Bushland Weeds Book

8/9/2019 Bushland Weeds Book

1/116

Bushland

WeedsA practical guide to their management

With case studies from the Swan Coastal Plain and beyond

Kate Brown and Kris Brooks

Ill ustrated by Libby Sandi ford


2/116

Published by and available from Environmental Weeds Action Network (Inc), PO Box 380 Greenwood, 6924, Australia.www.iinet.net.au\ ~ewan

©Individual artists for photographs and drawings©Environmental Weeds Action Network (Inc)

Parts of this publication may be reproduced for educational purposes with appropriate acknowledgment. A copy of anyreport or publication using these materials should be lodged with the Environmental Weeds Action Network (Inc).

National Library of Australia Cataloguing-in-publication entry

ISBN 0 9579001 1 2.

Disclaimer Any recommendations in this book do not necessarily represent the policy or support of the sponsoringorganisations. This book is intended as a source of information only. People should obtain independent advice before actingon information in this publication. It is the user’s responsibility to ensure that any methodology adopted or adapted from thisbook is suitable for the purposes intended. Read the label of herbicides for further information and registration status.Consult the NRA website www.nra.gov.au to determine the status of permits for your situation or state. The publisher or theauthors do not accept liability however arising, including liability for negligence, or for any loss resulting from the use of orreliance upon information in this book.

Design, typesetting and print management by West Print Management, Mosman Park, Perth, Western Australia

Cover photograph by Kate Brown – Late spring in the shrublands at Brixton Street Wetlands with Pink Feather Flower(Verticordia densiflor a ), Kunzea (Kunzea micrantha ) and Purple Flag (Patersonia aff. occidentalis ) in full flower.

Cover illustrations by Libby Sandiford – Some of the weeds starting to invade the Wetlands: From left to right; HarlequinFlower (Sparaxi s bulbifera ), Black Flag (Ferra ria crispa ), African Lovegrass (Eragrostis curvu la ) and Perennial Veldgrass(Ehrharta calycina ).


3/116

Acknowledgments

This manual is the culmination of a five year project that has been supported by the Environmental Weeds ActionNetwork. The project would not have been possible without the ongoing support of project manager, Margo

O’Byrne, and the project team Bob Dixon, Rod Randall and the extraordinary financial manager, Gary Matthews. LizWestern generously stood in as manager for a year, supervising the project while Margo was away. Peter Nash andPatrick Piggott also contributed to project management. The Swan Catchment Centre provided office space andadministrative support for three years. Sally Madden and Simone Tuten provided technical support in the earlystages of the project.

Production of the manual has been a collective effort. Kate Brown wrote the chapters on the project, managingweeds in bushland, the geophyte weeds, the other herbaceous weeds and with Kris Brooks, the mappingguidelines. Kris Brooks wrote the grasses, the woody weeds and the herbicides in bushland chapters. The mapswere compiled by Kris Brooks and Kate Brown.

Richard Groves, John Moore, Margo O’Byrne and Rod Randall provided valuable comments on various drafts, BobDixon on the woody weeds chapter and Bronwen Keighery the managing weeds in bushland chapter. Theircomments greatly improved the final manuscript. Neil Gibson, Richard Groves and Janice Marshall proof read andprovided comments on the final document.

Throughout the project Greg Keighery, Bob Dixon, John Moore and John Peirce have generously shared theirknowledge of bushland weeds with us. They also provided much of the information in the weed management tables.

Community volunteers and Friends groups carried out much of the weed management work that forms the basis of case studies outlined in the manual. They also helped us map weeds, set up trials and collect data. We would like tothank The Friends of Shenton Bushland, in particular Janice Marshall and Dani Boase-Jelinek, Friends of Blue GumLake especially Marian Watson, Friends of Brixton Street Wetlands particularly Regina Drummond, TrevorDrummond, Nick Buters and Elizabeth Buters, Friends of Gingin Brook in particular Richard Diggins and PaulineDiggins, and Friends of Talbot Road Reserve, especially Mark Gloyn and Helen Gardiner.

John Carter and Robyn Phillimore from Department of Conservation and Land Management, Grant McKinnon fromthe City of Swan, Wayne van Lieven from the City of Gosnells, Graznya Packowskya from the City of Melville andSteve McCabe from the City of Nedlands all went out of their way to provide invaluable on-ground support atbushlands under their management.

The project was largely funded through a Natural Heritage Trust Grant, with the Lotteries Commission and the CRC

for Weed Management also contributing.

Kate Brown and Kris Brooks

September 2002.

All illustrations are by Libby Sandiford and all photographs by Kate Brown, unless otherwise noted.

The Environmental Weeds Action Network (EWAN) is a community group formed in 1996 to:

• Promote understanding of the threat of environmental weeds to bushland.

• Provide information about weed control in native vegetation.

• Convince governments at all levels of the need for appropriate legislation and funding for weed control.

• Encourage research into methods of weed control.

• Encourage community participation.

i


4/116

ii


5/116

Contents

iii

Introduction ....................................................................................................................................................................iv

Chapter 1. The Project......................................................................................................................................................1

Funding and administration..................................................................................................................................................1

The sites ..................................................................................................................................................................................1

Chapter 2. Managing Weeds in Bushland: Some General Principles............................................................................5

Gathering area - specific information ..................................................................................................................................5

Regional information..............................................................................................................................................................6

Developing a weed management program..........................................................................................................................6

Implementation – taking an integrated approach..............................................................................................................6

Chapter 3. Grass Weeds....................................................................................................................................................9

Impacts ....................................................................................................................................................................................9

Biology – why are grasses successful weeds?..................................................................................................................11

Management and Control ....................................................................................................................................................14

Monitoring and follow-up ....................................................................................................................................................18

Case Studies ..........................................................................................................................................................................19Weed management table – grasses....................................................................................................................................28

Chapter 4. Corms, Bulbs and Tubers: The Weeds that Die Back to Fleshy Underground Storage Organs ............32

The underground storage organs; life-cycles and reproduction....................................................................................33


Key points..............................................................................................................................................................................42

Case studies ..........................................................................................................................................................................44

Weed management table – corms bulbs and tubers ......................................................................................................52

Chapter 5. Broadleaf Herbs, Sedges and Succulents ..................................................................................................56

Annual herbs ........................................................................................................................................................................56

Perennial herbs ....................................................................................................................................................................59

Case studies ..........................................................................................................................................................................60

Weed management table – annual broadleaf herbs ........................................................................................................66Weed management table – perennial broadleaf herbs....................................................................................................70

Chapter 6. Trees, Shrubs and Climbers: The Woody Weeds ......................................................................................74

Impacts ..................................................................................................................................................................................75

Mechanisms of spread ........................................................................................................................................................75

Resprouting, suckering and apical control ......................................................................................................................75


Case studies ..........................................................................................................................................................................84

Weed management table – trees, shrubs and climbers ..................................................................................................86

Chapter 7. Weed Mapping in Remnant Bushland ........................................................................................................90

Equipment required ............................................................................................................................................................90

What weeds to map?............................................................................................................................................................90

Mapping in practice ............................................................................................................................................................91Using weed maps..................................................................................................................................................................91

In conclusion ........................................................................................................................................................................93

Chapter 8. Herbicide Use in Bushland..........................................................................................................................94

Using herbicides in bushland ............................................................................................................................................94

Duty of care ..........................................................................................................................................................................94

Personal care ........................................................................................................................................................................94

Training..................................................................................................................................................................................95

Legislation ............................................................................................................................................................................96

Bushland care ......................................................................................................................................................................96

Using contractors ................................................................................................................................................................97

Some herbicides used in bushland....................................................................................................................................97

Final note......................................................................................................................................................................100

Appendix 1: Herbicide spraying contract ..................................................................................................................101

References ....................................................................................................................................................................102


6/116

Introduction

iv

The Banksia, Tuart and Jarrah woodlands, theshrublands and the species-rich heathlands of the SwanCoastal Plain and Darling Plateau are wonderful places.

They contribute to Perth’s unique natural landscapes,provide a window into the natural world, habitat fornative fauna and support an extraordinarily diverseflora. They are fast disappearing though, mostly underurban development. Those bushlands that remain facea range of threats; one of the most serious is invasionby environmental weeds.

The concept of environmental weeds is still relativelynew. Although the weediness of plants like BridalCreeper and Watsonia have become entrenched in thepublic psyche, there is some way to go before the widercommunity accepts the extent of the problem. Forexample, Harlequin Flower (Sparax is bulbifera ) threatensone of the last remaining clay-based wetlands on the

eastern side of the coastal plain and yet the species wasavailable this year from a fundraising catalogue in atleast one local primary school. Still, it is evident thatmany people do recognise the threats. Increasingly,community volunteers and state and local governmentsare involved in on-ground actions to protect bushlandfrom the impacts of environmental weeds.

