I

REMNANT NATIVE VEG ETATION

TEN YEARS ONA DECADE OF RESEARCH

AND MANAGEMENT

Published by: Dr Syd Shea, Executive Director, Department of Conservation and Land Management,50 Hayman Road, Como, Western Australia 6152.Managing Editor: Ray BaileyEditor: K.J. WallaceIlluskation: Kellee Merritt

o 1995 ISBN 0 7309 6174 5

This book is copyright. Apart from any fair dealing for the purpose of private study, research, criticism or revrew, aspermitted under the Copyright Act 1968, no part may be reproduced by any process without written permission.Enquiries should be made to the publisher.

@ DeOarLment of Conservation and Land Management

I would like particularly to acknowledge the work of Barb Kennington, David Mitchell and Anthony Sutton in

organising and managing many components of the workhop. Barb Kennington also assisted with the preparation of

manuscripts. Amanda Curtin prepared the references and assisted with editing. My thanks also to Marianne Lewis for

. proof reading a draft of the document

I am grateful to the following people who reviewed manuscripts published in these proceedin$s:

. Kenneth Atkins, Andrew Burbidge, Mike Fitzgerald, Kelly Gillen, Richard Hobbs, An$as Hopkins, Penny Hussey, Greg

i X.ighery, David Mitchell, Daryl Moncriefi Denis Saunders, Jeff Short, Anthony Sutton, and Roger Underwood.

ACKNOWLEDGMENTS

Ken Wallace

Seminar Convenor

CONTENTS

Setting th€ Scene - Research on Remnants during the Past Decade

Denis A. Saunders

Impact of Fire on Fauna in Remnant Vegetation- Research FindinSs and their Implications for Management

Gordon Friend

The Effect of Changing Hydrological Processes on Remnant Vegetation and Wetlands

R.J. George and D.J. McFarlane

Research into Remnant Management Issues: Nuts and Bolts and the Big Picture

Richard J. Hobbs

Gondwanan Botany: A Perspective on Remnant Manaliement in South-West Australia

Stephen D. Hopper

The Exotic Predator Problem

Jack Kinnear

The Use of Remnant Vegetation by N€ctarivorous Birds: Biodivetsity as a Model

for Management in the l{heatbelt of Western Auskalia

Robert J. Lambeck

Acquisition and Management of Conservation Reserves in Westem Australia

A.R. Main

Invertebrate Studies Performed through Curtin University

Jonathan Majer and Anne Brandenburg

MANAGERS' PERSPECTIVES

Sefting the Scene - Remnant Management during the Past DecadeK.J. Wallace

Remnant Management over the Last 10 Years - An Agricultural Viewpoint

G.R. Beeston

Managing My BushlandAlison Doley

l l

l5

2 l

25

29

33

37

4 l

5 l

i)l'

l t t

An Operational Perspective.,,........... .,.....,............ 6lSteve Gorton

The Dongolocking Corridors ProjectTerri Lloyd

A Farmer's P€rspective - Management of Remnant VegetationBob Thigg

65

b l

WORKSHOP REPORTS

Workshop Reports - IntroductionK.J. Wallace

How Do We Achieve Communication within and betr.teen Agencies?Ken Atkins and Helen Allison

How Do We Integrate Information into a Form Accessible for Land Managers and Researchers? ..,,........... ?zAnne Brandenburg and Greg Beeston

Ho$ Do We Int€grate Nature Conservation with Other Land Uses in the Context of theWestern Australian Wheatbelt? ...,...,Greg Keighery and David Mitchell

what Are the Philosophy and Goals for Managinp Remnant Vegetation within a Landscape context?..,...,.,. 8FDon McFarlane

How Do we Ensure that community Education and Extension Maximise Remnant conservation? ............ gTAnthony Sutton and David Bicknell

CONCLUSIONS

APPENDIX:List of Workshop Participants and their Addresses

81

89

lntroduction

K. J. WallaceDepanmenl of Conservationand Land Management,PO Box 1 00, Narrogin, WA, 6312

BACKGROUND TO SEMINARDuring the early 1980s, there were a number of changesthat had important implications for nature conservationin agricultural areas of Western Australia. The mostsignificant were:

.i. Sufficient rural land managers recognised theseriousness of land degradation to becomeinterested in taking action, and the landcaremovement was born in Western Australia, As ruralcommunities began to understand that clearing ofnatural vegetation was a major cause of landdegradation, many also began to accept thatbushland is ofvalue to agriculture. No longer was itconsidered merely a harbour for "vermin" and"noxious" species - useless, unproductivewasteland. Other factors also helped to changeattitudes. For example, recent generations offarmers were not, or were minimally, involved inclearing bushland to create farmland. Their experi-ence was not the same as that of older generationswho worked hard to convert bushland, the "enemy".

Barbara York Main (1993) discusses other factorsthat contributed to new, more positive attitudestowards bushland.

The Department of Fisheries and Wildlife establishedthe first rural-based management group for natureconseruation, at Pingelly in 1978. This was followedby a management team based in the KatanningDistrict in 1984. These two groups, now part of theDepartment of Conservation and Land Management(CALM), provided a core of rural-based landmanagers whose primarl concern was natureconservatlon.

The CSIRO Division of Wldlife and Ecology beganan intensive research program on natureconservation in the Wheatbelt in 1984. Theconjunction of this work u,ith research by theDepartment ofAgriculture, Water Authority ofWestern Australia, and CALM generated a largeamount of information useful for management.

Aside from work on nature conserration issues,research documenting hydrological changes in thelandscape, and their relationship to bushland andrevegetation, was vital in showing that solutions andtheoretical concepts must be developed at landscapeand regional scales.

* The seriousness and scale oflandcare and natureconservation issues led to greater recognition of theneed to manage across land ownership and tenureboundaries. (lnterestingly, the promotion of thisconcept in relation to nature conservation, by IanCrook, in about 1980 was probably too early foracceptance within the social and culturalenvironment of the time.)

Together, these changes stimulated a range ofimportant research and management actions. Researchactions, in particular, are summarised in a series ofexcellent workshops and conferences. For natureconsefration, these are best exemplified by the "Nature

Consenation" series and related workshops that beganwith "Nature Conservation: The Role of Remnants ofNative Vegetation", held at Busselton in 1985.

Although land managers and planners, as well asresearchers, were involved in these conferences, theprimary focus was on research activities. Whilerecognising the value of past research, Wheatbelt landmanagers within CALM, by the early 1990s, werebeginning to ask: "What are the actual implications ofth is research lor our day lo day managemenl?"Consequently, there were moves to convene a meetingof Wheatbelt managers and researchers to describe indetail the operational implications of the previousdecade of research work.

About the same time, during a multi-disciplinarymeeting at the CSIRO's office, Helena Valley, onintegration of research in the Wheatbelt, the desire tohold a meeting to discuss research in agricultural areaswas expressed.

Together, these two goals provided the impetus for aseminar at Dryandra in September 1993, aimed atdiscussing research and management in relation toremnants of native vegetation. Work from the previousdecade was to be a particular focus.

Specific objectives for the seminar were:

.:. to list significant research results for remnantmanagement;

.l to assess whether conclusions from research hadbeen implemented;

* to develop a list of actions to improve both therelevance of research and the implementation of itsIesults.

SEMINAR PROCESSTo achieve these objectives, it was decided to hold aresidential seminar at which a series of presentations byresearchers and managers would be followed byworkshops and plenary sessions. Those invited to theworkshop represented a cross-section of land managers(including farmers), researchers and administratorsworking with remnants of native vegetation in theWheatbelt. Thus, all those with a direct interest in theoperational management of remnant vegetation wererepresented.

Throughout the seminar, participants were asked tokeep in mind the following questions:

.t Over the last decade, what research results,including reviews and theoretical work, haveimportant implications for the management ofremnant vegetation in south-western Australia?

.i What do manapiers want from researchers?

.1. How well have managers implemented significantresearch findings?

* How well have researchers and managers liaised?

* How may research and management best beintegrated?

Apart from review presentations for research (DenisSaunders) and management (Ken Wallace), and aspecial presentation by Bert Main, presenters wereallowed only 10 minutes. Researchers were asked tospend about five minutes listing the management

implications oftheir research, and five minutes onwhether their research findings have beenimplemented. Conclusions were to include a listing ofthe three or four major issues which require resolutionin order to improve the management of remnantvegetation. Researchers were asked not to presentresearch data - that is, the seminar was intended toaddress the implications for management of pastresearch, and was not an opportunity to present newdata.

Managers were asked to spend five minutes discussingresearch results they have used in managing remnantvegetation, and a further five rninutes discussin6iwhythese actions did or did not work. Conclusions were toinclude a list of major issues which require resolutionin order to improve the management of remnantvegetation.

These constraints on presenters resulted in brief andpithy papers. They are not, nor were they intended tobe, detailed, formal papers.

Following the presentations, a plenary session was heldat which issues and workshop topics were developed.Participants then separated into workshop groups anddiscussed these topics. The workshop and plenarysessions are dealt with in detail on pages 69-88.

REFERENCEMain, B.Y, 1993. Social history and impact on thelandscape. pp. 23-58 in Reintegrating FragmentedLandscopes: Touards S ustainable Production andNature Conseruation, eds R.J. Hobbs and D.A. Saunders.Springer-Verlag, New York.

RESEARCHERS' PERSPECTIVES

o

Denis A. SaundersCSIRO, Division of Wildlife and Ecology, LMB 4, POMidland, WA, 6056

INTRODUCTIONAlthough I have been asked by Ken Wallace to reviewthe significant changes in research into remnants ofnative vegetation over the past decade, it is worthputting these changes into context by examining brieflythe changes in attitudes towards remnants over the pastthree or four decades. Early attitudes were usuallyhostile, because remnants were seen as harbours fornoxious weeds and vermin, particularly poison nativeplants, rabbits, and macropods (not necessarily in thatorder), or were regarded as wasteland covered withunsightly scrub. One only has to fly over the Wheatbeltand see how many remnants of native vegetation havegravel pits and rubbish dumps on them to gauge earlyattitudes towards remnants. These attitudes werereinforced by the feeling of many people that thelandscape was alien and did not resemble the sorts oflandscapes with which they felt most at home.

Barbara York Main (1993) discussed the effects ofsettlement by Europeans in the central Wheatbelt andpointed out the degrading consequences of thissettlement on the landscape and on remnants of nativevegetation. One simple and overlooked example was therole of domestic poultr) in changing native vegetation.Main wrote that "Farm fowls (as well as turkeys andother poultry) have undoubtedly contributed to thedestruction of the biological cohesiveness of theremnants associated with farmhouses".

Main also discussed the attitude at this period towardsthe land and its biota. This attitude is summarised inthe praise by Sutton (1952, in Main 1993) of the settlerswho conquered "forest wilderness" and turned "virgin

lands" into "well ordered farms and gardens". As Mainpointed out, by the 1960s the only ungrazed remnantsof native vegetation were the reserves set aside forpublic use, such as water reser.'res, townsites, and otherutilities, or a number on private property.

Then, in the 1970s, attitudes towards native biotachanged. Land degradation, the realisation that toomuch had been c leared in some areas. the s l ronginterest in native plants, the loss of native species, thestrong push for conservation through resen'res, and a

developing sense of "belonging to the landscape" allplayed a part in changing hostile attitudes towardsnative biota in general and remnants of nativevegetation in particular These changes wereinstrumental in a major push in the 1970s, by the WADepartment of Fisheries and Wildlife, to have set aside,for the consenration of flora and fauna, as many of thesignificant Crown reser.ues in the Wheatbelt as possible.This meant that many of the larger remnants becameflora and fauna reserves, or had conser'ration of floraand fauna added to their original gazetted purpose.

As part of this process in the early 1970s, theDepartment of Fisheries and Wldlife contracted the WAMuseum to carry out a series of flora and vertebratefauna surveys on 23 remnants of native vegetation theDepartment had recently acquired as conservationreserves, The aim of those surrteys was to assess theconser.ration importance of the remnants in theextensively cleared Wheatbelt of Western Australia. Theresults of the sur"ueys were published in a series ofpapers in Biological Conseruation, Australian Wildlifi:Research and Records ofthe Westem AustralianMuseum. The findings are interesting from bothscientific and management points ofview. The data theyprovided are stil l one of the best sets available onspecies area relationships in fragmented landscapes,and have been widely quoted in the internationalscientific literature during the debates on the theory ofisland biogeography, a theonr which has given little ofpractical importance to managers of conserration areas(Saunders el al 1991). While it does provide an idea ofhow many species of selected taxa one can expect tooccupy an area over l ime. i l does nol idenl i fu lhespecies involved, which is of much greater importanceand relevance to managers of conservation reserves,

The results of the sur"ueys by the Museum gave somereasons for managers to be optimistic about theimportance of the scattered conservation system in theWheatbelt for the conservation of much of theremaining vertebrat€ biota. For example, DarrylKitchener et al. (1980a, b) noted that remnants as smallas 30 ha have value as sanctuaries for lizards (althoughvaranids were not found on reserves under 272 ha) andspecific mammal species, and are valuable conservationareas for plants. They noted (1980a) that "although thehaphazardly spaced wheatbelt reserues are inadequateto preserre entire communities of large mammals and

apparently also of birds and snakes, we conclude thatthe reserve system is probably adequate to preserverepresentative lizard communities". They added (1980b)"that with careful management of the wheatbelt reservesystem most of the native animal species still extant inthis region should persist for periods within the timeframework considered by contemporaryconservationists ... (1 000 to 10 000 years)".

They also pointed out (1980b) that regardless of the sizeof the remnant, mammal species lost from particularremnants would alrnost certainly not re-establishthemselves by natural processes, because the isolationof remnants had disrupted movements which wouldfoster recolonisation in the event of local extiruation.Ki tchener et a l (1982)noted that the long- termpersistence of much ofthe avifauna of the Wheatbeltdepended on the 500 randomly scattered naturereserves, occupying 2.40/o of the area of the Wheatbelt.They were unable to indicate any loss of bird speciesfrom reser"ues during the period of land clearing in theWheatbelt (the 70 years prior to their surveys), whichthey took to imply that loss of species of birds fromreserves will be a slow process. They also pointed outthat small nature reserves in the Wheatbelt are ofvaluein the conservation of birds. They cited the example ofEast Yorkrakine Nature Reserve (north of Kellerberrin),which is 81 ha and during their surveys contained fourspecies of passerines ofvulnerable status, despite beingisolated from other native vegetation for at least 50years. They (1980b) did not believe that feral cats andfoxes would have as severe an impact on nativemammals in Wheatbelt remnants as cats have had onoffshore islands.

While research over the next decade was to changesome of these concepts dramatically, this early work bythe WA Museum provided the vital foundation ofinformation and ideas upon which new researchdeveloped the next generation of theories. AIso, ideasdeveloped by the Museum concerning issues such as therelationships between vegetation types and fauna, andthe separation ofthe bird fauna into various categoriesdepending on their tolerance of disturbance, urere ofreal value to land managers.

This early work had one other important value. Duringthe late 1970s and early 1980s, many people, includingsome land managers and academics, questioned the

value of smaller remnants ofvegetation. Whilepredictions in the Museum papers concerning viabilityof remnants proved optimistic, their work did confirmthe nature conservation values of small remnants andprovided a counter to those who argued that small areashad no value.

RESEARCH OVER THE PAST DECADEWhat has changed since the pioneering work of DarrylKitchener and his colleagues, and what is thesignificance ofthe changes from a management point ofview? In the first instance, remnants of nativevegetation have had their profile lifted markedly, with agreat deal of interest in them from researchorganisations, management agencies, agriculturaladvisers, funding agencies, and, most importantly, ruralcommunities, including farmers and landcare groups(see Saunders at a/. 1987, 1993; Saunders and Hobbsl99l; Hussey and Wallace 1993).

Research into remnants of native vegetation and theirecological role in the landscape over the past decade hasindicated that we do not have any cause for optimism.Several irrefutable facts of relevance to managementthat have come out of research in the recent past arethat remnants of native vegetation are degrading atrates that are measurable, and that species are stillbeing lost. Without active management applied over theentire landscape, remnants will continue to degradeuntil this relaxation phase results in species-pooreranimal and plant communities. Without major changesin management, there is no way that the collection ofremnants in our conservation system will conserve theremaining native biota of the Wheatbelt over the next100 years, let alone the I 000 to 10 000 year timeframethat Kitchener and his colleagues suggested. We cannotafford to treat remnants as islands. If we take thatapproach, we will ignore the major degrading processes,most ofwhich originate in the surrounding agriculturalmatrix,

Current State of RemnantsOne of the most worrying research reports I have readrecently is one written by Robert Lambeck and JeremyWallace (1993) on the assessment of the conservationvalue of remnants ofnative vegetation in the centralWheatbelt, using Landsat Thematic Mapper (TM)imagery. Their research showed that 70% of remainingnative vegetation in the study area is not typical of the

unmodified vegetation types that characterised thecentral Wheatbelt prior to settlement by Europeans. lnmany of the remnants they examined, particularly thesmaller ones, all of the vegetation within the remnantwas spectrally unlike any of the flora regarded asindicative of that found before settlement last century.They concluded that the conservation value of many ofthe remnants has been seriously jeopardised, and theyattributed the degradation to impacts of domesticlivestock, clearing, invasion by weeds, harvesting oftimber, mining of gravel and the dumping of rubbish(see also Arnold and Weeldenburg 1991). In theirsummary they stated that only 3% of the original pre-clearing landscape remains in what could be regardedas good condition. They added the important rider that"The probability of such a small component of thelandscape continuing to support the essentialecosystem processes that underpin regional biodiversity[is] extremely remote": I would have said it isimpossible.

Representation of RemnantsTypically, remnants are small, and none are largeenough to be driven by internal processes. All are nowdriven mainly by the ecological processes generated bythe surrounding agricultural matrix. Remnants do notrepresent the pre-clearing range of animal and plantassociations. The process of selection of land foragriculture and the process of fragmentation were notrandom. Plant communities were linked strongly to soiltypes, and because certain soil types were more suitablefor agriculture than others, those soil types are poorlyrepresented on remnants in agricultural areas. TheWheatbelt is no exception, and Graham Arnold andJohn Weeldenburg's (1991) study on the distributionand characteristics of native vegetation in the centralWheatbelt illustrates this point. For example, theyfound that the Merredin and Belka landforms, whichformerly were dominated by salmon gum woodlands,occupied about l67o and 5%, respectively. o[ lheIandscape, yet they occupied only 6% and 1%o ofremnants. On the other hand, rock outcrops occupiedonly 4%o of the area but represented 28%o of theremnants. In addition, they found that 77%o of remnantvegetation was privately owned. The implications ofthese results for management are serious and far-reaching. The soils regarded as indicative of gooda6lricultural land, and any associated uncleared biotaare poorly represented on conservation reserves.Woodlands are in this category, and if conservation of

woodlands is an aim of management, then private landmust be managed with that aim in mind. Private landnow contains much of what must be regarded as ourconservation estate. One of our challenges is to comeup with ways to ensure that part of the conservationestate is managed with conservation as the primaryfunction.

Native Vegetation Is ResistantResearch has shown that native vegetation is resistantto invasion by exotic vegetation, provided that theremnant vegetation is not disturbed or enriched withnutrients (Hobbs and Atkins 1988). Unfortunately, theagricultural matrix surrounding most remnantvegetation results in a wide range of disturbances andconsiderable nutrient enrichment. Ignoring the obviousand major disturbance of grazing by domestic livestock,and enrichment by fertiliser drift, how many times dowe see dead sheep disposed of by throwing the carcasesinto the bush? As the carcases decompose, nutrients arereleased and scavenging animals scraping around thecarcases disturb the area. Any seeds ofweed speciescaught in the wool have an ideal bed on which toestablish themselves. Similarly, the vegetation scrapedfrom the edges of roads or firebreaks is usually piled inthe bush and left. These piles, like the rotting caycases,are major foci ofweed invasion along road verges andthrough many other remnants of native vegetation.

Remnants Are Influenced by the Surrounding Matrixand Have an Effect on the Surrounding MatrixIt is well known that the extensive clearing of nativevegetation in the Wheatbelt has resulted in majorchanges in the hydrological balance, leading towidespread salination of susceptible areas. Nearly 20%of all cleared agricultural land may be useless for cerealcropping within the next 30 years because of increasesin salt levels in the soil (Nulsen 1993). Ecologicalprocesses do not stop at legal boundaries, and remnantsof native vegetation are being affected by increasing soilsalinity. Position in the landscape does not necessarilyguarantee immunity from degradation. For example,Durokoppin Nature Reserve is I 100 ha and located inthe highest point of the landscape at the top of twocatchments. Because of the higher water use byremnant vegetation than by agricultural vegetation,watertables are up to 7 m lower under the reserve thanunder nearby agricultural land (McFarlane el a/. 1993).Nonetheless, watertables are rising under the reserve as

water flows into the "hydrological shadow" and, if thiscontinues, will probably reach the root level of thesalmon gum woodland in the lower part of the reserve,with further loss ofan already severely restrictedvegetation type. At present, we do not know how manyother remnants are threatened in this way. However,George and McFarlane (this volume) believe that asignificant area of our conservation estate is underthreat.

Remnants Are Still Degrading and Losing SpeciesThe losses of native mammals from the Wheatbelt arewell known (Burbidge and McKenzie 1989) and stil lcontinuing (Hobbs. el aL 1993a). Birds aredemonstrating the same trends as the mammals. Forexample, of 192 species of birds recorded from theWheatbelt since 1900, 96 species (5070) have decreasedin range and/or abundance, and only nine (5%o) haveincreased (Saunders 1993). Species are stil l being lostfrom the region, from districts and individual remnants(Saunders 1989). Kitchener el al. (1,982) cited EastYorkrakine Nature Reserve as an example of theconservation value of small remnants for speciesdependent on native vegetation. In the period betweenthese researchers' surveys in 1974 and the surveys bythe CSIRO in 1988, three of the four species the formernoted the reserve was ofvalue for became extinct. Inthe Wheatbelt, a number ofspecies dependent on nativevegetation are located on remnants of native vegetationin populations which are too small to be viable. Inaddition, those populations are isolated from other suchpopulations, and they are gradually becoming extinctbecause of a range of stochastic events. In this situation,the effects of the fox and cat on these isolatedpopulations are much more severe than Kitchener andhis colleagues predicted. The work of Jack Kinnear ela/. (1988) and Tony Friend (1990) on relict populationsof endangered marsupials are good examples ofwhat westand to lose if we fail to control foxes and catsthroughout the Wheatbelt.

Rernnants Must Be Managed in a Total LandscapeCont€xtThe take-home message regarding the management ofremnants of native vegetation in the Wheatbelt must bethat they cannot be managed in isolation. Anymanagement which concentrates only on individualremnants, particularly those designated as conservationreserves, and ignores the surrounding agricultural

matrix is doomed to failure and will continue thedegrading processes which are leading to the loss of ourunique biota. We need to get away from the singlespecies-single remnant approach to management andconcentrate on developing an integrated landscapeapproach which involves all of the people and groupsengaged in management of elements of the Wheatbeltlandscape (Hobbs and Saunders 1991, 1993; Hobbs ela11993b). We need to work with all of these individualsand groups, to integrate our knowledge and use it toconstruct management models which aim to protectour remaining biological heritage within production-oriented agricultural landscapes. The aim of thisworkshop should be to come up with ways to assist usachieve that goal.

REFERENCESArnold, G.W., and Weeldenburg, J.R., 1991.The distributions and characteristics of remnantnative vegetation in parts ofthe Kellerberrin,Tammin, Trayning and Wyalkatchem Shires ofWestern Australia. Technical Memorandum No. 33.CSIRO Division of Wldlife and Ecology, Canberra.

Burbidge, A.A., and McKenzie, N.L., 1989.Patterns in the modern decline of Western Australia'svertebrate fauna: Causes and conservation implications.Biological Conseruation 50: 143-798.

Friend, J.A., 1990. The numbal Mgrmecobius fasciatus(Myrmecobiidae): History of decline and potential forrecovery. Proceedings ofthe Ecological Societg ofAustralia 16:369-377.

Hobbs, R.J., and Atkins, L., 1988. Effects of disturbanceand nutrient addition on native and introduced annualsin plant communities in the Western AustralianWheatbelt. Australian Joumal of Ecologg 13: 171-179.

Hobbs, R J., and Saunders, D.A., 1991. Reintegratingfragmented landscapes, a preliminary framework for theWestern Australian Whealbelt. Joumal ofEnuironment al Management 33: 761-167.

Hobbs, R.J., and Saunders, D.A. (eds), 1993.Reintegrating Fragmented Landscapes : TouardsSustainab Ie Product ion and Nature Corueruation.Springer-Verlag, New York.

Hobbs, R.1., Saunders, D.A., Lobry de Bruyn, L.A., andMain, A.R., 1993a. Changes in biota. pp. 65-106 inReintegrating Fragmented Landscapes : TowardsSustainable Production and Nature Consenation, edsR.J. Hobbs and D.A. Saunders. Springer-Verlag, NewYork.

Hobbs, R.J., Saunders, D.A., and Arnold, C.W., 1993b.Integrated landscape ecology: A Western Australianperspective. Biological Consenation 64: 237-2J8.

Hussey, B.M.J., and Wallace, K.J., 1993. Managing yourBushland. Department of Conservation and LandManagement, Perth.

Kinnear, J.8., Onus, M.L., and Bromilow, R.N., 1988.Fox control and rock-wallaby population dynamics.Australian Wildlife Research l5: 435-450.

Kitchener, D.J., Chapman, A., Dell, J., Muir, B.G., andPalmer, M., 1980a. Lizard assemblage and reserve sizeand structure in the Western Australian Wheatbelt -some implications for consewation. BiologicalConseru at ion 17 : 25-62.

Kitchener, D.J., Chapman, A., Muir, B.G., and Palmer,M., 1980b. The conser.uation value for mammals ofreserves in the Western Australian Wheatbelt.Biological Conseruation 18: 77 9-207 -

Kitchener, D.J., Dell, J., Muir, B.G., and Palmer, M.,1982. Birds in Western Australian Wheatbelt reserves -implications for consewation. Biological Corseruation22: 127-163.

Lambeck, R.J., and Wallace, J.Fi., 1993. Assessment ofthe conservation value of remnant native vegetation inthe central Wheatbelt of Western Australia usingt andsat TM imagery. Report for Australian NationalParks and Wildlife Service "Save the Bush', Program.Cuftin University of Technology, Perth.

Main, B.Y, 1993. Social history and impact on thelandscape. pp. 23-58 in Reintegrating FrogmentedLandscapes: Touards Sustainable Production andNature Consenation, eds R.J. Hobbs and D.A. Saunders.Springer-Verlag, New York.

McFarlane, D.J., George, R.J., and Farrington, p, 1993.Changes in hydrologic balance. pp. 146-186 inReintegrating Fragmented Landscapes : TouardsSustainable Production and Nature Conseruation, ed.sR.J. Hobbs and D.A. Saunders. Springer-Verlag, NewYork.

Nulsen, R.A., 1993. Changes in soil properties. pp. l0Z145 in Reintegrating Fragmented Landscapes: TowardsSustainable Production and Nature Consenation, edsR.J. Hobbs and D.A. Saunders. Springer-Verlag, NewYork.

