceptionofsixindividualsfromtheFeatherRiverHatchery(representing7.6%382

oftheSpringrun),whichwereheterozygous.WithinFallrun,90%(170individuals)were383

homozygousforthealternateallele,with8.5%(16individuals)beingheterozygous,and384

1.6%(3individuals)beinghomozygousfortheSpring/Winterassociatedallele.Twoofthe385

individualsthatwerehomozygousfortheSpring/WinterallelewerefromtheFeather386

RiverHatcheryandonewasfromMillCreek.Wefound87SNPlocithatalignedtothe203387

KbregioncontainingtheGREB1LandROCK1genesidentifiedbyNarumetal.(2018)on388

Ots28between11.022and11.225Mb.Wecalculatedtheallelefrequencydifferencesfor389

eachrunatthese87locitoevaluatethepatternsamongthedifferentruns(Figure4).The390

samelocusidentifiedbyaligningtothePrinceetal.(2017)scaffold(SNP#R008612)391

alignedtothisregionandclearlyshowsdifferencesamongthepre-matureandmaturerun392

timings.Nootherlocusshowedasclearofapattern.393

394

Assignmenttesting395

Assignmenttestingshowedhighprobabilityofcorrectassignmentusingthedatasetofloci396

withlessthan10%missingdata(7829loci).WhenFallrunpopulationsweregrouped397

Page 18 of 51



Draft

19

togetherinonegeneral“FallRun”reportinggroup,assignmentprobabilitieswerehighfor398

Fallrun(100%),butalsoFallrundrewalotofnon-Fallrunassignments(e.g.individuals399

fromLate-fallandSpringrun).Thevastmajority(96%)oftheLate-fallrunassignedtothe400

FallrunpopulationusingthisdatasetandassignmentofSpringrunpopulationsranged401

from40-100%(TableS2).However,whenweusedeachindividualFall-runpopulationasa402

uniquereportinggroup,assignmentforindividualFallrunpopulationsrangedwildly(0-403

100%),however,95%oftheassignmentsforFallrunindividualswereassignedtoaFall404

runpopulation.Notably,FallrunfromtheMercedRiver,ColemanHatchery,andNimbus405

HatcherydrewalmostallofassignmentsfromotherFallrundrainageswhenusing406

individualreportinggroups(TableS3).Basedonthesedata,wecannotdistinguish407

individualswhohadgeneticmis-assignmentstothewrongnatalstreamfromthosethat408

wereassignedcorrectlybutreturnedtonon-natalstreamstospawn(“strayed”),making409

themappeartobemis-assigned.Notably,weachieved100%accuracyofassignmentfor410

Late-fallwhenusingindividualFallrunreportinggroups.Assignmentaccuracywas100%411

whenwecombinedtheSpringrunfromMillandDeerCreeksintoasinglegroup(Figure5).412

ThesedatashowthatusingthefullSNPdatasetenableshighassignmentaccuracyof413

populations,withhighaccuracyofFallrunwhenusingageneralFallrunreportinggroup.414

415

LociUnderSelection416

OutlieranalysesperformedwithinFallrunfound829locithatweresignificantforshowing417

signaturesofdiversifyingselectionusingArlequin.654oftheoutliersalignedtothe418

Chinooksalmonlinkagemap,withthelocispreadthroughoutthegenome(between10-38419

outliersalignedtoeachchromosome)(Figure6).Twenty-threelocimappedtolinkage420

Page 19 of 51



Draft

20

groupsbutwerenotpositionedonachromosome.Forty-fiveoutliershadfunctional421

annotationsassociatedwiththem(TableS4).Wefound0outliersintherandomized422

dataset,providingreassurancethatourresultsfromanalysisoftherealdatasetare423

meaningful.424

425

TheDAPCmodelusingjusttheselociandretaining70principlecomponentsfoundthat426

K=1-3hadthelowestBIC,withK=4and5slightlyhigherbutwithclusteringthatshows427

biologicalrelevance(Figure7,FigureS5).AtK=2,auniqueclusterwasformedbyseveral428

individualsfromMill,Deer,andButteCreeks,theTuolumneandMercedRiverHatchery,429

andoneindividualfromtheMercedRiverpopulation.WhentheSpringandWinterrun430

populationsarealsoincluded,itbecomesclearthatthisuniqueclusterisformedwith431

SpringrunfromtheFeatherRiverHatchery(datanotshown).Therefore,itisverylikely432

thatthisclusterisdrivenbyintrogressionwithstraysfromtheSpringrunintheFeather433

RiverHatchery,ratherthanbeingauniqueFallrunlineage.TheFallrunfromColeman434

HatcheryshoweduniqueclustermembershipstartingatK=3.AtK=4,twoadditional435

groupsemergedwithDeer,Butte,andMillCreeksclusteringtogetheralongwiththe436

MercedRiverandMercedRiverHatcherypopulations,andtheNimbusandMokelumne437

HatcheriesandStanislausandTuolumneRiversclusteringtogether.AtK=5,theMerced438

Riverpopulationwasresolvedasitsownuniquecluster,withsomeindividualsinButte439

Creekdrawingmembershipfromthiscluster.Theseclusteringpatternsshowageneral440

geographicpatternofclustering,withmanyofthewildSacramentoRiverbasin441

populationsclusteringtogetherandmanyoftheSanJoaquinRiverbasinpopulationsin442

anothercluster.443

Page 20 of 51



Draft

21

444

TheoutlieranalyseswithinSpringrunfound940locithatweresignificantforsignaturesof445

diversifyingselection.776oftheoutliersalignedtotheChinooksalmonlinkagemap,with446

between12-46outliersalignedtoeachchromosome.Twenty-ninelocimappedtolinkage447

groupsbutwerenotpositionedonachromosome.Fifty-sevenoutlierlocihadannotations448

associatedwiththem(TableS4).Nooutliersweredetectedwhenanalyzingthe449

randomizeddataset.450

451

TheDAPCmodelusingjusttheselocifoundK=3-4hadthelowestBIC(Figure8,FigureS6).452

AtK=3,DeerandMillCreeksgroupedtogether,whileatK=4,eachtributarypopulation453

groupedseparately.101lociwereidentifiedasbeingpotentialoutliersinboththeFalland454

