No.ACIAR PROCEEDINGS

AUSTRALIAN CENTRE FOR INTERNATIONAL AGRICULTURAL RESEARCH

The Australian Centre for International Agricultural Research 91

FIRE AND SUSTAINABLE AGRICULTURALAND FORESTRY DEVELOPMENT IN EASTERN

INDONESIA AND NORTHERN AUSTRALIA

FIRE

AN

D SU

STAIN

AB

LE

AG

RIC

ULT

UR

AL

AN

D FO

RE

STR

YD

EV

EL

OP

ME

NT

IN E

AST

ER

N IN

DO

NE

SIAA

ND

NO

RT

HE

RN

AU

STR

AL

IAA

CIA

R

91

Fire and Sustainable Agricultural andForestry Development in EasternIndonesia and Northern Australia

Proceedings of an international workshop held at Northern Territory University, Darwin, Australia, 13–15 April 1999

Editors: Jeremy Russell-Smith, Greg Hill, Siliwoloe Djoeroemana and Bronwyn Myers

Australian Centre for International Agricultural ResearchCanberra 2000

© Australian Centre for International Agricultural Research, GPO Box 1571, Canberra ACT 2601Australia

Russell-Smith, Jeremy, Hill, Greg, Siliwoloe Djoeroemana and Myers, Bronwyn 2000. Fire andsustainable agricultural and forestry development in Eastern Indonesia and Northern Australia.Proceedings of an international workshop held at Northern Territory University, Darwin, Australia,13–15 April 1999. Canberra, ACIAR, Proceedings No. 91, 163p.

ISBN 1 86320 275 7

Production management: Janet LawrenceTypeset and layout: Design One SolutionsPrinting: Robeys, Canberra

2

The Australian Centre for International Agricultural Research (ACIAR) was established in June1982 by an Act of the Australian Parliament. Its mandate is to help identify agricultural problems indeveloping countries and to commission collaborative research between Australian and developingcountry researchers in fields where Australia has special research competence.Where trade names are used this does not constitute endorsement of nor discrimination against anyproduct by the Centre.

ACIAR PROCEEDINGSThis series of publications includes the full proceedings of research workshops or symposiaorganised or supported by ACIAR. Numbers in this series are distributed internationally toselected individuals and scientific institutions.

ContentsForeword 5Pendahuluan 6Workshop Opening and Regional Overview 7A Framework to the ACIAR Project on Fire Management in Eastern Indonesia and Northern Australia—and Welcome to the Workshop 8

Greg Hill

Message of Welcome from East Nusa Tenggara 12Siliwoloe Djoeroemana

Background to the Project ‘The Use of Fire in Land Management in Eastern Indonesia and Northern Australia’ 13

Bronwyn Myers, Greg Hill and Jeremy Russell-Smith

Shifting Cultivation and Fire: a Challenge to NTT’s Development 18Esthon L. Foenay

Atmospheric Issues for Fire Management in Eastern Indonesia and Northern Australia 21Nigel Tapper

Fire Management Issues in Eastern Indonesia 31Land-use and Land-cover Change in Nusa Tenggara Timur, Indonesia 32

Jeff Fox

Management of Fuel and Fire in Preparing Land for Forest Plantations and Shifting Cultivation 39Bambang Hero Saharjo, Endang A. Husaeni, Kasno and Hiroyuki Watanabe

Land and Forest Fire in East Kalimantan Province 45Riyanto

Forest Fire Management in Komodo National Park 48Bambang Hartono and Heru Rudiharto

The Implications of Fire Management and Reforestation for Economic Development in East Nusa Tenggara 52

Siliwoloe Djoeroemana, Haryono Semangun, Bungaran Saragih and Achmad Sulthoni

Fire and the Management of Agricultural Systems in East Nusa Tenggara 56Wayan Mudita

Forest Land and Fire Management in East Nusa Tenggara 62Slamet Riyadhi Gadas

Fire and Weeds: Interactions and Management Implications 65Colin Wilson and Wayan Mudita

Livestock Production and Fire Management in East Nusa Tenggara 69Abdullah Bamualim

Fire, Traditional Knowledge and Cultural Perspectives in Nusa Tenggara Timur 73Hendrik Ataupah

The Role of Fire in Swidden Cultivation: A Timor Case Study 77Tom Therik

Fire and Cultural Burning in Nusa Tenggara Timur: Some Implications of Fire Management Practices for Indonesian Government Policy 80

Andrew McWilliam

Forest Fragmentation and Water Catchment Hydrology—Case Study on Sumba 86Juspan

3

Fire in the Trans-Fly Savanna, Irian Jaya/PNG 90Neil Stronach

Fire Management Issues in Northern Australia 94Fire Management and Savanna Landscapes in Northern Australia 95

Jeremy Russell-Smith, Grant Allan, Richard Thackway, Tony Rosling and Richard Smith

Fire Management on Aboriginal Lands in the Top End of the Northern Territory, Australia 102Peter Cooke

The Role of Fire on Pastoral Lands in Northern Australia 108Rodd Dyer

Role of Remote Sensing in Bushfire Management 113R.C.G. Smith, R.L. Craig, M.T. Steber, A.J. Marsden, N.A. Raisbeck-Brown and J. Adams

Communication and Fire Management in Northern Australia 121Peter Jacklyn

Institutional Linkages and Cooperative Approaches 126Eastern Indonesian and Northern Territory Tertiary Institutional Linkages 127

Chris Healey

Nusa Tenggara Community Development Consortium 133Putra Suardika and Suhardi Suryadi

Development of Regional Synergies between Northern Australia and the Trans-Fly Region of Indonesia and Papua New Guinea 138

Michele Bowe

Workshop Breakaway Sessions 143Sesi Lokakarya yang Terpisah 151Referees 160Participants 160

4

Foreword

OVER THE past decade or so there has been growing recognition of the impacts, both environmental and political, ofbiomass burning in the wet forests of Sumatra, Kalimantan and Irian Jaya in Indonesia. This has translated intoconsiderable research effort in these regions, particularly in the last few years (Dennis 1999; Laumonier et al.1999). In contrast, much less attention has been given to annual, mostly prescibed burning practices and associatedland management issues, across the extensive savanna landscapes of the eastern Indonesian archipelago, theTransfly region of Irian Jaya and Papua New Guinea, and northern Australia. While there has been considerablepreliminary research undertaken into fire management and associated ecological issues in northern Australia overthe past couple of decades, scant documentation is available concerning the extent of burning in different regions,traditional and contemporary practices, and impacts and benefits of fire management in eastern Indonesia culturalsettings.

Building on the establishment of formal linkages in the 1990s between a number of eastern Indonesian andnorthern Australian institutions (universities especially), and recognition by all parties involved that firemanagement constituted a significant regional economic and social issue warranting cooperative applied research,in 1998 the Australian Centre for International Agricultural Research (ACIAR) offered support for a workshopwhere these problems could be defined and prioritised. The resulting workshop was held in Darwin, at theNorthern Territory University, between 13 and15 April 1999. The workshop attracted approximately 60participants, about half of whom were from Indonesia or further afield. The Tropical Savannas CooperativeResearch Centre (based in Darwin) and the World Wide Fund for Nature (WWF) provided additional funding foroverseas participants.

These proceedings comprise 26 papers presented at the Darwin workshop, summaries of the workshopdiscussion group sessions, and participant contact details. The papers are arranged in four groupings (Workshopopening and regional overview; Fire management issues in eastern Indonesia; Fire management issues in northernAustralia; Institutional linkages and cooperative approaches), not necessarily in order of presentation at theworkshop. In particular, the editors wish to compliment Indonesian authors for the outstanding quality of papersprepared and, in most instances, delivered in English.

As indicated in the workshop discussion group summaries, considerable progress was made in identifying andprioritising problems. Suggested resolutions include appropriate methodologies for undertaking demonstration-basedactivities, with associated training, educational and institutional linkage opportunities. In sum, we consider theworkshop provided an excellent forum for constructive interaction between eastern Indonesian and northernAustralian fire researchers, managers and policy-makers, establishing a sound platform for future cooperation. Wewish to acknowledge with thanks the contributions of Heather Crompton of ACIAR, the Northern TerritoryUniversity, the Tropical Savannas Cooperative Research Centre, all workshop participants, those involved withorganisation (especially Julian Gorman), and Janet Lawrence (ACIAR) for her excellent technical editing of theproceedings.

Jeremy Russell-SmithGreg HillSiliwoloe DjoeroemanaBronwyn Myers

January 2000

ReferencesDennis, R. 1999. A review of Fire Projects in Indonesia, (1982–1998). Center for International Forestry Research (CIFOR),

Bogor, Indonesia, 105 p.Laumonier, Y., King, B., Legg, C. and Rennolls, K. 1999. Data Management and Modelling Using Remote Sensing and GIS for

Tropical Forest Land Inventory. Rodeo International Publishers, Jakarta.

5

6

PendahuluanSelama lebih kurang satu dasawarsa terakhir orang mulai mengakui dampak lingkungan dan politik sebagai akibatdari pembakaran biomassa (seluruh hayati) yang ada di blok-blok hutan basah di Sumatra, Kalimantan dan IrianJaya, Indonesia. Hal ini menyebabkan banyak usaha penelitian dilakukan di daerah-daerah ini, khususnya dalambeberapa tahun terakhir (Dennis 1999; Laumonier et al. 1999). Sebaliknya, sangat sedikit perhatian yangdiberikan terhadap praktik pembakaran tahunan yang sebagian besar telah ditentukan dan persoalan yangberhubungan dengan pengelolaan lahan di savana yang luas di kepulauan Indonesia timur, daerah antar lintasudara di Irian Jaya dan Papua Nugini, dan Australia utara. Sementara itu, telah banyak penelitian pendahuluanyang dilakukan dalam bidang pengelolaan kebakaran dan yang berhubungan dengan masalah ekologi di Australiautara selama dua dasawarsa terakhir. Akan tetapi pendokumentasian mengenai luasan pembakaran di berbagaidaerah serta praktik-praktik tradisional maupun kontemporer masih terbatas. Hal serupa juga terjadi padapendokumentasian dampak serta keuntungan pengelolaan kebakaran pada latar budaya Indonesia bagian Timur.

Sebagai pengembangan dari penetapan hubungan formal pada tahun 1990an antar sejumlah lembaga diIndonesia timur dan Australia utara, khususnya perguruan tinggi, serta pengakuan oleh semua pihak yang terlibatbahwa pengelolaan kebakaran merupakan persoalan ekonomi dan sosial regional penting, yang menjaminkerjasama penelitian terapan, maka pada tahun 1998 ACIAR (Pusat Penelitian Pertanian Internasional Australia)menawarkan diri untuk menjadi tuan rumah lokakarya dalam rangka mendifinisikan dan memprioritaskanpersoalan tersebut di atas. Lokakarya itu akhirnya dilaksanakan di Darwin, tepatnya di Northern TerritoryUniversity, pada tanggal 13–15 April 1999. Lokakarya tersebut dihadiri oleh sekitar 60 peserta, separohnyaberasal dari Indonesia dan sekitarnya. Pusat Penelitian Kerjasama Savana Tropis yang berkantor di Darwin danLembaga penyandang dana Internasional untuk pelestarian satwa liar (WWF) telah menyediakan dana tambahanuntuk para peserta dari luar negeri.

Laporan ini terdiri atas 26 karya tulis yang disajikan dalam lokakarya di Darwin, dan disertai dengan ringkasansesi diskusi kelompok lokakarya serta daftar alamat serta nomor-nomor yang bisa dihubungi dari para peserta.Karya tulis tersebut disajikan dalam 4 kelompok (Pembukaan lokakarya dan tinjauan regional, Persoalanpengelolaan kebakaran di Indonesia timur, Persoalan pengelolaan kebakaran di Australia utara, Hubungankelembagaan dan pendekatan kerjasama), tetapi penyajiannya tidak sebagaimana disajikan dalam lokakarya. Timpenyunting ingin menyampaikan penghargaan khususnya kepada para penulis dari Indonesia atas karya tulismereka yang berbobot dan sebagian besar disajikan dalam bahasa Inggris.

Sebagaimana dinyatakan dalam ringkasan diskusi kelompok lokakarya tersebut, telah banyak kemajuan yangdicapai dalam mengidentifikasikan dan memprioritaskan masalah-masalah. Pemecahan-pemecahan yangdisarankan mencakup metodologi-metodologi yang tepat untuk melaksanakan kegiatan peragaan, dankesempatan-kesempatan yang berhubungan dengan pelatihan, pendidikan dan hubungan kelembagaan.Ringkasnya, menurut hemat kami lokakarya ini telah menyediakan forum yang luar biasa untuk interaksikonstruktif antara para peneliti kebakaran dari Indonesia timur dan Australia utara, para manajer dan pembuatkeputusan. Lokakarya ini juga telah menjadi wadah yang baik untuk kerjasama di masa mendatang. Kami jugaingin menghaturkan terima kasih khususnya kepada Heather Crompton dari ACIAR, Northern TerritoryUniversity, Pusat Penelitian Kerjasama Savana Tropis, para peserta lokakarya atas konstribusinya, dan semuayang terlibat di dalam pengorganisasian lokakarya ini (khususnya Julian Gorman) dan Janet Lawrence (ACIAR)atas penyuntingan teknis beliau yang luar biasa terhadap laporan-laporan ini.

Jeremy Russell-SmithGreg HillSiliwoloe DjoeroemanaBronwyn Myers

Januari 2000

7

Workshop Openingand Regional Overview

8

THIS CURRENT project had its beginnings at theInternational Conference on AgriculturalDevelopment in Semiarid East Nusa Tenggara, EastTimor and South East Maluku which was held inKupang from 11 to 15 December 1995. Arrangedthrough an organising committee headed by SatyaWacanna University, the conference attracted widesupport from within Indonesia and the region. TheNorthern Territory Department of Primary Industryand Fisheries (NTDPIF) and the Northern TerritoryUniversity (NTU) were core participants in theconference. In the case of NTU, participation waslinked to the Memoranda of Understanding (MOUs)the institution had established with five Indonesianuniversities in the region (refer paper by Healey, theseproceedings). The NTDPIF also had well establishedpartnerships in place and a long history ofinvolvement in the agricultural sector of easternIndonesia.