With few resources available, maximising the positiveoutcomes of these professional and voluntary efforts iscritical. For those working on the ground, informationon identification and control is available from a numberof good books: Managing Perth’s Bushlands (Scheltemaand Harris 1995), Southern Weeds and their control(Moore and Wheeler 2002), Bush Invaders of South-East

Australia (Muyt 2001), Western Weeds. A guide to theweeds of Western Australia (Hussey et al . 1997) andEnvironmental weeds: a field guide for SE Australia(Blood 2001). Importantly, this kind of informationneeds to be taken and applied in the context of particular bushland areas. There are no simpleformulas and effective management begins with anunderstanding and a knowledge of each site.

It was with this in mind that, in 1998, the EnvironmentalWeeds Action Network (EWAN), with funding from theNatural Heritage Trust, employed a project officer towork with community volunteers and local and stategovernment land managers at bushland sites acrossPerth’s Swan Coastal Plain. The underlying objective

was to help the various land managers to developstrategies for effective weed management in theirbushlands. This manual has grown out of that project.

The aim of the manual is firstly to bring togetherinformation on the biology and known control methodsfor the serious weeds of bushlands of the Swan CoastalPlain and J arrah Forest. Secondly, it is to illustrate,with examples and case studies, how this sort of information can be used to manage weeds in thecontext of particular sites. Hopefully this will providethe reader with the basic knowledge and theframework needed to begin effectively managing theweeds in their own bushlands.

Chapter one provides descriptions of bushlands fromwhere the case studies and examples throughout thetext are drawn. Chapter two covers some general

principles that should guide weed management inbushland, including the kind of area-specific informationrequired before setting weed management priorities.

The next four chapters each cover a different group orlifeform of weeds. The first three groups are all herbs –green non-woody plants: Chapter three covers thegrasses, chapter four herbs that die back to corms,bulbs or tubers over summer (geophytes) and chapterfive broadleaf herbs, sedges and succulents. The finalgroup, chapter six, includes all the woody plants and afew climbers that are technically herbs, but havecontrol strategies in common with woody climbers.

Individual chapters look at the general biology of eachweed group and how it relates to dispersal, spread,

control and management. Case studies examine controlof certain species at particular sites and discuss themanagement approach taken. They often describe theset-up of trials and monitoring programs. Each weedgroup has a corresponding table containing a summaryof information, gathered from a wide range of sources,on the biology and control of weeds occurring inbushland of the Swan Coastal Plain and Jarrah Forest.

The list is based on ‘A Checklist of the naturalisedvascular plants of Western Australia’ (Keighery 1999a)and includes species that can have an impact onbiodiversity and hamper restoration and regenerationefforts. Finally there are chapters on how to map weedpopulations in urban bushland, an integral part of weedmanagement and on the issues arising from the use of

herbicides in bushland.

Ideally, this manual will impart enough information forthe reader to devise a strategy for management of theirserious bushland weeds. At the same time the authorswant to highlight the complexities involved in workingin Perth’s bushlands; diverse and wonderful places toget to know and understand. The following chapterprovides detailed descriptions of some of thosebushlands along with the serious weeds that threatenthem and the people who are managing them.


7/116

1

Chapter 1 The Project

Over four years, project officers have been involved inwork at a series of bushlands across the Swan CoastalPlain and one in Western Australian Wheatbelt. The

major task over that time has been to work on site withthose involved in weed management and bushregeneration, providing both on-ground and technicalsupport. The idea was to select a series of sites thatwere representative of bushlands on a range of soils andland forms across the coastal plain, had regionallysignificant conservation values and were managed bypeople actively involved in on-ground works.

The underlying objective has been to work with thebushland ‘managers’, whoever they were, to help themprotect their bushland through carefully targeted andstrategic weed management. Initially work involvedidentifying those weeds that were the greatest threatsto the conservation values of each site, then mapping

their distribution across the bushland. The greatestefforts have been directed to consistent management of these locally serious weeds, taking an integratedapproach, addressing the causes of invasion andimplementing a range of control strategies. Often trialshad to be carried out to determine practical andeffective methods. Project officers regularly workedalongside Friends groups, council bushland workers,and others who carried out on-ground works.Sometimes they facilitated the employment of contractors and often worked alongside them,supervising and guiding their work. At all sitesmonitoring was put in place to measure theeffectiveness of control programs and the regenerationof native plant communities over time. For most

bushlands, field herbaria of all weeds and some nativespecies commonly mistaken for weeds were compiled.

Funding and administration

The project was funded through the Natural Heritage Trust and managed and administered by a voluntarysteering committee made up of members of theEnvironmental Weeds Action Network (EWAN): BobDixon from Kings Park and Botanic Garden, GaryMatthews community volunteer and EWAN treasurer,Margo O’Byrne from the Department of EnvironmentalProtection and Rod Randall from the WesternAustralian Department of Agriculture.

The sitesShenton Bushland

Shenton Bushland, a 21 hectare remnant of Banksia(Banksia menz iesii, B. attenuata ), Jarrah (Eucalyptus marginata ) woodlands lies on Spearwood dunesapproximately eight kilometres west of the centre of Perth. With only 18 % of the vegetation complexremaining uncleared the bushland is consideredregionally significant (Government of Western Australia2000). These species-rich woodlands occur on paleyellow to grey calcareous sands derived from Tamalalimestone. Rich in perennial herbs, Milkmaids(Burchardia congesta ), Vanilla Lily (Sowerbaea laxi flora ),

Leafy Sundew (Drosera stoloni fera ), Yellow Autumn Lily(Tricoryne elatior ) and the sedge Mesomelaena pseudostygia are common. The open sandy patches are

often colonised by annual herbs such as SlenderPodolepis (Podolepi s gracilis ) and after fire the annualgrass, Austrostipa compressa is noticeable, flowering

and seeding prolifically. Commonly occurring shrubsinclude Hairy Yellow Pea (Gompholobium tomentosum ),Daviesia nudiflora and Grass Tree (Xanthorrhoea preissii ). Around 120 species of natives (Marshallunpubl.) and 65 species of weeds (Brown and Brooksunpubl.) have been recorded. Approximately 50 % of the bushland is in good to excellent condition. Theremainder varies from good to degraded with areas of severe localised disturbance (Ecoscape 1994,Government of Western Australia 2000).

Weeds that threaten the undisturbed bushland includea number of South African geophytes (plants that dieback to bulbs, corms or tubers each year), including

Yellow Soldier (Lachenalia reflexa ), Freesia (Freesia a lba

x leichtlinii ), Watsonia (Watsoni a meriana ) and BlackFlag (Ferra ria crispa ). The perennial herbs GeraldtonCarnation Weed (Euphorbia terracina ) and RosePelargonium (Pelargoni um capitatum ) are seriousinvaders and many weedy annuals are common,particularly in the highly disturbed areas. Theseinclude Flat Weed (Hypochaeris glabra ), Ursinia (Ursinia anthemoides ), French Catchfly (Silene gallica ), andSlender Suckling Clover (Trifolium dubium ). PerennialVeldgrass (Ehrharta calycina ) is one of the most seriousinvaders of the bushland often establishing inpreviously intact areas following fire.

For the last nine years on-ground management of Shenton Bushland has been carried out by The Friends

of Shenton Bushland (Inc.), a community group formedafter the bushland was saved from development in 1993.In 2000, following lobbying from community groups, abushcare officer was employed to carry out on-groundmanagement of bushlands within the City of Nedlands,including Shenton Bushland. In the past, much of thefunding has come from government grants to theFriends. In recent times, however, the City of Nedlandshas started to fund the continuation of works programsinitiated through grant money. The Friends have a highlevel of input into management through organisation of works programs and supervision of contractors who domuch of the weed management work.

Friends of Shenton Bushla nd and vo lunteers hand-weedin g Yellow

Soldier in the Bank sia woodland.


8/116

2


Blue Gum Lake ReserveBlue Gum Lake is part of a chain of wetlands that lie on

the interface of the Bassendean and Spearwood dunesystems. Flooded Gum (Eucalyptus rudi s ) – SwampBanksia (Banksia l i t toral is ) woodlands fringe the lakewhile Banksia attenuata , B. menzi esii and B. ili cifolia woodlands occupy the drier uplands. Common in theunderstorey of the woodlands around the lake areCentella asiatica and Pale Rush (Juncus palli dus ). Thedrier Banksia woodlands of the uplands are veryspecies-rich, particularly in perennial herbs andshrubs. Phlebocar ya cili ata, Prickly Conostylis(Conostylis aculea ta ) and Purple Flag (Patersonia occidentalis ) are among the commonly occurring herbsand typical shrubs include Devils Pins (Hovea pungens ), Rose Banjine (Pimelea ro sea ) and PineappleBush (Dasypogon bromeliifoli us ). The perennial native

grass Microlaena stipoi des occurs occasionally in thebushland and after fire the annual grass, Austrostipa compressa , is very common. Around 62 species of natives (City of Melville 1992) and 80 species of weeds(Brown and Brooks unpubl. data) have been recorded.Weed invasion is a major threat with around 40 % of the bushland suffering from severe disturbance.