Springrunanalyses.455

456

Fstvaluesweremuchhigheramongpopulationswhenusingtheoutlierdataset,andall457

pairwisecomparisonsweresignificant(Fallrunmean=0.035andrange:0.008–0.061,458

Springrunmean=0.102andrange:0.036–0.166;TablesS5andS6).459

460

DISCUSSION461

462

Chinooksalmonexhibitsomeofthehighestgeneticandlifehistorytraitdiversityofthe463

Pacificsalmonids(Waples2001).Inparticular,thepresenceoffourdistinctspawningruns464

intheCentralValleyofCaliforniarepresentsthegreatestlifehistorytraitdiversity465

observedforChinooksalmon,andpresumablycorrespondswiththehighestbiocomplexity466

Page 21 of 51



Draft

22

andportfolioforthisspecies.Inreality,CVChinooksalmondisplaytheweakestportfolioin467

thespeciesrange(Griffithsetal.2014);however,ourresultsindicatethatthereisgreater468

geneticbiocomplexitythanpreviouslydescribed.Protectionofthisremaining469

biocomplexitymaybeimportantforsustainablemanagementoftheCVChinooksalmon470

stockcomplexandrestorationoftheportfolio.471

472

Thisstudyisthefirstgenomicinvestigationintothepopulationstructureandgenomic473

diversityofCentralValleyChinooksalmon.Bygenotypinghundredsofindividualsat474

thousandsofSNPsdistributedacrossthegenome,wefoundevidenceforgreater475

populationstructuringbeyondthelevelofrun,includingfine-scalepopulationstructurefor476

Springrunfishandapatternofisolation-by-distanceamongFallrunpopulations.Wealso477

findgeneticdistinctivenessofFallandLate-fallruns,despiteconclusionsfromprevious478

scientificstudyandcurrentmanagementasasingleESU(Lindleyetal.2004).Ourresults479

providemanagerswithinformationnecessarytonotonlyconservethediversityofrunsin480

theCentralValley,butalsotomaintainandenhancelifehistorydiversitywithinrunsfor481

maximalbufferingfromenvironmentalvariationandoverallportfolioperformance482

(Satterthwaiteetal.2017).483

484

WeshowthatSpringruncontainsfinerscalesub-structure,withevidenceforthreedistinct485

populations:1)MillandDeerCreek,2)ButteCreek,and3)theSpringrunintheFeather486

RiverHatchery.Thislatterfindingisquiteremarkable,aspreviousworkhasshownthat487

SpringrunintheFeatherRiverHatcheryisgeneticallyindistinguishablefromFallrun488

(Lindleyetal.2004;Garzaetal.2008).Pasthatcherypracticesincludeinadequate489

Page 22 of 51



Draft

23

separationofthebroodstockforFallandSpringrun,leadingtointrogressionbetweenthe490

tworuns,andtheprevailingwisdomhasbeenthattheserunsaregenetically491

indistinguishable(CaliforniaHatcheryScientificReviewGroup2012).O’Malleyetal.(2007;492

2013)foundnodifferentiationbetweenFallandSpringrunintheFeatherRiverHatchery493

basedonnineneutralmicrosatelliteloci,buttheydidfindcleardistinctionatthree494

circadianclockgenesthoughttocontrolruntiming.Ourresultsshowthatifthefull495

genomeissurveyed,additionaldistinctionsbetweentheFeatherRiverHatcheryFalland496

Springrunareobserved.WedoseeevidenceofpastintrogressionattheGREB1L497

associatedlocus,withtheSpringrunfromtheFeatherRiverHatcherybeingtheonlySpring498

runpopulationsthatcontainedheterozygousindividuals.Theseresultsshowthatsomeof499

thegeneticdiversityfoundintheFeatherRiverHatcherySpringrunisuniqueandhasnot500

beeneliminatedfromintrogressionwithFallrun.Additionally,theseresultssupportthe501

useofrevisedhatcherypracticesthataimtobetterseparatethetworunsinthehatchery502

(Baerwaldetal.2011).503

504

Ourresultsdemonstratethatwiththefull~12KSNPdataset,notonlycanwedistinguish505

SpringrunintheFeatherRiverHatcheryfromotherSpringrunpopulationsandfromFall506

run,wecanalsodistinguishtheLate-fallfromFallrunpopulations.Ourdemonstrated507

abilitytodistinguishindividualsfromtheserunsisasignificantleapforwardinourability508

tomonitorandmanagethesepopulationsseparately,enablingprotectionofbiocomplexity509

inthesystem.Itisvitalthatallofthegeneticdiversitypresentinthissystemisprotectedin510

ordertomaintainahealthyportfoliooflifehistoriesandgeneticdiversity(Schindleretal.511

2010;CarlsonandSatterthwaite2011).512

Page 23 of 51



Draft

24

513

Recently,ahigh-throughputCentralValleyspecific80-SNPpanelwasdevelopedthatgoesa514

longwaytowardimprovingourabilitytodistinguishthedifferentruns(Meeketal.2016).515

ThebenefitofthispanelisitisavailableasFluidigmSNPtypeassays(FluidigmCorp.San516

Francisco,CA)allowingveryfastturnaroundfromtissuesampletogenotypeandrun517

assignment.Thisfastturnaroundisrequiredforsomemanagementquestions(e.g.real-518

timetakedeterminations).Thetrade-off,however,isthisSNPpanelisunabletodistinguish519

theSpringrunintheFeatherRiverHatcheryfromtheFallrunpopulationsandithaslower520

assignmentaccuracies.Weshowthat,withalmost12,000SNPs,wecandistinguishthe521

differentpopulationswithhighaccuracy.Therefore,whenrunassignmentisnotneededin522

real-time,genotypingindividualsviaasequencebasedapproachwillallowfinerscaleand523

moreaccuratemanagementofthedifferentpopulations(MeekandLarson2019).Our524

workalsoshowsthatbylookingattheGREB1Lassociatedlocus,wecaneasilydistinguish525

theESAlisted(SpringandWinterrun)fromtheunlisted(FallandLate-fall)individuals.526

FutureworkwillinvolveexploringthisGREB1Llocusfurtheranddevelopingtheseloci527

intoRapturebaits(Alietal.2016),toallowveryhighthroughputandcosteffective528

genotypingviasequencingofthousandsofsamples.529

530

OurworkalsoshowsthattheWinterrunisthemostgeneticallydistinctpopulationinthe531

CentralValley,butitalsohasthelowestgeneticdiversity(e.g.observedandexpected532

heterozygosity,SeeTable1).WeshowthattheWinterrunhasaloweffectivepopulation533

size(NE=376)andmaybeexperiencinghighlevelsofgeneticdriftassociatedwith534

decreasedpopulationsizes.Frankhametal.(2014)suggestthatanNEbelow500istoolow535

Page 24 of 51



Draft

25

toretainevolutionarypotentialinperpetuity,andthatatargetofNE=1000forwild536

populationsismoreappropriate.Followingthisguideline,SpringrunfromtheMill/Deer537

Creekcomplexarealsoatriskofreducedlong-termviabilitygiventheirNE=591.These538

resultsarenotsurprising,especiallygiventheEndangeredandThreatenedstatusof539

WinterandSpringrun,butithighlightstheimportanceofmanagementeffortsaimedat540

protectingtheportfolioofdiversitypresentintheseruns,andCentralValleyChinookasa541

whole.542

543

Dispersalamongpopulationsisoneofthemainmechanismsforincreasedsynchrony544

amongpopulationsandweakeningoftheportfolioeffect(Liebholdetal.2004).Human545

activities,suchasartificialproduction,canincreasedispersalamongpopulations,546

decreasinglocallyadaptedlifehistoryvariationandincreasingsynchrony(McClureetal.547