A major outcome of the conference was theestablishment of an Information Centre on Semi AridAgriculture which has operated out of Kupang since1996 (refer paper by Djoeroemana, these proceedings).The partners in the original conference are the coremembers of the Information Centre. Fire managementwas raised as an issue at the 1995 conference and theneed for research in this area was further developed bythe Information Centre over the following year. When

in 1997 the massive forest fires that ravaged Indonesiaand other neighbouring countries grabbed the world’sattention, there was also the realisation that issues offire management in semiarid, wet/dry tropical regions,particularly in eastern Indonesia, were poorlyunderstood. This provided the framework forapproaching ACIAR with the concept of a firemanagement proposal. The Workshop we areattending today is the culmination of these activitieswhich have proceeded over the previous three plusyears. The discussions that have taken place over thistime have been collaborative, accepting that bothIndonesia and Australia have much to gain throughsharing of knowledge and joint research directedtowards the issue of fire management.

A notable aspect of the current proposal was that itwas developed through the Darwin-basedCooperative Research Centre (CRC) for SustainableDevelopment of Tropical Savannas. The TropicalSavannas CRC coordinates fire management researchacross the Northern Territory, Queensland andWestern Australia and a project dealing with firemanagement in northern Australia was part of the coreprogram when the CRC was opened in 1995. Thisproject has been particularly successful with regard toits research and extension activities (e.g. Jacklyn andRussell-Smith 1998; Russell-Smith et al. 1997 a, b;Schultz 1998).

The key CRC agency involved in this work is theNorthern Territory Bushfires Council, which haspioneered research into the use of satellite imagery for

A Framework to the ACIAR Project on Fire Management in Eastern Indonesia

and Northern Australia–and Welcome to the Workshop

Greg Hill1

1Science and Education Faculty Group, Northern TerritoryUniversity, Darwin 0909 Australia

9

detailing fire histories for savanna environments(Russell-Smith et al. these proceedings). Dataprovided by the Western Australia Department ofLand Administration are routinely used for mappingand monitoring fire activity in the Northern Territoryand the Kimberley region of Western Australia (Smithet al. these proceedings) and the program has recentlybeen extended to Queensland to complete anintegrated coverage for northern Australia. TheBushfires Council has also been proactive in engagingsavanna stakeholder groups in fire managementactivities and workshop programs. This hasincorporated integrating traditional indigenousknowledge into analysis of current and past fireregimes. The Bushfires Council has been an importantagent of change, alerting managers and landholders tothe implications of particular fire regimes and howthese can be manipulated to achieve desiredoutcomes.

In association with the work of the NT BushfireCouncil, a parallel development of infrastructure andexpertise has been occurring in the other core partnerin our current project, the Northern TerritoryUniversity remote sensing group. EnvironmentalRemote Sensing is recognised as an Area of ResearchStrength at NTU. Over the last 3 years the programhas attracted more than $500 000 in researchinfrastructure through funding agencies such as theAustralian Research Council (ARC), Research Infra-structure and Equipment Fund (RIEF). Laboratoryfacilities for computer-based processing and analysisof remotely sensed data are first class and recentfunding has provided sophisticated data collection andfield equipment, such as a digital aerial camera whichwas purchased this year. The group has also beensuccessful with research grants including thoseprovided by the ARC Large Grants Scheme and theARC Strategic Partnerships with Industry Researchand Training (SPIRT) Scheme.

The Northern Territory University has a wellestablished reputation for education and training,ranging from postgraduate research degrees throughto TAFE level courses in the vocational educationsector. The University is one of only a handfull ofAustralian universities that is a dual sector institution,offering the full spectrum of post-secondary educationopportunities from Certificate to PhD. In activitieswith a focus on technology transfer, this mix ofexpertise is a valuable combination. The BushfiresCouncil also has a well established training programfor its own staff which has the potential forapplication in Indonesia. There is establishedexchange of students between the University and itsMOU university partners in eastern Indonesia andboth the University and the Northern Territorygovernment agencies (e.g. NTDPIF, NT Departmentof Lands Planning and Environment) run training

programs in the region. Research partnerships withagencies such as the World Wide Fund for Nature(WWF) provide additional linkages within the region.

The Tropical Savannas CRCIn recent years the tropical savannas of northernAustralia have experienced a (long overdue) boost instatus. Factors responsible include the recognition of:aboriginal title to land; new trade developments withthe Asian countries (e.g. live cattle exports); therelatively pristine state of much of the region; and thegrowth in nature-based tourism. Factors such as these,combined with the distribution of the savannas acrossstate and territory boundaries and the comparativelysmall research effort expended on the region to date,were presumably a strong motivation for establishingthe CRC. The mission of the CRC is to undertakeresearch and education that will enhance for Australiaand its people the sustainable development of ourtropical savannas. The semiarid tropical savannascover more than 20% of Australia but support only5% of the continent’s population. Nevertheless, viamining, tourism and the pastoral industry this regionprovides more than 25% of the country’s exportincome. The need for an accelerated and innovativeresearch effort in the region, to optimise economic,environmental and social gains, has been the impetusfor the formation of the Centre.

The research program of the Centre is based aroundthe ecology, use and management of the savannas.Firstly, the CRC is interested in how savanna plantsand animals are affected by soils, storms, fires andtropical wet and dry seasons. Secondly, there is arequirement to know how the savannas are affected bythe uses people make of them. Above all, the CRCintends to work with users to understand the effects oftheir management of the land, so they can be confidentthat their enterprise will continue and developsustainably.

But there is more to sustainability than just ecologyand the effects of land uses on ecology. People willnot be able to manage their land properly if theirenterprise is uneconomic or does not meet their othersocial needs. So the Centre is looking at theeconomics of land use and people’s social aspirationsas well as ecology. It is examining managementoptions for the users of the savanna to ensure theirneeds are being met. The principal users of thesavannas that the Centre presently aims to serve arethe pastoralists/agricultural sector, the mining andtourism industries, the Aboriginal landowners andconservation interests. It is developing its programwith their advice to ensure it is doing research whichreflects their interests. These users produce verydifferent products from their use of their land. Butthey share a basic infrastructure—the savannas

themselves—and it is the development andmaintenance of that infrastructure for them, and for allAustralians, that is the concern of the Centre.

The consortium comprising the CRC encompassesthe major research and management agenciesservicing the tropical savannas together with theregion’s universities:

Participating Organisations (Core)Northern Territory Government

Parks and Wildlife Commission; PrimaryIndustry & Fisheries; Lands Planning andEnvironment

Queensland Government

Primary Industries; Lands

Western Australia Government

Agriculture, Conservation & Land Management

Federal Government

Environment Australia

CSIRO (Tropical Agriculture; Wildlife &Ecology)

Universities

Northern Territory; James Cook; AustralianNational University (NARU)

Participating Organisations (Supporting)Meat Research Corporation

Land and Water Research and DevelopmentCorporation

Northern Territory Power and Water Authority

Northern Land Council

Northern Territory Mines & Energy

Northern Territory Tourist Commission

Batchelor College (aboriginal education)

A perusal of these organisations indicates that most ofthe major research and management agenciesresponsible for usage and development of the tropicalsavannas are represented in the CRC structure. Thesegroups supply the staff who are responsible fordesigning and implementing the CRC’s programs.While collaborative research between theseorganisations has developed rapidly in recent years, ithas been mainly restricted to within State andTerritory boundaries. The Centre is committed toexpanding these activities and ensuring that industryand stakeholder requirements are met.

To achieve its mission, the Centre is bringing itsstaff together from right across Australia’s Top End ina way that has never been done before. Because theyare part of the Centre, they can cross the State andTerritory boundaries and work beyond theiremployer’s local interests. The program has started

from a base provided by the organisations which makeup the Centre. They all have well established researchprograms and we have drawn relevant parts from themto establish the starting point of the Centre program.The Centre is now building from that start by addingnew, high quality, value-adding components to theprogram which are directed specifically at the needsof savanna users. The program of the Centre is notcomplete and never can be. No research body willever have the resources to deal with all the problemsthat the savannas face. Nor could it deal with all thepossible uses to which the savannas are, or could be,put. Instead, it is a program based on priorities andopportunities that were seen at the time the Centrestarted. It will change with time as new problems andopportunities arise and as the aspirations of savannausers change.

The Tropical Savannas CRC and Eastern Indonesia

The activities of the CRC for Tropical Savannas are ofdirect interest to eastern Indonesia and many otherparts of the globe which experience similar climaticregimes and management problems. The link withIndonesia is particularly strong because of geographicproximity and the existing relationships betweenmajor partners in the CRC (e.g. Northern TerritoryUniversity, Northern Territory Department of PrimaryIndustry and Fisheries) and their counterparts in thisneighbouring country. The establishment of the CRCwill hopefully strengthen these existing ties andfacilitate greater cooperation between relevantagencies. Many of the issues being addressed by theCRC are of direct relevance and concern to easternIndonesia. This will undoubtedly strengthen the casefor collaborative research and information exchangewhich will be of great benefit to both countries.

Northern Australia is remote from the mainpopulation centres in the south of the continent andhas undoubtedly been neglected as far as investmentin research activity is concerned. It could be arguedthat a major encumbrance to accelerated developmentand growth is the lack of appropriate researchleadership to support the aspirations and activities ofthe existing participants and facilitate the entry of newones. A situation such as this would find sympathywith parallel organisations and personnel in easternIndonesia, which experiences similar geographic‘problems’.

ConclusionFire is a prime environmental driver in all wet/drytropical areas of the world and the mechanisms,

10

11

impacts and management issues related to fire in suchenvironments are unique, hence the need for focusedresearch. Likewise, there is an associated requirementfor appropriate technology transfer and educationalactivities so that research outcomes and knowledgeare effectively shared and applied, to enhance humanwelfare and environmental sustainability. ThisWorkshop provides an outstanding opportunity forinformation sharing on fire related issues betweenIndonesia, Australia and other neighbouring countries.The Workshop provides a forum to build on existinggovernmental, institutional and research relationshipsand create new ones. There is scope for highlightingthe key issues relevant to the participants and fordeveloping joint initiatives that contribute to firemanagement systems in both eastern Indonesia andnorthern Australia.

Welcome to the ACIAR Workshop on FireManagement in Eastern Indonesia and NorthernAustralia. I trust the next 3 days will be productiveand stimulating. I look forward to working with you toachieve our goals.

ReferencesJacklyn, P. and Russell-Smith, J., ed., 1998. Proceedings

from the North Australia Fire Management Workshop.Tropical Savannas CRC, Darwin, 24-25 March 1998.

Russell-Smith, J., Lucas, D.E., Gapindi, M., Gunbunuka, B.,Kapirigi, N., Namingum, G., Lucas, K., Giuliani, P. andChaloupka, G. 1997a. Aboriginal resource utilisationand fire management practice in western Arnhem Land,monsoonal northern Australia: Notes for prehistory,lessons for the future. Human Ecology, 25, 159–195.

Russell-Smith, J., Ryan, P.G. and DuRieu, R. 1997b. ALandsat MSS-derived fire history of Kakadu NationalPark, monsoonal northern Australia, 1980–94. Journalof Applied Ecology, 34, 748–766.

Schultz, D., ed., 1998. Fire on the Savannas. TropicalSavannas CRC, Darwin.

ON BEHALF of the Governor of East Nusa Tenggara, Iwould like to express my gratitude for the interest andthe invitation from the Northern Territory Universityresulting in our being able to attend and participate inthis workshop. I warmly welcome this workshop,because it is very strategic and urgent. The workshopwill give a great contribution to the development ofeastern Indonesia, especially the development of EastNusa Tenggara, and at the same time provide valuableinformation for the development of the NorthernTerritory in Australia. This activity constitutes thefirst step to further agricultural and forestrycollaboration between Indonesia and Australia in thefuture.

In this I feel that the Northern Territory Universityand the Australian Centre for InternationalAgricultural Research (ACIAR), along with theNorthern Territory Government, are very consistent,because this workshop represents a follow-up from anInternational Conference on Agriculture in theSemiarid Areas of NTT, East Timor and CentralMalukku, which was held in Kupang, from 10 to 16December 1995. At that conference it was decided to

create an Eastern Indonesia Semiarid AgricultureInformation Centre to follow up on the conferenceresults.

Because of this I request that NTU and ACIARwork closely together with the Information Centre tofollow up on the results of the Workshop for the sakeof development in East Nusa Tenggara and theNorthern Territory. This information centre is fullysupported by the local government of East NusaTenggara and a number of universities in East NusaTenggara. The local government of East Nusa Tenggaragreatly supports the holding of this workshop.

The last two days of visiting a few places in theNorthern Territory have left lasting impressions withus. Because of this we feel that basically we have thesame problems. With close collaboration in future wecan help each other to develop our respective regions.Let us continue to encourage the collaborationbetween East Nusa Tenggara and Northern Territorythrough the NTU and ACIAR by holding researchactivities that will increase inputs to the determinationof policies by the local government of East NusaTenggara.

12

Message of Welcome from East Nusa TenggaraSiliwoloe Djoeroemana1

1Universiti Satya Wacana, Sumba, Indonesia

13

Project Description

BackgroundThis project proposal had its origins with discussionsbetween ACIAR and staff of the CRC for TropicalSavannas during the severe and prolonged fire seasonin Indonesia in 1997.