Serious weeds of the drier Banksia woodlands includePerennial Veldgrass and South African geophytes suchas Freesia, Yellow Soldier and Black Flag. The SydneyGolden Wattle (Acacia longifolia ) is a major weed of thefringing vegetation, as is Vasey Grass (Paspalum urvillei ) and Kikuyu (Pennisetum clandestinum ). One of the serious weed problems at the reserve has arisen

from plantings of non-local species around and throughthe bushland. Many, including Sydney Golden Wattle,Melaleuca linearii folia , River Red Gum (Eucalyptus camaldulensis ) and Geraldton Wax (Chamelaucium uncinatum ) have become naturalised in the bushland.Additional to this threat is the planting of species thatoccur naturally in the bushland but have been grownfrom seed collected a long way from the reserve. Theearly flowering shrub-form ofBanksia menziesii, fromthe sand plains 200km to the north of Perth, has been

planted in the reserve. It makes a disturbing contrastto the later flowering form, a beautiful woodland treethat grows naturally around Blue Gum Lake. These

plantings threaten the nature conservation values of the bushland as much as any South African invader.

The Friends of Blue Gum Lake have played a major rolein managing the bushland for many years. They are notan incorporated group and, accordingly, acquiringgovernment grants is difficult. The little funding that isavailable for weed management work comes from theCity of Melville. The City has two staff dedicated to on-ground bushland management activities. Their work isspread across the City’s many reserves. For herbicidespraying the City is often obliged to use ‘preferredcontractors’.

Brixton Street Wetlands

The Brixton Street Wetlands lie 20 kilometres south eastof Perth at the foot of the Darling Scarp. A small remnant(19 hectares) on the winter-wet flats of Guildfordformation clays, the wetlands support many rare andrestricted plant species as well as threatened plantcommunities. They also support an exceedingly diverseflora of 307 native taxa (Keighery and Keighery 1995).

Most of the wetland soils are waterlogged through thewinter months and deep pools form in claydepressions. When the pools are full they support anumber of native aquatic plants but as water levelsstart to drop a series of annual and perennial herbsgrow and flower. In late winter Flannel Flowers(Tribonanthes species), Blue Squill (Chamaescilla

species) and Early Nancy (Wurmbea dioica ) areflowering and by late spring Swamp Wallaby Grass(Amphi bromus nervosus ) is in abundance. By earlysummer the pools are dry and the claypans covered inSundews (Drosera species) and Trigger Plants(Stylidium species). Slightly higher in the landscape, theflats surrounding the claypans also support a diverserange of native herbs, sedges and rushes. Shrubs,including Swish Bush (Viminaria juncea ) and FeatherFlowers (Verticordia species), are also common. On thesandy rises, Marri (Eucalyptus calophyll a ) woodlandsoccur and in spring, Red and Green Kangaroo Paw(Anigozanthos manglesii ) flower throughout theirunderstorey. With this kind of habitat almost entirelycleared on the Swan Coastal Plain the area is of

outstanding conservation value. Although more than 75% is relatively undisturbed there are areas of severelocalised disturbance. Weed invasion is one of the mostserious threats with 85 species recorded (Keighery andKeighery 1995, Brown and Brooks unpubl.).

Vegetatio n Profile of the Brix ton Street Wetlands (Illustration by Margaret Pieroni: from Keighery et al . 1996)


9/116

3


The most serious threats to plant communities acrossthe wetlands are South African geophytes. HarlequinFlower (Sparaxis bulbifera ) in particular is present on

the edges of the claypans, throughout the wet flats andup into the Marri woodlands. It produces vast amountsof viable seed that germinate each year. Other SouthAfrican geophytes that threaten the wetland floraincludeWatsonia species, One Leaf Cape Tulip (Moraea flaccida ), Babiana (Babiana angustifolia ), Wavy Gladiolus(Gladiol us undula tus ) and Freesia. The South Africangrass Tribolium uniolae is a relatively recent invader butfrequent fire in wetlands appears to be facilitating itsmovement into otherwise undisturbed areas. Annualweeds are prolific around the disturbed edges. Theannual sedge Isolepi s hystrix is a particularly seriousweed forming dense mats in low-lying wetter areas.

Most of the on-ground management at Brixton Street is

carried out by the Friends of Brixton Street Wetlands; acommunity group formed to save the wetlands fromhousing development in the early 1990s. The land is stillvested in the state housing authority, Homeswest, andhas been due to transfer to the Department of Conservation and Land Management (DCLM) for the lastnine years. Management guidelines for the wetlandshave been produced by the Wildflower Society of Western Australia (Inc.) with financial assistance from aCommunity Conservation grant (Keighery and Keighery1995). DCLM plays some role in management as the areasupports rare flora and threatened plant communities. Amanagement group with representatives from theFriends, the Wildflower Society, DCLM, and the City of Gosnells meets on an irregular basis.

Funding for weed management and bush regenerationwork in recent years has come from the Perth Branch of the Wildflower Society, DCLM and the City of Gosnells.

Talbot Road Bushland Talbot Road Bushland, a 95 hectare remnant, lies at thefoot of Darling Scarp on the soils of the Ridgehill Shelf.Marri, Jarrah and Wandoo (Eucalyptus wandoo )woodlands occur on the heavier soils and Banksiawoodlands and shrublands on the sandier soils. Thewoodlands support an exceedingly diverse understoreyof shrubs, herbs and sedges. Around 366 taxa, many of

them rare or restricted, occur across the bushland. It isthe most significant area of Ridgehill Shelf vegetationremaining and the floristic communities are considered

critically endangered (Keighery and Keighery 1993,Government of Western Australia 2000). Around 90 % of the bushland is in good to excellent condition but thereare areas of severe localised disturbance - around drains,paths and a gravel pit. Around 55 species of weeds occurin the bushland (Brown and Brooks unpubl.).

The most widespread serious weed in the Talbot RoadBushland is the South African geophyteHesperantha falcata . It occurs along path edges throughout thereserve, moving into undisturbed bushland on theheavier soils. Other South African geophytes are stillquite localised in their distribution. These includeFreesia, Babiana and Watsonia species. South Africangrasses are the other group of serious weeds with

Perennial Veldgrass widespread on the sandier soilsand African Lovegrass (Eragrostis curvu la ) occurringacross disturbed areas of the bushland.

Talbot Road is vested in the City of Swan and managedby a committee made up of representatives from theFriends of Talbot Road, the Department of EnvironmentalProtection, DCLM and the City of Swan. Much of the on-ground management is carried out by the Friends withDCLM and the City of Swan playing a significant role. ANatural Heritage Trust grant to the management group toimplement a management plan (Environs Consulting1999) has provided consistent funding for weedmanagement work over the last three years.

Gingin Brook The Gingin Brook, 150 kilometres north of Perth, is fedby perennial springs arising from the hills north east of the Gingin townsite. One of the last remaining patchesof fringing vegetation left along the brook where itcrosses the heavier soils at the base of the DandaraganPlateau is located in the townsite. For two kilometresalong the brook, Flooded Gum and Swamp Paperbark(Melaleuca rhaphi ophylla ) form a dense canopy over anunderstorey of native herbs, rushes, sedges and ferns.

The herbs, Centella asiatica , Persicaria sali cifoli a andCotula coronopifolia form ground cover in the wetterareas while Tassel Sedge (Carex fascicula ta ), Tall Sedge(Carex app ressa ) and the fern,Cyclosorus interr uptus ,dominate the understorey.

Elizab eth Buters, Frien ds of Brixton Street Wetlan ds.

Green Corps team hand-removi ng isolated plants of Perenni al Veldgrass from al ong paths in Talbo t Road Bushland.


10/116

4


Both Arum Lily (Zantedeschia aethi opica ) and Taro(Colocasia esculenta ) form dense monocultures indifferent parts of the fringing vegetation. Two exotic

pasture grasses, Para Grass (Urochloa mutica ) andReed Sweet Grass (Glyceria maxima ) have also becomenaturalised, smothering all native vegetation wherethey invade. In places woody weeds such as Edible Fig(Ficus cari ca ) and Brazilian Pepper (Schinus terebinthifolius ) are displacing the native Paperbarksand Flooded Gums.

The Friends of Gingin Brook, working closely with theShire of Gingin, have been responsible for allrestoration works carried out. In 1998 with supportfrom the shire, they received a grant from the NaturalHeritage Trust to restore the fringing vegetation in thetown site. They have carried out much of the on-ground works themselves, sometimes employing

contractors to assist them with weed management.Green Corps teams have made a major contribution tothe labour force over the life of the project.