Saunders, D.A., 1989. Changes in the avifauna of aregion, district and remnant as a result offragmentation of native vegetation: The Wheatbelt ofWestern Australia. A case study. BiologicalConseruation 50: 99-135.

Saunders, D.A., 1993. A community based observerscheme to assess avian responses to habitat reductionand fragmentation in south-western Australia"Biological Conseruation 64: 203-218.

Saunders, D.A., Arnold, G.W., Burbidge, A.A., andHopkins, A.J.M. (eds),1987. Nature Corceruation: TheRole of Remnants ofNatiue Vegetation Surrey Beattyand Sons, Chipping Norton (NSW).

Saunders, D.A., and Hobbs, R.J. (eds) , 1991. NatureConseruation 2. The Role of Conidors. Surrey Beattyand Sons, Chipping Norton (NSW).

Saunders, D.A., Hobbs, R J., and Ehrlich, P.R. (eds),L993. Nature Conseruation 3: Reconstruction ofFragmented Ecosgslens. Surrey Beatty and Sons,Chipping Norton (NSW).

Saunders, D.A., Hobbs, R.J., and Margules, C.R., 1991.Biological consequences of ecosystem fragmentation: Ar eview. Conseruat ion Biologg 5; 18-32.

Gordon FriendWildlife Research Centre, Departmentof Conservation and Land Management,PO Box 5.1, Wanneroo, WA, 6065

INTRODUCTIONThe following is a summary of the major findings fromongoing projects to research the impacts ofexperimental fires on small vertebrates andinvertebrates inhabiting remnant shrubland andwoodland vegetation in the Wheatbelt and South Coastregions of Western Australia. These studies have focusedon Tutanning, Durokoppin and East Yorkrakine NatureReserves in the Wheatbelt over the past seven years(1987-1993), and on the Stirling Range National parkbetween 1989 and 1992. For further detail on theseareas, study design, and results, the reader is referred toFrtend et al. (1989), Friend (1993), Friend and Williams(1993), Little and Friend (1993), and Strehlow (1993).

In addressing the primary aims put forward for theremnant vegetation workshop at Dryandra from theperspective ofa research worker involved in this area,I have adopted a quite specific approach which refers tothe first major objective: "To list, based on work in theSouth West over the last decade, significant researchresults and their implications for remnantmanagement".

RESEARCH FINDINGSTen major research findings and their implications arelisted in point form below, followed by theirimplications for management.

l. Research finding: The smallvertebrate andinvertebrate fauna inhabiting semi arid woodlandsand shrublands in Western Australia appears to berelatively resilient to single fire events of small scaleand low to moderate intensity (but see point ?). Theimpact of repeated fires, however, is unknown.

Management implicalrbru.. Single fires may not be agreat threat to fauna provided they are infrequentand of small scale relative to remnant size. It isessential that whole reser.ues are not burnt in oneevent, especially by a high intensity fire.

2. Research finding: Resilience may be greater in the

seasonally dry shrublands and woodlands than in themore mesic but Iess seasonal habitats (for example,South West forests). Adaptations for seasonal ariditymay impart considerable resilience to fire. However,despite this, fauna populations may change greatlyunder a regime of frequent fires. There are fire-sensitive species and habitats within the regionsstudied (for exampl e, Phascogale calura,Allocasuarina, mygalomorph spiders) that requirespecial consideration.

Management imp I icallor2s.. Managementprescriptions developed for mesic areas (forexample, jarrah forest, etc.) do not necessarily applyto semi-arid ecosystems. Some details of life historyand habitat preferences are needed for all species, inorder to decide appropriate managementprescriptions (see also point 5).

Research finding: Frequency and scale of burns areprobably the two most important factors to considerin fire management. There is no evidence from ourstudies (for example, in Stirling Range NationalPark, where both spring and autumn experimentalburns have been conducted) that season of burn is ofmajor significance, at least to invertebrates.

Management implicalronsr Burning should not becarried out too often, or on too large a scale (seealso points 6 and 10).

Research finding: Post fire response patterns ofmost smallvertebrate groups are closely tied to(and may be predicted from) their shelter, food andbr€eding requirements (life history parameters).

These patterns are as follows:.:. mammal responses are reasonably predictable;* reptiles somewhat less;. l amphib ia show l i l t le re la l ionship.These patterns are derived from large-scale wildfiresthroughout temperate Australia. Data from oursmaller scale and lower intensity mallee-heath fires,however, have not shown such clear trends for smallmammals and reptiles.

Management implicalrbrzs: A model to predict thebroad impact of fire on small vertebrates is thus

3 .

7.

possible, and indeed has been developed. A databaseof species' life history characteristics and probablepost-fire response patterns can be linked to such amodel. This needs further input of data and needs tobe made accessible to managers.

Research finding: Given the above, fauna can beconsidered and grouped in terms of Iife historycategories based on shelter and food requirements.Thus we do not have to worry about every species(for example, the 50 plus species in our Wheatbeltstudies reduce to 16 life history categories).

Management implicallons: The use of Iife historycategories greatly simplifies the amount ofinformation that managers need to consider whendeveloping fire management plans.

Research finding: Results from space-for-timestudies of potentially sensitive species in the StirlingRange National Park (for example, honey possum,Tarsipes rostratus) indicate clear trends with post-fire age of vegetation.

Management implicalroru. These data suggest aminimum time between burns in the Stirling RangeNational Park of l5 ,20 years {allowing someleeway). In the lower rainfall areas of the Wheatbeltregion, where vegetative growth is slower, theminimum is likely to be 20-25 years.

Research linding: For invertebrates, the level oftaxonomic resolulion (lhat is. orders versus specieslevel identifications) influences the results of studieson fire impact.

Management implicalrons: One cannot assume thatall is well if looking at broad-scale taxonomicgroupings, because responses are species-specific.Individual species may be markedly affected by fire(both increases and decreases), but at the order levelof identification these trends will tend to cancel outand thus not be apparent.

mygalomorph spiders) set the limits for fire regimes.

Management implicaftbns.' Invertebrates may provebetter indicators of seral status and appropriate fireregimes than higher organisms. Invertebratesshould be included in fire ecology studies and,indeed, in biological surveys in general.

9. Research finding: Invertebrate abundance andcomposition do not correlate well with floristics orvegetation structure. Patterns which may exist are ata fine level of resolution.

Management implicaltons. Categorising andprotecting areas on the basis of plant and vertebratespecies richness does noI necessarily ensureadequate conservation of invertebrates.

L0.Research finding: With respect to animal abundanceand composition, the synergiistic effects of season,locality and year-to-year variability in climate, andstochastic events like droughts and locust plaguesgenerally outweigh any changes attributable to firealone. An excellent example of such synergisticimpacts is the post-fire locust plague whicheliminated Allocasuarma from an area burnt atTutanning. This may lead to long-term impacts onstand structure and faunal abundance andcomposition, simply because these two even$happened in tandem.

Management implicaltor?s.' Fire managers need totake account of pre-fire conditions (for example, isdrought in force or imminent?) before burning, butmany post-fire events and outcomes are beyond themanagers' control.

CONCLUSIONGiven the importance of season, climate and stochasticevents in determining species abundance anddistribution patterns, the need for and use of fire as aroutine management tool to maintajn or increasefaunal diversity needs to be carefully evaluated. At thisstage in the development of our knowledge base, it isclearly better to err on the side of conservatism informulating any fire regime for remnant shrubland andwoodland habitats. As a general rule, larger scale blockburning should not be carried out except in special

8. Research finding: Particular groups of invefebrateswhich are longJived and have special habitatrequirements seem especially sensitive to fire. Such"indicator" species or groups (for example,

circumstances - for example, for specific regenerationpurposes, experimental research or where it contributesto a well-considered strategic fire managementobjective. Protection ofareas from large-scale, high-intensity wildfires through a system of internal andexternal low fuel zones should remain a high priorityfor managers of remnant vegetation.

REFERENCESFriend, G.R., 1993. Impact of fire on small vertebratesin mallee woodlands and heathlands of temperateAustralia: A review. Biological Conseruation 65l-99-114.

Friend, G.R., and Williarns, M.R., 1993. Fire andinvertebrate conservation in mallee-heath remnants.Final Report, Project P 144, World Wide Fund forNature Australia.

Friend, G.R., Smith, G.T, Mitchell, D.S., and Dickman,C.R., 1989. Influence of pitfall and drift fence design oncapture rates of small vertebrates in semiarid habitats ofWestern Australia. .4 ustralian Wildlife Research 16:1-10.

Little, S.J., and Friend, G.R., 1993. Structure ofinvertebrate communities in relation to fire history ofkwongan vegetation at Tutanning Nature Reserve.CALMScience l:3-18.

Strehlow, K.H., 1993. Impact of fire on spidercommunities inhabiting semi-arid shrublands inWestern Australia's Wheatbelt. BSc Honours thesis,Murdoch UniversiW.

The Hffect of Ghanging Hydrologice[Processes on Remnant Vegetatlonand Wetlands

R.J. Georgel and D.J. McFarlane2lDepartment of Agriculiure, PO Box 1231 , BunburywA,6231'?Department of Agriculture, 1 20 Albany Highway,Albany, WA, 6330

INTRODUCTIONRemnant vegetation and wetlands on both private andpublic land are rapidly being degraded by drylandsalinity, inundation, silting, nutrient enrichment andweed invasion. Remnants in the lower pads ofcatchments are most affected. Many carnot be savedwithout expensive and integrated programs to reducethe degradation of agricultural catchments in whichthey are located. There is an urgent need to identifuthose remnants which have the highest values and forwhich cost-effective recovery plans can be developed. Inmost cases, it is not possible to manage these remnantsin isolation from the surrounding catchment, andsociety will need to contribute to the cost of catchmentmanagement if the nature conservation values of theseremnants are to be retained. Innovative methods needto be developed which maintain the natureconservation value of the remnants at the same time asincreasing the soil conservation and farm productionbenefits that they provide for farmers. It is only whenthe on-farm remnants are seen as being important forland-holders that they will be properly managed.

This review concentrates on the hydrologic benefitsafforded by remnants and the threat posed byhydrological forms of degradation. We have opted to usethe case study approach to highlight the major issuesaffecting remnants in the Wheatbelt. Our case studiesare mainly from our current and past researchactivities. A more detailed account of our research ispresented elsewhere (George el a/. in press).

HYDROLOGICAL PROCESSES INREMNANT VEGETATIONSurface Hydrolog5rMallees have been shown to harvest water with theirvaseJike branches, directing stem flow to their bases,where it infiltrates deeply beside roots (Nulsen el al1986). This practice reduces soil evaporation andprovides the plants with water during the dry summerand autumn period. Salmon gums (Eucalgptussalmonophloia) and gimlets (E. salubris\ also appear to

harvest water, although no measurements have beenmade.

Water harvesting also appears to be occurringdifferentially in open areas between trees in theWheatbelt. Nulsen el al (1986) recorded up to 7.7 mmof overland flow coming from 4 m, plots in a bushcatchment during a 30.9 mm storm. However, therewas almost no runoff recorded from the catchmentfurther downstream. Local redistribution of wateramong the vegetation accounted for the lack of streamflow In contrast, organic crusts were suspected ofcausing low infiltration and sorptivity in four Wheatbeltsoils within bushland in comparison with adjacent soilswithin cleared catchments (McFarlane et aL.7992a\.Gravelly soils were also more water repellent beforeclearing. Soils in undisturbed remnants have low levelsof the radioactive fallout product caesium-137, furtherindicating that there is considerable local redistributionofwater in Wheatbelt remnants (McFarlane el a/.1992b). Infiltration must be concentrated at selectedpoints, as little runoff is observed leaving remnantvegetation.

The fate of macropores formed by tree roots afterclearing is poorly known. However, anecdotal evidencesuggests that they become clogged or sealed, resultingin increased waterlogging of surface soils. Some remainopen, as evidenced by the rapid responses by somewatertables to rainfall (Engel et al. 1989; George et al.1991) .

Native C4 grasses have different water use patterns fromintroduced (predominantly annual) C3 plants. Prior toagricultural development, soils were wetter in springand drier in autumn (Johnston 1993). Advantages ofthenative grasses include deeper rooting, greater wateruse, lower nutrient requirements and betterdistribution of feed throughout the year.

Groundwater HydrologYThere are numerous examples of groundwater levelsrising after clearing of native vegetation in theWheatbelt. Generally, water levels are rising by between0.2 and 0.5 m per year in the above 500 mm rainfallzone, and by between 0.05 and 0.2 m per year in thebelow 500 mm rainfall zone (George 1992). In somehigher rainfall areas, watertables have risen over 25 msince land clearing which occurred less than 80 yearsago.

Recharge occurs throughout most of the landscape(Ceorge et al. 1991). Processes may differ depending onthe soil type, and include:

.:. matrix recharge in soils with low water,holdingcapacities (for example, deep sands and gravels);

matrix recharge in arkosic sands below graniteoutcrops;

preferred pathway recharge in duplex soils onhillsides, and below sandplain seep discharges;

.i. preferred pathway and matrix recharge in inundatedvalleys after storms.

The proportion and total contribution of each form ofrecharge depend on the distribution of soils andlandforms in each catchment. No study has accuratelydetermined the relative contributions. From amanagement viewpoint, it makes most sense to reducethe recharge in that component which is most cost,effective or which offers other benefits (for example,drying out sandplain seeps, revegetating low-productivity rock outcrops, and draining duplex soilsprone to waterlogging).

EFFECTS OF REMNANT VEGETATIONON LAND DEGRADATIONSalinityTransects of boreholes were drilled into Durokoppin andKodj Kodjin Nature Reserves, to study the impact ofnative vegetation on dryland salinity in the adjoiningicatchments (McFarlane el aL 1992c). The watertables inboth reserves are up to 7 m lower than in equivalentlandscape positions in the adjoining cleared areas(Figure 1). Salinity affected only 0.170 ofthe farmlandin the agricultural catchment that contains thereserves. In contrast, over 2.8%o of the farmland in anextensively cleared adjoining catchment was salt-affected. Recent monitoring within the reservesindicates that watertables are rising at about 0.15 m peryear, suggesting that groundwater is moving into thereserves from surrounding land. High groundwaterlevels under cleared areas next to the DurokoppinNature Reserve are beginning to affect the south-western and south-eastern parts of the reserve,However, both Durokoppin and Kodj Kodjin Nature

Reserves appear to be safe from major salinisation, asthe reserves occupy most of their catchments.

EutrophicationDense native vegetation around streamlines is seen asone of the most effective means whereby phosphateattached to suspended sediment can be removed fromrunoff before it enters estuaries (David Weaver pers.comm.). Weaver (1991) noted that up to 60%o of thetotal phosphate washed into the Kalgan River in 1991was attached to mobile sediment.

Figure l. Lower groundwater levels under nativevegetation in nature reserves from George e/ a/.in oress.

A. DurokoppinNature Reserve

B. Kodj Kodjin Nature Reserve

--i.'* *T-*-i xm." .

Figure 2. Locations referred to in the text, from George €f a/, in press.

A YFNNU 3 EEEKEEPERS

^ D E P A R T M E N T O F A G R C I J L T U F EItf[\

a"",{,ffiF^iiiE3@

B I B B Y S P F I N G S

NAMBUNG NAIIONAL PAFK- Larqe Femnanh othe Oie na

Forcsl Cover

) "n**-a c'*.-, *-*o

LAND DEGRADATION AND REMNANTVEGETATIONSalinity - Public LandOlsen and Skitmore (1991) reviewed the effect ofsalinity, erosion, sedimentation, eutrophication andweed invasion on the state ofthe rivers in WesternAustralia. Some of these rivers form corridors of publicland through farmland (for example, Kalgan River),whereas others are privately owned. Few rivers arefenced to protect them from stock, although there areprojects under way to fence some riverine vegetation -

for example, pads of the Denmark, Avon and Kalganrivers,

The effects of high groundwater levels under clearedareas next to the Durokoppin Nature Reserve aredescribed in the previous section. This case study is anexample of the hydrological relationships that may beexpected for remnants high in the landscape, at least inthe eastern Wheatbelt. Effects lower in the landscapehave been, and will continue to be, more dramatic.

Toolibin Lake is located in the middle of a broad valleyin a catchment which is 92%o cleared of nativevegetation. Saline runoff and rising groundwatersthreaten the lake and its reserves (Stokes and Sheridan

1985; McFarlane el al 1989). Increased frequency ofinundation is thought to be contributing to the death offringing species (Mattiske 1982). Active inter.uention isbeing undertaken, and both a technical advisory groupand a recovery team have been appointed to begin thetask of protecting the lake and its flora and fauna(Toolibin Lake Recovery Team and Toolibin LakeTechnical Advisory Group 1994). Works to by-passhighly saline surface flows around the Lake have beencompleted in conjunction with catchment and reservedrainage schemes. Groundwater pumping and otheractions are planned. It is likely that the only way topreserve the Toolibin environment will be to isolate itfrom the catchment in order to prevent saline watersfrom entering it on all but very wet years. Issues raisedby work at Toolibin are as follows:

.i. Toolibin and its reserves will become badly degradedunless protected from saline groundwaters andsaline runofl

* The ecologicalvalue ofthe lakes needs to be clearlyenunciated by both the local communities and themanaglng agency.

.i. Protection carries large implementation andrunnlng costs.

.i. Restoration ofthe pre-clearing environment maynever be possible. Moreover, the preservation ofremnant ecosystems appears impossible inWheatbelt landscapes. Conservation of selectedcomponents may be all that is practical.

There are many other examples of threatened naturereserues. Capercup Nature Reserve is a wandoo (O.uandoo) woodland,located 5 km west of Duranillin.Recent drilling has shown that the reserve has verysaline groundwaters (2 000- 6 000 mS/m) within 2 m ofthe surface on its western side. The groundwater is asclose, but much fresher (300 mS/m), on its eastern side.Most of the adjoining catchment has been cleared in thelast 15 to 30 years, and groundwater levels are believedto be rising by between 0.3 and 0.5 m per year It islikely that the western side will begin dying in the nextfew years, and about 5070 of the reserve will be affectedwithin the next decade.

Kulikup Lake, another nature reserve, is located 15 kmeast of Boyup Brook. It is covered by reeds and fringedwith paperbarks, wandoo and flooded gums. At present,the lake appears to be safe from salinity for about 20years, as the watertable is over 9 m below the surfaceand has a moderate salinity (450 mS/m). TheDepartment of Conservation and Land Management(CALM) has monitored the surface water in the lake forover a decade. We believe groundwater monitoring isessential in all nature reserves.

The Qualeup lakes and associated nature reserves arelocated 25 km west of Kojonup, and range in cover fromreed beds and paperback swamps, to wandoo andflooded gum (4. rudrs) woodlands. In some cases, thelakes already have saline springs emerging within them(Wardles Bush Lake - not a reserve); in others, thewatertable is still 4-11 m below the lake's floor (WardlesGrassy Lake, catchment 60%o cleared). At Qualeup Lakeitself, the watertable is near the surface on the westernside and about 4 m below on its eastern side. The watersare saline (500-2 000 mS/m). Qualeup Lake will beginto deteriorate in the next five years

Groundwaters responsible for salinisation in theCapercup, Kulikup and Qualeup reserves are difficult tomanage, as they are often contained in fault zones orsmall artesian fluvio-lacustrine deposits. The lakes andwetlands contain some species of native, perennial

shrubs and grasses that may have agricultural value.

Issues that this section has raised are as follows:

.:. CALM and associated agencies must begin aprogram of risk evaluation for the reserves theymanaEe.

.1. Drilling observation wells in threatened reservesshould be a major target for the next five years.

* The reality is that many reserves may die beforemanagement plans are implemented. This should benoted from the risk evaluation program.

* Programs to preserve the genetic diversity ofthreatened reserves should be commenced (forexample, seed collection schemes).

Salinity - Remnants on Private LandSaline groundwaters are rising around the PallinupRiver as a result of land clearing. Saline seeps aroundthe river are affecting native vegetation on both privateand public land. Natural drainage in the Wellstead areais into yate (E. occidentalis) swamps that were perchedabove the regional groundwater system. Rates ofllroundwater rise of about 0.3 m per year have beenreported from the western south coast over the last 19years (McFarlane 1992), and many lakes are nowbecoming salt-affected.

Yate swamps throughout the south coast are undersimilar threat.

In some places, the yate swamps are the major form ofremnant vegetation. Their loss would greatly reduce thehabitat for fauna in the region. Unfortunately, it will bedifficult to prevent groundwaters from rising in theseareas without major changes to farming systems. Cropsand pastures are affected by waterlogging, and the mainoutlets for drainage of this land are the swamps. Therehave been no definitive studies of the long-term costsand benefits of drainage practices.

Valleys containing wandoo remnants north-west andsouth-west of the Qualeup lakes are beginning tobecome saline, and swampy areas have begun to die.Hillside seeps have developed above lakes on manyproperties, and saline water is collecting in them.

Gibbs Swamp is a small lake on private land about15 km west of Boyup Brook. The lake is covered byflooded gum and paperbark communities. The localcatchment group (Boree Gully) has successfullyestablished local vegetation near the swamp. Drillinghas shown that both the deep and shallow aquifers arehighly saline (1 200-2 000 mS/m) and that the shallowwatertable may already be adversely affecting theswamp. Monitoring has been started, to evaluatewhether the revegetation will be successful.

Drilling under wandoo vegetation on the Boyup BrookGolf Course located very saline groundwaters (7 000mS/m) which are discharging from a clayey regolith. Itis sobering to note that the golf course design,essentially based on 'alleys', with 30-40% of the lowercatchment under woody perennial vegetation, rsinadequate to prevent salinisation and death of thewandoo trees.

On properties near Duranillin, large areas ofvalleyremnants of wandoo and associated species have beenfenced from stock. However, rising watertables threatenmany of these remnants.

Many lakes and wetlands in the Unicup area are alsobeginning to degrade as a result of clearing in the 1970sand 1980s. Saline waters flowing from the Unicup lakesand surrounding farmland threaten Kodjinup Lake andits reserve, and the Buranganup Plain wetlands. Onsome farms, lakes are beginning to fill withSiroundwater, and overflow, their highly saline waters(2 000-6 000 mS/m) causing pollution. This is alsohappening in saline lakes within some reserves (forexample, Pindicup Lake). Both these lakes overflow intoadjoining reserves.

The points that these investigations have raised are asfollows:

.!. Farmers who have fenced and protected valleyremnants may not gain much value from theirinvestment. In particular, wandoo woodlands andyate swamps that occur in valleys are under threat.

.l Drainagle may be essential to maintain theproductivity of farmland and to preserve the lakesand wetlands.

.t In the Unicup area, lakes on both farmland andreserves are filling, and saline water is beginning toaffect other reserves downstream. Active drainagiesystems are required urgently.

RECOMMENDED ACTIONS* The beneficial and detrimental effects ofagricultural

drainage on the long-term viability of remnants inrepresentative areas of the South West need to beestablished. Studies which concentrate on the short-term effects of drainage on the remnant and ignorewhat is happening in the rest of the catchment arelikely to be misleading.

* There is an urgent need to identify those remnantsthat have the highest values and for which cost-effective recovery plans can be developed.Croundwater monitoring is needed, to identify areasthat are at risk and may be helped by earlyintel"rention.

Understanding the water relations of nativevegetation may be essential for its management andrehabilitation, particularly ifwater harvesting is asimportant as it appears in undisturbed remnants.Understanding how native plants manage water mayalso enable agricultural systems to be developedwhich result in less degradation. For example,understanding how moort (8. platgpus) grows sowell in sodic grey clays may help the management ofthese problem soils.

.i. Forest and woodland ecologists should becomemuch more involved in the development ofproductive agricultural systems that will reduce thenegative impacts of the current systems onremnants.

CONCLUSIONSDeveloping economic products from our remnantstands of native vegetation must become a prime rolefor the new generation of researchers. There is little orno practical use in studying plants and animals that arethreatened if a similar amount of time is not put intodeveloping solutions to eradicate the factors leading totheir demise. If we fail in this task, within our lifetimewe will probably lose over 80% of remnant ecosystemson private land and as much as 50% within publicreserves.

RE FERENCES Australian soils.,4us tralian Joumal of Soil Research J0Engel, R., McFarlane, D.J., and Street, c., 1989. Using 577 -532

Seophysics to define recharge and discharge areasassociated with saline seeps in southwestern Australia. McFarlane, D.J., Loughran, R.J., and Campbell, B.L.,pp. 25-39 tn Groundwater Recharge, ed,. MJ. Sharma. 1992b. Soil erosion of agricultural land in WesternA.A. Balkema Publishing Co., Rotterdam. Australia estimated by caesium-l37. Australian Joumal

of Soil Research 30: 533-546.

George, R.J., 1992. Hydraulic properties of groundwatersystems in the saprolite and sediments ofthe Wheatbelt, McFarlane, D.J., George, R.J., and Farrington, P, 1992c.Western Australia. -/oumal of Hgdrology 130].250-278. Changes in the hydrologic cycle. pp 146-186 in

Reintegrating Fragmented Landscapes, eds R.J. HobbsGeorge, R.J., McFarlane, D.J., and Lewis, M.R, 1991. A and D'A' Saunders Springer-Verlag, New York'review of recharge and implications for management insaline agricultural catchments, Western Australia. pp Nulsen, R.A., Bligh, K.1., Baxter, I.N., Solin, E.J., and193-197 in Proceedings Intemational Hgdrologg and Imrie, D.H., 1986. The state ofrainfall in a mallee andWater Resources Symposium, Perth. heath vegetated catchment in southern Western

Australia. Australian Joumal of Ecologg I l: 361-321.George, R.J., McFarlane, D.J., and Speed, R.J. (in press).The consequences of a changing hydrologic Olsen, G., and Skitmore,E., 1997. State ofthe Riuers ofenvironment for native vegetation in South-western the South West Drainage Diuislbn. Western AustralianWA.ln Nature Conseruation 4: The Role of Netuorks, Water Resources Council Publication 2/91.eds D.A. Saunders, l.L. Craig and E.M. Mattiske. SurreyBeatty and Sons, Chipping Norton (NSW). Stokes, R.A., and Sheridan, R.J., 1985. Hydrology of

Lake Toolibin. Report No. WH2. Water Authoritv ofJohnston, WH., 1993. Water use issues in temperate Western Australia, Perth.agriculture - a review. pp. 32-43 in Land Manogementfor Dryland Salinity Control, Bendigo. Conference Toolibin Lake Recovery Team and Toolibin Lakeproceedings. TechnicalAdvisory Group, 1994. Toolibin Lake recovery

plan. Unpublished report. Department of ConservauonMattiske, 8.M., 1982. NARWRC progress report - Lake and Land Management, Perth.Toolibin vegetation study. Unpublished report.Department ofFisheries and Wildlife, Perth. Weaver, D., 1991. South coast estuaries project

monitoring (1991). Draft report.McFarlane, D.J.,1992. Hydrogeology ofthe south coastof Western Australia.In Farming for the Future. LandConservation Districts, Southern ConferenceProceedings. Esperance Land Conservation DistrictCommittee, Esperance.