Eastern Indonesia has major problems of landdegradation and forest degeneration, particularly inthe driest districts (with wet/dry climate and annualrainfall <1000 mm) (Barlow et al. 1990). ‘Under-standing agricultural systems within Nusa Tenggaraand Maluku is essential for the management of nearly40% of the land and for improving the lives of most ofthe people of the region’ (Monk et al. 1997). Thedominant system of cultivation is the production oflow yielding crops by slash and burn agriculture—theswidden cycle. As the population has increased, landdegradation has become widespread, particularlywhere cultivation has extended onto land with slopes

Background to the Project‘The Use of Fire in Land Management in

Eastern Indonesia and Northern Australia’Bronwyn Myers1, Greg Hill1,3

and Jeremy Russell-Smith2,3

AbstractEastern Indonesia has a semi-arid climate with large areas of savanna vegetation, and consequently theregion is fire-prone. The land is used mainly for low-yielding subsistence agriculture and the regionfaces major problems of land degradation and forest degeneration. Fire is a fundamental component oftraditional and current land management. Prescribed fire is used in slash and burn cultivation, for landclearing, weed control and to increase nutrient status. Fire is also used to promote new grass growth forcattle grazing and to provide some protection from wildfire. Some fire regimes have undesirableeffects. Some fire and grazing regimes encourage the invasion of grazing lands by shrubs, threatenlives, property and forestry reserves, and contribute to regional atmospheric pollution. Fire has impactson soil and water conservation, which all affect the long-term productivity of the land.

Although the extent, severity and impacts of fire appear to be increasing, these changes are poorlydocumented. There is little understanding of the processes affecting fire behaviour in easternIndonesia and the steps necessary to better manage fire.

Current and past work by non-government organisations and international funding agencies ineastern Indonesia verify the importance of integrating fire management with animal husbandry andother farming practices when recommending modifications to land management practices. Anunderstanding of the cultural and economic restrictions within Indonesian communities is essentialfor the successful adoption of recommendations for change.

The objectives of the workshop were to: • review fire management in eastern Indonesia and northern Australia • develop closer regional and institutional links between researchers and land managers in northern

Australia and eastern IndonesiaThe workshop included position papers, describing the state of knowledge in a range of fire

management issues, followed by discussion sessions.Workshop tasks were to:

• identify and describe the key fire management issues for eastern Indonesia and northern Australia; • summarise recommendations and resolutions generated by the discussion sessions;• discuss methodology appropriate for adoption of recommendations and for research and training

programs in eastern Indonesia and northern Australia;• publish the proceedings of the workshop in the ACIAR Proceedings series.

1Northern Territory University, Darwin 0909, Australia2Bushfires Council of Northern Territory, PO Box 37346,Winnellie NT 0821, Australia 3Tropical Savannas CRC, Northern Territory University,Darwin 0909, Australia

14

greater than 30%. Also there has been degenerationfrom forests to secondary forests to savanna, and inplaces to ‘arid steppe’ (Ormeling 1957). Theseprocesses have been accelerated by large increases incattle numbers early this century and the use of fire forburning off scrub and old grasses (Barlow et al. 1990).The consequent effects on hydrology and increasederosion have threatened the viability/productivity ofirrigated areas downstream (Barlow et al 1990).

Fire is a fundamental component of current landmanagement practices in the wet/dry regions ofeastern Indonesia. Prescribed fire is used for landclearing, release of nutrients for crop establishmentand weed control at the beginning of the swiddencycle and for both weed control and to encourage newgrass growth for cattle grazing in savannas. Firemanagement, both the wise use of prescribed fire andeffective control of wild fire, has major impacts onretention of forest cover (only 6% of the originalforest remains on Sumba), the relative proportions ofgrasses and shrubs, and soil and water conservation.Although the extent, severity and impacts of fireappear to be increasing there is little understanding ofthe processes affecting fire behaviour and the stepsnecessary to better manage fire. Improved firemanagement practices would aim to use fire to controlweeds and promote regeneration of palatablevegetation whilst conserving topsoil and surface soilstructure, thus increasing water infiltration and waterholding capacity of the soil as well as preserving thenutrient status of the soil.

These factors are particularly important at thebeginning of the wet season when crops are beingestablished and also would affect productivity of bothcrops and rangelands.

Eastern Indonesia, with its semi-arid environmentand relatively low population density (by regionalstandards), contains the poorest provinces in thecountry. Annual rainfall and the length of the wetseason are highly variable and a particularly dry wetseason occurred in 1997–1998. The region’s‘capricious relief’ (Ormeling 1957) causes local rainshadows which exacerbate the divergence inagricultural potential. The area is remote from Jakartaand the government agencies promoting development.The need for international assistance in this region iswidely acknowledged. During the recent problemswith choking haze in the region, the TropicalSavannas CRC, through its partner agencies, hasprovided remotely sensed imagery of the fires ineastern Indonesia on a daily basis at the request of theEnvironment Ministry in Jakarta. Managementagencies within the region have listed firemanagement as one of the issues of high priority forscientific input (Information Centre for Semi-aridAgriculture, Kupang, pers. comm.). In 1997, theInformation Centre submitted to ACIAR and NTU a

list of projects of high priority. It included a‘Biophysical study of traditional fire management inshifting cultivation’ and other projects of which firemanagement would be an essential part (‘Aninventory of indigenous wisdom’ and ‘The integratedcontrol of Chromolaena odorata’). The NusaTenggara Regional Development Consortium(NTRDC), a collaborative network with over 100partners, has identified the need for investigation offire management to complement current activities inland management.

The NTRDC provides a network for collaborativework aimed at development in Nusa Tenggara. TheAgroforestry and Conservation of Natural ResourcesWorking Groups of NTRDC have fostered cooperativeprojects by a wide range of groups (government, NGO,farmers and educational institutions). Several of theseprojects, which are currently in progress in easternIndonesia, are highly relevant to fire managementissues. Works by various NGOs have attempted toreduce the extent of fires by encouraging the plantingof fire tolerant species around the farm, the use of firebreaks, closer control of the fires lit by farmers and theuse of controlled/‘polite’ fires on the farm. Burning isan integral part of traditional farming that has beenpractised for generations, however there has beensome success in modifying fire managementpractices. The impacts of changes to fire managementhave been poorly documented to date (S. Wodickapers. comm.).

The Tananua Foundation (a local NGO) is workingon improving farm management practices in Sumbawhere fire is used chiefly for improving feed for free-ranging cattle but also as part of hunting for wild boarand deer and scavenging wild tubers. Further work onfire management needs to be integrated with studies ofcattle husbandry (S. Wodicka, pers. comm.).

In Sumba there has been a steady invasion of thedominant Imperata grasslands by the woody shrubChromolaena, which burns easily and is unpalatable.This reduction in grazing land increases the pressureon farming land from free-ranging grazing animals.Fire management could play a vital role in modifyingvegetation change in grazing land such as east Sumba.

A group headed by Dr Jeff Fox (East West Centre,Hawaii) has begun a project aiming to develop localcapacity in spatial information analysis to assist theNusa Tengarra Regional Development Consortium(NTRDC), with funding from the Ford Foundation.

Fire is also a major land management issue for thesparsely populated savanna regions of northernAustralia. Savannas predominate in northernAustralia where annual rainfall exceeds 500 mm—northwestern Western Australia, the northern half ofNorthern Territory and north Queensland—representing 25% of the continent. Fire is extensive inAustralian savannas; for example, 50% of the Top

End of the Northern Territory was burnt in 1996. Fireis also frequent in Australian savannas. In humidsavannas (> 1000 mm annual rainfall), fires occur oneyear in two or two years in three (Braithwaite andEstbergs 1985; Press 1988; Russell-Smith 1995). Insemi-arid savannas (500–1000 mm annual rainfall)fire is a common land management tool in free-rangecattle grazing systems, the predominant land use.Prescribed fire is used to control woody weeds insemi-arid savannas (Grice 1997). Management ofgrazing and fire affects the long term productivity ofthe land via effects on (i) the soil condition and itsinfluence on distribution of nutrients and water, and(ii) the species composition of pastures, particularlythe relative abundance of perennial and annualgrasses, and therefore the value of this pasture forgrazing (Williams et al. 1997).

Fire in northern Australia is the subject of a majorcoordinated effort through the Cooperative ResearchCentre for Tropical Savannas, based at NorthernTerritory University (NTU) in Darwin (Russell-Smith1997). The CRC program involves: (i) thedevelopment of broad- and finer-scale satellitemonitoring technologies to assess the extent,seasonalities and frequency of fires in differentregions and across sectors (Hill 1997); (ii) theundertaking of focused research to assess the impactsof different fire regimes on natural and pastoralsystems; (iii) the development of regionalmanagement strategies involving all communitysectors as a means for developing effective, integratedapproaches to fire management (Russell-Smith 1996).Such an approach provides an appropriate frameworkfor addressing fire management issues in the easternIndonesian region also (Hill 1995).

The phenomenon of regional haze associated withsmoke and dust aerosols has long been observed inSoutheast Asia. Braak (1929) observed regional hazeover Indonesia and attributed this in part to dustaerosols from Australia. Since 1980, there has been‘an unprecedented amount of wildfire activity’ in theregion (Tapper et al. 1995).

The problem is particularly acute following severedroughts associated with the El Nino–SouthernOscillation (ENSO) events and due to the increasedflammability of drought-stressed vegetation, e.g.1982–83, 1987 (Goldammer 1993) and 1991–92,1994 (Tapper et al. 1995). Tapper et al. (1995)suggested that the periodic regional haze overSoutheast Asia results from a variety of causes:traditional Indonesian slash and burn agriculture andassociated break-away fires; clearing of primary andsecondary rainforests; fires in seasonally-drymonsoon savanna forests of mainland South Asia andnorthern Australia; fires in coniferous forests of SouthAsia; burning of agricultural residue, domestic woodfires and urban-industrial emissions. The

development of fire management strategies forIndonesia should include a consideration of theimpacts on atmospheric chemistry. The regional andglobal scale problems of fire in Indonesia andnorthern Australia are the subject of the South EastAsia Fire Research Experiment (SEAFIRE), which isan International Global Atmospheric Chemistry(IGAC) project, a core project of the InternationalGeosphere-Biosphere Program (IGBP).

Objectives and OutputsThe project aims to :a) overview existing knowledge of fire management

in eastern Indonesia and northern Australia;b) identify and prioritise the key fire management

issues;c) develop closer regional and institutional links

between researchers and land managers in northernAustralia and eastern Indonesia;

d)publish the proceedings of the workshop in theACIAR Proceedings series, documenting the stateof knowledge and disseminating the workshopresolutions and recommendations.The workshop facilitated the exchange of

information, including that which is not yet availablein written form, established a network forcommunication between workers in this area andpromoted the development of cost-effective researchand training programs.

Economic SignificanceFire has major impacts on two critical resources ineastern Indonesia—land and water. Prescribed fire hasdirect effects on slash and burn cropping andextensive grazing in savannas and secondary forests.In Nusa Tenggara, in terms of land area, 15% is foodand tree crop agriculture, 21% is forest that isencroached on by grazing and some slash and burn,and 64% is degraded secondary forest and savannaused for cattle grazing (1986 data, Barlow et al. 1990).

Fire management has direct impacts on agriculturalproduction in the long term. In Nusa Tenggara Timurduring 1995 (Nusa Tenggara Timur Dalam Angka1995) agriculture contributed about 40% of theregion’s gross domestic product (RGDP) with foodcrops constituting 57% and livestock constituting24% of agriculture’s contribution (945 Rp billion) toRGDP. Agriculture employed 70–80% of thepopulation directly (Barlow et al. 1990).

In Australia fire is also of major significance, as atool in pasture management for the tropical beefindustry but in a more negative way as a factor in landdegradation, also in loss of human life and property.

15

Benefits to Australia will be primarily related todeveloping closer international ties with our closestneighbours. More specifically, this project will (1)allow Australia to contribute to increasedunderstanding of regional environmental issues,including regional smoke haze, (2) increaseopportunities to export technologies, and (3) assist inthe sustainable development of one of Australia’sstrongest export markets, thereby improving theregional economic outlook.

Application of ResearchRecommendations for land managers, based on currentknowledge and effective means of implementation,will be published in the proceedings. The research anddevelopment priorities identified at the workshop willalso be published in the proceedings and will provide aframework for future activities for research fundingbodies. Policy makers in the Indonesian governmentwill be advised by the representatives of forestry andagriculture departments of both Central and ProvincialIndonesian governments who have participated in theworkshop and the proceedings will provide a furthersource of information for consideration in policydevelopment.

Environmental ImpactThe longer-term outcome of the project will be toassist development of more sustainable landmanagement in eastern Indonesia. Modification to firemanagement will ultimately improve the productivityof farmers via improvement to the condition of theland and conservation of water resources. Thoseinvolved will place special emphasis on design of themost effective mechanisms of adoption. Any researchand training projects proposals will promote thesustainability of productivity of the land and waterresources and enhancement of forest regeneration andnative biodiversity.

ReferencesBarlow, C., Gondowarsito, R., Birowo, A.T. and Jayasuriya,

S.K.W. 1990. Development in Eastern Indonesia: Thecase of Nusa Tenggara Timur. InternationalDevelopment Issues No. 13, Australian InternationalDevelopment Assistance Bureau, Canberra.

Braak, C. 1929. The Climate of the Netherland Indies;Volumes I and II. Verhandelingen No. 8, KononklijkMagnetisch en Meteorologisch Observatorium teBatavia.

Braithwaite, R.W. and Estbergs, J.A. 1985. Fire pattern andwoody vegetation trends in the Alligator Rivers regionof northern Australia. In: Tothill, J.C., and Mott, J.J.,ed., Ecology and Management of the World’s TropicalSavannas, Australian Academy of Science, Canberra.

16

Literature ReviewAlmost all the published work on fire in Indonesia isbased in the wet tropics, particularly on Java (e.g.Pickford et al. 1992, Christanty et al. 1997). Uplandareas (>1500 mm annual rainfall) have been the focusof a study of smallholder farming systems in Imperataareas in Southeast Asia that has provided evaluationsof land management practices, including use of fire(Grist and Menz 1996/7, Menz et al. 1996/7). Thescarcity of written information on fire in Indonesia isillustrated in the recently published ‘Biomass burningand global change’ (Levine 1996) in which the sectionon Southeast Asia comprises two papers neither ofwhich includes Indonesia.