Quairading Nature Reserve

The Quairading Nature Reserve is a 527 hectare crownreserve, located 164 kilometres east of Perth in theWestern Australian Wheatbelt. The diverse landscapeand soils of the reserve support a range of woodlandsand shrublands. Salmon Gum (Eucalyptus

salmonophloia ) and Wandoo woodlands cover thevalley floors, Acorn Banksia (Banksia prionotes ) andSand Plain Woody Pear (Xylomelum angustifoli um )woodlands cover the deep yellow sands while a seriesof shrublands occur on sands and gravel soils(Keighery et al . 2001). Project work at Quairading wasbased in the York Gum (Eucalyptus loxoph leba ) – Jam(Acacia acuminata ) woodlands that occur on the fertilebrown loams associated with the granites. Thesewoodlands support a rich annual flora. Through springPink Sunray (Rhodanthe manglesii ), Pink Everlasting(Lawrencella ro sea ) and Golden Waitzia (Waitzia nitida )carpet the ground. Later in the season OrangeImmortelle (Waitzia acuminata var. acuminata ) andflowering perennial grasses such as Aristida contorta

and Austrostipa species are noticeable. Shrubs areuncommon in the understorey.

The vegetation across the reserve is mostly inexcellent condition with only a few serious weedspresent. Around an old settlement, in the Wandoo

woodland, there are a few isolated populations of Freesia, and One Leaf Cape Tulip is starting to movedown creek lines. In other parts of the reserve a rangeof annual weeds occur around old carcass dumps,rabbit warrens, areas of nutrient run-off from adjoiningfarmland and along tracks. Wild Oat (Avena barbata )and Blowfly Grass (Briza maxima ) are the mostwidespread weeds across the reserve and can be foundon the more fertile soils of the York Gum – J amwoodlands, invading the understorey and displacingthe rich annual flora.

The Quairading Shire has had a temporary vesting of the reserve for the last five years (1998-2002). There isa reserve management committee with representation

from the Shire, the Quairading District High School, theLand Care District Committee, Rotary, Rural Youth, theGolf Club and the Tidy Town Committee. There are fewon-ground resources allocated to weed management orbushland restoration. The work carried out in thereserve was in conjunction with local farmers, localLandcare Coordinators and the District High School. Aworkshop was held in the reserve in August 2000involving EWAN, the Wildflower Society, the localQuairading community, DCLM and the local shire. Theaim was to foster interest in bushland restoration andweed management issues.

The work carried out in these bushlands over the lastfour years forms the basis for much of the informationand for most of the case studies. The descriptionshighlight how diverse and complex the bushlandsaround Perth and south west Western Australia can be.Managing them for nature conservation means gettingto know individual sites. There are, however, someuniversal principles and a general approach that canunderpin effective management of environmentalweeds where they are invading these areas. Thefollowing chapter discusses some of those principlesand provides an outline of the kind of approach thatcan be taken when starting out on a bushland

restoration/weed management project. Importantly, thekind of area-specific information that is vital toknowing and understanding individual bushlands isalso listed.

Pauline Di ggins, Friends of Gingin Bro ok.

Participa nts in a work shop held in th e reserve, August 2000.


11/116

5

Weed management in bushland is concerned with muchmore than simply the elimination of weeds. Theunderlying objective is always the protection andrestoration of diverse natural ecosystems.

South west Western Australia supports one of themost diverse floras in the world, occurring in intricatepatterns across a variety of landscapes and soils. Forthe southern Swan Coastal Plain alone, 1700 nativetaxa (species, subspecies and varieties) have beenrecorded occurring in at least 30 different plantcommunities (Gibson et al . 1994, Keighery 1999b).Along a 30 m long and 1 m wide transect through apopulation of Harlequin Flower (Sparaxi s bulbifera )

invading herblands in the Brixton St Wetlands, up to28 native taxa and 8 introduced taxa may be found.Bushland weed management in the region is oftenabout working in complex natural systems with a longhistory of varying disturbances. There is a need torecognise that effective weed management amongsuch diversity begins with knowing and understandingeach site; the distribution of the native plants and thenative plant communities, the patterns of disturbanceand the distribution of weeds. In particular it isimportant to recognise the locally serious weeds. Onlythen can weed control and management be strategic –carefully targeted, and tailored to site conditions andavailable resources.

Gathering area-specific informationNote: The informatio n listed below i s often compiled when bushland

areas are being identified for retention o r as part of a m anagement

plan. When such inform ation is not avail able it should be compil ed as part of the man agement program.

Vegetation maps

Vegetation maps of particular bushland areas provideinformation on the structure and patterns of nativeflora across different landscapes and soils (seevegetation map of Quairading Nature Reserve, Box 3.2).When carrying out bushland restoration, vegetationmaps, accompanied by a flora list, provide a vitalreference and an important guide to where in the

landscape particular native species occur and the soilson which they generally grow.

Flora list

A flora list aims to record all taxa known to occur at aparticular site. A comprehensive flora list is a vitalreference for ensuring species not known to occurnaturally at a particular site are never introduced as apart of bushland restoration – either through plantingor direct seeding.

In addition, flora lists will often indicate the plantcommunities and associated soils and landforms whereparticular species can be found growing. At BrixtonStreet Wetlands for example, such a flora list provided a

guide for appropriate selection of species for directseeding trials located along a degraded edge of herb-rich shrublands on damp, heavy clay soils (see direct

seeding case study 5.1). Flora lists should also includeweeds – it is important to know all flora and to be ableto recognise new weeds as soon as they arrive at a site.Lists will change over time. When our work began atShenton Bushland 40 species of weeds were recorded.After three years of working and getting to know thesite, 25 additional weed species had been identified.

Vegetation condition maps

Impacts of disturbances across a bushland can berecorded in vegetation condition maps. Fire intensityand frequency, weed invasion, soil disturbance, disease,rubbish dumping and past vegetation clearancesinteract to impact on the species composition, coverand structure of native vegetation. Vegetation conditionmaps aim to reflect the degree of those impacts.Combined with maps of the distribution of individualweed species, they are useful tools for carefullytargeting weed control work to protect relativelyundisturbed bushland (Box 2.1). Mapping of vegetationcondition requires knowledge of native flora as well asfamiliarity and understanding of the nature of aparticular undisturbed plant community (see Box 7.1 fordetails on criteria used to map vegetation condition).

Weed maps

Maps that provide a clear understanding of where theserious weeds occur across a bushland site are basic

planning tools. They allow for careful targeting of limited resources and provide the information requiredfor strategic weed management. They can provideuseful information on the spread of weed populationsover time and also provide basic information on theeffectiveness of control programs. Not all the weeds inthe bushland need to be mapped – only those that havea serious impact. See Chapter 7 for how to create anduse weed maps and for some of the criteria that can beused to determine the serious weeds at a particularsite.

Fauna information

Information on the area’s fauna is also useful.

Sometimes weeds can provide habitat or anopportunistic food source for native animals. A list of the known fauna in a bushland can help determine thisat a particular site. Gradual removal of such weeds overtime may be required while animals find alternativehabitat or food sources. For example, dense stands of Watsonia can provide important habitat for Bandicootsin the Perth area.

Chapter 2 Managing Weedsin Bushland

Some General Pri nciples


12/116

6

Chapter 2 Managing Weeds in Bushland

Regional information

Australia is divided up into a series of natural regions.

The Swan Coastal Plain and the Wheatbelt are twosuch regions within Western Australia from where casestudies in the text have been drawn. An understandingof a bushland’s natural values in relation to others inthe region provides information on the area’ssignificance. For example, a bushland may berepresentative of the more common or rare plantcommunities in the region. It is important to note thatthe rarest communities (Threatened EcologicalCommunities) and flora (Declared Rare Flora – DRF),are protected by laws, either at the state (DRF) or thefederal (communities) level. Bush Forever, Volume 2 –Directory of Bush Forever Sites (Government of Western Australia 2000) catalogues information onregionally significant bushland on the Swan CoastalPlain. It provides details on the area of the bushlandtype that remains uncleared in the region, how much isin conservation reserves, the quality and condition of those bushlands, and where similar bushland can befound. This type of information is important inproviding a focus for weed management work, and ingaining an understanding of bushland values thatrequire protection (Keighery et al . 1998).

Note: Bushland restoration carri ed out where ra re flora occurs

requires a permi t from the Department of Conservation and Land Management (DCLM).

Developing a weedmanagement program

Resources for bushland restoration work are generallylimited. It is critical that these resources are carefullytargeted through strategies that prioritise managementactions, based on a knowledge of the bushland areaand the weeds that are impacting on it. Central tothese strategies are a series of principles that arefundamental to successful weed management. Many of the case studies throughout this manual demonstratethe application of these principles:

• Contain the spread of serious weeds and protectintact bushland. Consider the impacts of seriousweeds on rare flora and rare plant communities.

• Prevent new weed species establishing.

• Consider restoration of degraded edges. Usuallythis is of lower priority than protection of goodareas. Often though, degraded edges harbourserious weeds, providing a source of propagulesthat continually disperse into intact areas.

Implementation – taking anintegrated approach

Prevent new weeds from establishing

• Clean tools, boots, equipment and machinerybetween jobs to reduce risk of spreadbetween sites.

• Practise soil hygiene.