2008;Mooreetal.2010).ItisestimatedthatstrayinginhatcheryrearedChinookcanbe548

above70%(CaliforniaHSRG2012;Palmer-Zwahlen,&Kormos,2015).Thisisadrastic549

increasecomparedtotheoftenlessthan5%strayingthatoccursinwildsalmonid550

populations(Quinn2005).Itisthoughtthatthisincreasedstrayingduetohatchery551

productionandthelackofmigrationpathimprintingfromtruckinghasledtothegenetic552

homogenizationanddecreasedportfolioeffectamongFallrunintheCV(Huberand553

Carlson2015;SatterthwaiteandCarlson2015).Infact,DedrickandBaskett(2018)found554

thatgenetichomogenizationintheCentralValleyFallrunplaysalargerroleindecreasing555

theportfolioeffectthandemographicsynchronization.Theroleoftruckingandincreased556

strayingcouldbeplayingalargeroleinthegenetichomogenizationweseeinFallrun.Our557

workhere,however,showsthatdespiteincreasedstraying,therearestillsomeremaining558

Page 25 of 51



Draft

26

geneticdifferencesamongpopulations.Thisisevidentinthestructuringweseewiththe559

putativelociunderselection.Thisworkalsodemonstratestheimportanceofaltering560

hatcheryandmanagementpracticesnowtoprotectextantdiversity,whileitstillexistsin561

thesystem.Maintenanceofbiocomplexityinasystemcancounteracthomogenizingforces562

andpromotepopulationasynchrony(Mooreetal.2014).563

564

Ourworkshowshowinsightsaboutpopulationdifferentiationcanbegainedbyevaluating565

genomiclocipresumablyunderselection.Whenweevaluatedpopulationstructureusing566

onlylocithatarecandidatesforselection,weobservedevenfinerscalestructuringofboth567

theSpringandFallrunthanwasobservedusingthefulldataset.AlloftheSpringrun568

populationsshowdifferentiation,anddifferencesamonggroupsofFallrunpopulationsare569

apparent.TheFallrunpopulationintheColemanHatcheryisveryclearlydistinctfromthe570

restofFallrunbasedontheoutlierdataset,suggestingthatsomeaspectofColeman571

Hatcherypracticesmaybedrivingdifferentiation.Additionally,theFallrunfromthe572

NimbusHatchery,MokolumneHatchery,StanislausRiver,andTuolumneRiveralso573

separateintoonegroupthatisdifferentiatedfromtherestoftheFallrunpopulations.Itis574

possiblethatfishinthesetributariesarefacingsimilarselectionpressuresdrivingthis575

pattern.AtK=5,wealsostarttoseetheFallrunspawningintheMercedRivershow576

distinctionfromtherestofFallrun.Interestingly,theFallrunspawningintheMerced577

RiverHatcheryshowverymixedclustermembership,withrepresentationfromother578

clustersfoundintheCV.Itispossiblethatthereisuniqueselectionpressureexperienced579

byfishspawningintheriverhabitatversusinthehatchery,whichdrivesthisdifference580

(Vasemägietal.2016),orthattheMercedRiverHatcheryreceivesalotofstraysfrom581

Page 26 of 51



Draft

27

othertributaries.Indeed,a2012studyofhatcheryreturnsfoundonly12.7%ofreturnsto582

theMercedRiverHatcherywerefromMercedRiverHatcheryraisedfish(Palmer-Zwahlen583

andKormos2015).584

585

Thereissomedebateregardingtheuseofadaptivelociinconservationplanning.Allendorf586

etal.(2010)cautionagainstfocusingonadaptivelociforconservation,statingthatloci587

identifiedasbeingadaptivemightnotbethoselocithatarecrucialforfutureadaptation.588

Additionally,theauthorswarnthatfocusingondetectableadaptivedifferencesmayresult589

inoverlooking,andpotentiallylosing,importantgeneticdiversityinotherregionsofthe590

genome.Funketal.(2012),however,highlighttheimportanceofusinginformationfrom591

bothneutralandnon-neutralgeneticdatatoidentifyanappropriatemanagement592

approachforspecies.WeagreewithbothFunketal.andAllendorfetal.andbelievethere593

isvalueinincorporatingbothneutralandnon-neutralgeneticinformationinconservation594

strategies.Webelieveboththesecautionshighlighttheneedtoconserveasmuchgenetic595

diversityasisfeasible,becauseitisverydifficulttopredictwhatdiversitywillbe596

“important”bothnowandintothefuture.Webelievethistobeparticularlytruefor597

migratoryspecies,asmigratoryspeciesaresubjecttoselectiveforcesacrossthemigratory598

pathway,itcanbeverydifficulttodecipherwhatisdrivingselectivedifferences,and599

selectiveforceswillcertainlychangeovertime(Dingle2014).Forexample,thedifferences600

weseeinselectedlociamongCVChinooksalmonpopulationsmaybeduetoselection601

differencesamongthespawninggrounds,seasonaldifferencesamongmigrationlife602

histories,ordifferencesinselectiveforcesinseparateregionsoftheocean.Additionally,603

ourstudyisnotanexhaustivestudyofallpossibleformsofbiocomplexityinthissystem.604

Page 27 of 51



Draft

28

Forexample,weknowthatthereisalsodiversityintimingofjuvenileout-migrationthatis605

certainlyimportantforbiocomplexity.Ourstudyhighlightstheimportanceofidentifying606

andprotectingasmuchdiversityaspossiblesothereisafullportfolioofgeneticdiversity607

presentinthesystem.608

609

CONCLUSION610

Collectively,theseresultshavestrongimplicationsforthemanagementofCVChinook611

salmonandprovideimportantlessonsforusinggenomicstoidentifybiocomplexityand612

applyresultinginferencestotheconservationofmigratoryspecies.Genomicsarean613

increasinglyimportanttoolformanagingspecies.Theyprovidetheabilitytoidentify614

biologicaldiversitypresentinasystem,andtoassignindividualstopopulationoforigin615

despitespatiotemporalmixing,suchasonmigratorypathways.Itisvitalthatweimprove616

ourabilitytoidentifyandprotectbiocomplexityinmanagementandconservationsowe617

canimprovespeciesandmetapopulationstabilityandresilience.618

619

ACKNOWLEDGEMENTS620

WewouldliketothankDr.MortenLimborgandDr.WesLarsonforhelpfuldiscussionsthat621

improvedthemanuscriptandDr.NadyaMamoozadehandanonymousreviewersfor622

suggestionsandedits.WethankDr.MikeMillerandDr.DanPrinceforsharingtheirdata623

andinsightsontheGREB1Llocus.WealsothankLeaKoerberandtheCaliforniaDept.of624

FishandWildlifeTissueArchiveforsupplyingChinooktissuesamples.Alltissueswere625

heldunderauthorizationofESASection10(a)(1)(A)permit(Pacificfish)#15926,a626

MemorandumofUnderstandingwithCaliforniaDepartmentofFishandWildlife,and627

Page 28 of 51



Draft

29

CaliforniaDepartmentofFishandWildlifeScientificCollectingPermits#8871,3561,8862,628