Monitors of fires in Indonesia now use satelliteimagery processed by the Western AustralianDepartment of Land Administration, a member of theTropical Savannas CRC. On a regional scale,monitoring of the fire theatre and implications foratmospheric pollution have begun with theestablishment of the South East Asian FireExperiment (SEAFIRE) in 1996. The broad aims ofSEAFIRE are ‘to identify the magnitude, patterns,quality and impacts of fire on the local terrestrialecology and regional atmosphere’ (Tapper et al.1995). However at this stage, data available throughSEAFIRE are mostly based in the wet tropics(Goldammer 1993), particularly Borneo (Goldammerpers. comm.), and include the ecological impacts ofwildfire with some information about fire inmaintaining agricultural and pastoral systems and theimplications for atmospheric pollution.

Some useful summaries of current fire research innorthern Australia include the proceedings of Bushfire97 (conference held in Darwin in July 1997), aworkshop on remote sensing in Kunnunurra on 6August 1997 and reports from a workshop organisedby the Tropical Savannas CRC on Fire Management,held in Darwin in March 1998.

Economic Impact of the Research and Development

The immediate economic benefits from this workshopwill be the efficient exchange of information and theopportunity for productive discussion betweenscientists (both Australian and Indonesian) andrepresentatives of the Central and ProvincialGovernments of Indonesia. Greater benefits willfollow from further research conducted as a result ofresearch priorities identified at this meeting. Theinclusion of socioeconomic considerations will ensurethat research is developed and knowledgedisseminated in a way that will be most readilyadopted by farmers.

Christanty, L., Kimmins, J.P. and Mailly, D. 1997. ‘Withoutbamboo, the land dies’: a conceptual model of thebiogeochemical role of bamboo in an Indonesianagroforestry system. Forest Ecology and Management,91, 83–91.

Goldammer, J. 1993. Indonesia: long-term integrated forestfire management programme initiated at Bandung.ECE/FAO International Forest Fire News, 8, 9–12.

Grice, A.C. 1997. Post-fire regrowth and survival of theinvasive tropical shrubs Cryptostegia grandiflora andZiziphus mauritiana. Australian Journal of Ecology, 22,49–55.

Grist, P. and Menz, K. 1996/97. Burning in an Imperatafallow/upland rice farming system. Imperata ProjectPaper 1996/97 (ACIAR/CIFOR).

Hill, G.J.E. 1995. New research initiatives for the semi-aridsavannas of northern Australia. Proc. InternationalConference on Agricultural Development in Semi-aridEast Nusa Tenggara, East Timor and South EastMaluku, Kupang 11–15 December, 9.

Hill, G.J.E. 1997. A new focus for remote sensing and GISactivities in northern Australia. Asian-Pacific RemoteSensing and GIS Journal, 10 (1), 15–25.

Levine, J.S., ed., 1996. Biomass burning and global change.Volume 2. (MIT Press, Canberra.)

Menz, K., Conroy C., Gunawan, A., Ellis, K., Grist, P. andBagnall-Oakley, H. 1996/97. Fire as an economicdisincentive to smallholder rubber planting in Imperataareas of Indonesia. Imperata Project Paper 1996/97(ACIAR/CIFOR).

Monk, K.A., de Fretes Y. and Reksodiharjo-Lilley, G. 1997.The Ecology of Nusa Tenggara and Maluku. TheEcology of Indonesia Series, Volume V, PerilpusEditions (HK) Ltd.

Nusa Tenggara Timur Dalam Angka 1995. Kantor StatistikBPS Propinsi NTT.

Ormeling, F.J. 1957. The Timor Problem. A GeographicalInterpretation of an Underdeveloped Island, J.B.Woltens, Groningen, and Martinus Nijhoff,S’Gravenhage.

Pickford, S., Suharti, M. and Wibowo, A. 1992. A note onfuelbeds and fire behaviour in alang-alang (Imperatacylindrica). International Journal of Wildland Fire, 2,41–46.

Press, A.J. 1988. Comparisons of the extent of fire indifferent land management systems in the Top End ofthe Northern Territory. Proceedings of the EcologicalSociety of Australia, 15, 167–175.

Russell-Smith, J. 1995. Fire management. In: Press, A.J.,Lea, D.A.M., Webb, A.L. and Graham, A.D., KakaduNatural and Cultural Heritage and Management.Australian Nature Conservation Agency/ NorthAustralian Research Unit, Darwin.

Russell-Smith, J., ed., 1996. Developing a coordinatedapproach to fire management and monitoring issuesacross northern Australia: common issues,opportunities. Report of the North Australia Rural FireManagers Technical Workshop, Darwin, October 1996.For the Land & Water Resources Research &Development Corporation, Canberra.

Russell-Smith, J. 1997. Developing a coordinated firemanagement and research program for northern Australia:the role of the CRC for Tropical Savannas. In: Proceedingsof Bushfires ‘97 Conference, Darwin, July 8–11. CSIRO,CRC for Tropical Savannas, and Parks Australia.

Tapper, N., Wain, A., Hooper, M., Vanderzalm, J. andGoldammer, J. 1995. An overview of the forthcomingSouth East Asia Fire Research Experiment.

Williams, R.J., Cook, G.D., and Ludwig, J.A. 1997. Torch,trees, teeth and tussocks: disturbance in the tropicalsavannas of the Northern Territory, Australia. In: Klomp,Nicholas and Lunt, Ian, ed., Frontiers in Ecology:Building the Links, Elsevier Science, Oxford UK, 55–66.

17

EAST NUSA TENGGARA PROVINCE (NTT) is one of thearchipelagic provinces in Indonesia. This archipelagoconsists of 566 big and small islands, of which 50 areinhabited by 3.5 million people who speak more than50 tribal languages and dialects. The total land area isabout 47 000 km2 or about 20% of the total land andwater area combined. This brings problems oftransportation and development management. Manyagricultural activities of traditional smallholderscannot be effectively controlled on the land areas, norcan the fishing and marine activities on the water orcoastal areas. As the closest neighboring country toAustralia, many fishermen from NTT enter Australianwaters and harvest the marine resources.

Agricultural ActivitiesThe livelihood of approximately 80% of the NTT

population depends on dryland cultivation andextensive cattle husbandry. Most still practice shiftingcultivation, and about 10% work on wet rice field andcash crops. This is largely a consequence of lowrainfall.

Average annual rainfall varies from 700 to 2900 mm(1992–1997) and is distributed unevenly among theislands. We have only two seasons, wet and dry. Thewet season begins in November and lasts until

February (normally ends in January) while the drytakes up the rest of the year, approximately 9 months.As agricultural activities depend mainly on rainfalland soil fertility these two natural phenomena have agreat effect on the variation of agricultural productsand the income of farmers. Foodstuffs such as maize,cassava, peanuts, various kinds of beans, green grainsand rice (paddy) are very dependent on rainfallintensities.

Too little rain disturbs the growth of food crops. Buttoo much rain also disturbs the harvest of maize,particularly at the moment, as La Nina brings morerainfall than usual. The maize fortunately ripenedwell, but farmers have less opportunity to harvest ontime. This means the rain has tended to damage thecorn. On the other hand the wet paddy farmers havehappily cultivated their wet land and prepared theirseedlings. By comparison, last season farmers failedto harvest because of the drought brought by El Nino.

Various agricultural efforts are carried out by thegovernment (at the national and local level) toincrease farmer income—for instance, instant creditfor agricultural land processing, and subsidies onchemical fertilisers, on supply of high yielding varietyseeds, and on the marketable price of the farmers rice.Such innovations have been successfully dispersedamong the wet rice farmers and cash crop planters, butless to the traditional dryland cultivators. Unfortunately,they adopted the innovation of the high yield varietyof maize (limited amount of fertilisers) and grew it formarket purposes rather than family food stocks.

18

Shifting Cultivation and Fire: a Challenge to NTT’s Development

Esthon L. Foenay1

AbstractEast Nusa Tenggara Province (NTT) is one of the archipelagic provinces in Indonesia, consisting of566 islands, of which 50 are inhabited by 3.5 million people speaking more than 50 tribal languagesand dialects. This paper gives an overview of the agriculture practised throughout the province andthe way in which fire is used in the cropping cycle and in the raising of livestock.

1BAPPEDA (Regional Development Planning Board of NusaTenggara Timur)

Agricultural innovations are adopted for cash whilethe farmers still practice their local knowledge forsubsistence.

On average, the per capita income of NTT’spopulation in 1988 (calculated at the current price)was Rp. 278 000, increasing to Rp. 466 000 in 1992and then to Rp. 877 000 in 1997. With the currentexchange rate of Rp. 9000 per US$ it meant thatNTT’s income per capita in 1997 was only aboutUS$97.00. This income figure shows the poverty rateof the people and the decreasing of their welfare asIndonesia’s currency drops in relation to the US$.

Land Cultivation, Cattle Husbandry and the Means of Fire

Agricultural activities of most farmers are a mixtureof dry field or wet land farming and cattle farming.Problems usually occur between July and October, thetime for preparing dry fields. Since soil fertility ismaintained by land rotation rather than crop rotationthe trees and bushes are cut down, left to dry for aperiod and then cleaned up by means of fire. In manycases fires become uncontrollable and burn allsurrounding bushes, forests and pastures. Thispractice is done repeatedly for 3–4 years, then farmersmove on when the productivity of the land decreases.

Local beliefs also lead people to choose a certaintype of land. Usually land areas on a steep slope arechosen rather than flat areas. It seems that cost savingsmust also be brought into consideration—no energyspent for weeding, and no environmentalconsiderations. Top soil is damaged by water runoff,the fertility decreases, and then the farmer leaves thedeteriorated area to repeat the process on a new area.

To avoid the types of mistakes in land cultivationthat were just mentioned, there have been manyefforts made by the local governments, e.g., lamtoro(Leucaena leucocephala) is introduced for landcovering and steeply sloped field terracing. Farmersalso benefit from lamtoro in another way and that isfor feeding their cattle. The 1960s was a period ofrejection of lamtoro because it was claimed to causehair loss in horses, because at that time horses had arole in transportation. But in the 1970s, coincidingwith the introduction of cattle fattening programs,people began to adopt lamtoro. Since then lamtoro hasbeen gradually adopted throughout the islands ofNTT, but fire damage is still a challenge. Besides theland clearing process, high temperatures in the dryseason also contribute to the burning of dry grass andforests, leading to even worse land deterioration.

It was reported that between 1995 and 1998 some13 000 ha of critical land were planted with variouskinds of trees. This excludes what was planted beforethen. An official report shows that in 1997 there were

some 39 000 ha of forests and pastures burnt out andin 1998, 7000. Therefore there is actually noguarantee that all the trees planted are still growing asthe figures above suggest. The challenge is stilldryness and associated fire.

An extensive contribution to the lifting of farmers’incomes has come from cattle raising, it also brings afairly serious problem of over grazing. It was reportedin 1997 that there were about 781 000 cows, 476 000buffaloes, 137 000 horses, 738 000 goats and 174 000sheep spread out in 12 districts of the main islands andthe surrounding satellite islands. Overgrazing cannotbe avoided, because the cattle graze freely throughoutthe forests and pastures. Unfortunately at the peak ofthe dry season when available stocks of cattle feed arelow, farmers intentionally burn pastures with theexpectation that young grass will sprout. But the fires,blown by strong winds, often grow bigger and sweepthrough the surrounding forests and pastures. If heavyrains fall then the flood will wash away topsoil andleave behind barren land.

There is no real effort to extinguish these burningforests and pastures. The poor pastures constitute non-plantable land and lack of water is the main problem.However it is not only that. The cultural value systemsalso play a dominant role. The highland people ofWest Timor (Helongese) for instance are drylandcultivators. Wet land cultivation was introduced in the1970s and extended in the 1980s. Growing wet ricewas new to their culture. After sowing wet riceseedlings no weeding or pest protection wereemployed (such practices are not used in drylandculture), but the farmers still expected an abundantharvest. One can easily imagine that insect attack anddamage from mice would affect the young rice plants.The uncleared bushes surrounding the rice fieldsprovide a favorable environment for mice and insectsto multiply rapidly.

In places where water was still available during thedry season there were no efforts by farmers toundertake farming activities, for example to growvarious kinds of vegetables or cultivate freshwaterfish. After this period there was a disparity betweenthe lowland Timorese and the Helongesecommunities. Small dams were built in the mentionedcommunities to provide a clean water supply, and thewater was also used for growing various vegetablesand beans during the dry season, implementing a newmethod of dryland cultivation with assured supply ofwater, introduced by agricultural extensionists. Thefarmers, of course, gain extra income withoutincurring extra costs. They need to be advised andencouraged for approximately 4 years by agriculturalextensionists.

The Timorese coastal inhabitants have notconsidered coastal waters as a potential food source.But this is actually changing slowly with the younger

19

generation. Once about 5 years ago I met two youngTimorese men in their small sailing boat with a net onhand to catch small fish. They were not confidentenough to sail more than 50 metres from shore. Afterthey returned to the shore, when I asked them abouttheir new experience in catching fish, they told methat their neighbour, an elderly Chinese man, hadtaught them.

AcknowledgmentsThat concludes the few thoughts that I wished to raiseat this opportunity. We would like to express ourdeepest thanks for your attention and interest. Onbehalf of the Governor of Regional Level One, EastNusa Tenggara, we would like to express our thanksto the University for their interest and for theinvitation enabling us to participate in this workshop.

20

21

THIS PAPER provides an overview of atmosphericissues of fire management in the tropics, withparticular reference to eastern Indonesia and northernAustralia (referred to hereafter as EINA). This broadregion, lying at the southern margin of the maritimecontinent, is characterised by its seasonally wetclimate and a fire regime dictated largely by seasonaldrought.