• Check paving materials before bringing on to asite, particularly limestone, for weed seed andonly acquire from accredited clean sources. BlackFlag (Ferra ria crispa ), Geraldton Carnation Weed(Euphorbia terracina ) and Pretty Betsy(Centranthus rubr a ) to name a few have beenintroduced to various bushland sites around Perth

in paving materials.• Avoid bringing soil or mulch from elsewhere into

bushland.

• Know the plants of your bushland (native andintroduced) and immediately remove infestationsof any new weeds. This is important at individualsites but also at a regional and state level.

See case study 5.2 on Holly-leafed Senecio(Senecio glastifoli us ) a recent invader to south westWestern Australia.

Limit the spread of established weeds

• Target small populations in good bush and the

outliers of dense infestations - use the maps(Box 2.1).

• Keep soil disturbance to a minimum. Disturbancefavours the establishment of many weeds. It bringsburied weed seed to the surface thereby releasingdormancy, and creates favourable conditions forthe germination of wind dispersed weed seed.

• Avoid working in areas where weeds are activelyshedding seed.

• Post-fire conditions (space, light and high nutrientavailability) often favour establishment of weeds.Weed control in the season immediately followingfire will prevent seed set in established weeds and

reduce germinating weed seedlings. It will limit theinevitable spread of many serious bushland weedsthrough the post-fire landscape.

See Box 4.2 on fire and cormous and bulbous weeds,Box 5.1 on the weeds that move in with soildisturbance, and Box 3.5 on limiting the spread of

Perennial Veldgrass following fire.


13/116

7


Box 2.1 Strategy for the management of Yellow Soldier (Lachenal i a ref lexa) invading aBanksia woodland at Shenton Bushland.

Th is st ra tegy is simp ly an i l l us t ra t ion o f how some o f the in fo r mat ion covered in th is chap te r can be

used to h elp se t p r io r i t ies and develop a n e f fect ive wor ks progra m.

Bio logy

Yellow Soldier flowers in winter and early spring, dying back to a bulb over summer. It sets prolific amounts of viable seed (around 40 seed per flower) and up to 1700 bulbs can occur in four square metres. Seed appears toremain viable in the soil for only a couple of years and may be dispersed by water. Plants are not killed bysummer fire, but flower well and are highly visible in the post-fire landscape. In addition high numbers of seedlings have been observed establishing on bare ground left following fire.

Suggested Contr o l Metho ds

For isolated plants in sandy soils use a knife to cut the roots and pull out with bulb in J uly to early August. Trialshave shown metsulfuron methyl at2.5-5 g/ha, applied just before

flowering, in late July, provideseffective control with little impacton co-occurring native species.Carefully spot spray.

Recommended str ategy for th e

effective ma na gement of Yellow

Sold ier a nd r egenerat i on of the

Banksia w ood land i t is invad ing

Hand-removal of entire populationsis time consuming (up to six hoursfor four square metres) and soildisturbance results in germinationof annual weeds. Specifically

targeted herbicide treatment isrecommended for denseinfestations and hand-weeding forisolated plants.

The populations on the easternside of the reserve and at thesouthern end of the reserve areboth relatively small, occurring invery good to excellent conditionbushland and should both be highpriority for eradication.

The larger population of Yellow Soldier on the western side

of the bushland is spreadingthrough a more degraded area andis not as high a priority if resourcesare limited.

Following fire, resources should beallocated for control. At this timeplants are an easy target forherbicide control and preventionof flowering and seed set willlimit establishment in thepost-fire landscape.

Allow indigenous species torecolonise the treated sitesunassisted (Yellow Soldierco-occurs with up to 25 nativespecies in a 2 m x 2 m plot). very good - excellent fair - good poor very poor

Figure 1. The distributi on of Yellow Soldier• and vegetation condition across Shenton Bushland .

Bushland condition


14/116

8


Understand the biology of a particular weedspecies – focus on how the weed reproducesand spreads

• When is it actively growing?

• When is it flowering and seeding?

• How long do seed or other propagules remainviable in the soil?

• How does it respond to fire?

• What are the vulnerable times in the life-cycle?

• Is there a preferred time for physical control?

• What is the preferred time for chemical control?

See case study 3.2 on learning about the biology of Tribolium, case study 3.4 on timing Perennial Veldgrass

control, Box 4.4 on corm exhaustion and herbicideapplication and case study 6.2 on timing Fig removal.

Consider all the control options

• Consider the impact of control options onco-occurring native flora including trees, shrubs,geophytes, grasses, and other native herbs andon fauna.

• Physical, chemical and biological methods are alluseful in managing bushland weeds.

• A combination of techniques is required forpractical and effective long-term outcomes.

• It is important to adapt control methods to

site-specific conditions and available resources.

See case studies 4.4 on methods used to controlHarlequin Flower (Sparax is bulbifera ) and 4.3 oncontrol of Taro (Colocasia esculenta ).

Assist natural regeneration

• Assist natural regeneration through carefullytargeted weed removal.

• Stimulate germination of the native soil seedbankwith smoke products or, where the nativeseedbank is depleted, consider direct seeding.Always use locally collected seed.

• Management practice that favours the

regeneration of native plants increasescompetition against weeds.

See case study 5.1 on direct seeding and Box 5.3 onassisting natural regeneration.

Keep a record of works programs over time

Keeping a record of works programs over time is an

important component of bushland restoration. Alogbook should record date, time and type of workscarried out including details of methods used, hoursworked, and site conditions.

Monitor the outcomes of weedmanagement work

Detailed monitoring of weed populations provides aquantitative record of the effectiveness of management programs, measures the impact of thecontrol programs on native plants and theregeneration of the native plant community over time.

The feedback provided can be used to adapt futuremanagement practices and to justify spending on

weed management works.How you monitor will depend on the changes to bedetected, the scale and distribution of the infestation,and the nature of the vegetation the weed/s areinvading. Setting up monitoring that is useful fordetecting detailed change over time is necessarilycomplex. For a particularly useful reference see‘Monitoring Plant and Animal Populations’ by Elzinga et al . (2001). ‘The Standard Operating Procedure forMonitoring Weed Control’ by the Department of Conservation, New Zealand (2000) is also a particularlyuseful reference and is designed to provide a detailedworking framework for field officers.

Case studies in the tuberous, bulbous and cormous

weeds chapter and the grass weeds chapter illustratesome of the methods used over the life of our projectto measure the effectiveness of weed managementwork, impacts on associated native plants and in somecases, regeneration of the native plant community.

See case studies 3.4 in the grass chapter and 4.1, 4.2,and 4.3 in the bulbs chapter.

The following chapters illustrate how the informationoutlined here underpins effective weed management inbushland and is vital in providing a framework for theprotection of unique bushland values. The managementof grass weeds is covered first, in the next chapter.


15/116

Chapter 3 Grass Weeds

9

The grass weeds as a group, particularly the SouthAfrican perennial species, are a serious threat tobushland around Perth. This chapter looks at grass

biology and the factors contributing to the invasion andestablishment of weed grasses, in particular fire andcanopy degradation. The influence of lifeform andgrowth form on management decisions is highlightedand the importance of growth stage on timing of control programs discussed. Finally, a table withinformation on the biology of individual species andavailable control measures is presented.

In terms of number of individuals, biomass, areacovered and diversity of habitat, grasses are one of themost successful plant families in the world (GibbsRussell et al . 1991, D’Antonio et al . 2000). Thecharacteristics that have contributed to this successhave also enabled many grasses to become aggressive

invaders of natural ecosystems.Almost one third of the 709 grass species recorded inWestern Australia are exotic (Western AustralianHerbarium 1998). In south west Western Australia,331perennial grasses were introduced and screened aspotential pasture species between 1943 and 1970 (Rogerset al . 1979). Of these, 14 % are now naturalised inWestern Australia (Western Australian Herbarium 1998)and many are now considered weeds of bushland of theSwan Coastal Plain and Jarrah Forest (Keighery 1999a).

These include serious invasive grasses such as AfricanLovegrass (Eragrostis curvul a ), Perennial Veldgrass(Ehrharta calycina ), Tambookie Grass (Hyparrhenia hirta ), Kikuyu (Pennisetum clandestinum ), Fountain Grass

(Pennisetum setaceum ) and the coastal invaders,Marram Grass (Ammophila arenaria ) and Pyp Grass(Ehrharta vill osa ). Horticulture has also been

responsible for the introduction of a number of seriousgrass weeds, among them Pampas Grass (Cortaderia selloana ) which is still seen in Perth gardens. However,it is important to identify unknown grass species andnot just assume that because it is a grass it is a weed(Box 3.1).