11381,and1751.ThisworkwasmadepossiblebyfundingfromtheCaliforniaDept.ofFish629

andWildlife,contractnumberP0740017,andMHMwassupportedbytheDavidH.Smith630

ConservationResearchFellowshipProgram.ThisworkusedtheVincentJ.CoatesGenomics631

SequencingLaboratoryatUCBerkeley,supportedbyNIHS10InstrumentationGrants632

S10RR029668andS10RR027303.633

634

DataArchivingStatement:635

Dataforthisstudyareavailableat:tobecompletedaftermanuscriptisacceptedfor636

publication.637

638

LITERATURECITED639

640

Ali,O.A.,O’Rourke,S.M.,Amish,S.J.,Meek,M.H.,Luikart,G.,Jeffres,C.,andMiller,M.R..641

2016.Radcapture(Rapture):Flexibleandefficientsequence-basedgenotyping.642

Genetics202:389–400.643

Allendorf,F.W.,Hohenlohe,P.A.,andLuikart,G.2010.Genomicsandthefutureof644

conservationgenetics.Nat.Rev.Genet.11:697–709.NaturePublishingGroup.645

Angeloni,F.,Wagemaker,N.,Vergeer,P.,andOuborg,J.2012.Genomictoolboxesfor646

conservationbiologists.Evol.Appl.5,130–143.647

Araki,H.,Cooper,B.,andBlouin,M.S.2007.Geneticeffectsofcaptivebreedingcausea648

rapid,cumulativefitnessdeclineinthewild.Science318:100–3.649

Baerwald,M.R.,Stephens,M.R.,Bork,K.S.,Meek,M.H.,Tomalty,K.,andMay,B.P.2011.650

Page 29 of 51



Draft

30

Spring-runChinooksalmongeneticmanagementplan--SanJoaquinRiverRestoration651

Program.124p.652

Baird,N.,Etter,P.D.,Atwood,T.S.,Currey,M.C.,Shiver,A.L.,Lewis,Z.A.,Selker,E.U.,653

Cresko,W.A.,andJohnson,E.A.2008.RapidSNPdiscoveryandgeneticmappingusing654

sequencedRADmarkers.PLoSOne3:e3376.655

Banks,M.A.,Rashbrook,V.K.,Calavetta,M.J.,Dean,C.,andHedgecock,D.2000.Analysisof656

microsatelliteDNAresolvesgeneticstructureanddiversityofchinooksalmon657

(Oncorhynchustshawytscha)inCalifornia’sCentralValley.Can.J.Fish.Aquat.Sci.658

57:915–927.659

Beacham,T.,Jonsen,K.,Supernault,J.,Wetklo,M.,Deng,L.,andVarnavskaya,N.2006.660

PacificrimpopulationstructureofChinooksalmonasdeterminedfrommicrosatellite661

Analysis.Trans.Am.Fish.Soc.135:1604–1621.662

Benestan,L.,Gosselin,T.,Perrier,C.,Sainte-Marie,B.,Rochette,R.,andBernatchez,L.2015.663

RAD-genotypingrevealsfine-scalegeneticstructuringandprovidespowerful664

populationassignmentinawidelydistributedmarinespecies;theAmericanlobster(665

Homarusamericanus).Mol.Ecol.,doi:10.1111/mec.13245.666

Benjamini,Y.andYekutieli,D.2001.Thecontroloffalsediscoveryrateunderdependency.667

Ann.Stat.29:1165–1188.668

CaliforniaHSRG(CaliforniaHatcheryScientificReviewGroup).2012.Californiahatchery669

reviewreport.PreparedfortheUSFishandWildlifeServiceandPacificStatesMarine670

FisheriesCommission.671

Camacho,C.,Coulouris,G.,Avagyan,V.,Ma,N.,Papadopoulos,J.,Bealer,K.,andMadden,T.L.672

2009.BLAST+:architectureandapplications.BMCBioinformatics10:421.673

Page 30 of 51



Draft

31

Carlson,S.M.andSatterthwaite,W.H.2011.Weakenedportfolioeffectinacollapsed674

salmonpopulationcomplex.Can.J.Fish.Aquat.Sci.68:1579–1589.675

Clemento,A.J.,Crandall,E.D.,Garza,J.C.,andAnderson,E.C.2014.Evaluationofasingle676

nucleotidepolymorphismbaselineforgeneticstockidentificationofChinookSalmon677

(Oncorhynchustshawytscha)intheCaliforniaCurrentlargemarineecosystem.Fish.678

Bull.112:112–130.679

Davidson,W.S.,Koop,B.F.,Jones,S.J.M.,Iturra,P.,Vidal,R.,Maass,A.,Jonassen,I.,Lien,S.,680

andOmholt,S.W.2010.SequencingthegenomeoftheAtlanticsalmon(Salmosalar).681

GenomeBiol.11:403.682

Dedrick,A.G.andBaskett,M.L.2018.Integratinggeneticanddemographiceffectsof683

connectivityonpopulationstability:thecaseofhatcherytruckinginsalmon.Am.Nat.684

192:E62-E80.685

Dingle,H.2014.Migration:thebiologyoflifeonthemove.SecondEdition.Oxford686

UniversityPress,NewYork,NY.687

Earl,D.A.,andVonHoldt,B.M.2011.STRUCTUREHARVESTER:awebsiteandprogramfor688

visualizingSTRUCTUREoutputandimplementingtheEvannomethod.Conserv.Genet.689

Resour.2–4.690

Etter,P.D.,Bassham,S.,Hohenlohe,P.A.,Johnson,E.,andCresko,W.A.2011.SNP691

discoveryandgenotypingforevolutionarygeneticsusingRADsequencing.Pp.157–692

178inV.OrgogozoandM.Rockman,eds.MolecularMethodsforEvolutionary693

Genetics.HumanaPress,NewYork,NY.694

Evanno,G.,Regnaut,S.,andGoudet,J.2005.Detectingthenumberofclustersofindividuals695

usingthesoftwareSTRUCTURE:asimulationstudy.Mol.Ecol.14:2611–20.696

Page 31 of 51



Draft

32

Excoffier,L.,Hofer,T.,andFoll,M.2009.Detectinglociunderselectioninahierarchically697