Regional fire regimes can be defined bycharacteristic fire intensities, fire frequencies andseason of burn. Atmospheric factors exert fundamentalcontrols on this regime, and also determine thetransport of products of burning. Around the globethere are various indices employed, mainly forpredictive purposes that represent aspects of a fireregime. In many areas of Australia the McArthurForest Fire Danger Index Mark 5 (McArthur 1968) iscommonly used to represent the daily fire danger,where the index is indicative of a fire starting, its rateof spread, intensity and difficulty of suppression. TheFire Danger Index (FDI) is defined mathematically as:

FDI = 1.275D0.987*[exp(0.0338T+0.0234V-0.0345H)]

where D is a drought factor, T is air temperature(°C), V is the 10 metre wind speed (km/hr) and H is

the relative humidity (%). Details of the equation andits derivation can be found in Noble et al. (1980).

As well as its regionally specific application inpredicting/quantifying fire danger on a day-by-daybasis (the purpose for which it was intended), the FDIis also a valid indicator of fire danger over the entirecontinent (and by extension the islands immediately tothe north of Australia that share a similar climate) andcan be used for comparison of regional, intra-seasonaland inter-annual fire danger variability. Themeteorological and climatological parametersrepresented in the FDI therefore provide a usefulconceptual framework for the discussion of the fireregime in the EINA region that follows in theremainder of this paper. However, it must beremembered that although fire regime characteristicsare highly coupled with weather and climate, therelationships are often complex. For example,variations in fuel also influence the properties of a fireregime such as fire frequency and fire intensity.

Luke and McArthur (1978), in their landmark texton Australian bushfires, produced a map based on firestatistics showing the generalised pattern of seasonalfire occurrence in Australia (Figure 1). Generally, thetrend is for the fire season to be progressively delayedfrom north to south across the continent. In northernAustralia (and by inference much of easternIndonesia), which has a true monsoonal climate, thelong austral winter dry season produces a late austral

Atmospheric Issues for Fire Management inEastern Indonesia and Northern Australia

Nigel Tapper1

AbstractUsing the McArthur Fire Danger Index as a framework for discussion, this paper provides anoverview of atmospheric issues of fire management in the tropics, with particular reference to easternIndonesia and northern Australia. After a brief discussion of the background meteorology of theregion, atmospheric influences on regional fire regimes that may operate over a range of time scalesare examined. Longer-term influences associated with predicted greenhouse warming suggest anincreased fire danger for the region. Inter-seasonal and inter-annual climate variabilities associatedwith the El Niño–Southern Oscillation phenomenon are also shown to impact on the regional fireregime. While impacts of seasonal drought associated with the north-south movement of the inter-tropical convergence zone (ITCZ) are well known to fire managers, general atmospheric transportpatterns associated with movement of the ITCZ are not. Some results of recent trajectory modellingare presented to elucidate these patterns. Intra-seasonal and synoptic-scale influences on regional fireregimes, especially as they impact southern parts of the region, are also discussed. Finally,suggestions are made about some priority areas for further research.

1School of Geography and Environmental Science and CRCfor Southern Hemisphere Meteorology, Monash University,Clayton 3168, Victoria

22

winter–early spring fire season. The fire seasonality ofnorthern Australia suggested by Luke and McArthuris supported by an analysis of fire ban records for theNorthern Territory (Figure 2). Declaration of a fireban is strongly influenced by fire weather warningsissued by the Bureau of Meteorology on the basis ofpredicted FDI (predicted FDI>45 provokes awarning). Over the Northern Territory as a whole, fireban days are at a minimum in April and at a maximumin the August–November period. However there areclear regional differences exhibited in Figure 2, withthe apparent fire season in the Top End (north of 16°S)extending from May to October (dry season), in thecentral region (16–21°S) from August to November,and in the southern region (21–26°S) from August toMarch. Clearly there are regional fire regimes in theNorthern Territory that can be documented indistinctive seasonal distributions of FDI.

The following sections of this paper place themeteorology of the EINA region in a global contextand discuss the major atmospheric influences on thefire regime of the region, particularly as they affect thecomponent parameters of the FDI. Atmosphericinfluences on the transport of the products of burningare also considered.

Background MeteorologyThe EINA region lies on the southern edge of themaritime continent region, one of three regions ofenhanced tropical convective activity located aroundthe world. Figure 3, showing mean outgoing long-wave radiation (deep tropical clouds have lowoutgoing long-wave radiation) identifies CentralAfrica and Amazonia as the other areas of intensetropical cloudiness, but both of these are continentalby comparison with the EINA region. Hence the term‘maritime continent’ coined by Ramage (1968). Theenhanced cloudiness in these regions representsmassive exchanges of energy that are fundamentallyimportant in the general circulation of the globalatmosphere. Of these regions, none is more importantin global climate dynamics than the Indonesianregion, because of its role in both the north–southtropical Hadley circulations and the east–west Walkercirculations (Sturman and Tapper 1996).

The other major characteristic of Indonesian climateis the seasonal reversal of wind flow across the regionas a pivotal part of the Asian–Australian monsoonwhich is an important component of the planetarymonsoon system. Figure 4, illustrating wind flow at500 m, clearly shows the seasonal reversal in low-levelmonsoon currents, particularly in the South China Sea(northeasterly trade winds in January and thesouthwest monsoon in July) and the Timor Sea region(moist westerly monsoon flow in January and drysoutheasterly trade wind flow in July). Winds in the

Figure 1.Patterns of seasonal fire occurrence in Australia(from Luke and McArthur 1978).

MonthJ

0

10

20

30

40

50

F M A M J J A S O N D

Num

bero

ffire

bans

Figure 2.Monthly distribution of fire ban days by latitudeband in the Northern Territory 1980–1990 (fromTapper et al. 1993).

40°N

20°N

0°

20°S

40°S

Figure 3.Mean annual outgoing long-wave radiation(W/m2) for December–February. Tropical areaswith outgoing long-wave radiation <240 W/m2 areshaded (from Sturman and Tapper 1996).

23

EINA region are relatively light (Figure 5),particularly in Indonesia and in the monsoon transitionperiods of March–April and October–November as theInter-Tropical Convergence Zone (ITCZ) moves northand south across the region. The ITCZ, whichrepresents the broad zone of convergence of tropicaleasterly flow from each hemisphere, achieves itsgreatest seasonal movement in the Asian region(Figure 6). As easterly flow from one hemispherecrosses the equator, earth rotation causes the flow to

recurve to become westerly monsoon flow in the otherhemisphere.

The climatological characteristics of regional airflow described above are present more or less everyyear, and are important both in the seasonaldistribution of rainfall and the dispersion of smokefrom biomass burning. Southern and eastern parts ofIndonesia and northern Australia generally have lowrainfall May–October under the influence of the drytrade winds blowing out of central Australia, with

Figure 4.Regional scale low-level winds at four times of the year, showing theseasonal reversal of the monsoon flow. The length of arrow denotes thestrength of wind (from McBride 1992).

Figure 5.Regional scale wind speed at 500 metres at four times of the year.Isotachs are at 2 m/sec intervals and winds >4 m/sec are shaded (fromMcBride 1992).

24

rainfall increasing into the transition seasons andduring the monsoon period. Weak low-level windsacross southern Indonesia during the transitionseasons mean that there is considerable air stagnation,leading to unacceptable levels of haze when forest andpeat fires burn uncontrolled. Further south, winds arestronger over northern Australia, particularly duringthe austral winter. Upper air inversions that arecharacteristic of most trade wind flow regimes(Sturman and Tapper 1996) are seen at heights of~2000 m over northern Australia, but are less distinctover Indonesia. Nevertheless characteristic maximumatmospheric mixing depths that help define the heightof dispersion of products of burning away from thesurface during the dry season would appear to beabout 1500 m over Indonesia (Nurhayati 1994unpubl.).

Over much of the EINA region it is probable thatseasonal drought is the greatest influence on FDI,since temperature and relative humidity tend to berelatively constant and high, and wind speedsrelatively low. Away from the coast, and further southin the region, temperature, relative humidity and windspeed are likely to become more important.

Atmospheric Influences on Fire Regimes ofEastern Indonesia and Northern Australia

While EINA regional climate exhibits a generaldegree of predictability, i.e. warm-tropical, seasonallywet with a seasonal reversal of prevailing low-levelwinds, there is a range of influences operating overvarying time scales that affect the region’s climate andfire regimes. These are illustrated in Figure 7 and arediscussed in the following sections, beginning withthose influences over the longer term.

Longer-term influencesThere is much uncertainty about the effects ofenhanced greenhouse warming on global and regionalclimates. Any climate change is likely to havesignificant effects on biosphere–atmosphereinteractions, including fire regimes. Recent work inour group (Williams 1997 unpubl.; Williams et al.1999) has sought to quantify the impact of climatechange on Australian fire regimes by estimatingchanges in fire weather parameters and the FDI for adoubled-CO2 climate. The CSIRO-9 level generalcirculation model (GCM) was used in thesesimulations. Figure 8 shows both the cumulative fireseason FDI (daily FDI summed over the entire fireseason defined as July–November for northernAustralia) associated with 2 × CO2, along withpercentage changes in SFDI from 1 × CO2 to 2 × CO2simulations. This indicates that fire danger increasessignificantly over all of northern Australia, with thelargest increases (30%) occurring on the northeastcoast and in the south of the region. The change in FDIacross the region is a result of increased temperatureand reduced humidity during the fire season, withwind speeds and rainfall showing no significantchange. Williams et al. also show that the frequencyof extreme FDI occurrence increases markedly acrossall of northern Australia.

As will be shown below, the El Niño–SouthernOscillation (ENSO) phenomenon has an importantimpact on fire regimes in the EINA region. Somerecent GCM simulations coupling atmosphere andocean have raised concerns about possible changes inENSO under enhanced greenhouse conditions. Meehlet al. (1993), Knutson and Manabe (1994, 1995) andSmith et al. (1996) all produced ENSO-like events intheir greenhouse simulations, while the work ofGordon and O’Farrell (1996) indicated weakenedENSO-like phases. This highly sophisticated

30°N

0°

0°

30° 60° 90° 120° 150° 180° 150° 120° 90° 60° 30° 0°30°S

30°N

0°

0°

30° 60° 90° 120° 150° 180° 150° 120° 90° 60° 30° 0°30°S

Figure 6.Mean position of the ITCZ during January and Julyand streamlines of near-surface windflow in thetropics (from Sturman and Tapper 1996).

Atmospheric influences on

N.Indonesian fire regimes

Inter-seasonal to inter-annual influences • Southern Oscillation (ENSO) • Quasi-biennial Oscillation

Longer-term influence • Greenhouse climate change

Seasonal influence • Plantary monsoon

Synoptic and meso-scale influences • Trade wind surges • Fronts/troughs • Land-sea/mountain-valley winds

Intra-seasonal influences • 40-50 day oscillation • Australian West Coast Surges • South China Sea Surges

increased time scale

increased time scale

increased time scale

increased time scale

Figure 7.Scales of atmospheric influence on the easternIndonesian–northern Australian fire regime.

25

modelling work must continue for us to be betterinformed about likely future impacts on ENSO.

Inter-seasonal and inter-annual influencesThe Walker circulation is a major circulation cellmoving air between the eastern and western sides ofthe southern Pacific Ocean. It is oriented north–westto south–east across the Pacific with a rising limbcentred over Indonesia and descending motion overthe eastern Pacific (Figure 9). Fluctuations in theintensity of this circulation occur during ENSO events(Figure 9) with a quasi-periodic time scale of 2–3 years.The Southern Oscillation refers to the associatedglobal-scale pressure fluctuations, while El Niñorefers to the rapid rise in sea surface temperatures in

the tropical eastern Pacific that occurs during thebreakdown of the normal Walker circulation. Thestrength of the Walker circulation is proportional tothe difference in pressure between Indonesia and theeastern region of the South Pacific. The SouthernOscillation Index (SOI), derived from the pressuredifference between Tahiti and Darwin, is a measure ofthe strength of the Walker circulation and its direction.ENSO has been shown to be the major control onrainfall variability across large parts of the globe(Nicholls and Wong 1990), with the strongestamplification of variability occurring at lowerlatitudes and in lower rainfall regions. This istherefore particularly relevant to the monsoonal areasof the EINA region. It has been well documented formany years that extremes in Indonesian andAustralian rainfall are strongly related to variations inthe Southern Oscillation (Braak 1919; Berlage 1927;Nicholls 1981, 1983; McBride and Nicholls 1983;Hastenrath 1987; Drosdowsky and Williams 1991;Yamanaka 1998). For Indonesia correlation betweenrainfall and SOI is strongest in the south and east, andduring the dry and monsoon transition seasons (Figure10). The major reason for rainfall suppression overeastern Indonesia during ENSO years is a delayedmonsoon arrival and early dry season onset (Kironoand Tapper 1999), allowing a much extended fireseason. However it should be remembered thatdrought may also impose limitations on fuelavailability. For northern Australia correlationsbetween SOI and rainfall are also strongest during thetransition (monsoon build-up) season (Figure 11).

There is strong documentary evidence linkingmajor fire episodes in Indonesia to ENSO. Five majorfire events have been recorded (Denis 1998); largefires occurred in Central Kalimantan in 1877, in Southand East Kalimantan in 1914, in East Kalimantan in1982–83, throughout Kalimantan in 1986–87, inSumatra, Java and Kalimantan in 1991–94, and inSumatra and Borneo generally in 1997–98. All of

EQ 30°S

(a) December-February

90°W 0 90°E 180 90°W

EQ 30°S

(b) December-February

90°W 0 90°E 180 90°W

Figure 8. Fire season cumulative FDI for 2 × CO2 (top) andpercentage changes in fire season cumulative FDIfrom 1 × CO2 to 2 × CO2 (bottom). The continenthas been divided into a northern and southern fireseason (from Williams et al. 1999).