McFarlane, D.J., Engel, R., and Ryder, A.T, 1989. Thelocation of recharge areas associated with secondarysalinity in the Lake Toolibin catchment, WestemAustralia. pp 255-268 in Groundwater Recharge, ed.M.L. Sharma. A. Balkema, Rotterdam8oston.

McFarlane, D.J., Howell, M.R., Ryder, A.T, and Orr, G.,1992a. The effect of agricultural development on thephysical and hydraulic properties of four Western

Research into Rernnant Management lssues:Nuts and Bolts and the Big Picture

Richard J. HobbsCSIRO, Division of Wildlile and Ecology, LMB 4, POMidland. WA. 6056

INTRODUCTIONPlant ecological research at the Kellerberrin study siteover the past nine years can be separated into four mainareas:

* survey and collection;.1. basic research on ecological processes;.1. review and synthesis;* "advancementof theory".

Here I outline the major components of each ofthese,indicate relevance to management where applicable,and discuss the ease with which results from each canbe utilised by managers. I suggest that only results fromthe review and synthesis phase can be transferredeffectively to management.

SURVEY AND COLLECTIONAn important and ongoing component of the researchhas been documentation of the flora and vegetation ofthe region. This has consisted of intensive specimencollection and identification, intensive vegetationsurveys of particular areas, and more extensive suF eysof the study area and beyond. In addition, the use ofremote sensing as a rapid assessment technique wasexplored (Hobbs el d/. 1989). None of this work has anydirect management application, apart from theprovision of information on new localities for rare andendangered flora. Nevertheless, the work has formedthe background to much of the research conducted inthe area. The taxonomic knowledge and information onplant distributions have also been utilised extensively byothers, and provided a basis for the development of arevegetation guide (see section on "Review andSynthesis").

BASIC RESEARCH ON ECOLOGICALPROCESSESThis work has focused on factors which impact onnative vegetation in remnants. In particular, fire, weedinvasion, grazing, and interactions between these havebeen investigated, and the relevance of these factors torehabilitation of depraded areas has been assessed.

Much of the work on fire is still in progress, but theexperimental fire program at Kellerberrin has indicatedthat patch burns within remnant areas are possible, andcan be conducted with little risk to surrounding areas.Initial results indicate that post-fire vegetation recoveryis good and that fire stimulates regeneration of a widesuite of species. The studies lack generality, however,and yield little useful information on the effects ofvariation in fire regimes.

Studies on weed invasion and control have highlightedthe importance of disturbance and nutrient additions tothe facilitation of invasion (Hobbs and Atkins 1988;Hobbs 1989, 199la). We have shown that fire need notenhance weed invasion, but that roadside vegetation isan important exception to this (Hobbs and Atkins 1991;Hester and Hobbs 1992). Weed control may enhanceregeneration of some tree species, but is expensive andtime-consuming (Hobbs el aL 1993a). Grazing studieshave highlighted the role of sheep grazing in promotingweeds (Scougallel al 1993). Indeed, the effects ofgrazing are extensive and pervasive, and affect not justthe vegetation but also soil characteristics. All thesefactors have important implications for rehabilitation.In particular, in many cases fencing is not enough toencourage regeneration, and more intensivemanagement is required (Hobbs 1991b).

REVIEW AND SYNTHESISIn conjunction with others, we have made severalattempts to synthesise results to date and to mesh thesewith other relevant work. The outcome ofworkshops oncorridors and landscape reconstruction has been asynthesis of existing information on these areas, and aheightened awareness of the issues involved on the partof the wider community (Saunders and Hobbs 1991;Saunders al a/. 1993). The effectiveness of this has beenincreased by these publications being used for materialby magaz ines such as Ecos and Rural Research, andbyother media exposure.

The most important synthesis work was, to my mind,the Rottnest workshop, which resulted in theproduction of the Springer-Verlag book ReintegratingFragmented Landscapes (Hobbs and Saunders 1993)and. the Reuegetatton Guide to the Central Wheatbelt(Lefroy et al. 1991). The first of these collected availableinformation, from numerous disciplines, on the historyof development in the central Wheatbelt, its current

problems and its potential solutions. It represents anattempt to provide an integrated approach to theregion's problems, and has been used extensively, bothon a local scale by catchment groups and universitiesteaching inland management courses, and more widelyon a national and international scale.

The aim of the revegetation guide was to provide a"user-friendly" version of information gathered in theSpringer-Verlag book. It was developed in consultationwith managers and was aimed directly at them. It waswell received when published, and a subsequent surveyhas indicated that it has had a reasonable rate of use. Of100 households contacted one year after the publicationof the guide, 50 remembered receiving it, 25 had used itand six stated that it had changed their managementmethods. The guide has also been used as a model forsimilar guides in the northern and southern Wheatbeltand elsewhere in Australia.

..ADVANCEMENT OF THEORY"An important part of our research serves not only themanagement community but also the broad scientificcommunity. Advancement of theory in the sciences ofconservation biology and Iandscape ecology is animportant step in ensuring that research carried out byscientists may be integrated and useful in amanagement context (Hobbs 1992a, b; Hobbs andHuenneke 1992; Hobbs 1993a, b; Hobbs et a/. 1993b).We have been emphasising that much conservationscience has little to do with real world situations, andthat there is little time to waste on research which doesnot address real world problems. We have had somesuccess in doing this, and our Western Australianlaboratory is recognised as a leading group inconservation biology and landscape ecology. This initself is important in promoting the transfer of researchfindings: acceptance is more likely if the promulgatorsare recognised leaders in their field

IMPORTANT ISSUES TO BE FLAGGEDFROM THE RESEARCHThe research to date has highlighted the following areaswhich deserve attention:

.i. External influences on remnants, and the need for alandscape context are important (Saundes et al.1991; Hobbs 1993b). Remnant managepent is (or

should be) mostly concerned with problems arisingoutside the remnant, but this is not necessarilvrecognised.

* The research brings into question the extent towhich rehabilitation of degraded remnants ispractical. Rehabilitation is possible, but is not easyand may be expensive. Can we expect extensiverehabilitation programs to occur?

* What is the correct emphasis between revegetationand remnant management? Both are valuable, but,given a choice, which is better to pursue?

.1. The latter questions are predicated on the largerquestion of "what are the conservation andmanagement priorities?". These are poorly definedexcept in broad "motherhood" terms, but areessential if research is to address the relevantquestions.

CONCLUSIONS: NUTS AND BOLTS ANDTHE BIG PICTUREInitially when writing this paper, I found it hard to placemuch of the research conducted over the past nineyears in a useful management context. It then becameapparent that, on their own, individual bits of researchresulting in individual research papers were seldom ofdirect use. They are an essential part of the process ofexamining the "nuts and bolts" of the problems beingaddressed, but do not necessarily plug straight into amanagement problem, except in a very broad sense. It isonly when a number of such individual pieces ofresearch are linked together, either by the individualresearcher, or in conjunction with other researchersand managers, that the relevance to managementbecomes more apparent. The assembly of a "big picture"enables all the individual bits of research to be placed incontext and to be made more readily accessible tomanagers. This process of assembling the "big picture"entails stepping back from the field quadrat or mist netor whatever, and taking stock ofwhat the informationalready gathered can tell us. This pursuit has, in thepast, not been accorded the same scientific merit asgoing out and discovering really exciting new things.However, this is changing, which is just as well, since Iconclude that the synthesis process is essential if thetransfer of research findings to management is tosucceed.

REFERENCESHester, A.J., and Hobbs, R.J., 1992. Influence of fire andsoil nutrients on native and non-native annuals atremnant vegetation edges in the Western AustralianWheatbell. Joumal ofVegetation Science 3: 101-108.

Hobbs, R.J., 1989. The nature and effects of disturbancerelative to invasions. pp.389405 in BiologicalInuasions: A Global Perspectiue, eds J.A. Drake,. H.A.Mooney, ED. Castri, R.H. Grooves, RJ. Kruger, M.Rejm6nek, and M. Wlliamson. Wley, New York.

Hobbs, R.J., 199la. Disturbance as a precursor to weedinvasion in native vegetation. Plant ProtectionQuarteflg 6: 99-104 ,

Hobbs, R.J., 1991b. Regeneration of native woodlands inWestern Australia. Report to World Wldlife FundAustralia.

Hobbs, R.J. (ed.), 1992a. Biodiuersity of MediterraneanEcosgstems in Australia. Swrcy Beatty and Sons,Chipping Norton (NSW).

Hobbs, R.J., 1992b. Corridors for conservation: Solutionor bandwagon? ?ands in Ecologg and Euolution 7:389-392.

Hobbs, R.J., 1993a. Can revegetation assist in theconservation of biodiversity in agricultural areas?Pacific Conseruation Biology l'. 29-38.

Hobbs, R.J., 1993b. Effects of landscape fragmentationon ecosystem processes in the Western Australian\Nheatbelt. Biological Conseruation 64: 193-201.

Hobbs, R.J., and Atkins, L., 1988. Effect of disturbanceand nutrient addition on native and introduced annualsin the Western Australian Wheatbelt.,4ustralian JoumalofEcologg 13: 171-179.

Hobbs, R.J., and Atkins, L., 1991. Interactions betweenannuals and woody perennials in a Western AustralianWheatbelt reserve. Joumal ofVegetation Science 2:643-654.

Hobbs, R.J., and Huenneke, L.F., 1992. Disturbance,

diversity and invasion: implications for conservation.Conseruat ion B io lo g g 6: 324-337.

Hobbs, R.J., and Saunders, D.A. (eds), 1993.Reintegrating Fragmented lnndscapes - TouardsSustainab Ie Production and Conseroahbr?. Springer-Verlag, New York.

Hobbs, R.J., Wallace, J.E, and Campbell, N.A., 1989.Classification ofvegetation in the Western AustralianWheatbelt using Landsat MSS data. Vegetatio 89:91,105.

Hobbs, R.J., Arnold, G.W, and Leone, J., 1993a. Weedcontrol and rehabilitation ofwoodland in the WesternAustralian Wheatbelt. Report to World Wildlife FundAustralia.

Hobbs, R.J., Saunders, D.A., and Arnold, c.W, 1993b.Integrated landscape ecology: A Western Australianpe'tspecti.re. Biological Corcenation 64: 237*238.

Lefroy, 8.C., Hobbs, R.J., and Atkins, L., 1991.Reuegetation Guide to the Central Wheatbelt. WesternAustralian Department of Agriculture, CSIRO andGreeningAustralia, Perth.

Saunders, D.A., and Hobbs, R.J. (eds), 7991. NatureCorsenation2: The Role of Corridors. Surrey Beattyand Sons, Chipping Norton (NSW).

Saunders, D.A., Hobbs, R.J., and Margules, C.R., 1991.Biological consequences of ecosystem fragmentation: Areyiew. Conseruation Biologg 5: 18-32.

Saunders, D.A., Hobbs, R.J., and Ehrlich, P.R. (eds),1993. Nature Consenation 3: Reconstruction ofFragmented Ecosgstems, Global and RegionalPerspectiues. Surrey Beatty and Sons, Chipping Norton(NSW).

Scougall, S.A., Majer, J.D., and Hobbs, R.J., 1993. Edgeeffects in grazed and ungrazed Western AustralianWheatbelt remnants in relation to ecosystemreconstruction. pp. 163-778 in Nature Consenation 3:Reconstruction of Fragmented Ecosgstems, Global andRegional Perspectiues, eds D.A. Saunders, R.J. Hobbsand P.R. Ehrlich. Surrey Beatty and Sons, ChippingNorton (NSW).

Gondwanan Botany: A Ferspeetive oc"lRemnant Management in South-WestAustralla

Stephen D. HopperKings Park and Botanic Garden,West Perth. WA. 6005

INTRODUCTIONBotanical research in vegetation remnants ofsouth-western Australia has been increasing over severaldecades. This has led to a new Gondwanan perspectiveon the flora, which has significant implications formanagement. In this exLended abstract, I aim tosummarise some pertinent research results, highlighttheir management implementation, and ionclude withrecommendations for achieving better integration ofresearch and management. Due to the required lengthof the paper, factual material is not referenced. Forfurther information, interested readers should consultreferences in Hopper (1979, 1992), Hopper et al.(1990), and the popular reviews ofAustralia's fossilrecord by White (1986, 1990) and Rich and Rich (1993).

SIGNIFICANT RESEARCH RESULTS _THE GONDWANAN PERSPECTIVEA new and deeper appreciation of the great antiquityand unique attributes of the south-western flora hasemerged in recent years. The present-day flora sits on alandscape that has not been glaciated since the Permian(260 million years ago), and has remained above sealevel, sometimes as an island, during all subsequentinundations ofthe Australian component of Gondwana.

The terrain ofthe South West, mountainous whenTriassic-Cretaceous dinosaurs dominated the fauna, hasbeen progressively eroded flat to the point where mostsoils are highly impoverished of nutrients, and drainageinland ofthe Meckering Line is largely uncoordinatedand incapable of flushing salt from the syslem.

The flora has responded to this unique environmentalhistory in ways quite different to those of the ve4rrecent, invasive, post-glacial floras of Europe and NorthAmerica about which most botanical science andteaching are based. For example, the gathering andstoring of nutrients from highly infertile soils hasplaced a selective premium on the evolution of novelroot systems - for example, the fine mat of subsurfaceproteoid roots of banksias, or symbiotic partnershipswith soil microorganisms such as mycorrhizal fungi.The diversity of such fungi has scarcely been

documented, but Syme's discovery of more than 300macrofungi in 1\.!o Peoples Bay Nature Reserve suggestsa complexity equal to that of the better documentedflowering plants.

The 50 million year old fossil banksias from theKennedy Range inland from Carnarvon indicate thegreat antiquity of some extant south-western genera,and suggest that much ofthe diversity in wildflowers wesee today originated when dinosaurs probably grazedthe then mountainous terrain of the South West.

Superimposed on the great antiquity of components ofthe flora have been opportunities for explosivespeciation among the relatively few genera thatsurvived the onset of aridity and great extinction of thesouth-western rainforests that dominated ourvegetation for much of the last 100 million years. Theprogressive aridity began some seven million years agoas Australia drifted northwards from Antarctica. AMediterranean climate became established, and theSouth West then entered a period of climaticturbulence during the ice ages of the past two millionyears. This precipitated differential soil erosion, withthe present-day Wheatbelt from Shark Bay to IsraeliteBay experiencing greater climatic and erosional changethan either the high-rainfall forests or the arid interior.

The flora responded dramatically, with rapid evolutionyielding one of the world's richest wildflower regions.The South West has an estimated 8 000 species, 75% ofwhich are to be found nowhere else,30% are yet to bedescribed by botanists, and possibly as much as 7570 areyet to be grown in cultivation.

Plant biodiversity, consequently, is concentrated incomplex patterns in Wheatbelt remnants, with rapidturnover across landscapes. Lateritic uplands, forexample, as little as half a kilometre apart, may haveless than 50%o of their species in common. Distinctivefloras are evident over short distances as one travelsthrough the Wheatbelt. Burgman's (1988) studiesbetween Ravensthorpe and Cape Arid National Parkindicated that nature reserves need replication everyl5 km in mallee, to capture most plant diversity.

Not surprisingly, threatened plants are concentrated inWheatbelt remnants, and are less frequent in the foreststo the south-west or in the arid zone inland.

Threatening processes, largely attributable to humanactivity, include land clearance, rising salinewatertables, nutrient poisoning through fertiliser usage,graz ing by exotic animals, inappropriate fire regimes,weed invasion. disease. harvest and eradication ofeconomic wild species, and loss of native animals andthe ecological processes they supported.

Moreover, it is evident that the recruitment biology ofnative plants is complex, and restoration of plantcommunities is very challenging. For example,Wheatbelt granite outcrops have exceptionally richfloras, with high numbers of fire-sensitive plants, whosemanagement has yet to be investigated in any detail.

Despite the above challenges, conservation of intactremnants is very cost-effective, and remains the highestpriority strategy from a biodiversity perspective.

MANAGEMENT IMPLEMENTATIONRemnant conservation programs on private and Crownlands should be managed in a way which recognises thehigh turnover of plant diversity in the landscape. Inparticular, a high degree of geographical replication iscalled for. However, a floristically complex vegetationand paucity of botanists have hampered the generalinventory of remnants, and have led to difficulties indeveloping local district management priorities tomaximise conservation of plant diversity.

Threatened plant distribution and abundance data havebeen effectively included in management considerationsin some districts, but poorly in others (dependent oninterest of local communities). More integrated effortsare required.

As would be expected, active management of remnantsto maintain plant biodiversity is embryonic, as isresearch. Management has quite rightly focused onstemming the tide of direct human destruction anddegradation of remnants. However, we have reached thepoint where active restoration management is needed.It is here that the Gondwanan perspective has much tocontribute.

RECOMMENDED ACTIONSFirstly, I firmly believe that botanical researchers needto focus more on communicating their findings in plain

concise Dnglish to the media and public, and especiallyto managers. Botanists need to monitor publicunderstanding and use of their results as much as howoften papers are quoted in the scientific literature.

Secondly, integrated research approaches are essential.A combination of process research (necessarily confinedto few sites) and descriptive research (widergeographical scope) is needed to yield cost-effective andefficient conservation of biodiversity.

Thirdly, it is vital that there be collaborative interactionof managers and researchers, combining theoperational and investigative skills of each group - forexample, in long-term monitoring and experimentalmanagement.

In terms of applicable prescriptions, in view ofthefloristic complexig to be found in remnants,conservative management and highly replicatedconservation activities across the lardscape arerecommended, focused on intact remnants with thehighest biodiversity (genetic, specific, community,landscape - for example, the Stirling Range, FitzgeraldRiver, and Lesueur national parks, remnants in thewestern central Wheatbelt, granite outcrops).

I would suggest that managers need to pay moreconsistent attention to ensuring biodiversityconservation in remnants - local activity must beintegrated with regional and global goals. This willoccur if clear objectives, integrated actions andperformance-measured outcomes are required for bothresearch and management of remnants.

Finally, we need to become Gondwanan botanists andmanagers. We are dealing with an incredibly diverseflora of great evolutionary antiquity. Many componentsare poisoned by conventional fertilising and wateringregimes. They respond to disturbance in new and oftenunpredictable ways. Their recruitment is triggered by arange of cues whose significance is only just becomingapparent. Even the ubiquitous impact of fire ongermination and flowering has remained enigmatic forlarge numbers of species, and has defied traditionalresearch tools developed on northern hemisphere floras(although recent experiments with smoke-inducedgermination by Kings Park and Botanic Garden staffshow significant promise). Few other regions ofthe

world demand such urgent and integrated collaborationof researchers and managers to meet conservation aims.

REFERENCESBurgman, MA., 1988. Spatial analysis ofvegetationpatterns in southern Western Australia: Implications forreserve design. Australian Joumal of Ecologg 13: 475-429.

Hopper, S.D., 1979. Biogeographical aspecl"s ofspeciation in the south west Australian flora. AnnualReuieut of Ecologg and Sgstematics l0:. 399422.

Hopper, S.D., 1992. Patterns of diversity at thepopulation and species levels in south-west AustralianMediterranean ecosystems. pp. 27-46 in Biodiuersitg ofMeditenanean Ecosystems in Australia, ed. R.J. Hobbs.Surrey Beatty and Sons, Chipping Norton (NSW).

Hopper, S.D., van Leeuwen, S., Brown, A.P., and Patrick,S.J, 1990. Westem Australia's Endangered Flora.Department of Conservation and Land Management,Perth.

Rich, PV, and Rich,T.H.,7993.Wildlife of Gondwana.Reed, Chatswood (NSW).

White, M.E., 7986. The Greening of Gondwana. Reed,Books, Fredrichs Forest (NSW).

White, M.E., 1990, The Nature of Hidden Worlds:,Animals and Plants in Prehistoric Australia and NeuZealand. Reed Books, Balgowlah (NSW).

The Exotic Predator Problem

Jack KinnearDepartment ot Conservation and LandManagement, Woodvale Research Centre, PO Box51 , Wanneroo, WA, 6065

INTRODUCTIONFifteen years ago, when I joined the Department ofFisheries and Wildlife, fire ecology was the dominantresearch topic of the day. Medium-sized marsupialswere either rare or extinct over much of mainlandWestern Australia, and those populations that werepreviously locally abundant had crashed. It wasspeculated that their demise was possibly due todrought and a variety of other causes - for example,inappropriate burning regimes, fragmentation ofhabitats beset by a variety of man-made disturbances,diseases, and climatic changes, to list a few. Clearly, itwas a situation that held little scope for optimism.

My own views were strongly biased towards anutritional explanation. Having worked with high-density populations of island macropods whichexperience seasonal nutritional stress, the idea ofdrought-induced mortality coupled with a reducedreproductive rate seemed to be an eminently reasonableexplanation, at least for the recent declines. Thepuzzling aspect was: why no recovery following the endofthe drought?

I abandoned this nutritional hypothesis after two yearsof research on Wheatbelt rock-wallabies (Petrogalelateralis\, because, contrary to expectalions, thesupporting evidence for nutritional stress was negativeeven during a declared drought year. Accordingly, wewere confronted with this question: given a fit, healthypopulation producing young, why the low numbers andwhy no population growth?

These questions were answered by my research on rock-wallabies living on remnant vegetation in the centralWheatbelt. In this area, we found, that fox (Vulpes,..r/pes) predation was the primary factor controllingpopulation numbers. By reducing fox populationsthrough baiting, we were able to reverse the parloussituation of the rock-wallaby.

Baiting programs, now operational on severalWheatbelt reserues, have produced comparable reversals

for other species. Tammats (Macropus eugenii), brush-tail possums (Trichosurus uulpecula), woylies(Bettongia penicillafa), and numbats (Mgrmecobiusfasciatus) have increased appreciably and can be sightedon most occasions, with little effort, where baitingprograms have been implemented. Elsewhere, chuditch(Dasgurus geoffroii) and ring-tail possums(Pseudocheirus peregrmus) have increased in responseto baiting regimes. Bandicoots (Isoodon obesulus'S havebeen successfully translocated to the Tutanning NatureReserve, from where they had disappeared (Friend, pers.comm.). There are some indications that thepopulations of brush wallabies (Macropus irma) mayalso be significantly affected by foxes.

To summarise, medium-sized marsupial populationspersisting on Wheatbelt nature reserves existed at lowdensities or, if locally abundant, crashed in the 1970sand did not recover spontaneously. Some populationsdeclined to extinction. Recovery to conspicuouslyhigher densities ensued under fox baiting programs.

ECOLOGICAL IMPLICATIONSThe message for management is this: in the absence ofexotic predator management, medium-sized marsupialsexperience great difficulties in maintaining viablepopulations on Wheatbelt reserves (and, for that matter,on other more extensive Crown lands as well). Thissuggests that the fox is acting as an efficient biocontrolagent affecting a suite of marsupial species and likewisesome ground-dwelling or ground-nesting birds (forexample, malleefowl Leipoa ocellata).

In situations where medium-sized marsupials persist inthe absence of predator control, population densitiesare invariably low. These sites should be viewed aspredation refuges that promote sur.uival but do notguarantee population viability.

Predator removal results not only in populationincreases but also in habitat shifts (that is, expansion ofthe realised niche) by affected species. Predator controlenables affected species to move out from refuge areasand utilise habitat denied to them in the presence ofpredators. Destruction or disturbance of thesepredation refuges - for example, through bushlandclearing or fire - greatly increases the risk ofextinction in the absence of oredator control.

In the past, lack of knowledge about the impact ofexotic predators has led to distorted ecologicalperceptions regarding the habitat requirements ofspecies affected by predators. Habitat requirementshave been equated to pr€dation refuges, which arecharacterised by protective cover with food nearby;these refuge sites are not necessarily typical of a species'habitat requirements (niche). Failure to realise thisleads to interpretations which promote the view thathabitat loss is the major factor causing the demise ofmany marsupial species. A corollary to this reasoningl isthe inference that medium-sized marsupials areintolerant to disturbance and thus can only persist innear-pristine environments. Clearly, these conceptsshould be re-examined in the light of current evidence.

IMPLICATIONS FOR MANAGEMENTThe objective of research into fox control was to test thehypothesis that fox predation is a limiting factoraffecting population dynamics of native marsupials -hence the need to nulliflr as far as possible the impact offox predation. This has led to baiting procedures forsmall nature reserves that may well be excessive formanagement purposes.

Nonetheless, it should be borne in mind that near totalremoval of foxes by intensive baiting provides a measureof the carryingi capacity of a reserve for medium-sizedmarsupials. This step is mandatory if other limitingfactors are to be identified.

Finding the optimum, cost-effective level for baitingwill not be a quick or easy exercise. The baiting effortwill need to be manipulated and the prey responsemonitored. Such research raises a difficult question: atwhat densities should recovered populations bemanaged? Fifty per cent of carrying capacig? More, orless?

Notwithstanding the above, and to complicate matterseven further, there is some evidence that the predationpressure exerted by foxes is variable. Data from therock-wallaby study (Kinnear unpublished) implies thatfox predation is likely to be episodic. The message hereis that short-term studies may well result in conflictingand confusinp outcomes.

INTERACTIONS WITH OTHER LIMITINGFACTORSInevitably, periods of drought can be expected to lowerthe carrying capacity of a reser"ve, leading to mortalityamong populations of medium-sized marsupials.Mortality could well be exacerbated due to invadingfoxes switching from collapsed populations ofalternative prey, such as rabbits, on adjacent farmland.Likewise, bushfires, naturally occurring or otherwise,may well increase the predation risk to medium-sizedmarsupials.

In both of these situations, a higher level of baiting maybe required, to avoid excessive losses of medium-sizedmarsupials. It is likely that a higher level of baiting willneed to be maintained during the recovery phase,particularly if the prey numbers are low.

THE FUTUREThe demonstration that control of exotic predatorsresulh in recovery of some populations of medium-sized marsupials on Wheatbelt reserves studied shouldput paid to the prevalent view that habitat loss ordegradation has been the single cause of their demise.That some of these species increase significantly innumbers when fox predation is eased shows that theyare adaptable, resilient species capable of persisting (asdo grey kangaroos) within fragmented, disturbedecosystems where large areas of bushland remain.Indeed, some species may, locally, achieve minor peststatus and thus create management problems.

These realisations provide a basis for greater optimismabout the future of some medium-sized marsupials.One can take some heart from the numerous islandpopulations of marsupials that have persisted for manythousands of years.

The ultimate solution to the fox problem is biocontrol.Such research is in progress, but it must be emphasisedthat this research is high risk and may not yield asolution in the near future. However, much will belearned about the fox as a predato! during! the course ofthis work. Unquestionably, some of this knowledge willgreatly improve and enhance conventional controlmethodoloEies.

FURTHER READINGFriend, J.A., 1990. The ntmbat Mgrmecobius fasciatus(Myrmecobidae): History of decline and potential forrecollery, hoceedings of the Ecological Societg ofAtrtralia 16: 369-377.

Kinnear, J.E., Onus, M.L., and Bromilow, R.N., 1988.Fox control and rock wallaby population dynamics.Australian Wildlife Research 15: 435-450.

Robert J. LambeckCSIRO, Division of Wildlife and Ecology, LMB 4, POMidland, WA,6056

INTRODUCTIONThe successful management of remnant vegetationdepends upon the formulation of clearly definedmanagement goals. Broad objectives, such as"management to maintain biodiversity", have beenidentified, but the implications for managers of suchsla lements have not been speci f ied.