Impacts

Introduced grasses are competitive in many nativeecosystems; they may displace the native understoreyand alter fire regimes (Humphries et al . 1991). Theimpact a particular grass species has is partlydetermined by the characteristics of the invaded site(Box 3.2). In south western Australia the annual

grasses, Wild Oat (Avena barbata ) and Blowfly Grass(Briza maxima ), threaten the herbaceous flora foundon the granitic soils of Wheatbelt reserves. PerennialVeldgrass is highly invasive in the sandy, nutrient-poorsoils of the Swan Coastal Plain - soils commonlyoccupied by Banksia woodland. Kikuyu, Water Couch(Paspalum distichum ) and Vasey Grass (Paspalum urvillei ) invade the more nutrient-rich wetlands and

Tambookie Grass tends to occur on the heavier soilsof the Darling Scarp. Grasses often colonise disturbededges or patches within bushland. This edgecolonisation allows the grass to take advantage of anydisturbance event within the bushland, establishingrapidly after fire, clearing or soil disturbances.

Box 3.1 Native grasses of the Perth Region

Of the numerous native grasses occurring in Perth’s bushland, many are mistaken for weeds. Knotted Poa (Poa drummondiana ) is superficially similar to Winter Grass (Poa ann ua ), while Kangaroo Grass (Themeda triandra )is often mistaken for Tambookie Grass. Some native grasses simply look ‘weedy’ to the untrained eye, includingSwamp Wallaby Grass (Amph ibrom us nervosus ), Clustered Lovegrass (Eragrostis elongata ) and Marine Couch(Sporobol us vir ginicus ).

Native grasses that occur in the Perth region include:

* Bold i ndicates that weedy species from the same genus that can be found i n the Perth area.

Agropyron scabrum Austrodanthonia occidental is Dichelachne crini ta Agrost is

avenacea Austrodanthonia pi losa Eragrost is

elongata Agrost is plebeia Austrodanthonia racemosa Glycer ia australis Agrost is preissi i Austrodanthonia setacea Hemarthria uncinata Amphibromus nervosus Austrostipa campylachne Microlaena stipoides Amphibromus vicker yae Austrostipa compressa Neurachne alopecuroidea Amphipogon avenaceus Austrostipa elatior Neurachne minor Amphi pogon amphi pogonoi des Austrosti pa eleganti ssi ma Poa drummondiana Amphipogon debi l i s Austrostipa flavescens Poa poiformis Amphipogon laguroides Austrostipa macalpinei Poa porphyroclados Amph ipogon str ictus var. h ir sutus Austr osti pa pycnosta chya Polypogon tenellus Amph ipogon stri ctus var. seti fer a Austro sti pa sem ibarbata Spin i fex hirsutus Amphipogon turbinatus Austrostipa tenui fol ia Spin i fex longifolius Aristida contorta Austrostipa variabi l is Sporobo lus virginicus Aristida ramosa Bromus arenarius Tetrar rhena laevis Austrodanthonia acerosa Deyeuxia quadriseta Themeda triandra

Austrodanthon ia caespitosa

List from Keighery ( 1999b)


16/116

10


Many grasses use the disturbance caused by bushfireto get a foothold within bushland areas. In turn, grassinvasions increase fuel loads which indirectly impact on

the native ecosystem by changing fire frequency,intensity, patchiness, size and timing (Humphries et al .1991, Mack and D’Antonio 1998). The introduction of Buffel Grass (Cenchrus cili aris ) as a pasture speciesthroughout arid and semi-arid regions of Australia issaid to have drastically altered the fire regime. In areaswhere the grass occurs, fuel loads are high andcontinuous. Dry water courses which previously actedas fire breaks are now bridged by fire and also act as

wicks, spreading the fire further (Humphries et al .1991). More locally, Perennial Veldgrass, Lovegrass andFountain Grass are all fire adapted, resprouting

vigorously and seeding prolifically after bushfires(Christensen and Abbott 1989, Walsh 1994, Milberg andLamont 1995, Benton 1997, Muyt 2001). Changes in thefire regime can have long term impacts on the structureand composition of the native communities (Williamsand Baruch 2000). Indeed, the change in fire regimebrought about by Buffel Grass invasion has convertedareas of the Sonoran Desert cactus forests of CentralAmerica into grassland (Van Devender et al . 1997).

Box 3.2 Ecosystem susceptibility: Resource availability, soil structure and Wild Oat

In the Wheatbelt of Western Australia Wild Oat (Avena barbata ) is often associated with York Gum

(Eucalyptus loxophleba ) – J am (Acacia acuminata ) woodlands. The soils of these woodlands are characterisedby high values of nitrogen, phosphorus and potassium. They are friable and lack a hard crust, making themmore susceptible to weed invasion (Hobbs and Atkins 1988, Arnold et al . 1998). The distribution of Wild Oatand Blowfly Grass (Briza maxima ) at Quairading Nature Reserve illustrates the association between thewoodlands and the two annual grass weeds (Figure 1).

The almost exclusive absence of Wild Oat and BlowflyGrass from other plant communities in the reservemay be attributed to a range of factors, poor nutrientlevels among them.

Heath and Tamar (Allocasuarina species)communities form dense thickets, competing stronglyfor resources and preventing the majority of lightfrom penetrating the low canopy.

Salmon Gum (Eucalyptus salmonophlo ia ) woodlandshave an open understorey with large areas of bareground and only dappled shade. However, even on thereserve edges, there is very little weed incursion. Thesoils associated with these woodlands form a hardcrust, relatively impervious to seeds not adapted toself-burial. Salmon Gums also have an extensive lateralroot system near the soil surface competing stronglyfor available surface moisture and inhibitingestablishment of grass weed species (Yates et al . 2000).

Wandoo (Eucalyptus wandoo ) woodlands are slightlymore prone to invasion. Like Salmon Gums, Wandoowoodlands have a relatively open canopy, but the soilcrust appears less impervious. Where Wandoo blendsinto Jam, patchy Wild Oat distribution often occurs.

Where Wild Oat occurs within other plantcommunities it is associated with reserve edges ordisturbance from rabbits or machinery. The onlyexception to this is two small patches growing underthe Christmas Tree (Nuytsia flor ibunda ). This parasiticplant may increase nitrogen within the soil byaccumulating the element from host plants. Thenitrogen is then released into the surrounding soilwhen flowers and leaves are shed (Hocking 1980).

Understanding the relationship between vegetationcommunities and weed invasions is important tomanagement. Susceptible areas can be targeted forweed prevention and early control. Vegetation mapscan be used as a rough guide to possible weeddistribution at other sites, enabling the estimation of resources required for control.

Figure 1. Vegetation compl exes occurring w ithin Quairading Nature

Reserve, and the distribution an d density of Wild Oat a nd Blowfly Grass acro ss the reserv e.

(Vegetation map adapted from Keighery et al . 2001)

Salmon Gum WoodlandWandoo Woodland

Wandoo Woodland on lateri teYork Gum - Jam WoodlandDegraded York Gum Wood landYork Gum - Jam - Rock Sheoa k Woo dlan dGranite outcropLithic ComplexTama r Woodlan d o n sandy gravelsTam ar Shrubland o n lateritic gravelTam ar Shrubland on sand y clayHeaths on deep g rey sandTam ar - Dryand ra - Erema ea Shrubla nd on crea m sandTa ma r - Eucalyptu s macrocarpa Shrubland o n yellow sandBanksia prion ot es - Sand Plain Woody Pear Low Woodlan dCleared

Patchy Wild Oat and Blowfly Grass coverDense Wild Oat an d Blow fly Grass cover


17/116


11

Fire is less important in the establishment of weedygrasses in wetlands and riparian zones. Instead,invasion of weed grasses is often encouraged by poor

land management practices that increase nutrient run-off into streams and wetlands. The disturbance causedby grazing along waterways can also promote thespread of weed grasses. The grasses that commonlyinvade wetlands are often rhizomatous, their rapidgrowth and mat-forming habit smothering native plantsand preventing further recruitment. Some semi-aquaticgrasses, such as Johnson Grass (Sorghum halepense ),Reed Sweet Grass (Glyceria maxima ) and Para Grass(Urochloa mutica ) grow in up to one metre of water.

These species can form large floating mats that reducethe habitat of waterfowl, slow and change the directionof water flow, alter silt deposition displace food sourcesand reduce nesting sites (Humphries et al . 1991).

Biology – why are grassessuccessful weeds?

The structural design of grasses helps make themhighly competitive. The presence of axillary buds atthe base of each internode allows most grasses toresprout vigorously when damaged (Gibbs Russell et al . 1991). This advantage is compounded by thedevelopment of crown tissue at the base of the grassplant. Crown tissue produces buds at or below groundlevel where they are largely protected from theenvironment. These buds are the source of rhizomes,stolons (Box 3.3) and tillers (new grass shoots orculms arising at or near the base of the primary culm).