structuredpopulation.Heredity.103:285–98.698

Excoffier,L.andLischer,H.2010.Arlequinsuitever3.5:Anewseriesofprogramsto699

performpopulationgeneticsanalysesunderLinuxandWindows.Mol.Ecol.Resour.700

10:564–567.701

FederalRegister.2005.EndangeredandThreatenedSpecies:FinalListingDeterminations702

for16ESUsofWestCoastSalmon,andFinal4(d)ProtectiveRegulationsfor703

ThreatenedSalmonidESUs.70:37160–37204.704

FederalRegister.1999.Endangeredandthreatenedspecies:Threatenedstatusfortwo705

Chinooksalmonevolutionarilysignificantunits(ESUs)inCalifornia.64:50394–50415.706

Funk,W.C.,McKay,J.K.,Hohenlohe,P.A.,andAllendorf,F.W.2012.Harnessinggenomics707

fordelineatingconservationunits.TrendsEcol.Evol.27:489–496.708

Garza,J.C.,Blankenship,S.M.,Lemaire,C.,andCharrier,G.2008.Geneticpopulation709

structureofChinooksalmon(Oncorhynchustshawytscha)inCalifornia’sCentralValley.710

FinalreportforCalFedProject“Comprehensiveevaluationofpopulationstructureand711

diversityforCentralValleyChinooksalmon.”712

Griffiths,J.R.,Schindler,D.E.,Armstrong,J.B.,Scheuerell,M.D.,Whited,D.C.,Clark,R.A.,713

Hilborn,R.,Holt,C.A.,Lindley,S.T.,Stanford,J.A.,andVolk,E.C.2014.Performanceof714

salmonfisheryportfoliosacrosswesternNorthAmerica.J.Appl.Ecol.51:1554–1563.715

Harvey,B.N.,Jacobson,D.P.,andBanks,M.A.2014.Quantifyingtheuncertaintyofa716

juvenileChinooksalmonraceidentificationmethodforamixed-racestock.NorthAm.717

J.Fish.Manag.34:1177–1186.718

Hess,M.A.,Rabe,C.D.,Vogel,J.L.,Stephenson,J.J.,Nelson,D.D.,andNarum,S.R.2012.719

Page 32 of 51



Draft

33

Supportivebreedingboostsnaturalpopulationabundancewithminimalnegative720

impactsonfitnessofawildpopulationofChinooksalmon.Mol.Ecol.21:5236–5250.721

Hilborn,R.,Quinn,T.P.,Schindler,D.E.,andRogers,D.E.2003.Biocomplexityandfisheries722

sustainability.Proc.Natl.Acad.Sci.U.S.A.100:6564–8.723

Huber,E.R.andCarlson,S.M.2015.TemporaltrendsinfallChinookhatcheryreleases724

fromCalifornia’sCentralValley.SanFr.EstuaryWatershedSci.13:1–23.725

Jakobsson,M.andRosenberg,N.A.2007.CLUMPP:aclustermatchingandpermutation726

programfordealingwithlabelswitchingandmultimodalityinanalysisofpopulation727

structure.Bioinformatics23:1801–6.728

Jombart,T.andAhmed,I.2011.adegenet1.3-1 :newtoolsfortheanalysisofgenome-wide729

SNPdata.Bioinformatics27:3070–3071.730

Jombart,T.,Devillard,S.,andBalloux,F.2010.Discriminantanalysisofprincipal731

components:anewmethodfortheanalysisofgeneticallystructuredpopulations.BMC732

Genet.11:94.BioMedCentralLtd.733

Larson,W.A.,Seeb,L.W.,Everett,M.V.,Waples,R.K.,Templin,W.D.,andSeeb,J.E.2014.734

Genotypingbysequencingresolvesshallowpopulationstructuretoinform735

conservationofChinooksalmon(Oncorhynchustshawytscha).Evol.Appl.7:355–69.736

Lew,R.M.,Finger,A.J.,Baerwald,M.R.,Goodbla,A.,May,B.,andMeek,M.H.2015.Using737

next-generationsequencingtoassistaconservationhatchery:asingle-nucleotide738

polymorphismpanelforthegeneticmanagementofendangereddeltasmelt.Trans.739

Am.Fish.Soc.144:767–779.740

Liebhold,A.,Koenig,W.D.,andBjørnstad,O.N.2004.Spatialsynchronyinpopulation741

dynamics.Annu.Rev.Ecol.Evol.Syst.35:467–490.742

Page 33 of 51



Draft

34

Lindley,S.T.,Grimes,C.B.,Mohr,M.S.,Peterson,W.,Stein,J.,Anderson,J.T.,Botsford,L.W.,743

Bottom,D.L.,Busack,C.A.,Collier,T.K.,Ferguson,J.,Garza,J.C.,Grover,A.M.,Hankin,744

D.G.,Kope,R.G.,Lawson,P.W.,Low,A.,MacFarlane,R.B.,Moore,K.,Palmer-Zwahlen,745

M.,Schwing,F.B.,Smith,J.,Tracy,C.,Webb,R.,Wells,B.K.,andWilliams,T.H.2009.746

WhatcausedtheSacramentoRiverfallchinookstockcollapse?NOAATech.Memo.747

NMFS-SWFSC447:1-61.748

Lindley,S.T.,Schick,R.,May,B.P.,Anderson,J.J.,Greene,S.,Hanson,C.,Low,A.,McEwan,749

D.,MacFarlane,R.B.,Swanson,C.,andWilliams,J.G.2004.Populationstructureof750

threatenedandendangeredChinooksalmonESUsinCalifornia’sCentralValleybasin.751

NOAATech.Memo.NOAA-TM-NM:70p.752

McClure,M.M.,Utter,F.M.,Baldwin,C.,Carmichael,R.W.,Hassemer,P.F.,Howell,P.J.,753

Spruell,P.,Cooney,T.D.,Schaller,H.A.,andPetrosky,C.E.2008.Evolutionaryeffects754

ofalternativeartificialpropagationprograms:implicationsforviabilityofendangered755

anadromoussalmonids.Evol.Appl.1:356–375.756

McKinney,G.J.,Seeb,L.W.,Larson,W.A.,Gomez-Uchida,D.,Limborg,M.T.,Brieuc,M.S.O.,757

Everett,M.V.,Naish,K.A.,Waples,R.K.,andSeeb,J.E.2016.Anintegratedlinkagemap758

revealscandidategenesunderlyingadaptivevariationinChinooksalmon759

(Oncorhynchustshawytscha).Mol.Ecol.Resour.16:769–783.760

Meek,M.H.,Baerwald,M.R.,Stephens,M.R.,Goodbla,A.,Miller,M.R.,Tomalty,K.M.H.,and761