Honeyeaters in the central Wheatbelt provide a modelfor understanding the functional significance ofbiodiversity, and for assessing the implications of areduction in regional biodiversity. This understandinS,in turn, can provide a framework for approaching thedaunting task of "managing for biodiversity".

Biodiversity encompasses the variation that occursbetween individuals, species, communities, landscapes,and broad biogeographical regions. The relationshipbetween honeyeaters and their food plants provides anopportunity to explore the consequences of diversity atall of these levels and to examine the functionalsignificance of this diversity. As pollinators, honeyeatersplay an important role in the maintenance of local floraldiversity. A decline in honeyeater numbers can affectplant communities by reducing the distance andfrequency of gene flow between discontinuouspopulations (Lamont et al- 1993), a problem that isexacerbated in fragmented environments. Changes infloral diversiiy may. in turn. affecl lhe persislenceprobabilities of other species associated lvith, ordependent upon. tha l f lora. By consider ing lheserelationships between plants and animals, we canidentify the importance ofvariety at variousorganisational Ievels in both plant and animalcommunities.

PLANT DIVERSITYIndividual DifferencesAlthough most members ofa plant species in one areaflower synchronously, individual differences result insome plants flowering earlier in the season, whileothers commence later. These phenological differencesmay result from variation in microenvironments as wellas from genetic differences between individuals. These

individual differences result in an increase in theamount of time that nectar is available for honeyeaters.

Species DifferencesDifferences in the flowering phenology of differentspecies also broaden the range of opportunities forhoneyeaters in a given habitat. An array of specieswhich flower synchronously in different locationsincreases the range of habitats that honeyeaters canutilise.

Community DifferencesEdaphic, evolutionary, disturbance and other factorshave resulted in Wheatbelt plants forming clearassociations in a pronounced vegetation mosaic. Theconstituent species within these associations haveparticular flo\!ering characteristics which result indifferent patches producing nectar at different times ofthe year. Provided the configuration of these resourcesis favourable, honeyeaters can move from one patchtype to another at different times of the year, furtherincreasing their resource options.

Landscape DiversityThe same vegetation communities at different positionsin the landscape often have different phenologies as aresult of regional variation in environmentalconditions, again expanding the resource base forhoneyeater species that can range widely across thelandscape.

Regional DiversityUnfavourable climatic conditions in a particular regionmay result in poor flowering and hence temporaryshortages of nectar over substantial areas. At suchtimes, mobile species are able to move to adjoiningregions urhere a different suite of plant species mayprovide the necessary resources. The persistence of aspecies in its preferred habitat may depend upon accessto resources in adjoining areas at times \rhen localresources are depleted. For example, white frontedhoneyeaters (PhAltdongris albifrons), pied honeyeaters(Certhiongx uariegalus) and black honeyeaters (C.niger) irrupt sporadically into the central Wheatbelt attimes of low rainfall in the more arid interior Nectar-producing plant species in the Wheatbelt may provide acritical resource for these birds at these times.

HONEYEATER DIVERSITYNot only is a variety of plants important for thepersistence of pollinators, but similarly a variety ofpollinators is essential for the maintenance of plantdiversity. The impact of fragmentation on plant-pollinator relationships will depend upon the capacityof pollinators to deal with the modified landscape. Arange of pollinators which respond to the landscape indifferent ways will enhance the prospect ofplantpopulation processes being maintained. The functionalsignificance of honeyeater diversity is expressedprimarily at the species level, but individual differencesmay also influence patterns of pollen dispersal.

Individual DifferencesWithin some species, some individuals establishterritories in which they remain for extended periods oftime. Other younger, or less dominant, individuals areunable to establish or maintain territories. They aremore mobile, and this results in different patterns ofpollen dispersal.

Species DifferencesDifferences exist between species in the way they movearound the landscape. Some species, such as white-eared honeyeaters (lichenostomus leucotis), arerelatively sedentary and tend to remain in territorieswithin remnants. Singing honeyeaters (L. uiriscens)commonly move back and forth between remnants andadjoining road-verge vegetation. Brown honeyeaters(Lichmera indistinclal move widely around thelandscape, commonly moving between remnants, andwhite-fronted, pied and black honeyeaters invade frommore arid regions to the north and east. Some speciesremain in close proximity to vegetation, while others,such as yellow-throated minerc (Manorina flauigula),routinely make extended flights over open paddocks.This array of different strategies for exploiting resourcesprovides various opportunities and scales for pollentransfer throughout the landscape,

IMPLICATIONS OF REDUCEDDIVERSITYThe value of greater biodiversity lies in the increasednumbers ofspecies that mediate important processes,thereby increasing the area and time over which a givenprocess proceeds. Species which play equivalent rolescan be viewed as functional analopues. The functional

significance ofa species in a system is inversely relatedto the number of its functional analogues. Loss ofbioticdiversity reduces the number ofalternative, functionalpathways and hence reduces the probability that aprocess will occur Species that have no functionalanalogues but make a major contribution to a givenfunction may become "keystone" species, and their losscould have serious implications for other associatedorganlsms.

The acorn banksia (Banksia prionotes)flowers in earlyautumn in the central Wheatbelt. No other nectar-producing plants flower in the Durokoppin area at thistime - it has no functional analogues. It is heavilyexploited by honeyeaters and, because it appears to be aIimiting resource, it may influence total honeyeaternumbers in the region for the remainder of the year.For these reasons, this species is a keystone species. Itdoes not play the same role in other parts of its rangewhere other species (for example,.B. m enziesii) flowerat the same time.

MANAGEMENT IMPLICATIONSHoneyeaters have been used here to explore thefunctional significance of biodiversity. This model mayhold for a wide yange ofspecies and communities,although the scale and number of participants mayvary. Several implications flow from this perspective.

Management for Function and Repres€ntationRepresentation of biodiversity is a necessary objectivefor the management of native biota, but it alone is notsufficient. It is also essential to consider how differentlevels of representation contribute to ecosystemfunction. The greater the functional contribution of aspecies or community, the more widely should it berepresented. Honeyeaters are not a high priority forconservation in their own right, but may be essentialfor the achievement of other objectives. Consequently,they must be well represented in the landscape. Forspecies that contribute less to community processes,representation in a few locations may be adequate.

The Importance of Scale and ConligurationManagement must be planned at an appropriate scalefor the species and processes being managed. Becausebiodiversity encompasses all scales, from individuals tolandscapes, it is necessary to have strategies that reflectthis diversity. While large-scale, regional assessments of

conservation value must be made, these must then becombined with local strategies which identify thoseactions that need to be implemented at particularlocations. Attempts to achieve an objective within asingle remnant will fail ifthat objective is dependentupon processes that operate at a scale greater than thatof the remnant. Management of remnants must beguided by their role in landscape processes and inachieving regional conservation objectives.

The Importance of ObjectivesThe above point presupposes that regional conservationobjectives and strategies exist. Conservation objectivesmust be identified at catchment, regional and nationalIevels, with local actions targeted towards thoseobjectives which are relevant for the area beingmanaged. For example, some areas may not encompassspecies or communities that are recognised as regionalor national priorities. This cannot be taken to meanthat nature conservation is not an important issue forsuch an area. It simply means that specific objectiveshave to be identified which address those values that doremain in that region.

Objectives, at any scale, will not be achieved in a shortperiod of time. It is therefore necessary to identify asequence of actions that will achieve nominatedobjectives. It is also important to identily thoseobjectives that need most urgent attention, and actionsthat will provide maximum conservation value forminimum effort.

BIODIVERSITY AS A MODEL FORSOCIAL MANAGEMENTIfa regional conservation strategy is developed, itsimplementation will require a coordinated effort by awide range of individuals and organisations. Diversity atdifferent levels within the human community must alsobe recognised, and the functional significance of eachcultural group identified in order to develop an effectivesocial process for achieving biodiversity conserration.In terms of the biodiversity model presented above,important components of diversity in the humancommunity are as follows:

.!. Individual differences: Different individuals will havedifferent capacities to contribute towardsmanagement of the native biota in a region,

depending upon inclination, financial resources,other commitments, etc. Local farm plans orcommunity projects should be designed to recognisethese differences, and should enable all individualsto feel that their contribution, no matter how largeor small, is an integral part of a larger conservationeffort.

.i. Species differences: Sometimes it appears that thevarious groups with vested interests in regions areindeed different "species". These "species" comprisethe land-holders, and the wide array of governmentand non-government agencies that are involved. Thedifferent perspectives and aspirations of thesegroups must be acknowledged and incorporated intoplanning. The functional role of each group must beclearly identified and their efforts coordinated so asto maximise conservation return.

.i. Community differences: Different areas may havedifferent conservation priorities, depending upontheir particular problems. If regional targets can beidentified, local communities, together with othergroups with appropriate information and expertise,must work jointly towards identi[ring how thosetargets can best be achieved in a manner whichmaximises both conservation return and primaryproductivity.

The use of "biodiversity" as a metaphor for socialorganisation will clearly have its limitations if taken toexttemes. For example, while a range of functionalanalogues performing similar roles in a naturalecosystem may enhance the stability of that system, thepresence of a number of government agencies allfulfilling equivalent functions will simply result inunnecessary duplication and conflict. Given thatgovernment agencies rarely "become extinct", the needfor functional equivalents is clearly not desirable.

In contrast, at the level of individuals who areimplementing recommendations, it is important toensure that multiple representation exists. Given themagnitude of the problems in agricultural landscapes,managernent actions will need to be implemented overa large propodion of the landscape. This requires aclear statement of objectives, identification of theactions to be taken, and the involvement of manyindividuals in implementing these actions. In the

absence of an integrated management strategy, therewill be only small localised responses which willgenerate limited benefit at a landscape scale.

The comparison of social organisational structures withbiodiversity may be more appropriate when consideringthe functional roles ofthese organisations. The value ofthe analogy lies in the recognition that a complexsystem comprising a number of interdependentcomponents needs to retain all of the essentialfunctions if that system is to persist. In order to achievesustainable agriculture and maintain biologicaldiversity, it will be necessary to integrate social,agricultural, land conservation and nature conservationobjectives.

Only by recognising the diversity within the communityand the different contributions that different sectorscan make will nature conservation succeed in what ispredominantly a freehold landscape. Ownership of, andresponsibility for, local conservation must be devolvedto the community, with government organisationsmanaging specifically identified, high-priority targetswhich are beyond the material and knowledge resourcesof local communities.

REFERENCESLamont, 8.B., Klinkhamer, PG.L., and Wtkowski,E.TR, 1993. Population fragmentation may reduceferlility to zero in Banksia goodii - a demonstration ofthe Allee effect. Oecologia 94: 446450.

Note that Robert Lambeck's work on honeyeaters,referred to extensively, has not yet been published.

#onservation Flesen'ves frm Western Austna!imAcquisition and l\llanagennent of

A.R. Ma inDepartment of Zoology, The University of WesternAustralia, Nedlands, WA, 6009

PREAMBLEI have been involved with conservation for a long time,first with acquisition of reserves and later with theirmanagement. For this paper, I have chosen to give anoverview of how and why my ideas developed.

INTRODUCTIONThe raw material with which managers have to work -

that is, the system of remnant vegetation - is largelyan historical accident. This is because what is selectedfor reservation is merely a remnant ofwhat wasformerly more widespread. But, in addition to thishappenstance, the success or otherwise of managementwill be affected by:

.t the manager's assumptions about ecosystemprocesses;

* the initial state ofthe ecosystem, which is the resultof past events;

* the operation of chance in the future.

HISTORYInitially, selections were based on personal knowledgeand what was available for reservation. Later, variouscommittees selected areas for reservation followingreviews of the knowledge available, or biological surveysof land in transition zones, or areas not yet representedin the conservation estate. This left the manager areasthat had, in many cases, less than all the desirablecharacteristics. For example, areas were small,boundaries extensive and irregular, and so on.

CRITERIA FOR SELECTIONEarly reserve selections were based on a desire toensure that reservations were representative in someway, either of regions or of biotic assemblages, andmanagement was expected to preser"ve these qualities.During this early period, habitat preservation was thevogue and minimal interference was the essence ofgood management. When ecosystems were consideredat all, they were thought of as being static, and whendisturbances did occur, it was assumed that over timesuccession would restore the biotic environment to itsformer state.

More recently, there has developed an internationalconsensus that conservation effort should be devoted toretaining biodiversity, and this adds another dimensionto management problems. Biodiversity is a diffuseconceptwhich can be considered to include:

Genetic diuersitg: where conservation goals are toretain as much of the genetic variability as possiblewithin any population that is being activelyconserved. Genetic diversity can be lost by:inbreeding in very small populations; reductions ofpopulation size resulting from breeding failure; orcatastrophic mortality resulting from disease,drought or fire.

Tatonomic diuersitg; this literally includes the fivekingdoms of organisms but is usually taken as totalspecies richness - for example, representation ofspecies within groups such as eucalypts, banksias,forest birds, raptors, burrowing ftogs, trapdoorspiders, or native bees. The range is endless and themanagement problems diverse.

Phglogenetic diuerslly: this includes the range ofdiversity within an evolutionary radiation forexample, within the marsupials, or a genus ofIizards, or the trigger plants. Some oftheseorganisms may be classified as rare or endangered,but also included are relicts such as cycads,echidnas, Lepidogalarr'as (the mud minnow),petipatus, Archaeocftlas (the granite rockchironomid) and the trapdoor spider Moggridgia.Such are to be considered particularly durablespecies.

Structural diuersitg: this category includes forest,woodland, shrubland, grassland and the structuralstages in regeneration of vegetation followingdisturbances such as fires or storms. Restricted orrelict habitats characteristic of former times fall intothis category of diversity.

These basic kinds of biodiversig may be combined todescribe more complex forms such as community,ecosystem or landscape diversity.

Such a wide-ranging focus suggests that management isconcerned with nothing less than whole ecosystemsconsidered as dynamic entities.

THEORYManagement based on empiricism cannot be expectedto cope with the diverse requirements imposed by theforegoing concept. Some theoretical basis is needed. Inthis sense, theory merely means a schema or systemwhich accounts for or explains a set of observations orphenomena. In terms of conservation management, weneed to explain what is there, the sequence of events bywhich it has been derived, and the way the system isrenewed or maintained. A comprehensive theoryprovides a framework within which the future may bepredicted. Thus the consequences of both managementpractices and chance events can be foreseen otanticipated. Research should provide the knowledge andfacts which form the basis of the theory provide tests ofcurrent theory, or provide new knowledge whichsuggests that theory should be modified. Monitoringthe consequences of management practices is a way oftesting theory and gaining new knowledge.

But there are five classes of people involved withconservation: the public, administrators, those whowrite management plans, managers, and scientists. Notall ofthese consider the managed system as beingdynamic; thus when their expectations are not met -for example, when the biota changes - there is aperception that preservation has not been achieved andmanagement has failed. This is of importance in theevent that the concept of accountability is extended tomanagement of the conservation estate. Such a literalextension is inappropriate when ecosystems aredynamic, process driven, non-static entities.Furthermore, administrators sponsoring, for example,Cooperative Research Centres look to research to solveproblems requiring complex and expensive technicalexpertise. On the other hand, perhaps in response topublic perceptions, the Australian Nature Conservationfuency and World Wildlife Fund for Nature see thestudy of rare or endangered species as a priority. Notunexpectedly, administrators of management-orientedorganisations see scientists as a skilled work force to beused in solving local problems as they arise. At alllevels, little attention is given to adding knowledge thatmight be useful theoretically.

NATURAL SYSTEMSThere is a tendency to view natural systems as theanalogue of agricultural, horticultural or forestrysystems. In such systems, any reduction in productivity

is disfavoured; consequently, diseases, pests, or anyfactor which reduces the productivity or the quality ofthe products is regarded as undesirable. However, innatural systems, diseases, pests and parasites reduce thevigour and may even lead to early death oftheir host.Such happenings make for less than optimal individualhealth, but conversely, when they affect the cornmon ormost vigorous species, they prevent the exclusion ofless vigorous species from the community. Thusdiseases, pests and parasites, along with predation andnatural disasters, are the mechanisms by whichdynamic ecosystems are maintained. They areintimately related to biodiversity but completelyionnrpd in i lc ct , rdu

Characteristically, disturbance, of whatever origin, leadsto the formation of gaps in the vegetation. It is in suchplaces that the availability of space permits thegermination of seeds and the initiation of populationchanges in the whole biota which sum to thesuccessional stages that follow. But disturbance mustnot be thought of solely as being caused by physicalenvironmental factors.

Gap creation, whether from branch fall or tree death, isalso effective in initiating regeneration. Most often,such deaths are caused by disease induced by fungi orother pathogens. Such mofality is usually scatteredthrough an ecosystem and is a potent source of minorhabitat diversity by providing sites for other plants,animals, fungi, protists and bacteria. Managers shouldappreciate that this sort of habitat diversity can only beprovided by natural events which vary in space and overtime. They are the basis of patch creation and thedynamics ofpersistence. Population sizes anddistributions are unlikely to be constant when suchfactors operate. Managers can only ensure that thelikelihood of persistence is as high as possible.

The contrasts between static and deterministic versusdynamic and stochastic interpretations of systemsmaintenance are not merely a matter of academicdisagreement. The general public and some biologistsfollow static, deterministic interpretations of nature -for example, as expressed in the phrases "balance ofnature" or "stable population". To those with suchbeliefs, patch creation by mortality from disease ordefoliation by insects is likely to be interpreted asfailure of management rather than as a demonstration

@

of the way that space, light and nutrients are madeavailable to different species or another generation.

Recently, it has become a legislative requirement thatthe public be involved when management plans arebeing prepared. Legislation or regulation requires thatthe aims, goals and requirements are specific - forexample, "preservation" may be a stated aim. Thisimplies a static universe which can be preserved.Moreover, management plans require funds, and itfollows that there is an expectation that those spendingthe money are accountable for achieving the legallyspecified goals. Yet, as indicated above, natural systemsare hostage to chance events that are not readilyaccommodated in a legal system couched in terms of anideal stable world. The potential difficulties should beanticipated by managers, who might sell their skills byemphasising that, even in a very ftagmented, dynamicnatural world, managers can retain the conservationvalues even if static preservation is not possible.Education to achieve a community awareness ofwhat isbiologically realistic is a prerequisite before sensibleaccountability can become a reality.

DISCUSSIONThe foregoing sets out what might be included in theconcept of biodiversity and so establishes an array ofreasons for having reservations. It also sets out a theoryof how the contained systems and communities mayfunction and be maintained as dynamic entities.Naturally, I did not develop these ideas as a completesystem at one time in a single paper. Initially, the size ofreservations was a critical issue, and Main and Yadav(1971)justified arguments for larger reservations thanhad hitherto been considered adequate by land useplanners. A general case for numerous reservesspanning a range of sizes and geographic Iocations wasthen developed (Main 1979). At this time, there was apublic perception that reserves would preserve nature,especially rare and endangered species. An argumentcountering this was that rare species were thus becausethey were on the way to extinction, so it was a waste toreserve useful land for an unachievable purpose. Theseissues were addressed in Main (1982, 1984).

Allied to the foregoing was the question of reservemanagement, especially in terms of nutrients (Main198Ia, 1987), fire (Main 1981b), the effect of insectgrazing on post-fire regeneration (Whelan and Main

1979), and the response to stress by vertebrates (Main1986). Many of the problems associated with reservemanagement were reviewed, and possible solutionstabulated (Main 1987). As soon as it became clear thatlandscapes needed to be managed as whole entities andthat management of nature conservation reser.resshould be integrated with other uses and landscapegoals, it was appropriate to reiterate the basicsimilarities in the resources required by all living things(Main 1993a). The problems associated with thepotential loss of biodiversity in reservations has beenaddressed by Main (1992a). Additionally, a possibleapproach to making management decisions when facedwith uncertaing was presented (Main 1992b). Theimplications of climatic change for restoration ecologyand management are dealt with in Main (1988, 1993b).

One would expect to see indications that the aboveinterpretations were having an influence on reserremanagement, in management plans and field responsesof managers, perhaps a decade after publication. Ingeneral, it is too soon to say whether the ideasdeveloped are, or have been, of use to those involved inthe practicalities of management, though I have hadsome verbal comments that the 1987 paper has beenuseful.

REFERENCESMain, A.R., 1979. The fauna. pp 77-99 in Enuironmentand Science, ed. B. O'Brien. University ofWesternAustraliaPress, Nedlands.

Main, A.R., 1981a. Ecosystem theory and management.Joumal ofthe Rogal Societg of Westem Australia4 : 1 4 .

Main, A.R., 1981b. Fire tolerance of heathland animals.pp.85-90 in Heathlands and Related Shrublands, ed,.R.L. Specht. vol. 9B,Ecosgstems ofthe World, ElsevierPublishing Company, Amsterdam.

Main, A.R., 1982. Rare species: Precious or dross. pp.163-174 in Species at Risk: Research in Australia, edsR.H. Groves and WD.L. Ride. Australian Academy ofScience, Canberra.

Main, A.R., 1984. Rare species: Problems ofconservation. Searcft l5: 93-97.

Main, A.R., 1986. Resilience at the level ofthe individualanimal. pp. 83-94 in Resilience in MeditenaneanlgpeEcosAstems, eds B. Dell, A.J.M. Hopkins and B. Lamont.Dr. W Junk Publishers, Dortrecht (Netherlands).

Main, A.R., 1987. Management of remnants of nativevegetation - a review of the problem and thedevelopment of an approach with reference to theWheatbelt of Western Australia. pp.7-73 in NatureConseruation: The Role of Remnants of NatiueVegetation, eds D.A. Saunders, G.W Arnold, A.A.Burbidge and A.J.M. Hopkins. Surrey Beatty and Sonsin association with CSIRO and Department ofConservation and Land Management, Sydney.

Main, A.R., 1988. Climatic change and its impact onnature conservation in Australia. pp. 361-374 inGreenhouse: Planning for Climatic Change, ed,. C.l.Pearrman. CSIRO, East Melbourne.

Main, A.R., 1992a. The role of biodiversity in ecosystemfunction: Ar overview. pp. 77-93 in Biodiversity ofMediterranean Ecosgstems in Australia, ed. R.J. Hobbs.Surrey Beatty and Sons, Chipping Norton (NSW).

Main, A.R., 1992b. Management to retain biodiversity inthe face of uncertainty. pp. 193-209 in Biodiuersitg ofMediterronean Ecosgstems in Australia, ed. R.J. Hobbs.Surrey Beatty and Sons, Chipping Norton (NSW).

Main, A.R., 1993a. Landscape reintegration: Problemdefinition. pp. 189-208 in Reintegrating FragmentedLandscapes: Towards Sustainable Production andNature Conseruation, eds R.J. Hobbs and D.A. Saunders.Springer-Verlag, New York.

Main, A.R., 1993b. Restoration ecology and climaticchange. pp. 27-32 in Nature Conseruation 3:Reconstruct ion of Fragmented Ecosgstems, eds D,A.Saunders, R.J. Hobbs and PR. Ehrlich. Surrey Beattyand Sons, Chipping Norton (NSW)

Main, AR., and Yadav, M., 1971. Conservation ofmacropods in reserves in Western Australia. BiologicalConseruat ion 3: 123-733.

Whelan, R., and Main, A. R., 1979. lnsect grazing andpost-fire plant succession in south-western Australianwoodland. Australian Joumal of Ecologg 4: 387-398.

Invertebrate Studies Performed throughCurtin University

Jonathan Majer and Anne BrandenburgSchool ol Environmental Biology, Curtin UniversityofTechnology, GPO Box U1987, Perth, WA,6001

INTRODUCTIONThe scope of this paper is to list invertebrate and otherrelated ecological research that has been performed inthe Western Australian Wheatbelt by Curtin Universitystaff and students. We have not included work that hasspecific relevance to invertebrates in rehabilitatedmines (for example, at Eneabba) or areas of primaryproduction (for example, farm paddocks). We have,however, borne in mind any management principlesthat can be derived from this and other researchconducted outside the Wheatbelt region where it hasimplications for the management of remnant vegetationin south-western Australia.

The major theses and papers resulting from theseinvestigations are listed in the References. In Table 1,these studies are grouped together under broadmanagement topics. For each of these topics, primaryresearch outcomes are briefly summarised, and theirmanagement implications and degree ofimplementation listed.

The major users of our research are considered to beland users, land managers, and those advising thesegroups.

RESEARCH TOPICSResearch undertalen on invertebrates in remnants andon other related topics through the School ofEnvironrnental Biology at Curtin University has focusedon the role of invertebrates in ecosystem function andrecovery following disturbance (refer to the researchtopics listed in Table 1). Some of this work has alsoemphasised the benefits of habitat diversity to theconservation of invertebrates. The general approach tothe work has been at an applied rather than atheoretical level.

MANAGEMENT IMPLICATIONSThe implications of research outcomes for landmanagers are listed in Thble 1. The implications areoften of a general nature and based on the results ofseveral different projects, some of which have been

performed by other research groups.

In some cases, fudher research is required before ourconclusions can be implemented by managers. Forexample, with respect to our work on corridors, whilewe have been able to demonstrate that corridors ofnative vegetation support greater numbers of antspecies than do corridors containing large proportionsof exotic plant species, we cannot describe theminimum number or cover of native plant speciesrequired to ensure that ant communities aremaintained. Our findings therefore need to beincorporated with those from other studies before finalrecommendations may be made.

It is important that the work of conservation managersbe guided by well-founded principles that are based onecological research. These principles should beconstantly upgraded in the Iight of new researchfindings. For this to happen, it is necessary for liaison tooccur between researchers and land managers so thatresults of research may be disseminated. This processwould also enable researchers to hear the needs of landmanagers and to plan research to fulfil theirrequtremen$.

DISSEMINATION OF FINDINGSWe have adopted a broad approach for thedissemination of research findings. For each researchproject undertaken, we aim to alert relevant personsand organisations to the thrust of research programsand their results. Up to five different methods may beused. These are shown in Table I under the heading"Dissemination of Findings". Details of the disseminationmethods are provided in the Note (see page 46).

Cooperating with other research institutions and landmanagement groups that have a role in theconservation of remnants is another important part ofthe information transfer process. Organisations thathave been involved in the research undertaken byCurtin University are shown in Thble 1

IMPLEMENTATION OF FINDINGSIt is difficult to know exactly whether, and if so how,research findings have been implemented at themanagement level, because some findings result in avery diffuse set of implications. It is even more difficult

to establish whether research findings have had somebearing on the development of ideas presented inmanagement plans. It is possible that research findingsmay have had some influence on a plan, even if there isno acknowledgment of the findings in the final product.

In endeavouring to demonstrate management adoptionof our invertebrate research, we have identified threelevels at which this might occur (Table 1). Firstly, thereare instances where research findings are known tohave contributed to management actions or thedevelopment of resource information for land managers(a "Confirmed Contribution"). In such cases, verbalacknowledgment or referencing of research materialhas been forthcoming.