Figure 9. (a) Normal Walker circulation and (b) itsbreakdown during ENSO events (from Sturmanand Tapper 1996).

Seasonal ΣFFDI for 2 × CO2 scenario

Seasonal ΣFFDI: % change from 1 × CO2 to 2 × CO2 scenario

26

these fire events were associated with strong ENSOevents (defined as having dry season SOI below –10).Interestingly, none of these large fire events, with thepossible exception of the 1997–98 event, saw muchlarge-scale burning in eastern Indonesia. This possiblyindicates that it is the drying out of the more heavilyforested areas further west, particularly in Kalimantanand Sumatra, that is most problematic. The linkbetween ENSO and fire activity in Indonesia is furtherdocumented in Figure 12. This figure indicates acurvilinear relationship between forest area burnt andSOI, with relatively little burning occurring in seasonswith SOI above –5.

Williams and Karoly (1999) document ENSOimpacts on FDI for a number of sites in Australia.Composites of cumulative fire season FDIs for the6 years with the most negative annual SOI and for the6 years with the most positive annual SOI between1960 and 1992 were compiled (Figure 13). AliceSprings, Roebourne, Normanton and Miles allrepresent northern Australian stations, and all exceptMiles show higher SFDI with more negative SOI.More detailed analysis indicated that the increasedFDIs are largely associated with reduced rainfall,lower humidities and higher temperatures. HoweverWilliams and Karoly warn that not all of the inter-annual variability of fire danger is associated withENSO and that actual fire occurrence depends onmany other factors including vegetation type and fuelload.

Apart from impacts on drought and fire occurrence,the two extreme modes of the Walker circulation alsohave implications for haze dispersion. During the‘normal’ phase enhanced vertical motion (convection)over Indonesia and stronger trade winds provide betterconditions for moving haze out of the region bothhorizontally and vertically, as well as providing rain-out of particulate matter. During ENSO years the tradewinds slacken and convection is reduced, particularlyover Indonesia, resulting in greater stagnation of haze.

A quasi-biennial oscillation (QBO) signal has alsobeen seen in EINA rainfall (Yasunari 1981), but inturn the QBO also appears to be at least partly relatedto fluctuations in the Walker circulation. HoweverENSO remains the greatest source of year-to-yearvariability in regional rainfall.

Seasonal influencesThe most important seasonal influence on EINAmeteorology and its fire regime is clearly theplanetary monsoon circulation. Broad characteristicsof the circulation relevant to the region were discussedabove. The monsoon is responsible for a largecomponent of annual rainfall over much of the region,with much of the rainfall falling over the land andislands which provide a number of the mechanisms

120°E10°S

20°S

30°S

40°S

10°S

20°S

30°S

40°S

10°S

20°S

30°S

40°S

10°S

20°S

30°S

40°S

130°E 140°E 150°E 120°E 130°E 140°E 150°E

120°E 130°E 140°E 150°E 120°E 130°E 140°E 150°E

Figure 11. Correlation of the SOI with precipitation overAustralia by season (from Sturman and Tapper1996, after Drosdowsky and Williams 1991).

Figure 10. Spatial distribution of stations in Indonesiaexhibiting a significant simultaneous correlationsbetween monthly rainfall and SOI. Data for1950–1998 (from Kirono and Tapper 1999).

27

required for air mass uplift. The retreat northwards ofthe ITCZ and associated monsoon flow in the australautumn initiates the drying necessary for annualburning to commence, while the southward movementlate in the year (and the thunderstorm activityassociated with the build-up) normally means the endof the burning season. These atmospheric influencesare generally well known to fire managers of northernAustralia and will not be discussed further here. Inter-seasonal monsoon variability associated with ENSOof significance to the particular seasonal fire regimeswas discussed previously.

One of the most significant seasonal influences isthe impact of characteristic wind flow regimes on thetransport of the products of biomass burning. The lastthree years have seen many countries in SoutheastAsia, Central and South America severely affected bytrans-boundary smoke haze originating fromuncontrolled forest fires burning mainly in tropicalforests severely affected by drought. Recentdevelopments in numerical weather prediction (NWP)and atmospheric transport modelling (ATM) provide

an ability to predict transport patterns, impact area andpollution concentrations downwind of fire sourceareas. Our research group is currently utilising theHYSPLIT_4 trajectory-dispersion model coupledwith Bureau of Meteorology NWP output to developan atmospheric transport climatology for the entireSouth East Asian–northern Australian region. Figure14 shows monthly trajectory composites (~30 dailyforward trajectories each extending out for 5 days) forJuly 1994 (EINA dry season) and February 1995(EINA wet season) for three locations in the EINAregion—eastern Indonesia on the equator (0.0°N,125.00°E), Jabiru in the Top End (12.86°S, 132.35°E)and east–central Northern Territory close to TennantCreek (20.00°S, 135.00°E). These give characteristicregional transport patterns for the respective seasons.During the wet season transport from the northernlocation is dominantly toward the southeast in thenorthwest monsoon flow. Many of the 5-daytrajectories impinge on mainland Australia. Transportfrom the mainland Australian locations is morevariable, possibly reflecting monsoon and monsoon-break conditions. However a large amount of re-circulation occurs, consistent with persistentcontinental low pressure at this time of the year(Sturman and Tapper 1996).

Dry season flow is less complex. The Australianlocations exhibit consistent transport to the northwestin the southeasterly trade wind flow, but some re-curvature toward the northeast is seen as trajectoriescross the equator. Clearly there is some potential fortransport of products of biomass burning into theIndonesian region. Interestingly a Dutch colonialmeteorologist raised the possibility of transport of dryseason haze from Australia into the Indonesian regionback in the 1920s (Braak 1929). Transport from thenorthern station is dominantly toward the northeast asthe trade winds re-curve over the equator to becomethe southwest monsoon. The position of the ITCZ thushas a major influence on the transport patterns, as wasseen during the major Southeast Asian haze episode ofSeptember–October 1997. Through much of thisperiod the ITCZ lay well north over Indo-China andthe Philippines, allowing smoke from fires inKalimantan and Sumatra to be transported northwardover Singapore and Peninsula Malaysia. Smoke hazecleared from these areas as the ITCZ movedsouthward.

One outcome of the recent global forest fire activityhas been a move by the World Health Organization toestablish guidelines relating to health effects ofexposure to emissions from biomass burning (WHO1999). Almost certainly implementation of theseguidelines will involve the use of atmospherictransport models similar to the one described above tocalculate ground level concentrations of pollution.Current atmospheric transport models (ATMs)

Figure 12. The relationship between SOI and area burnt inIndonesia (after Badawi et al. 1998).

FFD

I

-VE

SOI

+V

ESO

I

-VE

SOI

+V

ESO

I

-VE

SOI

+V

ESO

I

-VE

SOI

+V

ESO

I

-VE

SOI

+V

ESO

I

-VE

SOI

+V

ESO

I

-VE

SOI

+V

ESO

I

-VE

SOI

+V

ESO

I

0

500

Sale Hobart Mildura Katanning Alice Springs

Roebourne Normanton Miles

1000

1500

Mean+SD Mean-SD

Mean+SE Mean-SEMean

2000

2500

3000

3500

4000

4500

5000

5500

6000

* * *8 SITES

Figure 13. Differences in fire season cumulative FDI betweenextreme positive and extreme negative SOI years1960–1992 (from Williams and Karoly 1999).

El Niño Southern Oscilation Index

Off

icia

llyR

epor

ted

Bur

ntA

rea

(ha

x10

00)

28

utilising numerical weather prediction are welladvanced in terms of their meteorology andcomputation of dispersion characteristics, but therequired level of information on emission rates fromforest fires and smoke deposition rates for input to theATMs is lacking (Tapper and Hess 1999). This is anarea that must be quickly addressed if a capability toissue smoke predictions for the guidance of therelevant authorities is to be rapidly achieved.

Intra-seasonal influencesA range of intra-seasonal meteorological phenomenaplay an important role in rainfall processes in theregion, particularly as they influence monsoon onsetand the timing of major rainfall periods within themonsoon season. Northerly surges of low-level windfrom the South China Sea associated with fluctuationsin the intensity of the Siberian high pressure systemare important in regulating convective activity andrainfall over northern Indonesia. The surges areconsidered a mechanism for monsoon onset oversouthern Indonesia and northern Australia (Mantonand McBride 1992). Australian west coast surges aresimilarly thought to be important. Another influenceon monsoon onset over the region and on intra-seasonal rainfall is the 40–50 day oscillation inequatorial wind and rainfall first identified by Maddenand Julian (1971, 1972). This large-scale wave ofenhanced convection is first seen in the Indian Oceanand moves from west to east with a propagation speedof ~20 metres per second.

Synoptic influencesOver time scales measured in hours to days, synoptic-scale influences can have an important role indetermining levels of fire danger. Although the EINAregion is characterised by extended periods ofmoderate to high fire danger during the dry season,two distinct synoptic-scale weather patterns provideextreme fire weather situations, at least over theregion’s southern parts (Tapper et al. 1993).

The passage of mobile pre-frontal troughs throughsouthern and central parts of the Northern Territory isone such weather situation. These troughs occur inassociation with the eastward passage of higherlatitude cold fronts during the austral spring and earlysummer and are associated with hot windy conditionsand sometimes ‘dry’ thunderstorms with lightning.Tapper et al. (1993) provide a detailed discussion ofthe impacts of pre-frontal troughs on fire danger in theTerritory.

Of more significance in the northern tropics is theinfluence of intense subtropical ridging over the southof the continent during the austral winter and earlyspring. This strong ridging can be associated with asoutheasterly trade wind surge of sufficient strengthand dryness to generate extreme fire danger over thenorthern region. Curing is greatest in the north at thistime of year and fuel loadings can be high, particularlyif there has been heavy rainfall during the precedingwet season. The weather pattern associated with tradewind surges is typically quite slow-moving, soextreme fire danger may persist for several days.Figure 15 illustrates a typical synoptic pattern that inthis case was associated with the strong trade windsurge of 8–13 August 1987. Although temperatureswere generally below average, surface winds of

Figure 14. Daily 950 hPa forward trajectories (5-day runs) from three locations in the EINA region.Upper boxes are for February 1995, lower boxes are for July 1994 (from Wain 1999 unpubl.).

29

40–60 km/hr were widespread and relative humidityreadings were generally below 20%.

These meteorological factors, combined withseasonal drought, generated extreme FDI values in therange of 45–60 across the north. Trade wind surgesappear to be associated with degenerating cold frontspenetrating the tropics and it is the strong outbreak ofdry air of southern origin that is instrumental inelevating fire danger. The rapid drying of the air massfollowing the passage of a trough–front system is clearlyshown in Figure 16, which illustrates a mixing ratio timeseries obtained from Mt Isa during September–October1991. A deep moist air mass precedes the passage of alltrough–front systems, but the air mass dramaticallydries out behind the system and remains dry for severaldays. Although subtropical ridging and associated tradewind surging is an important mechanism in thegeneration of high fire danger over northern Australia, itis unlikely to be of significance in the eastern Indonesianregion. Synoptic situations associated with high fire

danger here require investigation.

ConclusionsIn providing an overview of atmospheric issuesrelating to fire management in the EINA region, thispaper has identified a number of areas worthy offurther research. Hopefully this forum will identifysome of these areas as important research priorities:1. Very little is known about the likely impacts on

regional fire regimes of predicted global warming.The small amount of available evidence suggests asignificant increase in fire danger in northernAustralia, but this would need to be confirmed andextended into the eastern Indonesian region.

2. Links between fire activity and ENSO climatevariability have also been tentatively established forthe region, but need to be more firmly established.In particular the relative impacts of increased(meteorological) fire danger and (probable) reducedfuel loadings during ENSO warm events must beexamined further.

3. Work continues on developing regional transportclimatologies to link possible source areas of smokefrom biomass burning to likely impact locations.However this work should be extended if possibleto examine variability in transport patterns andprocesses between warm and cold ENSO events,since warm events appear to be associated withmore burning, especially in more heavily forestedregions.

4. Little appears to be known about synopticinfluences (if any) on fire danger in easternIndonesia. This information plays an important rolein fire management practises in Australia and wouldbe most useful for the eastern Indonesian region.

5. More generally, ATMs have the capability toprovide useful information to fire managers andpublic health authorities on the transport and levelof pollution from fires. However the usefulness ofthese models is limited by current poor level ofknowledge about the relevant fire emission ratesand atmospheric deposition rates for input to themodels. This is an area that must be quicklyaddressed by research.

Figure 15. Daily sea level pressure analyses for the tradewind surge event of 8–13 August, 1987—units inhPa (from Tapper et al. 1993).

Hei

ght(

km)

0.01 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38

1.0

2.0

3.0

4.0

5.0

6.0

event 1 front

event 2 front

Days from September, 1991event 3

depressionevent 4 trough

Figure 16. Time series of mixing ratios obtained from radiosonde ascents at Mt Isa,September–October 1991—units in g/kg (from Tapper et al. 1993).

30

ReferencesBadawi, W., Walsh, T. and Jhamtani, H. 1998. Forest and

Land Fires in Indonesia, Impacts, Factors andEvaluation, Volume 1. State Ministry for Environment,Republic of Indonesia; United Nations DevelopmentProgramme, 185 p.

Berlage, H.P., Jr. 1927. East monsoon forecasting in Java.Verhandelingen No. 20, Koninklijk Magnetisch enMeteorologisch Observatorium te Batavia.

Braak, C. 1919. Atmospheric variations of short and longduration in the Malay Archipelago and neighbouringregions, and the possibility to forecast them.Verhandelingen No. 5, Koninklijk Magnetisch enMeteorologisch Observatorium te Batavia.

Braak, C. 1929. The Climate of the Netherland Indies;Volumes I and II. Verhandelingen No.8, KoninklijkMagnetisch en Meteorologisch Observatorium teBatavia.

Denis, R. 1998. A Review of Fire Projects in Indonesia,1982–1998. Centre for International Forestry Research,Bogor, 110 p.