May,B.2016.Sequencingimprovesourabilitytostudythreatenedmigratoryspecies:762

GeneticpopulationassignmentinCalifornia’sCentralValleyChinooksalmon.Ecol.763

Evol.6:7706–7716.764

Meirmans,P.G.andVanTienderen,P.H.2004.GENOTYPEandGENODIVE:twoprograms765

Page 34 of 51



Draft

35

fortheanalysisofgeneticdiversityofasexualorganisms.Mol.Ecol.Notes4:792–794.766

Miller,M.R.,Brunelli,J.P.,Wheeler,P.A.,Liu,S.,Rexroad,C.E.,Palti,Y.,Doe,C.Q.,and767

Thorgaard,G.H.2012.Aconservedhaplotypecontrolsparalleladaptationin768

geographicallydistantsalmonidpopulations.Mol.Ecol.21:237–49.769

Moore,J.W.,McClure,M.M.,Rogers,L.A.,andSchindler,D.E.2010.Synchronizationand770

portfolioperformanceofthreatenedsalmon.Conserv.Lett.3:340–348.771

Moore,J.W.,Yeakel,J.D.,Peard,D.,Lough,J.,andBeere,M.2014.Life-historydiversityand772

itsimportancetopopulationstabilityandpersistenceofamigratoryfish:steelheadin773

twolargeNorthAmericanwatersheds.J.Anim.Ecol.1035–1046.774

Moyle,P.B.,Lusardi,R.A.,Samuel,P.J.,andKatz,J.V.E.2017.Stateofthesalmonids :status775

ofCalifornia’semblematicfishes.CenterforWatershedSciences,Universityof776

California,DavisandCaliforniaTrout,SanFrancisco,CA.777

Myers,J.M.,Kope,G.H.,Bryant,G.J.,Teel,D.,Lierheimer,L.J.,Wainwright,T.C.,Grant,W.778

S.,Waknitz,F.W.,Neely,K.,Lindley,S.T.,andWaples,R.S.1998.Statusreviewof779

ChinooksalmonfromWashington,Idaho,OregonandCalifornia.U.S.Deptartment780

Commer.Commer.NOAATech.MemmorandumNMFS-NWFSC-35443.781

Narum,S.R.2006.BeyondBonferroni:Lessconservativeanalysesforconservation782

genetics.Conserv.Genet.7:783–787.783

Narum,S.R.,Banks,M.A.,Beacham,T.D.,Bellinger,M.R.,Campbell,M.R.,Dekoning,J.,Elz,784

A.,Guthrieiii,C.M.,Kozfkay,C.,Miller,K.M.,Moran,P.,Phillips,R.,Seeb,L.W.,Smith,C.785

T.,Warheit,K.I.,Young,S.F.,andGarza,J.C.2008.Differentiatingsalmonpopulations786

atbroadandfinegeographicalscaleswithmicrosatellitesandsinglenucleotide787

polymorphisms.Mol.Ecol.17:3464–3477.788

Page 35 of 51



Draft

36

Narum,S.R.,DiGenova,A.,Micheletti,S.J.,andMaass,A.2018.Genomicvariation789

underlyingcomplexlife-historytraitsrevealedbygenomesequencinginChinook790

salmon.Proc.R.Soc.BBiol.Sci.285:20180935.791

Nazareno,A.G.,Bemmels,J.B.,Dick,C.W.,andLohmann,L.G.2017Minimumsamplesizes792

forpopulationgenomics:AnempiricalstudyfromanAmazonianplantspecies.Mol.793

Ecol.Resour.17:1136-1147.794

O’Malley,K.G.,Jacobson,D.P.,Kurth,R.,Dill,A.J.,andBanks,M.A.2013.Adaptivegenetic795

markersdiscriminatemigratoryrunsofChinooksalmon(Oncorhynchustshawytscha)796

amidcontinuedgeneflow.Evol.Appl.6:1184–94.797

Oksanen,J.,Blanchet,F.,Friendly,M.,Kindt,R.,Legendre,P.,McGlinn,D.,Minchin,P.,798

O’Hara,R.,Simpson,G.,Solymos,P.,Stevens,M.,Szoecs,E.,andWagner,H.2017.vegan:799

CommunityEcologyPackage.Rpackageversion2.4-4.https://CRAN.R-800

project.org/package=vegan.801

Ouborg,N.J.,Pertoldi,C.,Loeschcke,V.,Bijlsma,R.,andHedrick,P.W.2010Conservation802

geneticsintransitiontoconservationgenomics.TrendsGenet.26,177–87.803

Palmer-Zwahlen,M.andKormos,B.2015.Recoveryofcoded-wiretagsfromChinook804

salmoninCalifornia’sCentralValleyescapement,inlandharvest,andoceanharvestin805

2012.CalifornidaDepartmentofFisheriesAndWildlifeFisheriesAdministrative806

Report2015-4.66p.807

Pritchard,J.K.,Stephens,M.,andDonnelly,P.2000.Inferenceofpopulationstructureusing808

multilocusgenotypedata.Genetics155:945–59.809

Quinn,T.P.2005.ThebehaviorandecologyofPacificsalmonandtrout.Universityof810

WashingtonPress,Seattle,WA.811

Page 36 of 51



Draft

37

RDevelopmentCoreTeam.2005.R:Alanguageandenvironmentforstatisticalcomputing,812

referenceindexversion2.2.1.RFoundationforStatisticalComputing,Vienna,Austria.813

ISBN3-900051-07-0,URLhttp://www.R-project.org.814

Reich,D.,Thangaraj,K.,Patterson,N.,Price,A.,andSingh,L.2009.ReconstructingIndian815

populationhistory.Nature461:489–494.816

Satterthwaite,W.H.andCarlson,S.M.2015.Weakeningportfolioeffectstrengthina817

hatchery-supplementedChinooksalmonpopulationcomplex.Can.J.Fish.Aquat.Sci.818

72:1860–1875.819

Satterthwaite,W.H.,Carlson,S.M.,andCriss,A.2017.Oceansizeandcorrespondinglife820

historydiversityamongthefourruntimingsofCaliforniacentralvalleychinook821

salmon.Trans.Am.Fish.Soc.146:594–610.822

Schindler,D.E.,Hilborn,R.,Chasco,B.,Boatright,C.P.,Quinn,T.P.,Rogers,L.A.,and823

Webster,M.S.2010.Populationdiversityandtheportfolioeffectinanexploited824

species.Nature465:609–12.825

Sommer,T.R.,Nobriga,M.L.,Harrell,W.C.,Batham,W.,andKimmerer,W.J.2001a.826

California’sYoloBypass:Evidencethatfloodcontrolcanbecompatiblewithfisheries,827

wetlands,wildlife,andagriculture.Can.J.Fish.Aquat.Sci.58:325–333.828

Sommer,T.R.,Nobriga,M.L.,Harrell,W.C.,Batham,W.,andKimmerer,W.J.2001b.829

Floodplainrearingofjuvenilechinooksalmon:evidenceofenhancedgrowthand830

survival.Can.J.Fish.Aquat.Sci.58:325–333.831

Vasemägi,A.,Kahar,S.,andOzerov,M.Y.2016.GenesthataffectAtlanticsalmongrowthin832

hatcherydonothavethesameeffectinthewild.Funct.Ecol.30:1687–1695.833

Waples,R.S.2001.CharacterizingdiversityinsalmonfromthePacificNorthwest.J.Fish834

Page 37 of 51



Draft

38

Biol.59:1–41.835

Weir,B.S.andCockerham,C.C.1984.EstimatingF-Statisticsfortheanalysisofpopulation836