Soon after a seedling germinates, adventitious rootsdevelop from the crown tissue, firmly anchoring theplant to the ground. These rootsstore excess carbohydrates andcan permeate large volumes of soil very efficiently, makinggrasses highly competitivefor moisture and nutrients(Hannaway et al . 2000, GibbsRussell et al . 1991). Grassesare broadly grouped intotwo lifeform categories,annuals and perennials:

Annual grassescomplete

their life-cycle in a singlegrowing season, storing alltheir excess photosynthate(plant food) within the seed(Hannaway et al . 2000).Annual grasses like WildOat, Blowfly Grass,Annual Veldgrass(Ehrharta longiflora ),Silvery Hairgrass(Aira caryophyllea )and Barnyard Grass(Echinochloa crus-galli ) all

rely entirely on seedling recruitment to establishwithin an area. Annual grasses are usually tussockforming or caespitose (Box 3.3), although occasionally

tillers of prostrate grasses may root at the nodes.

Perennial grasses reproduce vegetatively as well as byseed, storing excess photosynthate within rhizomes,stolons, seed and occasionally corms (swollenunderground stem bases). Perennials develop winter-or summer- hardy buds capable of resuming growththe following season (Hannaway et al . 2000). Perennialscan be sterile and still reproduce vigorously, spreadingby stolon, rhizome and dispersal of grass rootfragments in water and soil, Kikuyu and Giant Reed(Arundo donax ) being just two examples. Perennialgrasses generally fall into one of three descriptivegrowth forms, caespitose or tussock-forming,rhizomatous and stoloniferous (Box 3.3).

Reproduction, dispersal and persistence –implications for limiting spread

Sexua l reproduct ion

Seed is important for the spread and establishment of many perennial grasses. For the annual lifeform,seedling recruitment is essential.

Dispersal: An understanding of seed dispersalmechanisms allows us to limit further spread and re-infestation, an important, although often overlooked,component of any management program.

• Wind plays a central role in dispersal and many

grasses occupy open habitats subject to frequentwinds. Small, lightweight seed, suitable for winddispersal, is the most common form found in thegrasses. To further aid wind dispersal the outerbracts may be covered with long soft hairs(Davidse 1986). Seed from adjacent degraded areascan be carried into the bushland by prevailingwinds, providing seed rain after a disturbanceevent such as fire. Depending on the species thisdeposited seed may establish as a seedbank, readyto germinate given the appropriate conditions.

Once r ip e, w in d di sper sed seed is easi l y

d is lod ged an d car e less removal of grass weeds

a t th i s stage wi l l fa c i l i ta te d ispersa l .

Prefera b ly, seed heads shou ld b e removed an d

bagged pr ior to th i s stage.

• Water can disperse large numbers of seeds. Thelight weight of many seeds allows them to floateasily. Seeds may have aerenchymous tissue, thetiny air pockets aiding flotation. Some seeds havea long awn that helps embed the seed when itreaches a suitable embankment (Davidse 1986).Run-off from rainstorms can carry grass seeddownhill and into creeks and storm drains,concentrating vast numbers of seeds in thewetlands they feed into.

Upst ream an d uph i l l sour ce popu l a t ion s need

to be man aged . Dra i n ou t le ts can have

sumps inco rpor a ted to a l low weed seed to sett l e and col l ect .

inflorescence

internode

tiller buds location of

crow n tissue

adventi t ious roots

axillary budsor nodes


18/116

12


Box 3.3 Grass growth forms

Caespitoseor tussock grasses are the dominantgrass growth form around the world. Examplesinclude the bushland weeds Perennial Veldgrass(Ehrharta calycina ), Pampas Grass (Cortaderia selloana ), Tambookie Grass (Hyparrhenia hirta ) andLovegrass (Eragrostis curvula ). Reproducing by seedand/or tillers they form dense, usually erect, clumps.As clumps age, each year’s old leaf materialaccumulates, creating large fuel loads (Briske andDerner 1998, Hannaway et al . 2000). Caespitosegrasses typically occupy resource-poor habitats. Byaccumulating soil organic carbon and nitrogendirectly beneath clumps, they monopolise resources(Gibbs Russell et al . 1991).

Stoloniferousgrasses are creeping or mat-forminggrasses that spread locally by stolons and includePara Grass (Urochloa mutica ) and Queensland BlueGrass (Digitaria didactyla ). A stolon is a segmented,horizontal stem, which runs predominantly alongthe soil surface. Adventitious roots and aerial shootsarise from stolon nodes. Stolons themselves arisefrom adventitious buds in the crown tissue.Propagation is both vegetative, from stolonfragments, and from seed. Although commonly

occupying moist, high nutrient areas, some foredunespecies such as Spinifex (Spin ifex sericeus ) andSaltwater Couch (Paspalum vaginatum ) arestoloniferous. Many stoloniferous grasses areconsidered invasive species (Hannaway et al . 2000).

Rhizomatousgrasses form dense mats, extendingtheir coverage by producing below-ground lateralrhizomes. A rhizome is a modified underground stem

capable of rooting and shooting at nodes to developdaughter plants. They can serve as storage tissuefor vegetative propagation and, being underground,are protected from fire. These grasses, includingspecies like Reed Sweet Grass (Glyceria maxima ),Pyp Grass (Ehrharta villosa ) and Johnson Grass(Sorghum halepense ), propagate vegetatively, butmay also spread by seed. Rhizomatous grassesdominate moister, more nutrient-rich habitats andare often invasive (Hannaway et al . 2000).

Many grasses can be both rhizomatous andstoloniferous, Couch (Cynodon dactylon ) and Kikuyu(Pennisetum clandestinum ) being just two examples.

tiller buds

noderhizome

advent i t ious root

adventi t ious shoot

stolon

n o d e

adventi t ious shoot

African Lo vegrass(Eragrostis curvula)

Para Grass (Urochloa mutica)

Reed Sweet Gra ss(Glyceria maxima)


19/116


13

• Animalscan carry grass seeds on their skin andfur. Seed bracts and/or awns may have stickyhairs, hooks or spines that aid adhesion. Ingestion

also plays a role in grass dispersal. Grassesevolved with grazing and several authors havenoted seeds are voided intact (Davidse 1986).

Dogs, hor ses, na t i ve an i ma ls an d hu mans

(amon g other s) r eadi l y d isperse seeds al ong

bush t ra cks . Hor ses can spread gra ss weeds

f rom paddocks in to bush la nd , deposi t in g seed

w i th the i r manu r e.

• Human activities including inappropriatemanagement practices, provide additionalmechanisms for grass weed dispersal:

Grading drags seed (and tillers) along road verges.

Slashing during flowering spreads seed with an

explosive action.

Lawn clippings dumped in bushland oftencontain grass seed.

Soil seedbanks: Seedbanks consist of dormant seed,which is ready to germinate given the appropriateenvironmental cues. Dormancy allows seed to persistin the soil in the absence of further seed rain.

As a general rule grass seeds are not considered to bepersistent within the soil. However, there areexceptions. Numerous grasses have a soil seed lifegreater than five years and not surprisingly many of these are widespread weeds (eg. Fountain Grass andBarnyard Grass). Poa pratensis and a number of Setaria species are known to remain viable for at least39 years (Baskin and Baskin 1998). Environmentalfactors, including temperature, moisture and lightintensity can induce dormancy in seeds. For example,deep burial induces dormancy in Wild Oat seed,increasing its longevity from six months to anything upto ten years (Baskin and Baskin 1998, Nugent et al .1999). Dormancy is broken when the seed is returnedto the surface through soil disturbance.

For many species, seed longevity is not understoodand yet it is invaluable information when makingmanagement decisions. The length of a controlprogram is largely determined by seedbankpersistence. The shorter lived the seed, the morerapidly a grass infestation can be controlled(provided the source is no longerpresent or is also managed).

Vegetative r eprod ucti on

Vegetative reproduction as a strategy in grasses is

most successful in moist, nutrient-rich conditions(Briske and Derner 1998). Under these conditionsrhizomatous and stoloniferous grasses can easilyestablish roots and rapidly disperse rhizome andstolon fragments in soil and down streams. In ReedSweet Grass (and possibly other species), shoots of young plants can be vegetative or flowering. However,once established, the majority of new shoots producedare vegetative. This strategy allows the plant to quicklycolonise new areas by seed, whilst increasing thedensity of established plants (Department of PrimaryIndustries Water and Environment 2001).

Tussock-forming grasses may also propagatevegetatively. Tillering is responsible for each season’s

new growth and the expansion in tussock area(Hannaway et al . 2000). Tillers are also referred to assprouts, shoots and daughters and may break off whendisturbed.

I f resour ces ar e scarce, man agement shoul d focus

on con t r o l l in g the invad ing edge(s) and new or

i sola ted popu la t i ons .

Grass growth stages

As with all weeds the timing of grass control measuresis important. Active growth is required for herbicideuptake and slashing can enhance leafy growth or resultin defoliation, depending on the growth stage. In mostgrasses (sterile grasses are the exception) there are

three main stages of active grass growth: Vegetative,transition and flowering.