Drosdowsky, W. and Williams, M. 1991. The SouthernOscillation in the Australian Region. Part I: Anomaliesat extremes of the oscillation. Journal of Climate, 4,619–638.

Gordon, H.B. and O’Farrell, S.B. 1996. Transient climatechange in the CSIRO Coupled Model with dynamic seaice. Monthly Weather Review, 125(5), 875–908.

Hastenrath, S. 1987. Predictability of Java monsoonanomalies: a case study. Journal of Climate andApplied Meteorology, 26, 133–141.

Kirono, D. and Tapper, N. 1999. Inter-annual rainfallvariability in Indonesia and its impacts on cropproduction. Physical Geography (submitted).

Knutson, T. and Manabe, S. 1994. Impact of increasing CO2on simulated ENSO-like phenomena. GeophysicalResearch Letters, 21, 2295–2298.

Knutson, T. and Manabe, S. 1995. Time–mean response overthe tropical Pacific to increased CO2 in a coupledocean–atmosphere model. Journal of Climate, 8,2181–2199.

Luke, R. and McArthur, A. 1978. Bushfires in Australia.Australian Government Printing Service, Canberra,359p.

McBride, J. 1992. The meteorology of Indonesia and theMaritime Continent. Extended Abstracts, FourthInternational Symposium on Equatorial ObservationsOver Indonesia (4th ICEAR Symposium), Jakarta,Indonesia, November 10–11.

McArthur, A. 1968. Fire behaviour in eucalypt forests.Leaflet 107, Commonwealth of Australia Forestry andTimber Bureau, Canberra, 36p.

McBride, J. and Nicholls, N. 1983. Seasonal relationshipsbetween Australian rainfall and the SouthernOscillation. Monthly Weather Review, 111,1998–2004.

Madden, R. and Julian, P. 1971. Detection of a 40–50 dayoscillation in the zonal wind in the tropical Pacific.Journal of the Atmospheric. Sciences, 28, 702–708.

31

Fire Management Issues in Eastern Indonesia

32

THE LONG-TERM EFFECTS of changes in land use andland cover on human welfare and the physicalenvironment may be as significant, if not more so,than those associated with climate change. Unlikeglobal warming, however, land use and cover changesare undisputed aspects of global environmentalchange. Notwithstanding the significance of thepotential effects of land-use and land-cover changes,our understanding of the scale and pace of thesechanges, their human and biophysical origins, andtheir linkages to other global change is inadequate.Nusa Tenggara Timur is a region of the world wherewe know little about land-cover change.

NTT is part of an arc of islands stretching from theeast of Bali, to the north of Australia’s NorthernTerritory. The main islands are the western half of theisland of Timor, and the Alor archipelago—Sumba,Roti, Sabu, Flores. The province has a population ofaround 3.75 million and a low population density ofaround 75 persons per km2. Higher populationdensities are found in a few places, namely in theSikka region in Flores, and around Kupang, theprovince’s capital in Timor. NTT is one of the poorestand least developed of Indonesia’s eastern provinces,with the majority of the population living a

subsistence agricultural lifestyle. Indeed, NTT had thelowest per capita gross domestic product of all ofIndonesia’s 27 provinces in 1997.

Most islands in the region have rugged topography,but while the islands of the northern chain are coveredby volcanic soils, uplifted marine deposits form thesouthern islands or the ‘outer arc’. In climate theislands of this arc are affected by their proximity toAustralia. Characteristic features are a long dryseason, low rainfall, and consequent semiaridconditions in comparison with the more tropical areasof Indonesia. The eastern and western ends of the arcswing upward to the north; thus west Sumba and eastTimor are less dry than areas in the southerncircumference of the arc.

Biro Pusat StatistikThe Biro Pusat Statistik (PBS) in Indonesia compilesand publishes statistics on a variety of subjects for allof Indonesia and for some subjects by province. Onehas to be careful in reviewing land cover data,however, since most fields and even many forests inNTT have not been mapped. In addition, because PBScollects data on agricultural and forestlands fromdifferent line agencies (i.e., agriculture, estates,forestry, etc.) these data do not always agree with each

Land-use and Land-cover Change in Nusa Tenggara Timur, Indonesia

Jeff Fox1

AbstractLand-use/land-cover change (LUCC) research is central to the science of global environmentalchange. LUCC influences, and is affected by, climate change, loss of biodiversity, and thesustainability of human–environment interactions, such as food and fibre production, fire regimes,and human health and quality of life. Little has been written about LUCC in Nusa Tenggara Timur(NTT), Indonesia. This paper reviews annual statistics published by the Biro Pusat Statistik on landutilisation in NTT, reviews the literature on land use in NTT, and discusses a project in NTT beingconducted by the East-West Center. Preliminary results suggest that traditional land uses in NTT(swidden cultivation and livestock raising) and land cover (secondary vegetation) have remainedfairly stable over the last 20 years. But land-use change is occurring at an increasingly rapid rate asfarmers switch from subsistence farming to commercial crops. Both the market potential of cashcrops and the deterioration of environmental conditions that constrain the cultivation of traditionalcrops are driving these changes. The large amount of land zoned ‘state forest land’ is also affectingland cover and land use. These changes have long-term implications for biodiversity, environmentalstability, and fire regimes in the region.

1Environmental Studies, East-West Center, 1601 East-WestRoad, Honolulu, Hawaii 96848-1601

other. Yet these data provide perhaps the best regionalpicture we have of what is happening to land use andland cover in NTT and may be useful if we examinethem in terms of the magnitude and direction ofchange and not in terms of absolute numbers.

Table 1 shows data on land utilisation for NTT in1979 and 1995. Note that in 1995 the figures ondifferent land covers add up to 118% of the total landarea of the province! The most likely explanation forthis error is that the agriculture and forestrydepartments double counted some land. This error issignificant but a second figure is perhaps moreimportant. Between 1979 and 1995 ‘state’ forest(hutan negara) grew from 22 to 66% of the landscape.While this figure undoubtedly contains error, therecan be no doubt that during this 16-year period, thenational and provincial forestry departments laidclaim to a much larger portion of the province.

Among the remaining land categories it is importantto note the following changes. Fallow and grass landsdecreased by approximately 7% of the land area whilewoodlands (non ‘state’ forest areas covered by trees)increased by almost 3%, agricultural lands (drylandand rice) increased by over 3%, and estate crops alsoincreased by approximately 2%. While the totalamount of land devoted to growing vegetables is still

less than 1%, this category has grown by more than10% every year since 1979.

Table 2 summarises data on dryland crops. In 1995,maize accounted for 70% of the land devoted to thesecrops and the amount of land devoted to maize hasincreased by 2% per year since 1979. The amount ofland devoted to cassava has decreased at a rate of 1%per year. Perhaps more importantly, the amount ofland devoted to soybeans, while still insignificant, hasbeen growing at a rate of 18% per year.

Table 3 summarises data on vegetable crops. The totalamount devoted to growing vegetables is still small buthas been growing at a rate of 10% per year since 1979.The amount of land devoted to potatoes, cabbage, andmustard greens has been growing even faster, at rates of12–14% per year, throughout this period.

Table 4 summarises data on livestock. Cattleaccount for the largest number of big livestock andhave increased at a rate of 4% per year since 1979.The number of horses has decreased at a rate of almost1% per year. Goats, sheep and pigs have all increasedat rates of 5–6% per year.

These data suggest a trend away from less intensiveland uses (fallow and grass) towards more intensiveand commercial land uses (agriculture, estates, andwoodlots). In addition, a considerably larger

33

Table 1. Land use in NTT in 1979 and 1995.

1995– Annual rate of Land use 1979 (ha) 1995 (ha) 1979 (%) 1995 (%) 1979 (%) change (%)

House compoundPekarangan/lahan untuk bangunan dan halaman sekitarnya 72414 119220 2 3 1 3

Dryland agriculture Tegal/kebun/ladang/huma 387708 464555 8 10 2 1

Rice—irrigated and uplandLuas panen padi (padi sawah + padi ladang) 117643 159823 2 3 1 2

VegetablesLuas panen tanaman sayur-sayuran 516 2490 <1 <1 <1 10

Estate crops (coffee etc.)Perkubenunan negara/swasta 101934 196131 2 4 2 4

GrasslandsPadang rumput 551934 498663 12 11 -1 -1

Fallow landLahan yang sementara tidak diusanhankan 869684 589060 18 12 -6 -2

WoodlandLahan untuk tanaman kayu-kayuan 298438 434726 6 9 3 2

ForestsLuas hutan berdasarkan tata guna hutan kesepakatan 1063000 3135000 22 66 44 7

Water ponds/dykesKolam/tebat empang/tambak 15260 5841 <1 <1 <1 -6

Total 3478531 5605509 73 118Total land area in province 4735000 4735000

proportion of the landscape was classified as stateforests in 1995 than in 1979, suggesting that thepotential for conflict between local land users andprovincial and national forest departments may havealso grown proportionally.

Literature ReviewThere have been three major works on land-usepractices in Nusa Tenggara since the late 1970s. Theseinclude Jim Fox’s work on Roti and Savu (Fox 1977),and Joachim Metzner’s books on Flores (1982) andEast Timor (1977) (East Timor is not part of theprovince of NTT). In addition, there have been severalsmaller works by both Indonesian and western authors(e.g., Dewa 1995; Molnar 1994; KEPAS 1986; Siskel1986).

Fox (1977) argued that the ‘traditional’ agriculture ofthe present day — especially the swidden areas ofSumba and Timor — is in large measure the creation ofthe period following the arrival of the Dutch and the

introduction of new food sources. The most prominentof these was maize; of lesser importance were sweetpotato and cassava, and along with these, squash,onions, garlic, eggplant, and in the 1970s tomato. Priorto European arrival, the seven most important cropswere rice, millet, sorghum, Job’s tears, green gram,pigeon pea, and sesame. Fox maintains that it is notswidden agriculture per se that accounts for theprecarious subsistence base of the peoples of Timor andSumba, but the historical creation of a monolithic formof swidden overly dependent on maize. The KEPAS(1986) study also points out that in most drylandswidden systems agrochemical inputs are rarely utilised,and the use of organic fertilisers is also minimal.

About 80% of the farm households in NTTmaintain trees around their houselots or in abandonedswidden fields and in recent years commercialcultivation of a variety of tree crops has grown inimportance. This is particularly true on the island ofFlores (Metzner 1982). These systems have developedas marketing channels have improved and thetechnologies and capital essential for theirdevelopment have become more readily available.Tree crops given particular attention by the region’sfarmers include coffee, coconuts, cocoa, cloves, andcandlenut. KEPAS (1986) presents case studies fromaround the province that show increasing marketorientation. Better communication systems andimproved access to consumer goods and services haveincreased the demand for cash. However, the intensivemanagement of major cash crops is a newphenomenon in much of this region and problemsboth in their cultivation and marketing are common.

In Flores high population density puts a limitationon the available land per family and the use of land forshifting agriculture (Metzner 1982). Dewa (1995)argues that as population density has increased andland has become scarcer, the fallow period has oftenbeen shortened and many fields have come undersedentary cultivation.

Metzner (1982) estimated that grassland savannascovered 50% of the islands of Sumba, Alor, Floresand Timor in the late 1970s and early 80s. Grasslandsare often owned communally and used as fallowedswidden fields or grazing lands. Cattle are the most

34

Table 2. Dryland crops in NTT in 1979 and 1995.

1995− Annual rate of Land use 1979 (ha) 1995 (ha) 1979 (%) 1995 (%) 1979 (%) change (%)

Maize 179108 254189 60 70 10 2Cassava 96472 81396 32 22 -10 -1Sweet potatoes 13988 11963 5 3 -2 -1Peanuts 8485 10540 3 3 0 1Soybeans 410 5627 0 2 2 18

Total 298463 363715 100 100 – 1

Table 3. Vegetable crops in NTT in 1979 and 1995.

Annual rate of Vegetables 1979 (ha) 1995 (ha) change (%)

Onions 24 67 7Shallots 293 978 8Potatoes 56 331 12Cabbage 33 271 14Mustard greens 110 767 13Carrots 0 76 51

Total 516 2490 10

Table 4. Livestock in NTT in 1979 and 1995.

Livestock 1979 (no.) 1995 (no.) Annual rateof change (%)

Cattle 414203 785100 4Buffalo 147373 191100 2Horse 192831 170600 -0.7Goat 225181 612200 6Sheep 47419 111500 5Pig 696738 1538000 5

important livestock in this system. In the past, elitegroups representing only 2–10% of the populationowned large cattle herds and monopolised commonproperty pasturage. In recent decades a semi-intensivefattening system has developed in some areas inTimor. Since a family can rarely tend more than a fewcattle, the system has improved the equity and freedpasturelands for small farmers to cultivate (KEPAS1986). Other important animals are horses in Sumba,water buffaloes, goats and sheep. Pig raising isubiquitous and practiced by almost every household.

Increases in animal population particularly inTimor and Sabu have put more pressure on landavailable for agriculture. For swidden farmers fencingis essential to protect their crops from strayinganimals, requiring a larger investment in labor forfence construction and maintenance. In some areas upto 30% of farm labor is allocated to the building offences around individual fields. In other areascommunal fences are built to protect the agriculturalarea with maintenance responsibility divided amongthe members. In Sabu, the use of ‘living fence’substantially reduces labor demand for maintenance(KEPAS 1986; Fox 1977).

As land is increasingly fenced and thereby zonedfor grazing and agriculture, it becomes more difficultto practice extensive shifting cultivation. In Sumbazoning is now enforced through forestry regulationsthat ban the use of forestlands for agriculture,including swiddening (KEPAS 1986).