structure.Evolution.38:1358–1370.837

Williams,J.G.2006.CentralValleysalmon:AperspectiveonChinookandSteelheadinthe838

CentralValleyofCalifornia.SanFr.EstuaryWatershedSci.4:416p.839

Williams,J.G.2012.JuvenileChinooksalmon(Oncorhynchustshawytscha)inandaround840

theSanFranciscoEstuary.SanFr.EstuaryWatershedSci.10:24p.841

Williamson,K.S.andMay,B.P.2005.HomogenizationofFall-runChinooksalmongene842

poolsintheCentralValleyofCalifornia,USA.NorthAm.J.Fish.Manag.25:993–1009.843

Yoshiyama,R.M.,Fisher,F.W.,andMoyle,P.B.1998.Historicalabundanceanddeclineof844

ChinooksalmonintheCentralValleyregionofCalifornia.NorthAm.J.Fish.Manag.845

18:487–521.846

847

Page 38 of 51



Draft

39

TABLES848

849

Table1:Sampleinformation.SamplesizecolumngivestotalnumberusedinRAD-850

sequencing/Totalnumberusedinanalysesafterfiltering.He=expectedheterozygosity,851

Ho=observedheterozygosity,Ar=averagenumberofalleles,Fis=inbreedingcoefficient852

(95%confidenceinterval)853

854

Location Years Sampled

Location Abbreviation

Sample Size

He Ho Ar Fis

Fall Run

Butte Cr. 2002-2004 F_BUT 32/16 0.246 0.259 1.844 -0.054 (-0.063 - -0.044)

Coleman Hatchery (Battle Creek)

2002 F_COL 30/30 0.245 0.261 1.915 -0.064 (-0.072 - -0.056)

Deer Cr. 2002-2004 F_DER 30/13 0.245 0.259 1.826 -0.056 (-0.065 - -0.046)

Feather R. Fish Hatchery

2001, 2007-2010

F_FRH 28/21 0.244 0.259 1.870 -0.061 (-0.069 - -0.052)

Merced R. 2001, 2004, 2008

F_MER 34/27 0.245 0.262 1.910 -0.072 (-0.08 - -0.063)

Mill Cr. 2001-2004 F_MIL 35/14 0.249 0.264 1.847 -0.061 (-0.07 - -0.051)

Mokelumne R. Fish Hatchery

2004-2005, 2008

F_MKH 30/18 0.244 0.258 1.854 -0.058 (-0.068 - -0.049)

Merced R. Fish Hatchery

2001-2004 F_MRH 30/21 0.245 0.258 1.877 -0.05 (-0.059 - -0.041)

Nimbus Fish Hatchery

(American River)

2002-2005 F_NIM 30/25 0.243 0.259 1.886 -0.068 (-0.076 - -0.06)

Stanislaus R. 2001, 2002, 2004

F_STN 30/12 0.243 0.259 1.791 -0.067 (-0.077 - -0.058)

Tuolumne R. 2004, 2008 F_TOU 33/18 0.246 0.260 1.862 -0.061 (-0.07 - -0.052)

Late Fall Run

Upper Sacramento R.

2003-2005 L_USR 30/24 0.245 0.26 1.881 -0.062 (-0.071 - -0.054)

Spring Run

Butte Cr. 2008-2009 S_BUT 30/22 0.241 0.250 1.827 -0.034

(-0.043 - -0.026)

Page 39 of 51



Draft

40

Deer Cr. 2002, 2005 S_DER 31/20 0.246 0.254 1.864 -0.031 (-0.039 - -0.022)

Mill Cr. 2002, 2004, 2005

S_MIL 32/18 0.247 0.261 1.858 -0.056 (-0.065 - -0.048)

Feather R. Fish Hatchery

2009-2010 S_FRH 30/30 0.247 0.263 1.910 -0.068 (-0.076 - -0.06)

Winter Run

Upper Sacramento

R. 2001, 2002 W_USR 30/30 0.194 0.199 1.717 -0.03

(-0.04 - -0.02) 855

Table2:ResultsofAMOVAanalysis856

Source of Variation Nested in %variance F-stat P-value Within Individual -- 1.006 F_it -- Among Individual Population -0.051 F_is 1 Among Population Run 0.005 F_sc 0.001 Among Run -- 0.039 F_ct 0.002 857

858

859

860

861

862

863

864

865

866

867

868

869

Page 40 of 51



Draft

41

Table3:Fstvaluesamongpopulationsusingfull,filtereddataset.Valuesbelowthediagonal870

areFst.Valuesabovethediagonalarethesignificancevalue.Bolded,italicizedFstvalues871

aresignificantatp<0.015.872

873

F_COL F_MIL F_DER F_BUT F_FRH F_NIM F_MKH F_STN F_TOU F_MER F_MRH L_USR S_BUT S_DER S_MIL S_FRH W_USR

F_COL -- 0.002 0.009 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001

F_MIL 0.002 -- 0.441 0.309 0.013 0.002 0.039 0.007 0.038 0.002 0.22 0.001 0.001 0.001 0.001 0.001 0.001

F_DER 0.002 0 -- 0.864 0.01 0.022 0.112 0.037 0.033 0.024 0.106 0.001 0.001 0.001 0.001 0.001 0.001

F_BUT 0.003 0 -0.001 -- 0.299 0.306 0.589 0.977 0.528 0.534 0.489 0.001 0.001 0.001 0.001 0.001 0.001

F_FRH 0.003 0.001 0.001 0 -- 0.001 0.001 0.001 0.003 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001