Secondly, there are instances where it is suspected thatresearch findings have contributed to the developmentof a general management ethos, or the adoption ofcertain management practices, but these suspicions areunconfirmed (an "Unconfirmed Contribution"). Thisoccurs when the relevant literature does notacknowledge work, and personnel with responsibilityfor the management of the relevant issue may beunaware ofthe work. Yet, it is virtually inconceivablethat management approaches taken are not influencedby the research outcomes, although this influence maybe three or more times removed from the researchersand their work.

The third level at which research findings may haverelevance to managers are those instances where themanager is simply not aware of the research, or therelationship between the research findings and the landuser or manager is less obvious. An example is theimportance of unpredictable, infrequent disturbanceevents, such as one in one hundred year flood or fireevents. While these may have an enormous impact onspecies and communities, they are poorly understood bymanagers. As a consequence, we have identified groupsof people or organisations who could potentially benefitfrom such information and listed them under theheading "Potential Contribution". Dissemination ofinformation to these groups probably requires directconsultation as a means of ensuring that themanagement implications of research findings areadequately disseminated.

Despite efforts to ensure that research findings are

appropriately disseminated, we still experienceproblems in getting our message to users. Reasons forthis are thought to include:

.1. insufficient time for the preparation of scientificpapers and conference presentations;

.1. an absence of mechanisms which support thedissemination of research information to targetaudiences;

.1. an absence ofmechanisms that provide feedback onthe adoption by managers of research findings;

* inadequate knowledge of relevant research findingsamong users, including landcare officers, trainingstaff and land managers;

t an apparent misconception by users that scientificwork is undertaken for academic rather thanpractical reasons.

CONCLUSIONResearch undertaken at Curtin Universi$ has relevanceto the conservation, management and restoration ofremnants in south-western Australia. While there aresome instances where research findings are known tohave contributed to management programs andrecommendations, we believe that much informationnever reaches target audiences. This occurs despite asubstantial effort to inform the community aboutcurrent research programs and outcomes.

Inadequate communication between researchers andmanagers is perceived to be a major cause for the poordissemination of research results. Actions that will helpovercome this problem include:

* databases ofresearchers, managers and their specificinterests and projects, to encourage communicationof information and ideas between both groups;

* developing feedback mechanisms so that theresearch community may check that research dataare reaching relevant land managers, and also sothat land managers can suggest suitable topics forfuture research;

setting up a centralised computer system or sortinghouse for research outcomes that are related to thevalues and management of remnants for natureconservation, sustainable agriculture, and otheruses. This central information base could act as aformal centre for remnant networking. It could beresponsible for the development of mechanisms thatfacilitate the dissemination of information betweenland managers, researchers and other users. Thecentre'could provide a focus for the development oflandscape models integyating nature conservationand agriculiural production values;

dealing with the perceived misconception thatresearch is undertaken for academic rather thanapplied reasons;

ensuring that research findings are taught topotential land managers (for example, traineefarmers) by circulating information to coordinatorsofagricultural courses at secondary and tertiarylevels.

We have also identified two other factors that requireattention if the nature conservation values of remnantsare to be maintained. These issues, which are worthy ofdiscussion at this workshop, are:

* Remnants on pfivate property require protectionfrom ongoing degradation processes. Protectionprograms must be accompanied by a managementprogram that ensures the ongoing survival oftheseremnan$.

* The irnportance of invertebrates andmicroorganisms in the maintenance of remnantecosystems must be realised by land managers.

THIS IS A BLANK PAGE

Table l: Summary of research undertaken by the authors and their associates in the Western Australian Wheatbelt, includingthe broad management irnplications of each topic.

Research Topic* Summary olResearCh Outcomes Management lmplications

Dissemination oI Findings

b 3

6 Ei iO E

x t

o =

.zi ;EF $

PAST RESEARCHA. Seed removal by ants

(1 ,6 ,10 ) .

Herbivory by nativecaterpi l lars (I I , l2, l3).

Role of termites in nutrientcycling (7).

D. Use of ants as bioindicators ofenvironmental healih (4,8,9).

Conservation poteniial oftemnants and corridors ofnatil)e vegetation (1,8,9).

Ii Influence of prescriptionburning on invertebrates(2,3,5).

NEW RESEARCHC. Abundance of arthropods and

birds on trees in paddocks,corridors and native vegetation.

Re-establishment of biota andnutrient cycling in r€vegetatedhabitats.

Direct se€ding operations may beimpeded by:. certain species of ants. temporal factors. si techaracterist ics.

Magnitude of damage by bag sh€ltermoths varies lrith:. temporal factors. si iecharacterist ics,

Drepanotemes tamminensis has amajor role in nutrient cycling withincertajn ecosystems.

Ant community composition is aus€ful indicator of the degree of:. habitat degradation. habitat restoration. habitat similarity.

The biotic composition and ecologyof remnant habitat is influenced byits:. physical feaiures. managem€nt.

Prescribed burning of remnantsinflu€nces invertebrate:. abundance and diversity. recovery time,

Pot€ntial of trees in different spatialalrangements to supportinvertebrates, avifauna and food webprocesses.

Degree to \Nhich differentrevegetation strategies facilitate there-establishment of nutrient cyclingprocesses. Value of revegetatedhabitat to the maintenance ofregional biodiversity.

Maintain ligilance for problem antspecles.Avoid direct seeding at vulnenble times.Treat vulnerable seeds to avoid seedtemoval.

Depending on the extent of the outbreak,control of leaf defoliation may berequired in:. outbreak years. nutrient enriched ar€as (e.9.

road verges and remnant edges).

Maintenance of remnants should include:. the ret€ntion of dead plant

material as habitat for soil faunaand the maintenance ofnutrient cycling

. the protection of termitestructures liom damage,

Ass€ssment ofthe ant community canprolide a rapid indication of the state ofthe environment.

Corridors of native vegetation will havegreater conservation potential if theyj. are wide rather than narrow. comprise native rather than exotic

plant species. are adjacent to native habitat.

Edge effect impacts on remnant habitatmust considerl. the ecosystem component for

conservation. the imposing disturbance

reglme.

Remnants should be fenced to excludefarm livestock.

Restoration of remnants willrequire amelioration of:. alteredmicrohabitats. soil physical changes. soil nutrient imbalances. altered plant and animal

composition and dynamics.

Findings may have relevance to decidingthe timing and periodicity of burning'regimes.

Ttees remaining in cor dors andpaddocks may form a nucleus forlandscape restoration.

Re-establishment of soil and litter biotaand nutrient cycling may be influencedby revegetation practices.

CSIRO

CSIROCALM

CSIRO

CSIROCALM

MiningCo's

MRD

CSIRO

:l__"1**l

to on page

Confirmed Conlribution lJnconfirmed Contribution Polential Contr ibui ion

Adoption of seed predation controlmeasures (e.g. pelletisation) byrelegetation organjsations (e.9. MRD,mining companies).

Control measures (e.9. shire councils)have not been required to date.

Other research and publicationsinvolving res€arch on ecosystem statusand habitat restoration.

Contributed to the justification ofwidening of road verges (MRD).

Strengthened the need to plant or seednative species on road verges (MRD).

Literature providing guidance on themanag€ment of remnants (Hussey,B.J.M., and Wallace, K,J,, 1993.Managing Your Bushland Departmentof Conservation and Land Management,Perth).

Auaiting results

Atuaiting results

General recognition of the problem indirect seeding operations.

Could influence choice of tr€es lorroad verge plantings.

Maintenance of ecosystem ptocessesby temnant and reserye managets.

Protection of termite habitat andstructures from destruction.

The incorporation of ants asbioindicators of environmental healthin every State and territory ofAustralia.

Supporting the thrust ofthe RemnantVegetation Protection Scheme.

Findings have stimulated furtherresearch on the impact of prescribedburning outside the jarrah foyest.

Au)aiting results

Auaiting results

Consenration organisationsMining companiesShire councilsSchoolsRevegetation organisations

MRDShire councilsConservation groups

Consenation managersFarmers managing remnantson private propertyRevegetation organisations

Conservation organisationsMinjng CompaniesShire councils

Conservation organisationsShire Councils

Technical education oflandcare advisers (e.9. LCOS)Managers of remnants onpnva@ properryLCDS and other runl groups

Conservation managersManagers of remnants onpnvare properlyManagers of road verges

Managers of remnants onprivate propertyConseNation managers

Managers of remnants onprivate propertyLCDsRevegetation organisationsConservation managers

REFERENCESThis reference list covers invertebrate studiesundertaken in the Wheatbelt excluding areas of primaryproduction and mined lands. Numbers with eachreference relate directly to those used in the Table.

1. Keals, N., and Majer, J.D., 1991. The conservationstatus of ant communities along the Wubin-Perenjori Corridor. pp. 387-393 in NatureConsenation 2: The Role of Corridors, ed,s D.A.Saunders and R.J. Hobbs. Surrey Beatty and Sons,Chipping Norton (NSW).

2. Little, S., 1989. A study ofsoil and litterinvertebrates in semi-arid kwongan vegetation withspecial reference to fire age. Honours thesis, CurtinUniversity of Technology, Perth

3. Majer, J.D., 1980. Report on a Study of Inuertebratesin Relation to the Kojonup Fire Management Plan.Department of Biology Bulletin No. 2. WesternAustralian Institute of Technology, Perth.

4. Majer, J.D.,1983. Ants: Bioindicators of minesiterehabilitation, land use and land conservation.Enuironmental Management 7 : 375-393.

5. Majer, J.D., 1985. Fire effects on inveftebrate faunaof forest and woodland. pp. 103-106 in Fire Ecologgand Management in Westem AustralianEcosgstems, ed. l.R. Ford. Western AustralianInstitute of Technology, Perth.

6. Majer, J.D., 1990. The greening ofAustralia: Tbkingthe animals into account. pp. l7-25 in Souing theSeeds: D[rect Seeding and Natural RegenerationConference. Conference proceedings. GreeningAustralia Ltd, Deakin.

7. Park, H.C., 1993. The role of the Wheatbelt termite,Drepanotemes tamnmensls in nutrient cyclingwithin native woodland and shrubland. PhD thesis,Curtin University of Technology, Perth.

8. Scougall, S.A., 1991. Edge effects in fenced and non-fenced remnants ofjam-York gum woodlands in theWestern Australian Wheatbelt. Honours thesis,Curtin University of Technology, Perth.

9. Scougall, S.A., Majer, J.D., and Hobbs, R.J., 1993.Edge effects in grazed and ungrazed WesternAustralian Wheatbelt remnants in relation toecosystem reconstruction. pp. 163-178 in NatureConsenation 3: Reconstruction of FragmentedEcosystems , eds D.A. Saunders, R.J. Hobbs and P.R.Ehrlich. Surrey Beatty and Sons, Chipping Norton(NSW).

10. Van Schagen, J., 1986. Ant-plant interactions in aWesternAustralian Wheatbelt reserve. GraduateDiploma in Natural Resources thesis, WesternAustralian Institute of Technology, Perth.

ll.Van Schagen, J., 1989. The biology and ecologicalimpact of the bag-shelter moth in the WesternAustralian Wheatbelt. MSc thesis, Curtin UniversityofTechnology, Perth.

12. Van Schagen, J.J., Hobbs, R.J., and Majer, J.D., 1992.Defoliation of trees in roadside corridors andremnant vegetation in the Western Australian\Nheatbelt. Joumal ofthe Rogal Societg of WestemAUSITAI0 /5: /5-6 i -

13. Van Schagen, J.J., Majer, J.D., and Hobbs, R.J., 1992.Brology of Ochrogaster lunifer Herrich-Schaeffer(Lepidoptera: Thaumetopoeidae), a defoliator ofAcacia acuminata Bentham, in the WesternAustralian Wheatbelt. Australian EntomologicalMagazine l9t 79-24.

NOTE: APPROACHES USED BY CURTINUNIVERSITY TO DISSEMINATERESEARCH FINDINGS

The methods used by Curtin University to disseminateits research results, shown in Table 1, are outlinedbelow.

Scientific PublicationPublication in scientific journals, conferenceproceedings and books (for example, AustralianEntomological Magazine, Australian Society for SoilScience and the "Nature Conservation" book series).This is done either during, or within one to three yearsof completing, a project.

Conference ParticipationOral and poster presentations at scientific andmanagement oriented conferences (for example,Australian Entomological Society, Australian MiningIndustry Council Environmental Workshops,Enviromine).

Popular ArticlesPublication ofpopular articles in magazines such asEcos, Curtin Gazelle or publications aimed atschoolchildren.

Semi-formal DiscussionsSemi-formal and informal discussions andpresentations to local people and potential end-usersduring the course of field study programs. Talks anddiscussions with rural land users (for example, Men ofthe Trees at Dowerin, opportunistic discussions withlocal farmers, talks to catchment group Annual GeneralMeetings, Lions Club talks).

General MediaPublicity through local and State newspapers (forexample, articles in lhe Countrgman\, radio (forexample, The Science Show) and,television networks.

MANAGERS' PERSPECTIVES

@

Setting the Scene-RemnantManagement during the Past Decade

K.J. WallaceDepartment of Conservation and LandManagement, PO Box 100, Narrogin, WA, 6312

INTRODUCTIONIn this short paper, I summarise ways in whichmanagement of native vegetation remnants haschanged over the decade 1983-93. My aim is to sketchthe more significant changes and thus provide a contextfor later papers that will necessarily have a narrowerfocus,

At Busselton in 1985, I outlined (Wallace and Moore1987) five elements crucial for effective management:

i an adequate research database;

* an adequate technical database;

.i. an informed and sympathetic public;

.i. adequate staff and financial resources;

.i. an accepted philosophy/methodology for drawingtogether the above and implementing management.

These provide a useful framework for examining thechanges over the decade; however, the term"knowledge" is more appropriate than "database", andthe inadequacy of the third point will be discussedbelow.

Ofthe five points listed, the first is discussed in detailwithin other workhop papers, and is therefore notconsidered further here. Also, no attempt is made tocomprehensively describe changes; rather, the aim is tosketch the major shifts and issues.

CHANGE AND THE ELEMENTS OFEFFECTIVE MANAGEMENTTechnical l{nowledgeThe technical information and knowledge among, andavailable to, managers has improved over the decade.The many advances have included: the more effectivefield use of herbicides, better techniques for fox control,improved availability and use of computers, and

refinements concerning the safe use of fire, particularlyin coastal and near coastal ecosystems of the SouthWest. Also, the quality ofvehicles and other equipmenthas generally improved.

However, the value of new technologies has not alwaysbeen realised, or been accessible to managers. Forexample, the promise of Geographical InformationSystems and remote sensing has not been generallyrealised (Burbidge and Wallace 1995), and the outlay ofresources in these areas should be justified in relationto alternative projects.

Despite these concerns, managers have mostly gainedfrom changes in technical knowledge and equipment.Computers, in particular, have been an important factorin counterbalancing some resource constraints,especially with regard to clerical aspects ofmanagement,

An Informed and Sympathetic PublicSince 1983, growing concern due to land degradationhas resulted in greater support for management ofremnant vegetation. This change is paralleled byincreasing public concern for the environment, at leastin the United States (Dunlap 1987, 1991). Dunlap'swork is of particular interest as it shows that althoughpublic support for environmental issues has increased,this is not necessarily reflected in electoral votingpatterns. While I know of no Australian equivalent toDunlap's work, surveys by the Australian Bureau ofStatistics (Castles 1993) show significant communityconcern about environmental issues. Furthermore,although 75%o of people sampled were concerned aboutenvironmental problems, only 1070 had acted toregister their concern during the preceding 12 months.Tianslating concern into effective action is animportant and challenging issue.

Despite changing attitudes, uncertainty among keygroups as to the value of remnants in landscapefunctions, together with the rural recession, hascountered the positive social changes and detrimentallyaffected the translation of community concern intoaction. Specifically, we have failed to convince thepublic that conservation of remnants is crucial to asatisfactory lifestyle for each human. In my view,nothing less will allow us to achieve our natureconservation goals with remnant veSletation.

Therefore, as a management goal, the heading for thissection does not go far enough. The community mustnot only be informed and sympathetic; it must bepersonally committed to acting in ways that conserueremnants.

If conserving remnants of native vegetation is notcrucial to a satisfactorl lifestyle, then we must reviewour goals. Although there have been some attempts todescribe the values of remnants for a wider audience(for example, Hobbs and Wallace 1991), the communityhas not been convinced.

Furthermore, while aspects of communication andeducation have greatly improved over the last decade -for example, some of the educational materials andpublications produced by the Department ofAgriculture, Western Australia (DAWA), and theDepartment of Conservation and Land Management(CALM) - there have been notable failures. Inpar l icu lar . our inabi l i ty to ef fect ive ly educale lhecommunity about the complexity of land managementissues and the personal values of remnants has been ablow to their protection. Organisational constraints(Wallace 1992; Burbidge and Wallace 1995), politicalconstraints, and the lack of a coherent philosophy (seebelow) are some of the many factors contributing tothis failure.

At the same time, the economic recession, with itsattendant cultural change including increased illegalactivities and a more fundamentalist approach toeconomic and political issues, has not helped landmanagers. While essential, social changes involvingAboriginal issues are also complicating management.

Therefore, although there have been modest gains ineducation and communication over the past decade,these have not achieved the social change necessary toensure conservation of remnant vegetation.Additionally, social issues complicate the managers'task and present new challenges for which they aregenerally il l equipped. Some of these issues and theirsolution are discussed in Burbidge and Wallace (1995).

ResourcesCeneralOver the decade, resources allocated to management ofremnants have increased through a number of sources

including:

.i. increased allocation and better distribution ofresources with the establishment of CALM;

.i. increased interest in remnant vegetation by DAWA,reflected most significantly in the RemnantVegetation Protection Scheme and its management;

{. operation for sixyears ofthe State's RemnantVegetation Protection Scheme;

* allocation of Federal resources to protection ofremnant vegetation * for example, through Savethe Bush, the States Cooperative Assistance Scheme,funding for threatened communities and speciesprograms, the National Estate Program, parts ofboth the National Landcare Program and the OneBillion Trees Program;

.i. growing interest in, and assistance with, remnantvegetation protection through the activities ofgroups such as Greening Western Australia;

.i. greater protection of remnants by freehold land-holders themselves, independent of any externalfunds;

.l involvement of the Water Authority of WesternAustralia in remnant protection within higherrainfallareas;

* greater research by CSIRO and tertiary institutionson nature conservation issues in agricultural areas;

* the advent of the landcare movement and theformation of land conser.'ration district committees.

While these increased resources have helped, they arenot sufficient to prevent further losses of biodiversityfrom remnant vegetation in agricultural areas. Theneed for public and community support is againapparent.

Management ExpertiseA second, often forgotten, aspect of resources lies in thequality of managers. Well-trained, enthusiastic

managers achieve much more than untrained orunenthusiastic managers.

As a managet I cannot claim objectivity in discussingthis issue, However, in my estimation, our currentmanagers are on average more technically competentand better informed than they were 10 years ago. This iscounterbalanced in part by increased communityexpectations of managers, and the overwhelmingamount of important information that is poorlyintegrated and largely inaccessible to managers.

Also, a major concern is the difficulty of attractingprofessional people inland. This has always been so, andaffects all government agency managers with whom Ihave discussed the topic. Given the economic and bioticimportance of inland agricultural areas to the State,resolution of this matter is vital. The issues concernedare well beyond the scope of this paper However,conservation goals will be achieved more often if alarger group of effective professional officers is attractedto rural communities to work and participate ascommunity members. Perhaps it is time to examine theincentives used to encourage people to move to thenorth of Western Australia. The same technioues mavbe relevant to southern rural areas.

Finally, the lack of a clear management philosophy toguide managers and help them integrate complex issuesinto effective management strategies is of concern.

Management Philosophy and Effective Decision-rnaking

One of the needs for successful management I definedin 1985 was an effective management philosophy. Whilethere have been papers that have considered some ofthe relevant issues (for example, Fairweather 1993;Lefroy and Hobbs 1993; Burbidge and Wallace 1995),they represent work on the foundations of a philosophyrather than a completed structure.

If management is to be coherent, driven by a vitalvision, and characterised by effective setting ofpriorities and decision-making, then the developmentofan explicit management philosophy is veryimportant. To fully achieve such a philosophy, whichideally should reflect community ethics and cosmology,will be a challenging task in such a young nation withso many divergent, cultural strands.

CONCLUSIONSManagement of remnant vegetation is undoubtedlymore effective and better resourced than it was in 1983.However, management is neither sufficiently effectivenor well enough resourced to conserve the existingbiodiversity at a district or regional level. Furthermore,resourcing is likely to go into decline pliven the currenteconomic climate and the associated re-ordering ofsocial, cultural and political priorities. An importantmeans of countering these trends is to show people thatvegetation remnants and their biodiversity are crucialto them as individuals.

I am not sure if it is increasing age or increasingunderstanding, but more and more I perceiveconservation problems as requiring socio-culturalchanges as much as, if not more than, biologicalresearch and operational management. Biologicalresearchers and land managers are not equipped to dealwith social issues. It is time that we sought the help ofsocial scientists and related workers.

REFERENCESBurbidge, A.A., and Wallace, K.J., 1995. Practicalmethods for conserving biodiversity. pp. 11-26 inConseruing Biodiuersitg: Threats ond Solutions, ed,sR.A. Bradstock, TD. Auld, D.A. Keith, R.T. Kingsford, D.Lunney and D.P Siverstsen. Surrey Beatty and Sons,Chipping Norton (NSW).

Castles, I.,1993. Enuironmental Issues: People's Vieusand Practices. Auslralian Bureau of Statistics CatalogueNo. 4602.0. Australian Government Publishing Service,Canberra,

Dunlap, R.E., 1987. Polls, pollution, and politicsrevisited: Public opinion on the environment in theReagan era. Enuilonment 29(6): 6-37.

Dunlap, R.E., 1991. Public opinion in the 1980s: Clearconsensus, ambiguous commitment.Onuironmant33(8) :10-37.

Fainiveather, PG., 1993. Links between ecology andecophilosophy, ethics and the requirements ofenvironmental management. Australian Joumal ofEcologg 18: 3-I9 .

Hobbs, R., and Wallace, K., 1991. Remnant vegetationon farms is a valuable resource. WA Joumal ofAgr icult ure 32 (2\ : 4345.

Lefroy, E.C., and Hobbs, R.J., 1993. Some humanresponses to global problems. pp. 33-39 in NatureConsenation 3: Reconstruction of FragmentedEcosgstems, eds D.A. Saunders, R.J. Hobbs and PR.Ehrlich. Surrey Beatty and Sons, Chipping Norton(NSw).

Wallace, K.J., 1992. Services - a barrier to achievingland conservation objectives? pp. 459+66 in 7th ISCOConference Sgdneg - People tuotecting Their Land,eds PG. Haskins and B.M. Murphy. Proceedings, vol.2.International Soil Conservation Organisation, under theauspices of the Department of Conservation and LandManagement, Sydney.

Wallace, K.J., and Moore, S.A., 1987. Management ofremnant bushland for nature conservation inagricultural areas of Southwestern Australia -operational and planning perspectives. pp.259-268 inNature Conseruation: The Role of Remnants of NatiueVegetation, eds D.A. Saunders, G.W. Arnold, A.A.Burbidge and A.J.M. Hopkins. Surrey Beatty and Sonsin association with CSIRO and Department ofConservation and Land Management, Sydney.

Remnant Management over the Last TenYears-An Agricultural Viewpoint

G.R. BeestonDepartment of Agriculture, Baron-Hay Court,South Perth, WA, 6151

RESEARCH IMPLEMENTEDFencing of RemnantsThis form of remnant management is by far the mostwidespread in the agricultural zone ofthe State.Farmers accept that results can be obtained by fencing,and either do it as part of normal farm operations ormake application for special grants. The continual over-subscription of the Remnant Vegetation ProtectionScheme is one positive indication ofthe farmingcommunities' attitudes.

Protection of Remnants by Buffer PlantingThis form of management is necessary, to protect nativeremnants from rising watertables, which causewaterlogging and salt encroachment. While nowherenear as popular a management method as fencing, it isbeing increasingly used in the State by farmersinterested in preserving their remnant vegetation areas.

The choice of species in tiese plantings is veryimportant, with the preferred option being speciesnative to the district and hopefully of types similar tothose in the remnant vegetation.

Econonic Value of Remnant Vegetation in Farm andCatchment PlanningThe incorporation of remnant vegetation managementin the catchment planning process has been widelyaccepted. This will increase in future years, with agreatly increased demand for technical information onnature conservation issues.

The utilisation of managed remnant vegetation toprovide products for commercial return is a growingpractice. Wildflower and seed picking, apiary sites, andfirewood gathering are uses that most readily come tomind. There is also a valuable role for remnantvegetation in the farm stay industry. This industry isalready widespread in the lower South West of theState, but has not yet expanded to the same degree intothe Wheatbelt.

WHY HAS OTHER RESEARCH NOTBEEN IMPLEMENTED?EconomicsMany farming communities have planned for land andnature conservation at farm and catchment levels, butthis has not been translated into works on the ground.The main reason for this is the continued depression ofthe agricultural commodity markets.

Revegetation is the one form of remedial action that hascontinued through the depfessed times. The reasons forthis are rnany, and vary from district to district, but,above all, probably indicate that the community hasaccepted the extension message on this topic.

Landscape ImportanceThe other factor which hinders implementation islinked to economics, but is mainly about education. Therole of remnant vegetation in the dynamics of the rurallandscape has not been fully appreciated by the majorityof the farming community. This may reflect the way theconcept has been presented.

Education materialwhich targets all sectors of the ruralcommunity is needed so that the rural communityeventually sees the ecosystem, or "big picture" of thelandscape in which they farm. Specific targeting ofschoolchildren and women is suggested as a way ofreaching receptive and influential sectors of thecommunlty.

Technical ExtensionIncreased awareness ofthe importance of the "big"

landscape picture by the farming community, togetherwith the trend to more broad-scale planning by landconservation groups, could increase the need fortechnical information in the future. Imoroved economicconditions will expand this need.

Currently, there is a lack of detailed information aboutthe biology ofAustralian native species that live inremnants. Also, while there is much information onland management in scientific papers and manuals, thisneeds to be collated and presented to managers inreadily accessible forms. Weed scientists present us witha model on how this should be tackled.

THE FUTUREThe role of remnant vegetation as an importantcomponent of the landscape will continue to grow inthe perception of the rural community. The challengefor all scientists and extension specialists will be to havemeaningful answers that enable rural communities toretain and expand both the areas and the speciescomposition of the remnants.

fl.rianag n mg ilVTy ffi p*s$r nae'nel

Alison Doley"Koobabbie", Coorow, WA, 6515

INTRODUCTIONMy brief from Ken Wallace was to outline what has beendone to preserve remnant vegetation on our farm, andto outline the information on which we base ourmanagement,

John and I farm "Koobabbie", which is part of the WaddiForest district of south-east Coorow. The land wasselected by my grandparents in 1906, and developed bythem and by my parents. The development was carefullyplanned, and waterways, granite, diorite and lateritichills, and areas showing signs of natural salinity werenot cleared. Strategically placed belts of woodland wereretained for shelter. Unfortunately, only one rocky hillwas permanently fenced from livestock.

My parents died prematurely, and by 1966 John and Iwere farming "Koobabbie" on our own account. We nowown 6 855 ha, of which 4 028 ha is arable or used forgrazrng.