• Vegetative growth involves the production of shoots, mostly leaf blades but also stolons andrhizomes. Herbicides are best applied at this stage,while slashing usually results in increasedproduction of leafy material.

veg e ta t ive f lo w eringt ra nsit io n b o o t st a g e


20/116

14


• Transition occurs when the growing points stopproducing vegetative material and start developingflowers and the tissue between nodes, the

internode, elongates. Because of internodeelongation, slashing or mowing at this stage canremove many axillary buds, thereby reducing leaf production in the regrowth.

• Flowering begins when the seed head is justemerging from the leaf sheath (boot stage) andcontinues through to seeding. Some grass-selectiveherbicides are only effective if applied prior to theboot stage. Slashing or mowing during floweringmay facilitate the spread of seed.

The period of active growth depends largely onwhether the plant is a cool season (C3) or warmseason (C4) grass. Cool season and warm seasongrasses differ in the way they use carbon dioxide. Coolseason grasses (Wild Oat, Blowfly Grass, PerennialVeldgrass) photosynthesise more effectively, and thusare most actively growing, during the cooler periods of winter and spring (15-25° C). Warm season grasses, orsummer grasses, photosynthesise best with full lightsaturation and consequently their period of mostactive growth is late spring and summer (25-40° C).

The summer growing Couch (Cynodon dactylon) ,Kikuyu and Lovegrass are all C4 grasses.

For su ccessfu l herb i c ide upta ke gr asses shou ld be

sprayed in the season a ppr opr ia te to thei r

pho tosyn the t ic pa thw ay and wh en act ive ly grow ing .

Fire and the spread of grass weeds

Many grasses, especially caespitose species, have long,narrow, vertical leaves that are efficient in strong lightallowing sunlight to penetrate deep inside the clump(Gibbs Russell et al . 1991). This structure allows theproduction of a large biomass in a small space. Stemsdie each year and new ones develop. Over time a largebiomass, much of it dead material, creates a significantfire hazard. As a result, grass weed invasions can leadto changes in fire frequency and intensity.

These changes are self-perpetuating as the disturbancecaused by fire tends to promote germination andestablishment of seedling grasses (Cheplick 1998,Williams and Baruch 2000). At Shenton Bushland,

Perth, Perennial Veldgrass was mapped in the yearfollowing an intense fire that burnt half of the site. Acomparison between a pre-fire bushland condition mapand the post-fire Veldgrass map, highlighted asignificant post-fire expansion in the population.Previously healthy bushland now recorded 75-100 %cover of Veldgrass (Brown and Marshallunpublished data).

Fire contributes greatly to grass invasion in low-nutrient soils, where post-fire increases in nutrients,light and space availability are more effectivelyexploited by invasive grasses than by the native flora.

I t is impor tan t to r educe the poss ib i l i t y o f f i r e in

bush la nd ar eas a t r isk o f gra ss invas ion . Where

bush f i r e does occur ta r get resprou t i ng p lan ts and

seedl in gs for contr o l befor e they become

estab l ished.

Management and control

Prevention and early intervention

Degradation of our native bushland remnants throughclearing, fire, disease and mismanagement may be of particular relevance to the establishment and spreadof grass weeds. Often an observed decline in thecanopy cover within Banksia woodlands around Perthcorrelates with an increase in the occurrence of Perennial Veldgrass. Reduction in canopy cover leadsto an increase in light availability, a condition favouredby many grass weeds (Williams and Baruch 2000,D’Antonio et al. 2001).

Weed grass invasion can be limited by:

• Maintaining the bushland canopy. Many grassesprefer open sunny sites and do not establish orcompete successfully in the shade.

• Reducing the potential for bushfires. Asmentioned, fire can significantly increase theestablishment and spread of grass weeds.

• Identifying unfamiliar grasses. It is important todetermine if grasses are introduced or native.Early identification of grass weeds allows you toassess the likelihood of invasion and preventsmistaken eradication of native grasses!

• Establishing weed barriers.

Weed bar r iers

A common grass weed problem in urban remnants isthe spread of Kikuyu, Couch or Buffalo Grass(Stenotaphrum secundatum ) from adjacent parks androad verges. A successful ‘weed break’ can be createdby embedding a physical barrier such as weed mesh orconveyor belt vertically into the ground and usingconcrete kerbing above it. This prevents rhizomescreeping into the bushland and provides a mowingedge for council mowers (Box 3.4). Some grasses haverhizomes 50 to 60 centimetres below ground and anyroot barrier must allow for this.

The bush itself often provides an excellent weed break(Box 3.2). However, the sandy soils and openvegetation of Banksia woodland or the friable soils of

York Gum–Jam woodlands are more susceptible.Establishing a dense cover of locally-occurring low,medium and tall shrubs along the edges of degradedsites and paddocks could act as a sieve, preventingentry of wind and water carried seed. Knowing theprevailing wind direction, position in the landscapeand relationship to waterflow can help to determine if the adjacent paddock, weedy rail reserve or over-run,semi-rural backyard, is the weed source. When plantingor direct seeding, use native seed collected from thearea of bushland you are protecting and only plantwhere run-off or prevailing winds are definitelycarrying weed seed in.


21/116


15

Box 3.4 Spread of turf grass at Blue Gum Lake

At Blue Gum Lake in the City of Melville, parkland often adjoins bushland. In most of these sites Couch(Cynodon dactylon ) and Kikuyu (Pennisetum clandestinum ) have been planted as turf. On one boundarybetween turf and bushland a large infestation of the woody weed Geraldton Wax (Chamelaucium uncinatum )was removed. The site revegetation project was undertaken by the Friends of Blue Gum Lake who were contin-ually faced with Couch that kept creeping back from the grassed area. After spraying the Couch several timesan alternative solution was sought.

Creating a physical barrier by using kerbing, was an attractive option. Kerbing was multipurpose: It provided aneat and tidy mowing edge which pleased the local residents and council, prevented the incursion of mowersinto the bushland and also limited the spread of Couch. Several stories were circulating regarding the use of rubber conveyor belts buried into the ground vertically and used as barriers to stop rhizome and stolonspread. When a pre-used conveyor belt was donated by Wesfarmers, and the Lotteries Commission agreed toprovide funding for the above-ground kerbing, the project was underway.

An outline was created using a rope and marked with stakes. Making an early start, the City of MelvilleBushland Works Crew, an enthusiastic Work for the Dole crew and EWAN project officers dug the trench, layand buried the one metre wide conveyor belt. On several occasions there was grateful acknowledgment thatthe digging was in sand. In all, it took twelve people four hours. The only hiccups were several large Banksia roots, which were dug around, and the conveyor belt cut to fit. A contractor was brought in to do the kerbing,providing the finishing touch. With the donation of the conveyor belt, the labour or in-kind support, the projectcost $1164 for 100 metres of barrier – this was all in the kerbing contractor’s fees.

While the amenity grass already in the revegetation area still needs to be controlled, once it is eradicated itshould be an easier job to keep new incursions out!


22/116

16


Control techniques

Ann ua ls – pr event seed set and li mi t

seedl in g estab li shment

Highly disturbed areas are often dominated by annualgrasses. Continual removal by physical or chemicalmeans maintains the status quo. In degraded areas thecontrol of annual grasses over the longer term requiresthe establishment of desirable vegetation (Refer tocase study 5.1). However, undisturbed native plantcommunities are also susceptible to annual grassinvasion, such as the understorey of York Gum –Jamwoodlands.

To control annual grasses it is essential to exhaust thesoil seedbank and limit further seed recruitment. Themost effective way of doing this is to prevent seed set

or destroy seed before it is shed. The length of thecontrol program will depend on the seed longevitywithin the soil. The following approaches can besuccessful if thoughtfully adopted:

• Hand-weed small infestations occurring in goodcondition bushland. Remove weeds prior to seedset. Be aware that seeds can be spread during theweeding process and as contaminants of clothingand tools.

• Slash prior to seed set. Timing is important; slashplants after the flower head has emerged butbefore seed is dry and ready for release. If slashedtoo early, plants will flower again. Flower headsmust be bagged and removed as seed can continue

to develop after removal from the plant. Slashingcan be labour-intensive in bushland as it must becarefully done by hand to avoid damaging nativeplants.

• Spray germinating weed grasses at the three tofive leaf stage with the recommended rate of agrass selective herbicide. This approach may bedifficult to apply to species that have staggeredgermination. Spray such species four to six weeksafter rains begin to ensure maximum germinationhas occurred. Higher rates may be required and aneven coverage harder to obtain.

Research indicates Fusilade® does not impact

seriously on established perennial native grasses,although flowering is often inhibited. However,many native grass seedlings appear to besusceptible (Hobbs and Atkins 1988, Hitchmoughet al . 1994, Davies 1997, Arnold et al . 1998, Brownand Brooks unpublished data).

Note: For many years Fusila de ® has been the on ly grass selective

herbi cide registered for use in bushlan d in Western Australia a nd some ann ual gra ss species(Poa annua, Vulpia spp) are resistant to the ‘fops’ group of herbi cides to which it belo ngs. With these

probl

Bushland Weeds Book

Documents

Transcript of Bushland Weeds Book