F_NIM 0.004 0.002 0.001 0 0.002 -- 0.713 0.558 0.231 0.234 0.029 0.001 0.001 0.001 0.001 0.001 0.001

F_MKH 0.004 0.001 0.001 0 0.002 0 -- 0.524 0.37 0.442 0.038 0.001 0.001 0.001 0.001 0.001 0.001

F_STN 0.004 0.002 0.002 -0.002 0.003 0 0 -- 0.703 0.593 0.276 0.001 0.001 0.001 0.001 0.001 0.001

F_TOU 0.004 0.001 0.001 0 0.002 0 0 0 -- 0.231 0.16 0.001 0.001 0.001 0.001 0.001 0.001

F_MER 0.004 0.002 0.001 0 0.002 0 0 0 0 -- 0.037 0.001 0.001 0.001 0.001 0.001 0.001

F_MRH 0.005 0.001 0.001 0 0.003 0.001 0.001 0.001 0.001 0.001 -- 0.001 0.001 0.001 0.001 0.001 0.001

L_USR 0.009 0.007 0.007 0.006 0.009 0.008 0.007 0.007 0.008 0.007 0.008 -- 0.001 0.001 0.001 0.001 0.001

S_BUT 0.037 0.035 0.038 0.038 0.035 0.041 0.04 0.041 0.04 0.04 0.041 0.042 -- 0.001 0.001 0.001 0.001

S_DER 0.02 0.019 0.02 0.021 0.018 0.024 0.023 0.024 0.023 0.023 0.024 0.025 0.017 -- 0.002 0.001 0.001

S_MIL 0.021 0.019 0.021 0.021 0.019 0.024 0.024 0.024 0.024 0.023 0.024 0.026 0.023 0.004 -- 0.001 0.001

S_FRH 0.01 0.007 0.008 0.009 0.005 0.012 0.011 0.012 0.01 0.01 0.011 0.015 0.033 0.019 0.019 -- 0.001

W_USR 0.14 0.148 0.152 0.15 0.145 0.152 0.153 0.161 0.154 0.15 0.154 0.151 0.151 0.14 0.146 0.135 --

874

875

876

877

878

879

880

Page 41 of 51



Draft

42

FIGURELEGENDS:

Figure1:CentralValleyChinooksalmonsamplinglocations.Thepointsrepresenttheriverssamples

werecollectedfrom,nottheexactsamplinglocation.FiguremodifiedfromMeeketal.(2016).

Figure2:DiscriminantAnalysisofPrincipleComponents(DAPC)clusteringresults,testingdifferent

numberofgroups(K=3-8)usingthefullSNPdataset.Eachbarrepresentsanindividualandthex-axis

showssampledlocation(locationabbreviationsdefinedinTable1).Thecolordistinguishesclustersand

they-axisquantifiesclustermembership.

Figure3:ProportionofindividualsshowingeachgenotypeattheGREB1LassociatedSNP,byruntiming.

Figure4:Heatmapofallelefrequenciesfor87SNPsalignedtotheGREB1LtoROCK1regionofOmy28

identifiedinNarumetal.(2018).Eachrowisoneofthe87SNPs.ThearrowpointstoSNPR008612

showninFigure3thatalignstotheGREB1LregionidentifiedinPrinceetal.(2017).Thecolorscalebar

denotesthefrequencyofthefirstalleleateachlocus.

Figure5:Assignmentaccuraciesusing7892SNPswith<10%missingdata.Pointsarescaledtothe

assignmentvalue.Thevalueontopofeachpointisthepercentassigningtoeachgroup,thevalueunder

eachpointisthenumberofindividualsthatwereassignedtoeachgroup.

Figure6:ManhattanplotofoutliersofSNPsmappedtotheChinooklinkagemap(McKinneyetal.2016).

EachpointrepresentsaSNP.X-axisshowsthepositionoftheSNPonthelinkagemap.Alternating

chromosomesarecoloredinblackorgrey.ThetoppanelhastheFallrunoutlierlocihighlightedinred,

thebottompanelhastheSpringrunoutlierlocihighlightedinred.Colorfigurefoundonline.

Page 42 of 51



Draft

43

Figure7:DiscriminantAnalysisofPrincipleComponents(DAPC)clusteringresultsforFallrun

populationsusing852candidateoutlierlocifordiversifyingselection.Eachbarrepresentsanindividual

andthex-axisshowssampledlocation(locationabbreviationsdefinedinTable1).Thecolor

distinguishesclustersandthey-axisquantifiesclustermembership.

Figure8:DiscriminantAnalysisofPrincipleComponents(DAPC)clusteringresultsforSpringrun

populationsusing940candidateoutlierlocifordiversifyingselection.Eachbarrepresentsanindividual

andthex-axisshowssampledlocation(locationabbreviationsdefinedinTable1).Thecolor

distinguishesclustersandthey-axisquantifiesclustermembership.

Page 43 of 51



Draft

44

Figure1:

H

Page 44 of 51



Draft

45

Figure2:

K=3

Mem

bers

hip

prob

abili

ty

F-CO

L

F-M

IL

F-DE

R

F-N

IM

F-M

KH

F-BU

T

F-FR

H

F-M

ER

F-M

RH

F-TO

U

L-U

SR

S-BU

T

S-DE

R

S-FR

H

W-U

SR

S-M

IL

POPULATIONS

F-ST

N

K=3

K=4

K=5

K=6

K=7

K=8

Page 45 of 51



Draft

46

Figure3:

Fall Late-fall Spring Winter

Homozygous (T/T)Heterozygous (C/T)Homozygous (C/C)

SNP R008612

Run

Proportion

0.0

0.2

0.4

0.6

0.8

1.0

Page 46 of 51



Draft

47

Figure4:

Fall Late-fall Spring WinterRUNS

LOCI

Page 47 of 51



Draft

48

Figure5:

Fall

Late_fall

Sp_Butte

Sp_FRH

Sp_MillDeer

Winter

Fall Late_fall Sp_Butte Sp_FRH Sp_MillDeer WinterActual

Assigned

Value0.00

0.05

0.50

0.75

0.90

1.00

94.9%

100%

100%

96.7%

100%

100%

0.05%

4.7%

3.3%

204

1

10

24

38

22

1

29

30

Page 48 of 51



Draft

49

Figure6:

Chromosome

Fst

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34

0.0

0.1

0.2

0.3

0.4

0.5

0.6

Chromosome

Fst

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34

0.0

0.1

0.2

0.3

0.4

0.5

0.6

Page 49 of 51



Draft

50

Figure7:

F-CO

L

F-MIL

F-DE

R

F-NIM

F-MKH

F-BU

T

F-FRH

F-MER

F-MRH

F-TO

U

F-STN

POPULATIONS

K=2

K=3

K=4

K=5

Mem

bershipprob

ability

Page 50 of 51



Draft

51

Figure8:

S-MIL

S-DE

R

S-BU

T

S-FRH

POPULATIONS

Mem

bershipprob

ability

K=4

K=3

K=5

K=6

Page 51 of 51



Draft - University of Toronto T-Space · 71 runs of Chinook salmon, each named for the timing of...

Documents

Transcript of Draft - University of Toronto T-Space · 71 runs of Chinook salmon, each named for the timing of...