The remaining 2 857 ha, or 41.50/o of the farm, can bedescribed as uncleared salt Iake country not grazed bystock (1 864 ha), rocky hills, most of which are fencedfrom stock (521 ha), and timber belts, most ofwhichare fenced from sto ck (472 ha) .

There are three major soil and vegetation types:sandplain on lateritic soils above the Darling Scarp,woodland on duplex soils, and saline lakes (Beard 1976).

MANAGEMENT OF REMNANTSThe property has approximately 196 km of fencing. Ofthis,43 km is fencing necessary to fence the bush fromlivestock, and 5 km protects salt-affected land fromgrazing by sheep.

The fenced-off remnant vegetation is in 47 separateayeasJ ranging from 2 to 1 682 ha in area. Many of thefenced remnants are only separated because of thenecessity to have a Ianeway for stock and vehicles toaccess paddocks. Of the 47 fenced remnants, only sixare islands in cleared paddock. The remnant vegetationfencing program began in 1970.

We still have 12 remnants that urgently require fencingbut, unfortunately, they require long lengths of fencingfor small total areas of bush.

The only financial assistance received has been $625towards fencing, and I 200 seedlings as part of aGreening Western Australia grant to the Waddi ForestLandcare Group.

The program to fence off remnant vegetation was givengreat impetus in the early 1970s, when I read BetweenWodjil ond Tor by Barbara York Main (1967). This bookclearly demonstrates the profound changes wrought onthe natural ecosystem by even minimum interferenceby agriculture. Although depressed by the realisation ofthe degradation of our bush, I resolved we shouldendeavour to preserve all the remnant vegetation, nomatter how degraded. As a result, we have given firstpriority to fencing remnant vegetation, and our treeplanting endeavours are aimed at reconstructing themore degraded areas, in preference to planting incleared paddocks. Ideally, the bush would not needmanaging, but because of the impact of weeds, feralanimals, and chanples in indigenous flora and fauna,intervention is required.

At about the same time, I met the late CharlesChapman, who farmed at Winchester (32 km north-west of "Koobabbie"). He was a dedicated amateurbotanist, and on that first meeting named three speciesof plants. Because of his encouragement, I became veryinterested in learning about the local flora.

In 1976, Roy Casey came to work for us. At last, we hadan expert fencer, and the fencing program began inearnest,

It was in approximately 1980 that Brian Jack purchasedan area of bush west of Coorow and established a plantnursery. Until the early 1980s, the idea of tree plantingin our district was to make an annual purchase fromthe Forests Department nursery at Dryandra, usually ofgoldfields species or Eucalgptus camaldulensis (rivergums), and 50 was a large order.

Brian quickly convinced me of the importance ofgrowing local flora, preferably using seed collected onthe farm, and ideally from the site where it was to be

replanted. Since the mid 1980s, I have been collectingseed which Brian grows for me. He pasteurises his soil,and supplies plants free of soil-borne diseases andweeds. We no longer purchase or receive plants fromany other source.

In the early 1980s, Brian and Victoria Syme wereinstrumental in establishing a wildflower group in thedistrict. As the emphasis of the group changed tonatural history in general, many interesting peopleaddressed our meetings, and often stayed at"Koobabbie". I cannot stress too much the importanceof having people like Roger Edmiston, Penny Hussey,Denis Saunders, Barbara and Bert Main, etc. cominginto the area, providing ideas and encouragement, andrecommending sources of information. Fortunately,Roger Edmiston was one ofthe first speakers, inapproximately 1982. We were using the post. holedigger in our clay to make tree planting easier. SinceRoger's visit, we have ripped the soil before treeplanting. After reading about a specially developedauger attachment for tree planting inRural Research,we purchased one. It is useful for planting amongexisting trees, but requires a strong armed man withplenty of time to operate it.

In 1987, an area of 2 208 ha, consisting mainly of saltlake countrl 2 km west of our house, came on themarket and we purchased it. We farm the 348 ha ofarable land, and exclude stock from the remainder It isa very interesting area with a wide variety of habitats,and is regenerating well. Nine hundred plants of a newspecies of Halosarcla grow around one lake, and willeventually be described as H. koobabbiensis. The otherexciting discovery was Ptilotus caespttuloszzs, collectedby James Drummond in 1842 and not collected since.An area like this provides little long-term grazing, andshould be preserved in perpetuity. Excising the area as areserve has management problems. I would like to see alegal mechanism developed so that land-holders can, ifthey wish, place permanent preservation orders onareas of bush they have preserved, which transfer withthe land title.

Some areas of bush we have fenced have Gastrolobiumpaniflorum (box poison) regenerating. Because thisspecies is poisonous to introduced species ofvertebrates, it provides some protection to the bush.Legal protection will still be necessary to prevent

clearing, but the poison will ensure farmers do notallow stock to graze on a short-term basis.

In 1987, an article appeared in the Cozrntrgmon, animportant line of communication to farmers, in whichDenis Saunders called for volunteers to keep a checklistof birds sighted each week. We still keep the list, andhave learnt a lot about birds in the process.

About the same time, Denis spoke to our wildlife group,and alefted us to the plight of the Carnaby's cockatoothat nests in the timber belt that runs past our house,and on the salt lakes.

In 1990, we commissioned Dr Stephen Davies toundertake a survey of the wildlife of "Koobabbie".

Stephen spent two weeks here in autumn and spring. ltwas my first experience of using pit and Elliott traps,and mist nets. We already had some information onreptiles present, because Betty Wellington has made anumber ofvisits here. Our CSIRO bird checklist was thebasis for the bird list, and I made a plant list.

At present, the list of reptiles recorded for "Koobabbie"

stands at 18, bird species total 93, and the veryincomplete plant list numbers 313. There is notsufficient information on frogs to make a list, but theyare here.

In the late 1980s, the Waddi Forest Landcare Group wasestablished. At one stage, with Viv Read and PhilBellamy as advisers, our group conducted somevaluable workshops on mapping, salt land revegetation,and remnant vegetation. However, the office bearerstend to be overwhelmed by the bureaucracy and thepaperwork the landcare movement generates. For thelittle financial support provided by governments, it hasbeen suggested we would be better off incorporatingand financing ourselves. At present, our group islacking momentum and ideas.

The work on remnant vegetation done in the lastdecade is now reaching farmers in publications likeManaging Your Bushland by Penny Hussey and KenWallace (1993), and Reuegetation Guide to the CentralWeatbelt by E.C. Lefroy, R.J. Hobbs and L.J. Atkins(1991). A similar publication for the northernagricultural areas is currently being prepared.

For John and me, long-term farming goals are verl

important, We rarely use insecticides, are involved in aDepartment ofAgriculture biological program for thecontrol of insect pests in pastures, and endeavour tominimise the use of herbicides and fertilisers. We striveto ensure that farming on "Koobabbie" is sustainable inthe long term environmentally and economically, andaim to keep our farm as an interesting and pleasantplace to live and work.

REFERENCESBeard, J.S., 1976. Vegetation Survey of WesternAustralia, 1:250 000 Series. a) Dongara, b) Geraldton,

c) Ajana. Vegmap Publications, Perth.

Hussey, B.J.M., and Wallace, K.J., 7993. Managing YourBushland. Department of Conservation and l,andManagement, Perth.

Lefroy, E.C., Hobbs, R.J., and Atkins, L.J., 1991.Reuegetation Guide to the Cetztral Weatbelt . BulletinNo. 4231. Department ofAglriculture, Western Australia.

Main, B.Y., 1967 . Betueen Wodjil and Tor. JacarandaPress and Landfall Press, Melbourne and Perth.

r&m Sraes'a'tfr onal Perspectiwe

Steve GortonDepartment ol Conservation and LandManagement, PO Box 100, Nanogin, WA,6312

INTRODUCTIONfu a manager, I've witnessed a number of changes overthe last 10 years in the management and research ofremnant bushland.

Initially, I was involved solely with the management ofDryandra, a relatively large area of State forest, wherethe conservation values are obvious. Following theformation of the Department of Conservation and LandManagement (CALM), my management role expandedto include many small reserves that, in my mind, hadvery little value. However, with time and the assistanceof research scientists, I now view these areas as beingcritically important for the protection and conservationof different species and communities of the Wheatbelt.

In addressing the topic, I will list the main researchprojects that have occurred in the Narrogin District inthe last 10 years, state examples ofwhere researchresults have been incorporated into management,suggest how liaison between managers and researcherscould be improved, and, finally, outline futuredirections for research in the Narrogin District.

RESEARCH PROJECTSNarrogin District, which covers much of the centralWheatbelt, has been the focus for a variety of researchprojects. This diversity is illustrated by the following listof projects that have commenced in the last 10 years:

.1. fire ecology studies, including regeneration andfauna recolonisation, in Dryandra and Tutanning, byboth CALM and tertiary institutions;

.!. flora and fauna studies - for example, TonyFriend's work on the biology ofthe numbat(Mg rmecob ius fasci at us);

.:. tertiary institution work, including wholecommunity monitoring, such as JonathanMajer's study of invertebrates in Dryandra;

arboretum and tree establishment plantings,completed by departmental staff;

fauna ecology and other studies by CSIRO in theKellerberrin area;

the establishment of the Tree Research Centre(now defunct), which primarily studied theregeneration of woodland communities;

rare flora, species management, diseasemanagement and weed control research by CALMstaff. This research in some instances was conductedby local staff;

wetland monitoring - in particular, boremonitoring at Lake Toolibin;

* taxonomic reviews;

t continued tertiary monitoring and research;

.i. sandalwood research and development, includingseed collection and trial plol eslablishment;

* research into tree establishment by CALM and theDepartment of Agriculture.

INCORPORATION OF RESEARCH INTOMANAGEMENTOverall, the above projects have provided the NarroginDistrict with a broad research base for management.Specific examples ofwhere research results have beenincorporated into management are listed below:

.1. preparation of area management and recoveryplans on the basis of research and operational work(for example, the Draft Dryandra Management Plan);

* amendment of fire regimes in the DryandraWoodland in accordance with researchro .^nnon. ia l i ^nc .

.1. introduction of predator (fox) control on anoperational basis;

* development ofdirect seeding techniques by the

Tree Research Centre, which have been tried withmixed results;

{. successful translocation of numbats to BoyaginNature Reserve. District staff assisted with this work;

J. establishment offurther sandalwood seedresources in Dryandra, following research anddevelopment by scientists and local staff.

MANAGER AND RESEARCH LIAISONWe at Narrogin have been fortunate that importantresearch work has happened on our doorstep and theresearchers involved have always made time available tospread their message. Consequently, our whole group,from district manager to maintenance worker, isexposed to the latest research results. Other researchhappening elsewhere is not as accessible, so we rely onrelevant "readable" publications such as Landscope,Managing Your Bushland (1993), and rare florapublications, to provide us with up-to-date information.

I believe liaison with researchers is the regionalecologist's role. Hopefully, he or she can keep districtpersonnel informed of relevant research and itsapplication to management. This task could be achievedby providing district managers with periodic reviews ofcurrent practices. Recommendations in such reviewsneed to be clear and have measurable outcomes if theyare to be incorporated into management.

To further improve liaison, researchers should briefdistrict staff when a research program is to beimplemented in their district. Some ownership andinvolvement by the district personnel are essential. Icommend Tony Friend in this area for his past liaisonwork with the Narrogin District staff.

Finally, all recommendations generated by researchhave a cost. It is important that researchers address thecost of incorporating results into managementprograms.

FUTURE RESEARCH DIRECTIONfu a managet I now take the opportunity of listingareas that we in the Narrogin District believeresearchers should investigate further.

FireFire ecology, particularly with respect to woodland andheath regeneration, remains an important researchtask. The Naffogin Tree Research Centre was, I believe,breaking into that area of research when it closed.

R€mnant ManagementMinimum specifications for the successful managementof communities and species within small reser"ues arerequired. Guidelines are also required so that managersmay evaluate whether communities or species are indecline.

CorridorsSpecifications for the development of corridors arerequired. For example, what is the minimum width ofan effective corridor?

Plant Regeneration by SeedInformation on seed production and the natural storageof propagules from understorey species in Wheatbeltreserues is required.

CONCLUSIONMy conclusions are:

* Some research projects have provided the NarroginDistrict with a sound scientific basis formanagement.

.i. The most successful research projects have includedclose liaison between researchers and managers,with operations staff assisting researchers andgaining first-hand understanding of the project andits implications for management.

* The regional ecologist has an important role to playin reviewing local operations and synthesisingresearch results into management prescriptions.

.!. New research - for example, on corridors andregeneration - could assist my district to manageremnants more effectively and to regenerate thedegraded areas better.

Our district looks fonrrard to the continued goodworking relations with research personnel, and hopesthe feeling is mutual.

REFERENCEHussey, B.J.M, and K.J. Wallace, 1993. Managing YourBushland. Department of Conservation and LandManagement, Perth.

Terri LloydPO Box 52, Dumbleyung, WA, 6350

INTRODUCTIONDongolocking is in the Dumbleyung Shire and lies 22km north of the Dumbleyung township andapproximately 210 km south-east of Perth. The districtof Dongolocking and the Dongolocking Nature Reservesare high in the landscape, and lie roughly halfwaybetween Toolibin Lake and Lake Dumbleyung.

The Dongolocking Nature Reserves are made up of 13separate blocks which total 2 269 ha in area. They arehome to at least 82 species of birds, seven species offrogs, 23 species of reptiles and 10 species of nativemammals. The Dongolocking Corridors Project aims tolink these separate blocks, other public reserves, bushon farms, and roadside vegetation with substantialpassages of revegetation consisting of local species.

The Dongolocking Corridors Project was set up while Iwas employed by Greening Western Australia. I hadworked and lived in the general area for 11 years.During this time, seminars on Wildlife on Farms,Ribbons of Green, and Fox Control had been held in theGreat Southern Region, and bus trips had taken placewith the general theme being remnant vegetation, andits plants and animals, among farmland.

ACTIVITIESBefore stading our corridor fencing, we held a publicmeeting for all farmers in the Dongolocking district,with short talks, "cuppa" tea, discussion and then a fieldtrip. Most farmers were willing to cooperate, with onlytwo farming families declining the offer of subsidisedfencing for the protection of bush.

Our first step was in 1987, when we planted nativespecies on the road verge ofthe newly made TincurrinRoad, which runs parallel to the Dwelyerdine Road andintersects the Dongolocking Road to the south. In 1991,with the help of schoolchildren, the local Dongolockingcommunity planted several kilometres of the oldDwelyerdine Road, shortly after it had been realigned bythe Dumbleyung Shire Council. This corridor connectstwo Dongolocking Nature Resenr'es (Reserr,e Nos 19082and 19083) with the Dongolocking Road.

ln the same yeal a Save the Bush grant enabled 14Dongolocking farmers to fence 32 km of unformed roadreser"ue and several more kilometres of connectingcorridors to bush on farms, roadsides, public reservesand Dongolocking Nature Resel.ves. This linked 11reserves into the corridor system. Within thesecorridors on private property, we are now replantingand direct seeding local gene pool plants to recreatebush for the movement of plants and animals from oneremnant to another Wth supplementarl plantings ofunderstorey this year, most plantings will have includedboth canopy and understorey species.

RESEARCHWe have just become the volunteer "legs" for scientificresearch, or "barefoot biologists", to answer our ownquestions ol

.! "How wide do we make our corridors?"

* 'lvhat is going to use our corridors?"

To do this, we have begun a six year study of three ofthe corridors planted in 1992 and 1993, under theguidance of Drs Graeme Arnold and Graeme Smith,Senior Research Scientists, CSIRO Division of Wildlifeand Ecology, Perth.

We began monitoring the movement of birds frorn onepatch of bush to another in March of this year (1993),by listening and looking for birds at either end of, andalong, our corridors. We record species and numbersseen and will continue to sur.;ey bi-monthly until 1999.We are about to set up pit traps to see what smallinvertebrates and vertebrates are using our man madecorridors, and we will monitor these for a week at atime in December, January and February for thefollowing six years.

We are fortunate in having Graeme Arnold's guidance,expert advice and teachings and Graeme Smith'sdirection and support. We also have support and helpfrom the friendly and enthusiastic people who work inthe Narrogin and Katanning offices of the Departmentof Conservation and Land Management. We have beenhelped and encouraged by easy access to people dealingin remnant vegletation research.

We have a reasonable reference library, including theimportant publication by the WA Museum (Chapman el4L 1978) that provides the basic biological reference onthe Dongolocking area. It is an invaluable guide for ourrevegetation work.

Our efforts have taken a lot of time, money and "hard

slog". The project requires seed collecting and sorting,weed and rabbit control, site preparation, directseeding, planting and fencing, and we are still far fromrecreating bush with all of its diversity ofplants and agegroups - only time and plant succession will do that.The bush corridors we are growing are small in area -19 ha in total - and compare sadly with the 240 ha ofremnant vegetation adjacent to the DongolockingReserves cleared in the past 10 years.

We cannot make long-term advances without adequateextension programs to publicise our results and theirpractical applications. For example, at a recentDumbleyung Land Conservation District Committeemeeting, a motion was passed to support a local farmerin his intent to clear 197 ha ofvirgin bush. This isdisappointing in a Shire where only 6.3%o ofprivate landoccurs as remnant vegetation (Coates 1987).Furthermore, less than 1% of remnant vegetation in theShire is of a size and condition to be of ecologicalsignificance (Environmental ProtectionAuthority1993), and 197 ha accounts for a considerableproportion of this meagre total.

The amount of public land under vegetation in theDumbleyung Shire is estimated at 4.1% (Coates 1987).

CONCLUSIONSIt is heartening to see a positive change of attitude tobush and its worth among our volunteer work force atDongolocking. We have enjoyed extensive coverage ofour activities in the local, regional and State wide press.Regrettably, we still have a hard road to hoe to convincethe local farming community of the importance ofremnant vegetation.

More emphasis needs to be placed on extension workpromoting the retention, management and appreciationof the little remaining bush that we have.

REFERENCESChapman, A., Dell, J., Kitchener, D.J., and Muir, B.G.,7978. Biological Surueg of the Westem AustralianWheatbelt, Part 5: Dongolocking Nature Reserue.Records of the Western Australian Museum.Supplement No. 6.

Coates, A.M., 1987. Management ofNatiue Vegetationon Farmland in the Wheatbelt of Westem Australia.Report from the Voluntary Native Vegetation RetentionProject supported by Conservation Council of WesternAustralia, Department of Agriculture, Department ofConservation and Land Management, Department ofRegional Development and the North-west, Depadmentof the Arts, Sport, the Environment, Tourism andTerritories, Environmental Protection Authority,Creening Australia, Land Resource Policy Council(Department of Premier and Cabinet), and WesternAustralian Farmers Federation, Perth.

Environmental Protection Authority, 1993. Clearing of197 hectares of Land for Agriculture - WilliamsLocation 13895 North Kukerin. Bulletin No. 689.Environmental Protection Authority, Perth.

A Farmer's Perspective-Management of Remnant Vegetation

Bob TwiggPO Box 86, Jerramungup, WA, 6337

INTRODUCTIONThroughout the Jerramungup Shire, to my knowledge,the only research finding relevant to remnantconservation that has been implemented is thenecessity to fence stock out. Certainly, on our farm thishas been the case.

No doubt all or most ofyou would have read the bookManaging Your Bushlarzd by Penny Hussey and KenWallace (1993), who list fencing as virtually the firstaction in the first year of a program to conserveremnants.

However, if we look into paddocks when driving aroundthe country, it is immediately evident that only a verysmall percentage - pevhaps syo, or even only 0.5% -

of remnants are actually fenced. Does this suggest thatthe research finding is wrong, or is it irrelevant? Andwhat do we mean or conjure up in our mind when wetalk about remnants? Do we think of once pristine areasnow slightly, or greatly, degraded?

To address the second question first, I believe the needto fence remnant vegetation applies equally to allremnants that are to be conserved, irrespective oftheircondition. But what of the first question? My opinion isthat, generally speaking, fencing remnants fromlivestock works well. However, there are costs anddifficulties simply because there is no such thing as a"free lunch".

COSTS AND DIFFICULTIESFor today's exercise, I have identified five costs anddifficulties.

Costs of Erecting FencesThe first and obvious cost is dollars - $l 400.&m formaterials and $450/km for erection to completeconventional fencing, or $700/km plus $300,&m forerection of electric fencing, not counting the energiserand other initial costs. Bank managers and farmconsultants are usually interested in "the bottom line".Despite the long-teym benefits that may arise fromconserving remnants, there is no immediate benefit interms of dollar returns from fencing them. Also, once

the fence is there, it has to be monitored andmaintained lorever if lhe remnant is going to besuccessfully conserved. This adds an ongoing annualcost.

Costs of Damage to Fences

Damage to fences is an additional cost. This can happenin a few ways:

Pressure of stock (literallg): Modern uopping

techniques require the elimination of grasses beforesowing cereals. Therefore, before sowing, pasturepaddocks need to be heavily grazed during spring,summer and autumn. If remnants are within thesepaddocks, then fence damage is likely due to "the

grass being greener on the other side of the fence".To reach green vegetation, sheep will push onfences, thus causing damage.

Flooding: lf the remnant adjoins a waterway, healysummer or autumn rainfall may result in fencesbeing washed over or away.

Farm machinery: All farm fences are subject todamage by careless use of machinery or vehicles,especially during the cropping phase.

Wldlife: Kangaroos or emus caught in fences canbreak wires and cause considerable fence damage.

Flre.'Apart from wildfires, deliberately lit fires arealso a factor causing fence damage - for example,where stubble burning is being practised on thefarm. This controversial practice has begun loreappear as farmers experidnce difficulty seedinglegume crops into heavy cereal stubbles. Higheryielding crops, and their associated heaw stubbles,are occurring due to the rapid adoption of newtechnology.

FirebreaksThese are costly in many ways. In some shires duringsummer - Jerramungup, for example - firebreakmust be constructed and maintained around theboundaries ofbush. Like all firebreak, those aroundremnants suffer from water erosion, which may becomecostly.

Encroachment by Annual Pastures or Weedslf remnants adjoin a waterway, it is virtually impossibleto prevent pasture residues and seeds being washed intothem. To prevent extra contamination with weeds, thesepatches would need to be sprayed annually with aherbicide.

Loss of Shelter for StockThis applies particularly to protection of lambing ewesor freshly shorn sheep. Remnants, at least in theJerramungup area, cannot be used as shelter for stockunless the farmer is prepared to eradicate native poisonplants and accept the loss of the understorey throughgrazing. Grazing will, if it is intensive enough and overa sufficiently long period, result in the loss of a remnantdue to ringbarking of trees and shrubs by stock. Somepristine or partly degraded remnants have been fencedto provide some protection to stock, but these are farfewer than those that are unfenced and partly degraded,and where stock can get right in among the bush.

SUMMARYIn summary, then, I wish to make just three points:

Only one research finding has been generallyimplemented by farmers, and even that in a minuteproportion, although the finding works well.

Protecting remnants is expensive and very time-consuming and shows no immediately obviousannual return in terms of income.

* Perhaps future research needs to address the lowadoption of a basic finding.

REFERENCEHussey, B.J.M, and Wallace, K.J., 1993. Mmaging YourBushland. Department of Conservation and LandManagement, Perth.

a-

WORKSHOP REPORTS

r_-!@

Workshop Reports-lntroduction

K.J, WallaceDepartment of Conservation and LandManagement, PO Box 100, Narrogin, WA,6312

PROCESSFollowing the presentation ofpapers by managers andresearchers, a plenary session was held. During thissession, participants identified ideas and issues crucialto the effective management of remnant vegetation.These ideas were then grouped into categories andtanked for consideration by small discussion groups.

After discussing their topic, groups reported theirconclusions to a final plenary session. Each groupprovided a written summary of its discussion, togetherwith any relevant points that arose during the finalplenary session. Reports from the small groupworkshops are presented in the following pages. Theplenary session points that stimulated each topic arelisted with the relevant repot.

ISSUES LISTED. BUT NOT DISCUSSEDDue to time constraints, a number of topics and manyissues were not discussed, despite their importance.Issues that were not covered by small groups are listedbelow. This list supplements the workshop reports byproviding insight into the range of matters pertinent tothe effective management of remnant vegetation.

Research Issues.i. Start a "gene collection" from those remnants that

will become entirely degraded.

Remote sensing is not widely used by theDepartment of Conservation and Land Management(CALM). Why?

Describe fauna diversity in the same manner asStephen Hopper's paper (this volume) on floradiversity.

.i' A database on Crown remnants is required.

.:. Research and management program requiredconcerning buffers for remnant vegetation onfarmland.

Problem: as rare fauna increase - for example, as aresult of fox baiting - there may be a detrimentaleffect on flora.

Need to understand fauna and flora dynamics. Donot necessarily want to keep "maximum" populationnumbers of animals. Animals do not necessarilyneed to be visible.

How do you cater for cumulative and synergisticepisodic events?

.!. What level of fox baiting do we require?

District trials on fox baiting are required - forexample, to determine the most effective baitingregime.

.! How degraded is too degraded?

How do we make sufficient seed available forrevegetation?

.i. Disturbance: how do you get the right disturbance?

.i. What is the functional hydrologic unit for a farmer?

.i. Definition of functional units for managementtargets.

Integrating values in assessing land clearingproposals.

The role of native pests in revegetation (for example,twenty-eight parrots, kangaroos, insects, etc. may allhave a detrimental effect on revegetation wofk).

The need to develop commercial products to ensuresustainable agriculture.

List, on a species basis, constraints on regeneration.What is the role of shade, space, moisture, fire,nutrients, etc.?

Water harvesting in remnants: how does it occur?What are implications for revegetation? Implicationsfor use of local species of plants?

*

{.

'i' What is the relative importance ofcats and foxes as explaining to key decision-makers the degree to whichpredators? Iand degradation has already occurred, and just how

much worse it is likely to become..i. Understanding of soil processes, including the role

offungi. .!. Who is going to be the messenger? Who is credible?

How Do We Make Land Management Decisions? t Equity issues in catchments - for example, who is.i. What is an easy way for farmers to monitor the going to tell the politicians that there is a problem?

health of remnant vegetation? other Issues

.:. Networking within organisations is a more effective t Minimise resources required to implement research.

way of getting things done than using formalhierarchies. .:. Agricultural systems require an overhaul ofthe

production system.

{. In relation to rising watertables, there is a need for"risk evaluation" for remnant vegetation * Nintendo numbat game - an example of a way of(determining management priorities). promoting some ofthe issues.

* The,,bottom up" approach is possibly the most * "sNAG Farmers" are required.

effective means of decision-making * Databases of managers and their interests. Butwho

* Issues of male-female relationships in decision- maintains? Must be well packaged'

making. .:. How do we overcome community prejudices agalnstinvertebrates?

.! Who makes decisions and how?.i. Database of "facts" for managers? Who maintains?

.i. Are expert systems an option? Should develop a proforma that forces people tofocus on implications of research.

What Are the Values of Remnant Vecetation?.l what is the dollar value to farmeis of retaining

{' South west is biggest island from cretaceous - we

remnants and of revegetation? must get across to people the extreme history of theflora.

'i. Does the community accept nature conservation as * The role of drainage in agricultural production anda worthwhile goal in its own right? land conservauon.

* [s "uniqueness" a value of remnants? '] Is the "groundwater" message getting through tothe public?

* we need a better description of remnant value * what about clearing contrors?

.t Does real conservation lead to increased production? .1. There is a delay in the implementation of researchresults by managers. Researchers should quantify

.1. Does the value of the remnant equal price? the management implications of their work _ this

.:. Ecotourism: does it have value for remnants? may enhance management adoption

.1. Is rare species consenration valuable, given theThe Messen$er urgent need to expend resources on fixing bigThe "messenger" is the person charged with the task of processes?

How Do We Achieve Gommunicationwithin and between Agencies?

Ken Atkinsl and Helen Allison,lDepartment of Conservation and LandManagement, PO Box 104, Como, WA, 6152'?Department of Agriculture, Baron-Hay Court, SouthPer ih, WA,6151

Workshop GroupHelen Allison, Ken Atkins, Gordon Friend, RichardGeorge and Ken Wallace.

Points from Plenary SessionThe following points, raised during a plenary session,stimulated this workhop topic:

.i. Regional ecologists - are they the appropriateinterface between research and management?

.:. How do you get researchers and managers to talk toeach other?

.1. Do we need a central focus group? What should itbe?

.1. Formal networking on remnants. Who organises it?

.i. Technical extension and remnant management.

.1. Technical advice on species for advisers - forexample, "facts" sheet on proposed speciesintroductions.

* Wanton oversights by end-users.

STAKEHOLDERS AND PROCESSBefore discussing particular issues and actions, thegroup first listed stakeholders and dealt with aspects ofthe communication process.

StakeholdersThe main agencies involved in remnant vegetationresearch and management are:

Federal.i. Australian Nature Conservation Agency{. Department of Primary Industry and Energy* CSIRO.i. Land and Water Resources Research and

Development Corporation

.1. Rural Industries Research and DevelopmentCorporation

* tertiary institutions.

State.i. Department of Agriculture, Western Australia

(DAWA)

* Department of Conservation and Land Management(CALM)

* Department of Environmental Protection (DEP)

* Department of Planning and Urban Development.:. Integrated Catchment Management Group* Kings Park and Botanic Garden* Office of Catchment Management* Water Authority of Western Australia* Waterways Commission.

Non- g ou emm ent O r g an is at i on s.i. AustralianConservationFoundauon* Conservation Council of Western Australiat Greening Western Australia* Men of the Trees* World Wildlife Fund.

Communication Flow DiagramThe interactions within and between researchers andmanagers are generalised below. The arrows indicatethe direction in which information and ideas arecommunicated.

Means of CommunicationCommunication between individuals and groups may bethrough one or more of the following channels:

.1. paper formats (letters, notes, published,unpublished, etc.);

* verbal (including teleconference);{. computer

Processes for CommunicationLinkages between groups and individuals who share acommon interest may be maintained by using a rangeof methods including:

* informal network;* formal network;* research working groupi.! workshops and technical workshops;.1. computer network (for example, that maintained by

DAWA);.l register of contacts;.i. information sheets and bulletins.

DISCUSSIONTtansfer of Information bet$€en Researchers andManagersWthin and between agencies, there is a need foreffective internal and external communication ofresearch information. This communication shouldinvolve both researchers and managers. It was agreedthat communication within agencies is not adequate.Furthermore, a specific problem was identified withregard to accessing information from universities,where many graduate and undergraduate projects havethe potential to contribute to the knowledge base.Unfortunately, this latter information is not widelyavailable.

One model to resolve this problem is provided byDAWA. It has a computer facility which lists, for eachresearch scientist, their research projects and results.This allows ready access, by scientist name or topic, tothe current work being undertaken, and also allowsaccess to unpublished results. It was considered that itwould be beneficial to link all institutions undertakingresearch - at least in the natural sciences and, inpadicular, those related to land management -

through a common database. Institutions that shouldbe involved include CALM, CSIRO, DAWA, DEP anduniversities. A problem was seen, however, inmaintaining these databases. In DAWA,, it is eachresearcher's responsibility, but this may be harder toachieve in the universities.

ActionCALM (Ken Wallace) to write to Jim Armstrong,suggesting that CALM's Science and InformationDivision develop a computer network system similar tothat used by DAWA.

ActionPromote the use of a similar database in tertiaryinstitutions.

ActionDevelopment of a network across organisations. Withregard to access to computer networks in rural centres,it was noted that DAWA and CALM are linked at SouthPerth, and DAWA has a network to major rural centres.It is therefore feasible to have CALM and DAWA regionalcentres linked through the existing networks.

Information IntegrationA problem was identified regarding the integration ofinformation into a form suitable for managers to use.Options were developed that would improvecommunication to managers. It was recognised thatoptions must be tailored for the particularcircumstances of each agency and management orresearch group.

ActionAgencies to adopt a suitable strategy forcommunicating research information to managers,using one or more of the following options:

* employ a regional person to integrate information- for example, a regional ecologist;

.i' employ an organisational person to integrateinformation - for example, an information officeras used by CSIRO;

* expect research scientists to integlrate informationand produce prescriptions for managers based ontheir research;

.i. develop collaborative projects between managersand researchers, including university studentswhere practicable, to achieve specified targets. Suchprojects involve integration from the outset, and aredriven by feedback from project results.

It was commented that, to ensure that research isapplicable to management requirements, managersneed to drive research projects.

Manager Involv€ment in Research ProjectsIn line with the fourth option presented above (and thecomment), managers should be proactive in identifuingresearch requirements and seeking means to haveresearch undertaken.

ActionManagers to identi! and describe research projects thatare a high priority. They should then circulate theirproposals to all research bodies and promote fundingunder, for example, Federal grants, and the researchprograms of agencies and tertiary institutions.

It was pointed out that CSIRO has a model operatingwith universities, whereby research projects are listedwith the intent of them being taken up as studentprojects. This model should be developed as a means forother agencies to promote their research requirements.It was noted that the provision of some funds by thesponsoring agency was not essential, but it helped. Also,it was noted that joint agency submissions couldimprove the chances of getting funds.

ActionJ. Majer and CSIRO to circulate to managers andagencies the model or procedure for the listing ofresearch projects at universities.

Coordination of Research ActivitiesResearch steering committees in agencies are intendedto establish protocols and policies for interagencyresearch cooperation. Thus this Ievel of committee doesnot provide for communication or interaction betweenresearchers and managers.

The objective of a research coordinating committee isto ensure that research within an area or topic is

appropriate and coordinated. An example of thisfunctional level is the Agroforestry Working Group. Thislevel of interaction was seen as being more applicablefor communication between researchers and managers.

Action

CALM to set up a research coordinating group forremnant vegetation management.

Extension to ManagersIt was agreed that field based, topic driven technicalextension workshops were verl effective for providinginformation to managers. These are best organised asinteragency meetings where all relevant workers areable to contribute. Results from field days should becollated and distributed to participants and otherrelevant agencies, to confirm agreements and to providea basis for planning future research and managementactions.

ActionAgencies to promote interagency technical extensionworkshops and technical review days from whichproceedings can be produced.

Networking for Information ExchangeIt was agreed that there was a continuing need forinformal and formal nehvorking. While inforrnalnetworking (between officers) provides valuablecontact, regular "formal" meetings are required, toensure that all agencies are aware of currentmanagement and research actions and to provide aforum for the exchange of information and ideas. This is(at the time ofwriting) being provided through the DEPNative Vegetation Network Group.

ActionAgencies to be aware that the Native VegetationNetwork Group is to continue meeting at DAWA.

ltl*w F.)CI bVe lnntegrate lrufns'ma'tlsm Sntoi.r F*!'l'yl Aueessfrbfie ferr' fl-amd &/fiaa"tm,gers, "r'q$ ffies€ft il"uiner,-q

"i

Anne Brandenburgl and Greg Beeston2lSchool of Environmental Biology, Curtin Universityof Technology, GPC Box U 1987, Perth, WA, 6001'?Depanment of Agriculture, Baron-Hay Court, SouthPerth, WA,615'1

Workshop GroupGreg Beeston, Anne Brandenburg, Steve Gorton andTerri Lloyd.

Points from Plenary SessionThe following points, raised dufing a plenary session,stimulated this workshop topic:

* Who creates the "Big Pitcher"? Does this really help?

* Looking at "nuts and bolts" (see paper by RichardHobbs, this volume ) can help set priorities, butsynthesis is required.

.i. "Big Pitcher" requires an integrated andinterdisciplinary approach.

.! Problem: management implications ofresearch arevery site specific.

.i. Remove land from production? Ifso, then re-allocate priorities for land use.

* Integrating information into packages.

.l. Is there too much information?

INTRODUCTIONResearchers and managers of remnant vegetation insouth-western Australia have indicated that theprotection and management of remnants must befurther improved if the conservation value of remnantsis to be maintained and enhanced. Presentations byresearchers and managers at this meeting have pointedout that there is a lot of information available onremnants, but this is not necessarily beingimplemented by managers. Furthermore, remnantmanagers are not necessarily being supported byappropriate research and information support systems.Poor communication between researchers andmanagers of remnants was considered a major cause ofthese problems.

Integration and communication of researchinformation and management needs must be enhancedif problems are to be overcome. The sheer bulk ofinformation on remnants and site specificity of researchresults compound problems of inadequatecommunication systems, Furthermore, conservationefforts must also involve an interdisciplinary approach,to achieve better management of remnants. This reportsummarises a workshop discussion and pointsemanating ftom the plenary presentation whichspecifically addressed these problems and asked thequestion: "How do we integrate information into a formaccessible to land managers and researchers?".

THE "BIG PITCHER"The synthesis of research information and thedevelopment of a unified, cohesive vision for remnantconserration (affectionately termed the "Big Pitcher")were identified by the meeting as integral tomaintaining the conservation potential of remnants.The workshop group took the concept of the "Big

Pitcher" on board. and at all times discussed theintegration and presentation of information for landmanagers and researchers within the context of a "Big

Pitcher".

DEVELOPMENT OF THE "BIG

PITCHER"Development of the "Big Pitcher" was conceptualised asbeing through a small organisation that would providea base and framework for the coordination andcommunication of information between individuals andgroups with an interest in remnant conser"uation. Itwould also act as a base for the strategic planning forremnant conserration. This would include such aspectsas conservation planning, policy development, and thepromotion of an improved consenration ethos forremnants which must be instilled in the widercommunity if remnants are to be conserved in the longterm. Hence, the "Big Pitcher" would be operational atboth applied and strategic levels.

SPECIFIC ROLES OF THE "BIG

PITCHER'' AGENCYApplied OperationsAt an applied level, the role of the "Big Pitch€r" agencywould be to undertake a series of tasks relating to themanagement of research results, identifying

management needs, and stimulating informationtransfer and extension. Outlines of some informationmanagement tasks follow.

Research* Synthesise research outcomes.

* Develop the management implications ofresearchoutcomes if these have not been done by individualresearchers.

N. Develop holistic databases and landscape modelsthat can be used and implemented at the on-ground,remnant manag€ment level,

.t Disseminate synthesised research information to theappropriate land managers.

Management.i. Identify the information needs of remnant

managers.

.1. Synthesise management needs into informationpackages for remnant researchers,

.l Disseminate information on management needs tothe appropriate research institutions.

.l. Collate a database of management needs for the "Big

Pitcher" management committee to use whenconsidering the future direction of remnantresearch,

Information Tlansfer.1. Coordinate the design of information packages on

remnant conservation. These must be appropriatelytargeted, presented and packaged for users.Interactive packages with interesting, informativeand prescriptive materialwritten in "simple

English" were considered key elements in the designof good information packages (for example,Managing Your Bushland by B.J.M. Hussey and K.J.Wallace, Department of Conseruation and LandManagement, Perth, 1993).

t Identib, the most receptive and influential group ofland managers for receipt of remnant information

packages. Children and rural women were perceivedas being ideal initial targets.

* Disseminate remnant information to target groups.Such information should include a bibliography ofresearch work done in south-western Australia and amailing list of persons from the various interestgroups and government organisations involved inremnant research and management. For thelandowner, dissemination of information wouldinclude one-to-one interactions wherever possible.All avenues for information dissemination should beexplored (for example, television, radio andaudiovisual material).

* Feed the synthesised information that identifiesresearch outcomes and management needs forremnant conservation to the "Big Pitcher"management committee and, thus, facilitatestrategic planning for remnant conservation.

Extension.1. Develop and coordinate remnant conservation

extenslon programs.

* Employ the skills of specialist extension personnel(for example, have nature conservation officers thatoperate in parallel to landcare technicians andproject officers).

The workshop discussion stressed that informationmanagement must involve two-way communicationbetween researchers and remnant managers. Whileresearchers require an avenue for dispersing theirresearch findings, managers must also have an avenuefor identifying their research needs.

Strategic OperationsThe strategic planning role of the "Big Pitcher" wouldinclude identifying conservation objectives, developingremnanI conservation policy. and promoLing aconservation ethos among the wider community. Someideas of what this work would involve included:

.1. developing avision for remnant conservation insouth-western Australia. This would involve definingconservation goals and objectives for the region, aswell as developing principles and guidelines for

remnant management. This process would befacilitated by the provision of synthesisedinformation collected by the "Big Pitcher" agencyand would also include inputs from the widercommunity;

developing the concept that remnant conservationis a land use in its own right;

promoting the integration of various land useswithin the agricultural landscape;

supporting economic incentives that facilitatelandscape and remnant conservation (for example,taxation benefits for nature conservation purposes);

developing and promoting legal rnechanisms for thevoluntary protection of remnants in peryetuity;

.:. providing support to persons with a "bent" forremnant consen/ation and a positive influence onothers;

t identi[,ing the direction of remnant research.

STRUCTURE OF THE "BIG PITCHER''AGENCYIt was considered that the "Big Pitcher" agency shouldbe an autonomous, non-government agency comprisinga committee that broadly represents remnantresearchers and managers. The committee would beresponsible for managing a group of skilled personnelemployed to undertake the specialists tasks of theorganisation (refer to "Applied Operations" above). Theorganisation should be located at a site in WesternAustralia which experiences the problems we are tryingto solve.

Other suggestions for an operational format for a "Big

Pitcher" agency which were discussed during theworkhop or plenary session included:.:. creating a remnant conservation section within a

current government agency;

* extending the current briefofthe RemnantVegetation Steering Committee;

* expanding the current Save the Bush scheme;

adopting the Roadside ConseFr'ation Committeetype model;

extending the conservation commitment of on-ground land managers (that is, land consenr'ationdistricts and local government agencies). Discussionof this option highlighted the importance ofrecognising land ownership as a critical factorinfl uencing adequate remnant conservation;

adopting a management authority approach (forexample, Kings Park Board, Rottnest Island Board);

adhering to the current situation, with interestedpersons doing more work on remnant conservationwithin their individual work environments.

A series of problems with setting up any type of newconservation agency or management authority wereidentified during the plenary session. The developmentof a new bureaucracy would require the injection of alarge amount of funding, which could prove difficult toobtain. It may also pose a threat to current conservationagencies and land managers, and would probably provea politically sensitive issue with which to deal.Furthermore, assuming that an organisation such asthat proposed would be affiliated with at least onegovernment authority as a consequence of fundingarrangements and the need for organisation credibility,it would be virtually impossible for the "Big Pitcher"agency to remain completely autonomous and not becommitted in some way to the expectations of itssupervising authority. A better approach to attaining anintegrated and accessible information transfer systemmay be to facilitate the networking process that iscurrently in place. However, given the constantreminder that there are a series of problems workingagainst remnant conservation, there seems to be atrend for a more integrated, interdisciplinary approachto conservation planning developin6l.

I'ile',v ffio Vde [ntegrmte Nature #onsenwatlonuli;tFl ffitftlelr Land Uses nn'l rthe Gomtext sf ttr'leffr*stea'm Aruetra$ Hmre W&?effitB:leBt?

Greg Keigheryl and David Mitchell,rDepartment of Conservation and LandManagement, Woodvale Research Centre, PO Box51 , Wanneroo, WA, 6065'?Department of Conservalion and LandManagement, PO Box 1 00, Narrogin, WA, 6312

Workshop GroupPenny Hussey, Greg Keighery Jonathan Majer, DavidMitchell and Denis Saunders.

Points from Plenary SessionThe following points, raised during a plenary session,stimulated this workshop topic:

.! Clearing and values of resultant remnants in thecontext of land conservation and natureconservation (incremental degradation is occurring,only 10% left, etc.).

.i. The farmer requires solutions from advisers, notjustinformation or problems (value returns need to bequick).

{. Is "fixing" the agricultural system a prerequisite fornature conservation? Is it a prerequisite forsustainable agriculture?

.i. Is fencing of degraded remnants worthwhile?

.!. Local versus nonJocal and exotic species inrevegetation - which should be used and when?

* What is the role of economic plants in revegetation?

INTRODUCTIONIn the Wheatbelt, we have a matrix of cleared landenclosing scattered remnants of uncleared nativevegetation (Crown and private) of varying size andshape. The remnants are the responsibility ofnumerousdifferent managers. Their management responsibilitiesare imposed over, rather than aligned with, a mosaic ofsoil and drainage patterns.

Research indicates that the current system of naturereserves will not sustain the biota of the region withintheir boundaries, and that degradation of these areasand their biota is occurring at an increasing rate.

Management staff are beset by "day to day" problemsand are therefore unable to manage these areas with along-term consemation goal.

What is required is a sustainable (this impliesprofitable) system of land use that does not compromisenature conservation values. It was noted that it may bepossible to develop a "long-term system" that does nottake into account nature conservation; however, manywould question the long-term viability of such a system.

The necessity of change is generally accepted, withmodels having been developed on how such a systemwould work (see, for example, paper by E.C. Lefroy, R.J.Hobbs and M. Scheltema in ly'alare Conseruation 3: TheReconstruction of Fragmented Ecosgstems, ed.s D.A.Saunders, R.J. Hobbs and PR. Ehrlich, Surrey Beattyand Sons, Chipping Norton, 1993), and there isresearch aimed at widening the range of agriculturalproducts and systems. However, much of the above isoccurring as an uncontrolled experiment, with "bits"

being undertaken by individuals, groups and agencies.The timescale of change is more than 20 years, theprocess is unstructured, methods and results areunrecorded, and information transfer is essentially byosmosis.

How do we assist, hasten and direct this experiment?

PROCESSESTo coordinate and enhance a sustainable system of landuse, the following processes are required:

* multilayered education;* economics (commitment of resources);t coordination and integration of land management;.! transfer and communication of information.

What actions may participants in this workshop groupundertake to increase their effectiveness?

EDUCATIONThis topic was dealt with by another workshop group.Howevet our group highlighted the need for amultilayered effort, including formal education(schools, agricultural colleges, adult education) andinformal education (popular media, workshops,

meetings). There is also a need to define what it is wewant to teach. Are we raising the general (biological andecological) knowledge base of land managers so thatthey can make informed decisions, or at leastunderstand decisions made for them? Or do we requirereliable, informative and "user friendly" summaries and"recipes" from research results?

ActionAnyone (scientists, managers, etc.) who is committed tochange needs to be able, and perhaps required, toallocate a large proportion (15-30%) of their time tocommunicating with communities concerning natureconservation issues. This involves talking to both theconverted and the unconverted. This commitmentshould include writing research summaries, medialiaison, participation in workshops, talks to landcaregroups, etc. Currently, our focus is still too directed topublishing in the scientific literature and talking to theprofessional land management community.Communication beyond these groups will foster a two-way flow of experience and ideas between all landmanagers.

ActionPass on the results of this workshop to decision-makers,by proceedings, media releases, personal contact, andany other available means.

ECONOMICSA profitable, self-funding system of agricultural land useis a basic requirement. The establishment of such asystem will require input from all Australians. It isbeyond the scope of this workshop to recommend theallocation of funding to this change; however, theactions recommended will assist in highlighting thisneed to the public and politicians.

ActionAs well as allocation of money, there needs to be a re-allocation of priorities and resources withinbureaucracies. All workshop participants are now betterplaced to enunciate these changes within their groupsor organisations.

COORDINATION AND INTEGRATION OFLAND MANAGERSThere is a need to increase effective communicationbetween farmers, shires, Department of Agriculture,Western Australia (DAWA), Department of Conservationand Land Management, Main Roads Department,Westrail, non-government organisations, etc., to ensurethat a sustainable land use system is attained.

This and other actions will require additional resourcesin terms of funding. The scale at which action shouldtake place, and the primary body for integration aredifficult to determine. The Environmental ProtectionAuthority and DAWA have defined 84 ecological units insouth-western Australia; there are 114 LandConservation District Committees (LCDCs), and manyshires, agencies and individuals (who will do most ofthe work).

The workshop group felt that the importance of local"ownership" made LCDCs (or similar "grass-roots"

based groups) the best bodies for integration. However,they will require additional resources to hire and selectpersonnel, and to maintain a capability to train others.

ActionA focus on interagency cooperation is required. It isessential to speed this with a commitment of additionalfunding, including recruitment of new staff. These extrajobs cannot simply be "dumped" onto any existingperson's or group's current workload. Workshopparticipants must assist in ensuring that people outsidethe region (cities and Eastern Australia) realise thatfunding these changes is a "whole community"responsibility. This will eventuate in the establishmentof a self-funding network in the Wheatbelt in the future.

TRANSFER AND COMMUNICATION OFINFORMATIONThere is a need for a system that encourages andrecords the current unplanned experirnents. Currently,there exists a growing knowledge base - not enough,but enough to warrant more action than at present.There are several useful models and experiments tobuild upon. However, managers cannot spend all theirtime attending open days and workshops.

It would appear that the best action workshopparticipants can undertake is the first listed above -namely, to spend more time in communicating theirresearch and management results in a wide variety offormats such as local papers, radio, television,meetings, etc. This should ensure that the "message"

will reach the target audience and be adopted by someof them, who in turn will spread the message to others.

ht

, {

What Are the Philosophy and Goals forManaging Remnant Vegetation within aLandscape Context?

Don McFarlaneDepartment of Agriculture, 1 28 Albany Highway,Albany, WA, 6330

Workshop GroupMike Fitzgerald, Kelly Gillen, Angas Hopkins, RobertLambeck, Don McFarlane and Barbara York Main.

Points from Plenary SessionThe following points, raised during a plenary session,stimulated this workshop topic:

* What are the goals for land and natureconservation? How are these goalsintegrated?

.1. Researchers should recognise that theyare servers.

* What are the conservation objectives andtargets for the Wheatbelt? We can't Iookafter everything, so what and how do wechoose what to save?

.i. There is a strong need to develop a landcareethic.

.l. People need to develop a vision for thefuture of agriculture and nature conservationon private property.

.l Better definition ofgoals and objectives.

DISCUSSIONPhilosophy Needed for Managing RemnantVegetation

Society needs to develop a deep love for the land(landcare ethic) similar to that held by indigenouspeople, making abuse of the land unacceptable. Societyalso needs to accept its responsibility for managing theconservation estate on both private and public land. Thestrong affinity for the land needs to override issues ofboundaries and ownership, and landowners must becustodians for future generations. In this regard, thereis a dilemma between the rights of the individual whoowns the land and the responsibilities of that individualto the community and future generations.

Conser.uation was seen as being possible once the basicneeds of individuals, families and society were met. futhe benefits of remnants are shared by the community,there needs to be government assistance for owners,particularly when times are hard. Assistance should notbe through compensation, but in the form of technicaland financial help, to ensure that the nature and landconservation values of remnants are maintained. Privateowners should be encouraged to voluntarily enter intoagreements to manage remnants, with thiscommitment being registered on land titles. Imposingvalues and management constraints on individuals isunlikely to result in the proper management ofprivately owned remnants. The maintenance of privateownership supports good management throughpersonal identification with the piece of land.

The intrinsic value of biota in the Wheatbelt must berecognised in its world-wide context. While it may bepossible to develop sustainable farming systems withoutretaining nature conservation values, this would resultin a loss of biodiversity, ecosystem function, aestheticsand an Australian identity.

Primary GoalStakeholders need to develop their own goals,objectives and strategies so that they have ownership ofthem. To assist in this process, a first attempt was madeby the group. The primary goal proposed was: "To havea long-term, profitable land use, while maintainingecosystem function and biodiversity and minimisingany off-site deleterious effects".

It needs to be accepted that some parts of the landscapemay have highly profitable uses, whereas other partsmay not. The overall system needs to be profitable fornature conservation values to be protected.

ProcessThe primary goals need to be accompanied by sets ofobjectives, strategies and actions that are specific at theState, regional, catchment and enterprise level. Some ofthese objectives have already been set (for example, theState Conservation Strategy, State Salinity Strategy,Decade of Landcare Plan, landcare district andcatchment group objectives, strategies and actions).

Key ecological parameters need to be identified fordeveloping management strategies (for example,

hydrologic water balance, energy and nutrient cycles,key species in ecosystems). Some of these will becomethe means of monitoring the health of the agro-ecological system, providing feedback for managers. Itwas accepted that both the native vegetation andfarming systems would be in a state of dynamic flux.

ActionsRepresentatives of the major stakeholder gyoups needto be drawn together to reach consensus on goals,objectives and strategies. This may take the form ofidentifying and coordinating existing structures at theState, regional, catchment and enterprise level. It wasthought necessary to raise awareness of natureconservation values in some groups that were notcurrently giving these values sufficient emphasis.

A specific example of this process would be for theDepartment of Conservation and Land Management (forexample, Ken Wallace) to liaise with the Department ofAgriculture's Program Leader for the Great Southernand the Blackwood Catchment Coordinating Group(BCCG), to see how the groups could integrate farmingand nature conservation objectives to everyone'sadvantage. This could become a model fot other areas.It was stressed that we should not try to come in overthe top of existing initiatives being taken by groupssuch as the BCCG, which are already developing anature conservation strategy,

Implementation

The integration of production and nature conservationobjectives may require legislative changes and changesto statutory procedures. These details were notdeveloped.

How do We Ensure that CommunityEducation and Extension MaximiseRemnant Conservation?

Anthony Suttonl and David Bicknell,lWilson Inlet Management Authority, pO Box 359,Denmark, WA, 6333'?Department olAgriculture, 1O Doney St, Narrogin,wA, 6312

Workshop GroupDavid Bicknell, Mal Graham, Steve Hopper, Bert Mainand Anthony Sutton.

Points frorn Plenary SessionThe following points, raised during a plenary session,stimulated this workshop topic:

.t The importance of direct personal contact (forexample, researchers devoting a fixed percentage oftheir time to direct communication).

* Networking using local people at a face to face level.

{. Promote nature conservation through logos, etc.(that is, start nature conservation movemenr ar alocal level).

{. Tangible examples of the value of remnantvegetation are required (for example, positivestories, pamphlets, natural histories, walk trails).

.:. Take advantage of simple stories which capturepeople's enthusiasm.

.i. Is one to one contact with a researcher the only(most) successful extension method?

.i. Are there usefulsocial models to help us interestrural people in nature conservation?

* Issues ofmale-female relationships incommunication and information.

.i' Education of new farmers in tertiary institutions.

* Technical extension and remnant management.

{. Should we have "bushwatch,' for schools?

Following detailed discussion ofthese points, a vision

for education and extension was created, the majorissues were identified, and actions were drafted.

VISIONWorkshop participants considered that communitysupport and action were required ifareas of remnantvegetation were to be conserved. To achieve this, it wasagreed that the long-term vision for education andextension programs was: "To foster people,s interest tothe point where the community actively seeksinformation to conserve remnant values,,.

ISSUESFrom both management and research perspectives, theparticipants then identified key issues that needed to beresolved in order to achieve the above vision. Specificaction items stemmed from these discussions.

Direct Personal ContactWorkshop participants agreed that direct ,,face to face,,contact between researchers, managers and thecommunity was an effective means of communication,especially in rural areas. However, as this method ofcommunication is extremely time-consuming, smallgroup meetings addressing a series of issues wereconsidered the preferred medium. The participants alsonoted that people with the most knowledge were notalways the best people to deliver a message. In ruralcommunities, it was considered that the idealcommunication method was to have local people givinginformation to a local audience.

Action

Identifu and support the good communicators withinagencies and local communities.

Promotion of Nafure Conservation at a Local LevelWorkshop participants quickly agreed that thepromotion of nature conservation within ruralcommunities required that the uniqueness of localenvironments be highlighted. It is also important toraise the awareness of each local community to thevalues of its landscape. Through these means, it wasconsidered that local people would develop a sense ofownership for local bushland and consequently pyotectthe values of these areas.

Actions

* Develop simple nature conservation stories that arepertinent to local people, and include themes orelements that are unique to their landscape. Storieswith a local historical perspective should be sought.

1. Document and promote specific examples ofpeople's nature conservation efforts in rural areas -

for example, the Doleys' story(see page 57). Provide"hands on" experience of the values of remnantvegetation so people may internalise the natureconservation ethic - for example, through guidedbush walks or the involvement of local communitiesin fieldwork.

.l Support people's nature conservation efforts bypromoting their achievements in the mass media.

Nature Conservation and EducationThe basics of nature conservation are not taught inschools or in agricultural training courses, such aslandcare courses and the Bachelor ofAgriculture. Theparticipants believed that the integration of natureconser.r'ation units into existing school and universitycourses was an essential step towards the long-termconserration of remnanl bushland.

Actions.1. Ensure researchers and managers involved with the

conseruation of remnant vegetation have input intolandcare courses.

* Introduce a natural science course into teaching andIand management degrees. This course would formpart of the first year syllabus and cover geology,zoology, botany, geography and hydrology.

.! Provide information for a basic textbook whichcovers the natural history of Western Australia.

.i. Support community conser.ration programs, such as"bushwatch", which have an important educationalrole.

Technical Extension and Remnant ManagementThe workshop group acknowledged that technicalinformation on the management of remnants should bereadily accessible and interpreted at different levels of

expertise, ranging from newspaper articles to journalpublications. The padicipants then resolved that aneffective technique for explaining remnant managementto the rural community was to use agriculturalanalogies. For example, the basics of remnantmanagement could be illustrated by explaining that theprocesses underlying agricultural systems (the nutrientand water cycles, and the flow of energy and geneticmaterial) also sustain natural ecosystems, but on adifferent timescale.

ActionEnsure information on the management of remnantvegetation is readily accessible. is interpreted atdifferent levels of expertise, and concentrates onc v c f a m c : n d

Women in Remnant ConservationWthin the group, anecdotal evidence was presented toshow that, in some cases, women were more caring landmanagers, and had better long-term vision, than men.However, with further discussion, the group resolvedthat for successful long-term conservation of remnants,it was essential that the whole community be involved.This includes all sectional groups, lor example:

.i women, children and men;{. Aboriginal people;.:. land managers;* rural and urban communities;.i. local people.

ActionSeek to involve the whole community in remnantconservatlon.

CONCLUSIONIn summary, the workshop participants considered thatfor education and extension to be successful, they must:

* be relevant;* be credible;.t be interesting;{. result in actionj.i. lead to success;.1. be "followed-up".

ffisne iln*sioms

K.J. WallaceDepartment of Conservation and LandManagement, PO Box 1 00, Narrogin, WA, 6312

The seminar objectives provide a useful structurearound which to summarise the conclusions from theseminar. Specific objectives for the seminar were:

{. to list significant research results for remnantmanagement;

a to assess whether conclusions from research havebeen implemented;

.1. to develop a list ofactions to improve both therelevance of research and implementation of itsresults.

Each of these is considered below. In writing theseconclusions, I have not tried to summarise all therecommendations and research outcomes. Rather, Ihave aimed to provide an overview, to highlight priorityissues, and to develop conclusions based on the seminarpapers as a whole. Throughout this section, the word"manager" refers only to those responsible formanaging remnants of native bushland, includingfarmers and government officers with directresponsibility for managing land.

SIGNIFICANT RESEARCH RESULTSFOR REMNANT MANAGEMENTThe significant research outcomes, as identified by bothmanagers and researchers, may be separated into threecategories: development of theory, proof of operationalhypotheses, and results with direct application. Whilesome work contributes in all three categories, mostemphasises only one.

Development of TheoryAs pointed out by Bert Main, the success or otherwise ofmanagement is partly determined by managers'assumptions about ecosystem processes. Theseassumptions are based on the theoretical constructsthey adopt, and thus the development of effectiveecosystem theories is crucial to the work of bothmanagers and researchers.

Research results presented in this seminar havecontributed to the fund of ideas, theories, and modelsthat form the essential platform on which effectivemanagement is based. The papers by Bert Main, RichardHobbs, Stephen Hopper and Robert Lambeck bestexemplify the development of concepts and "big

picture" contexts. Whether emphasising the importanceof taking a Gondwanan perspective of our biota, or therole of disease, parasites and predators in maintainingbiodiversity, or the need to better assemble these andother ideas into a bigger picture, the theory andconcepts described in these papers give valuableguidance to managers. Although some managers arereluctant to study and understand theory, in practice,unsound theory often results in ill-directed andineffectual management. This suggests either thatmanagers should be better trained in theory, or thatthey require more comprehensive guidelines andprescriptions. While both would enhance managementperformance, I note that adaptive management, whichshould be a goal, requires that managers are competentin ecosystem theory.

Even when research work does not itself develop theoryideas generated may help to create a socio-culturalclimate that influences political and managementdecisions. In their paper, Jonathan Majer and AnneBrandenburg provide examples of how this may occur.Such ideas and "bits" of information must besynthesised into theory or a "big picture" if they are toprovide maximum benefit. Unfortunately, this task isoften ignored, or given too low a priority (seecomments in papers by Hobbs and Main).

Ideas and theories change with new knowledge. One ofthe values of Denis Saunders' review paper is that itdescribes conceptual changes that have occurred inrelation to remnant management. It is important thatthe history of theory and research is documented * wecannot afford to waste resources unnecessarilyrevisiting ideas that have proved unproductive. Also,such reviews show that ideas change slowly. This is animportant lesson for anyone who aims to make a majorcontribution in the land management field: they mustbe prepared to commit 10 to 15 years to effect changesin attitudes and management practice. To document thehistory of management practice is also an importanttask.

Proof of Operational HypothesesA second category of research tests ideas based onunquantified obseruations and anecdotal evidence.Often, these have been, or are proposed to be,implemented operationally. For example, the value offencing remnants and linking isolated areas of bushlandwith corridors had been operationally "known" beforeresearch quantified the advantages of thesemanagement actions. Some of the research outcomesreported by Jonathan Majer and Anne Brandenburg fallinto this category.

This category of research is important. While anecdotalwisdom is valuable and should be better documented, itis unwise, and may be very expensive, to use suchinformation uncritically.

Ideally, research in this category should also contributeto the development of theory and managementprescriptions. Unfortunately, this does not occur asoften as it might.

Results with Direct ApplicationThis is the category of research most sought bymanagers. Managers want concrete prescriptions theycan implement immediately and effectively to solve anexisting problem. Examples reported in this seminar arerepresented best in the papers of Gordon Friend,Richard George and Don McFarlane, and Jack Kinnear.From a manager's viewpoint, the work by Jack Kinnearis ideal: it provides a practicable, operational techniquethat has a rapid, positive effect on populations ofendangered animals. This is not a common outcomefrom research,

Although the work of Friend, and George andMcFarlane contains elements that may be immediatelyand usefully implemented, valuable collaborativeprojects between managers and researchers are also alogical consequence. In the case of the work by Georgeand McFarlane, this is happening with the directapplication of management action at Toolibin Lake,with valuable help from Richard George. Not only isRichard's contribution essential for effectivemanagement, he has an opportunity to test first handthe validity of his predictions. The value of this on-ground, collaborative translation of concept into actionis inestimable.

It is appropriate here to note Bert Main's observationsconcerning the dangers of creating unrealisticcommunity expectations for biodiversity management.While strictly not a research result, his commentsconcerning "static and deterministic versus dynamicand stochastic interpretations of systems maintenance"should be read by all managers. Bert's comments are anexample of practical wisdom based on a comprehensiveunderstanding of theory, bureaucracies and politics.

SummaryEach of the research categories described is important.In the context of research as a whole, one categoryshould not be emphasised at the cost of the others:there should be a balanced contribution from each.Looking at the results from this seminar, there is,certainly from a manager's viewpoint, a need to givehigher priority to research that explores the actualimplementation of theory and research that directlyaddresses the needs of private land-holders so that theyare able to combine economically the synergisticbenefits of land and nature consel.ration. The challengeis for researchers to consider more thoroughly theapplication of their work. Such consideration shouldnot be an appendix to tesearch; it must also occurduring the development and implementation phases.

IMPLEMENTATION OF RESEARCHRESULTSAmong the presentations by managers, it is notable thatall three farmer presentations focus on fencingremnants as the predominant, and necessa4r,management action. In Terri Lloyd's paper, fencing isspecifically linked to revegetation and the developmentof corridors. Bob TWigg's paper asks several verypertinent questions about implementation - inparticular, why fencing of remnants is not morewidespread. There is a need for researchers andgovernment agencies to address the benefits ofremnants and revegetation, and to determine why thereis so little implementation of fencing as a basicmanagement action.

Apart from fencing remnants and creating corridors -activities that have been validated, rather thanstimulated, by research - the best examples of researchresults being implemented are fox baiting to protectpopulations of medium-sized marsupials (see paper byKinnear), and the fire research represented by the work

of Burrows el aL (1987, not presented at this seminar)and, more recently, Gordon Friend. The fire researchproduced specific guidelines and also showed that firemanagement principles should not be transferreduncritically between regions. Accordingly, the researchhas affected management plans and action. SteveGorton also points out the direct application of some ofthe revegetation research. Unfortunately, this researchgroup has now disbanded.

Apart from these aspects, and the operational actionsdiscussed by Ken Wallace, only a comparatively smallamount of research results has been imolemented. Whvis this so?

In part, the answer is probably that a critical mass oftheory and ideas must be developed before writingguidelines and prescriptions becomes a commonactivity. If this is so, then we can look forward to awelter of helpful guidelines, or at least more actionoriented research in the next decade. However, thereare other important barriers to the transfer of researchresults into management action. From the workshopdiscussions, and comments during presentations, it isapparent that implementation of results is also impededby inadequate communication and liaison betweenresearchers and managers.

ACTIONS TO IMPROVE RELEVANCEAND IMPLEMENTATION OF RESEARCHRESULTSA common theme running through many of thepresentations and workshops is the need for better two-way communication of ideas and information betweenresearchers and managers, Numerous ways are listed forachieving this end - see, for example, the workshoppapers by Ken Atkins and Helen AIlison, and AnneBrandenburg and Greg Beeston. Proposals includeestablishment of network groups, committees,computerised databases, and a new, integrating agency.

These ideas should be explored and those that are cost-effective implemented. Nevertheless, one cannot helpfeeling that, worthy as many of these ideas are, somecould become a barrier to achieving change where it ismost required - at the individual level. Researchers donot need organisations to enable them to liaise andcommunicate with other groups, and neither do

managers. That this is so is shown through three of thefive manager presentations specifically mentioning thegreat importance for their work of one-to-one contactwith researchers. (The latter point raises anotherchallenge * one-to-one relationships are important butvery expensive of an individual's time. How do weextend new ideas to managers in a cost-efficient way?David Bicknell [pers. comrn.] has suggested supportnetworks for early implementers so they may use theirproperties as demonstration sites and themselvesbecome extension sources.)

To improve the relevance and implementation ofresearch. the mosl urgent need is to increasecollaboration between individuals and groups, betweenmanagers and researchers, across agencies, and acrossdisciplines. There are many examples of effectivecollaboration between managers (including farmers)and researchers, and across disciplines and agencies.Although this process will occur more readily in asympathetic organisational environment, such a milieuis not essential. Thus there is a challenge to individuals,irrespective of their organisational environment, tocreate collaborative projects and networks.

Not discussed in detail, and perhaps it should havebeen, was the responsibility for converting researchresults into management guidelines. Some see this as aspecialist task - for example, for regional ecologists.Others view it as the responsibility of each researcher,and others argue that managers have the skills andshould use them. I suggest that the best method forconverting research results into managementguidelines is through collaborative efforts betweenmanagers (including planners) and researchers. This isan efficient means of bringing together the skillsnecessary to achieve implementation of researchresults. For best results, researchers need to collaboratewith land managers at all stages of project development.

Many other ideas to improve the relevance andimplementation of research results were presented atthe seminar, and these should be explored. Of particularnote is the suggestion by Richard Hobbs that werecognise and reward researchers who synthesise andliaise effectively to achieve implementation of results.Similarly, we need to recognise and reward managerswho build successful partnerships with researchers, todevelop and implemenl effective management.

ACTIONS RECOMMENDED DURINGTHE SEMINARA number of actions were recommended during theseminar.

Workshop ActionsNearly 30 recommendations for action are listed in theworkshop section. Most of these recommendationsfocus on communication and education within andbetween the many groups involved in conseruation ofremnant vegetation. The better integration ofmanagement and research is the subject of many actionstatements and is covered in the previous section. Threeother issues should be stressed in these conclusions.

Firstly, the development ofvisions, objectives andpriority actions for the conseruation of remnantvegetation and their biota has been inadequate. Thisdeficiency should be redressed immediately, and theworkshop report by Don McFarlane provides ideas onhow this should progress.

Secondly, education has been better addressed in recentyears, but our efforts could, and should, be substantiallyimproved. The workshop report by Anthony Sutton andDavid Bicknell gives some guidance in this area.

Finally, all seminar padicipants agreed that a credible"messenger" was required, to inform key decision-makers of the degree to which land degradation hasalready occurred, and just how much worse it is likelyto become. Given that seminar participants generallybelieved that a "messenger" was likely to be "shot", it isnot surprising that there were no offers for this task,and no groups wanted to address the issue in theirworkshop discussion!

Research ActionsIdentifuing new lines of research was not a major focusof the seminar. Some 20 ideas were listed in one of theplenary sessions, and other suggestions were developedin presentations. These diverse ideas defu readydescription - the reader is encouraged to scan the liston pages 71-72. Most urgent for remnant vegetation isthe need to tackle the problems of landscape hydrology(see paper by George and McFarlane). Although therising watertable and related hydrological problems area potential tragedy, they also provide a compelling

process through which to integrate productiveagriculture and nature conservation. It is vital that weseize this opportunity.

Consequently, it is crucial that researchers andmanagers work hard, and with urgency, to develop:

.!. Practical means and prescriptions for controllinggroundwater that protect nature conservation valuesand are, at the same time, economically viable forfarmers.

* Systems of sustainable agriculture, particularlythose that advance the synerglism betweenconservation of natural biodiversity and profitableagriculture.

Such work should be undertaken as collaborativeprojects, and should involve landscape scaleexperiments. There is sufficient theory to guide actionbut insufficient resources to permit the luxury offurther theory development in isolation from actualaction.

Another important research area raised during theseminar is the value of addressing cultural and socialissues (see workshop reports and the papers byLambeck, Saunders and Wallace). While not discussedduring the seminar, the social models we use forinteraction and decision-making are a fundamentalfactor in land management. These models, oftenimplicit rather than explicit, may assist or impedecollaboration and decision-making.

For example, it has been fashionable to talk of "top

down" and "bottom up" approaches to planning anddecision-making. More recently, the concept of"empowering" people has again become widely used.While these concepts can be useful, when dealing withcomplex land management issues it is more helpful toview planning and decision-making as involvinginteraction between a number of subcultures, whoserelative powers vary from situation to situation. In thisview, it is the responsibility of each subculture toconsider and describe its own goals and priority actions,and then to neglotiate and debate these with othersubcultures. Decisions which result will reflect a rangeofsituational factors, including the personalities of theparticipants.

The accuracy of this model is open to debate and shouldbe researched. However, from a manager,s viewpoint, itdescribes the fluid, somewhat chaotic, interactions anddecision-making processes. Such a model provides aworkable basis for planning and interpretinginteractions.

There are many other examples of research topics thatwould benefit through the involvement of socialscientists. Researchers and managers should aim toincrease the involvement ofsocial scientists in theirwork. While this will only be appropriate in specificprojects, it is time we acknowledge through our actionsthe importance of social processes identified in this andother seminars.

T\.^ro other areas of research merit specific comment.Firstly, Bob lhigg in his paper stressed that landmanagers do not understand the values of remnantvegetation. While there has been some work in thisarea, it is vital that the economic and other values ofremnant vegetation be clearly detailed (see Wallace1994 for a starting point). Similarly, the values ofbiodiversity are not sufficiently explained. It is oftenassumed that maintaining natural biodiversity is eithera prerequisite for achieving sustainable agriculture, orthe most profitable means of doing so. Such statementsshould be tested and the outcomes explained in termsthat all land managers understand

FINAL REMARKSFrom my management viewpoint, I had hoped that theworkshop would generate far more specific implicationsfor managers, based on research from the past decade.That this did not occur is partly because the period hasbeen one of developing ideas and theory rather thanmanagement prescriptions. However, there are otherreasons for the infrequent transfer of research resultsrnto management action. During this seminar, the mostimportant barrier identified, at least at the individualIevel, was the poor communication and liaison betweenresearchers and managers.

Despite outstanding examples of collaboration, thegeneral failure of researchers and managers to worktogether effectively has decreased the values of bothresearch and management. While this was of lessconcern during a period when new ideas and theorydevelopment were pre-eminent, ifallowed to persist, it

will severely affect nature conservation in the future.We now have an opportunity to develop projects ofcollaborative, experimental managlement that effectivelytackle the major challenges that the next decade holds.We are at the leading edge of a wave of at best local, atworst global, extinctions and losses in agriculturalproduction. The value ofworking together has neverbeen higher Not only should we look to corporatechange, we should act now as individuals, to ensure thateffective collaboration occurs. For maximum effect,projects should address the nexus between hydrologicalproblems, revegetation and remnants, farm profitability,and nature conservation.

If collaboration is so important, how, then, should weachieve it? Actions to achieve collaboration may, forconvenience, be divided into those at the corporatelevel, and those at the individual level.

Although there is a continual, and impofant, need toreview corporate cultures critically, to ensure they aredesigned optimally to achieve community andcoryorate goals (see, for example Norgaard 1992 andWallace 1992 for discussion of some of the issues), thisis usually not the most crucial factor. Ultimately, thesuccess or failure of collaborative groups depends onthe individuals involved.

In my experience, successful projects involving differentgroups work because a "netuork,' of people with acommon interest has come together of their volition, toachieve a task that meets their individual needs. Thisdoes not mean that all involved must have the sameneed; rather, their needs must be sufficiently congruentto drive each to attain the collaborative goal. No doubtthere are other collaborative methods that work;however, that described is the most effective in myexperience,

Therefore, the challenge is for land managers,researchers and others themselves to form highlymotivated project groups driven by common goals andcompatible personalities (see the Note on the next pagefor one formula for collaboration). At the same time,these goals should involve tasks that develop the nexusdescribed above, and either test, or add to, the structureoftheory and unifying concepts.

REFERENCESBurrows, N.D., McCaW W.L., and Maisey, K.G., 1987.Planning for fire management in Dryandra Forest. pp305-312 in Nature Conseruation: The Role ofRemnants ofNatiue Vegetation, eds D.A. Saunders, A.A.Burbidge and A.J.M. Hopkins. Surrey Beatty and Sons,Chipping Norton (NSW).

Norgaard, R.B., 1992. Co-ordinating disciplinary andorganizational ways of knowing. Agriculture,Ecosgstems and Enuironmmt 42: 205-276.

Wallace, K.J., 1992. Services - a barrier to achievingland conservation objectives? pp. 459466 in 7th ISCOConference Sgdneg - People Protecting Their Innd,eds PG. Haskins and B.M. Murphy. Proceedings, vol. 2.International Soil Conservation Organisation, under theauspices of the Department of Conservation and LandManagement: Sydney.

Wallace, K.J., 1994. Remnant native vegetation andsustainable agricultural systems. pp. 138-145 inLandcare '93 - A Conference for Landcare , eds D .A.Hills and J.S. Dufl Department ofAgriculture, WesternAustralia.

NOTEA formula I recommend for developing collaborativeprojects is given below. For simplicity, it is writtenaddressed to a researcher, but is, with minoradjustment, applicable to any group involved in landmanagement.

Forming a NetworkFundamental to developing a collaborative project isalready having a network and a knowledge ofpracticalissues. Suggestions in the following two points will helpindividuals achieve these oulcomes:

.! Involve yourself in at least one committee or othergroup that has an interdisciplinary and interagencycomponent. Such a group should involve managersand researchers. Working with a committee will notonly help you to build a network across disciplinesand agencies, it will help you to gain otherperspectives on land management issues.Committee members should develop a workingrelationship, and this often leads to empathy. This is

a more effective way to develop a cross-culturalnetwork than the most common means, theconference. The latter, while valuable, tends to buildnetworks within subcultures rather than acrossthem, and does not provide the "forced empathy"aspects of a good committee. If there are nocommittees that are appropriate, consider joining anorganisation or group that contains diverse views. Inrural areas, organisations such as Apex, Rural Youth,etc. may provide this contact.

Make opportunities to visit people from otherdisciplines or agencies. It is particularly importantto do this in the field where problems may bediscussed and related to a physical example. In thesecircumstances it is often much easier to understandthe specific viewpoint of others. The value of"kicking the dust" with others of different views andbackground cannot be over-emphasised.

Developing a Collaborative Project1. When you have a project that is strategically

important for remnant conservation, and personallyimportant (a project for which you have realenthusiasm), then use your broad network of peopleto develop the project and initiate design. This phasewill include those who have time to advise, but nottime to participate as project team members. Whenyou have your project in a reasonable form, castaround for a group who are prepared to participateas part of the project team. These will be people

who:

t are interested in the same project outcome (theydo not have to have the same ethos or the samegoals - but theirs must be compatible withyours);

.:. will work effectively with you;

.r will commit to and contribute effectively to theproject.

The project team should be small - say, fourpeople. Maintain good contact within the group, andmeet only as necessary. The responsibility of eachteam member to the project should be clear. It isimportant that any meetings:.:. be held for a good reason;* consider important issues;* result in concrete action.

Notes on meetings should be kept and circulated toall those whom you wish to involve in the project(this will often stimulate ideas and will ensuremisunderstandings are clarified). There is no need tobe formal, but it is impodant to be accurate.

Very early in the development ofthe project, discussthe project with managers and include at least oneon your project team. It is particularly importantthat their needs are met, and that their ideas are,where appropriate, incorporated.

Maintain contact with managers throughout theproject, and feed information back to your broadernetwork as appropriate.

When the work is finished and you are clear on theoutcomes, turn them into management guidelinesor ideas in conjunction with a manager. Do not waitfor the publication of results before implementingthem. Participate in extension activities.

Always acknowledge the involvement of those whohave helped - this is not only courteous; it buildsownership.

Review the success ofyour projecl both in terms ofapplying research results and as a collaborativeexerctse.

"'f'-'

@

Appendix: Workshop Participants

Ms H. AllisonDepartment of AgricultureBaron-Hay CourtSOTNH PERTH WA 6151

Dr K.J. AtkinsWldlife BranchDepartment of Conservation and Land ManagementPO Box 104COMO WA 6152

Mr J. BartleDepartment of Conservation and Land ManagementPO Box 104COMO WA 6152

Mr G. R. BeestonDepartment of AgricultureBaron-Hay CourtSOUTH PERTH WA 6151

Mr D. BicknellDepa|tment of Agriculture10 Doney StNARROGIN WA 6312

Ms S. A. BrandenburgCurtin UniversityGPO Box Ul987PDRTH WA 6001

Ms A. Doley"Koobabbie"

COOROW WA 6515

Mr J. Doley"Koobabbie"

COOROW WA 6515

Mr M. FitzgeraldDepartment of Conservation and Land ManagementPO Box 332MERREDIN WA 6415

Dr G. R. FriendWoodvale Research CentreDepartment of Conservation and Land ManagementPO Box 51WANNEROO WA 6065

Dr R. J. GeorgeDepatment of AgriculturePO Box 1231BUNBURY WA 623I

Mr K. GillenDepartment of Conservation and Land Management44 Serpentine RoadALBANY WA 6330

Mr S.C. GortonDepafment of Conservation and Land ManagementPO Box 100NARROGIN WA 6312

Mr M. S. GrahamDepartment of Conservation and Land ManagementPO Box 811KATANNING WA 6317

Dr R. J. HobbsCSIRODivision of Wildlife and EcologyLMB 4MIDLAND WA 6056

Mr A. J. M. HopkinsWoodvale Research CentreDepartment of Conservation and Land ManagementPO Box 51WANNEROO WA 6065

Dr S. D. HopperDirectorKings Park and Botanic GardenPERTH WA 6000

Ms B.M.J. HusseyWildlife BranchDepartment of Conservation and Land ManagementPO Box 104coMo wA 6152

Dr G. KeigheryWoodvale Research CentreDepartment of Conservation and Land ManagementPO Box 51WANNEROO WA 6065

Dr J. E. KinnearWoodvale Research CentreDepartment of Conservation and Land ManagementPO Box 51WANNEROO WA 6065

Dr R. J. LambeckCSIRODivision of Wildlife and EcologyLMB 4MIDLAND WA 6056

Mrs T. LloydPO Box 52DUMBLE}IJNG WA 6350

Mr D. J. McFarlaneDepartment of A€riculture128 Albany HighwayALBANY WA 6330

Professor A.R. MainDepartment of ZoologyUniversity of Western AustraliaStirling HighwayNEDI.ANDS WA 6907

Dr B. York MainDepartment of ZoologyUniversity of Western AustraliaStirling HighwayNEDI.ANDS WA 6907

Dr J. D. MajerCurtin UniversityGPO Box Ul987PERTH WA 600I

Mr D.S. MitchellDepartment of Conservation and Land ManagementPO Box 100NARROGIN WA 6312

Dr D. A. SaundersDivision of Wldlife and EcologyCSIROLMB 4MIDLAND WA 6056

Mr R. ThiggPO Box 86JDRMMUNGUP WA 6337

Mr A.X. SuttonWlson Inlet Management AuthorityPO Box 353DENMARK WA 6333

Mr K.J. WallaceDepartment of Conservation and Innd ManagementPO Box 100NARROGIN WA 6312

Mr C. YatesDepartment of Environmental ScienceMurdoch UniversityMURDOCH WA 6150

4105-0695-500