This literature review suggests that in most of NTTtwo forces are increasingly determining land-usesystems. First, market pressures — the commercial-isation of subsistence resources and the substitution ofcommercial crops for subsistence crops — are providinga pull factor encouraging farmers to engage in new anddifferent forms of commercial agriculture. And second,the growth of land extensive livestock managementsystems, as well as national forest and land tenurepolicies — the nationalisation of forest lands and effortsto increase control over upland resources by the centraland provincial governments — are providing a pushfactor that makes it increasingly difficult for farmers tomaintain their traditional swidden land-use practices.Combined these forces will eventually cause a majorchange in land-use practices from swidden agriculture tocommercial crops, and a change in land cover fromsecondary vegetation to monoculture commercialagriculture.

Work at the East-West CenterAt the East-West Center in Hawaii we have beencollaborating with the Nusa Tenggara UplandsDevelopment Consortium to improve local capacity toincorporate spatial information into community-basedresource management projects. Specific objectives of

this project include providing training in acquiringand analysing spatial information and developingcapacity to use spatial information to producepractical and policy-relevant insights into resourcemanagement issues.

Working in collaboration with KOPPESDA(Coordination Team for Natural Resources Researchand Analysis), one set of activities has focused onunderstanding land use and land cover change for anarea in and around Wanggameti ProtectedArea/National Park in eastern Sumba. Projectmethodology has included incorporating informationcollected from interviews with key informants andfarmers with the development of a spatial databasebased on satellite images.

There are 14 villages around the conservation areaand two ‘enclave’ villages located within theconservation area. EWC and KOPPESDA researchershave mapped villagers’ views of their customaryboundaries in order to further negotiations with thepark/forest officials on boundaries for the protectedarea/national park.

Maps showing changes in land cover forWanggameti were developed from two satelliteimages — October 4 1973 and October 17 1997. Theamount of vegetation in the images was calculatedusing a normalised difference vegetation index(NDVI). The NDVI compares the amount ofreflectance in the near infrared (NIR) and red (R)spectral wavelengths. An area with a lot of vegetationhas a high reflectance value in NIR and a lowreflectance value in R and therefore a high NDVI. Anarea with little vegetation, or unhealthy vegetation,has a lower reflectance in NIR than in R, and hence alow NDVI. Studies of NDVI ratio results have shownthat forests, and tree-covered areas in general, havehigher NDVI values than grass and shrub coveredareas. Also burned areas have very low NDVI values,but soon recover as grasses and other vegetation growback. Because the same wavelengths are used tocalculate the NDVI from images obtained fromdifferent satellites, the calculations are fairlycomparable between satellites and over time.

Preliminary analysis of the NDVI data is shown inFigure 1 and Table 5. Figure 1 shows vegetation coveraround Wanggameti in 1973 and 1997. These images(as summarised in Table 5) suggest a shift from moredensely vegetated categories towards less densevegetation. This analysis supports the hypothesis thatpopulation growth and shortened fallow periods areleading to less biomass.

Figure 2 shows a supervised classification of the1997 image. This classification is based on relatingground cover located with a GPS in the field withspectral reflectance values from the image. We did nothave sufficient ground truth to conduct an accuracyassessment of this classification. This image provides

35

support for the amount of area mapped as low orlowest vegetation cover in the NDVI analysis. Thisimage, however, also suggests that a large part of themost vegetated class is a mix of species similar tothose found in village tree gardens. In other words, theimage suggests that land cover the Department ofForestry classifies as ‘forest’ is, in reality, vegetationmanaged for subsistence purposes by the people wholive in and around the park.

Our preliminary conclusions hence support theview of east Sumba as a dry environment whereswidden is the main land-use practice. Land cover in

this environment may be slowly degrading fromgreater to lesser biomass. On the other hand, however,a large part of the new conservation area/national parkmay actually be home gardens planted and maintainedby generations of villagers from surrounding villages.

Discussion and ConclusionsBoth national statistics and the case studies reviewedsuggest that land-use in NTT is slowly shifting awayfrom less intensive uses (fallow and grass) towardsmore intensive and commercial uses (agriculture,estates, and woodlots). These trends are stronger inthe wetter portions of the province (western Sumba,Flores, and eastern Timor) and less strong in the dryerportions (eastern Sumba) (note we have not reviewedthe lontar economies of Roti and Sabu). In terms ofthis conference it is important to note that the lessintensive land uses (fallow and grass) weretraditionally managed by fire, whereas fire plays asmaller role in the management of the more intensiveand commercial land uses. In addition, the fact that aconsiderably larger proportion of the landscape wasclassified as ‘state forest’ in 1995 than in 1979

36

Table 5. Percentage changes in land cover in Wanggameti,Sumba as observed from Landsat images.

Wanggameti, Sumba (124 000 ha)

Annual rate NDVI NDVI of change

Land cover 1973 1997 1973–1997

Most vegetation(tall) 48 30 -2.0Some vegetation (medium) 24 13 -2.5Less vegetation (short) 12 33 4.3Little or no vegetation 1 9 7.8

Figure 1. Comparison of vegetation cover around Wanggameti protected area in 1973 and 1997.

37

Figu

re2.

Prel

imin

ary

land

cove

rmap

ofW

angg

amet

ipro

tect

edar

ea,O

ctob

er19

97.

ReferencesChazdon, R. 1998. Tropical forests — log’em or leave’em.

Science, 285, 1295–1296.Dewa, R. 1995. Change and Continuity in Customary Land

Use in Ndu’aria, Flores, Indonesia. Honolulu, East–WestCenter, Program on Environment Working Paper No. 44,19 p.

Fox, J. 1977. Harvest of the Palm: Ecological Change inEastern Indonesia. Cambridge, Harvard UniversityPress, 290 p.

Janzen, D. 1998. Gardenification of wildland nature and thehuman footprint. Science, 279, 1312.

KEPAS 1986. Agro-ekosistem Daerah Kering di NusaTenggara Timur. Jakarta, Departemen Pertanian, 119 p.

Metzner, J. K. 1977. Man and Environment in Eastern Timor.Canberra, Development Studies Centre, AustralianNational University, Monograph No. 8, 380 p.

Metzner, J.K. 1982. Agriculture and population pressure inSikka, Isle of Flores: A contribution to the study of thestability of agricultural systems in the wet and drytropics. Canberra, Development Studies Center,Australian National University, Monograph No. 28,355 p.

Molnar, A. K., 1994. Biodiversity Conservation Project inFlores and Sieberut. Manila, Asian Development Bank,Loan No. 1187-INO, 104 p.

Pimentel, D., Stachow, U., Takacs, D., Brubaker, H., Dumas,A., Meaney, J., O’Neil, J., Onsi, D. and Corzilius D.1992. Conserving biological diversity inagricultural/forestry systems. BioScience, 42, 354–362.

Siskel, S. E., 1986. Lembor Baseline Socio-Economic Survey.Jakarta, GOI/UNDP/IBRD Project INS/83/013, 61 p.

38

suggests that the potential for conflict between localland users and provincial and national forestdepartments may have grown proportionally. These‘potential’ conflicts may be reflected in an increase inthe ‘accidental’ burning of disputed lands.

The slow switch from subsistence swidden tocommercial agriculture will provide an incentive tolocal people to prevent uncontrolled fires. On theother hand, tropical biodiversity conservation iscurrently undergoing a conceptual transition whereisolated forest fragments, logged forests, andsecondary growth forests (swidden fallows) are nowbeing recognised for their value in the conservation ofbiological diversity (Chazdon 1998). These studiesreinforce a new paradigm in the management oftropical biodiversity that extends conservation tohuman-impacted lands (Pimentel et. al. 1992; Janzen1998). The commercialisation of agriculture andconcomitant control of fires may hence lead to lessbiodiversity. Planners in NTT should consider thatswiddening may be the most ecologically appropriateand culturally suitable means available for preservingbiodiversity in many upland areas. Rather thenencouraging the uniform spread of commercialagriculture, it may be useful to begin investing inmethods of maintaining the biodiversity associatedwith swidden fallows while increasing theproductivity and soil-sustaining properties of theselands.

39

IN 1997 AND 1998, when fires destroyed forests andland not only in East Kalimantan but also in otherprovinces in Indonesia, shifting cultivation wasblamed as the source of smoke and associatedenvironmental problems. Subsequently it was provedthat the sources of the fires there were mostly fromland preparation areas for forest plantations controlledby forest concession holders and farm estatecompanies. It was shown that between 65% (Anon.1998) and 80% (WWF 1998) of the forest area burnedin East Kalimantan was within the forest concession.From January to May 1998, fires in East Kalimantandestroyed 507 000 ha of forest and 11 000 ha of landbelonging to the people. The total area burnedcomprised 315 000 ha in the forest concession areaand 84 400 ha in the forest plantation; 71 000 ha inKutai national park; 10 600 ha in an ex-forestconcession area; 8800 ha of protected forest; 4200 hain Taman Hutan Raya Bukit Soeharto; 2200 ha in theWanariset Samboja and conservation area, and10 800 ha on private land (Anon. 1998).

Fire is used because it is easy, cheap and simple.Much of the haze came from huge conversion burnsused to prepare land for pulpwood and oil palmplantations (Schindler 1998). Use of fire is officiallyforbidden although every company uses it, becausethis is the only viable and economic method ofreducing the huge biomass. The underlying cause is,therefore, the policy that plans to convert 500 000 haof forest into plantations every year (Schindler 1998).The government has licensed and encouraged manycompanies to develop new industrial plantations ofrubber, oil palm and pulpwood, as well astransmigration sites (CIFOR 1998). These activitiesrequire the clearing of hundreds of thousands ofhectares of land, and fire is their cheapest option. Thetraditional method of claiming forested land, as inmany parts of the world, has been to burn and thenplant.

Shifting cultivators burned the land in order toprepare it mainly for agricultural purposes. Suchactivities have been done for thousands of years. Freeminerals are released from the ash after burning—phosphorus, magnesium, potassium, sodium, andorganic carbon (Daubenmire 1968; Viro 1974; Laland Cummings 1979; Pritchett and Fisher 1987;Saharjo 1995). The increase of these minerals afterburning is temporary, with decreases recorded in the

Management of Fuel and Fire in Preparing Land for Forest Plantations and

Shifting Cultivation Bambang Hero Saharjo1, Endang A. Husaeni1,

Kasno1 and Hiroyuki Watanabe2

AbstractSources of Indonesian fires in 1997 and 1998 are believed to be in land preparation areas where firewas used for forest plantations or agricultural purposes (oil palm and rubber plantations), logged-overforest areas and shifting cultivation. In fact, it has been recognised for several years that logs in landpreparation areas and shifting cultivation have caused the conditions to deteriorate, although thereasons for this in logged-over forests and planted forests have not emerged.

Fuel and fire have important roles in land preparation areas for forest plantation and shiftingcultivation. Burning is easy, cheap, and very simple and free minerals result from the ashes.Unfortunately, high quantities of fuel in overly expansive land preparation areas will burn, causingheavy smoke and widespread damage, sometimes crossing into neighboring countries as well. Forthese reasons, the use of fire and fuel in land preparation areas for forest plantation and shiftingcultivation should be carefully managed, reducing the negative impact on the environment. Thepotential and methods for such management are examined here.

1Faculty of Forestry, Bogor Agricultural University, P.O. Box168, Bogor 16001, West Java, Indonesia

2Laboratory of Tropical Forest Resources and Environments,Division of Forest and Biomaterials Science, Graduate Schoolof Agriculture, Kyoto University, Kyoto 606-8502, Japan

40

second year and thereafter (Garren 1943; Jordan 1985;Whitmore 1990).

Although it should be remembered that shiftingcultivation activities have occurred for thousands ofyears, unfortunately it has recently become a seriousenvironmental problem. Observations reveal thatillegal shifting cultivation was behind most of thesmoke damage and these people are not true shiftingcultivators (Saharjo and Husaeni 1998). They arenewcomers from other regions that usually work inthe paddy fields. True shifting cultivators have theknowledge of how to burn and prevent the fires fromjumping to other regions. This experience andknowledge cannot be learned from only one or twoburnings without prior practice.

Observers in the field believe that shiftingcultivators will never stop using fire in landpreparation, and this situation will remain in theagricultural and forest plantation sectors (as fire is thecheapest method) while government laws are notenforced. One possible solution then should be toallow the use of fire but to eliminate problems ofsmoke and environmental damage.

For that reason, the objective of the researchdescribed below is to find patterns and methods forforest plantation and shifting cultivation that enablesafe burning on land preparation areas withoutaffecting the environment.

Research MethodsStudies were carried out in three time periods anddifferent locations. The first burn took place atSubanjeriji, South Sumatra, from August toSeptember 1994, in a shifting cultivation areabelonging to 13 families with a total area of 20 ha. Thesecond burning was done from August to September1995 in an area belonging to one forest concessionarea located in Subanjeriji with a total area of 4 ha.The third (last burning) involved a test on a 1 ha areabelonging to one person in Jasinga, West Java.Owner/renters of the land and their families undertookthe first and second burnings, while the authorscarried out the third test. Parameters monitored werefuel load, fuel bed depth, rate of the spread of fire,flame temperature and fire intensity.

Parameter measurements and calculations

Estimation of fuel loadIn each quadrat of 1 ha in the South Sumatra and WestJava plots, 10 subplots of 1 × 1 m were chosen for theestimation of fuel load and moisture content. Fuelload was estimated by collecting all the materials,both dead and alive, in the subplot. These materials

were then brought to the laboratory to measure weightand moisture content. Fuel moisture content wasbased on weights calculated before and after placingin an oven for 24 hours at 105°C.

Estimation of fuel bed depth Fuel bed depth was the average height of theassociation of living and dead plant materials ofvarious sizes and shapes that extended from mineralsoil to the top of the vegetation layer.

Experimental burning procedure Burning was conducted in the quadrat in the afternoonbetween 13.00 and 16.00. A study in Manitoba andSaskatchewan showed that the burn must be made ona day when the humus layer has a moisture content ±20% (Williams 1960). Torches were used as thesource of fire. Fires were set on the sides of thequadrate at the same time and allowed to propagatenaturally. Burning was allowed until most of the fuelwas spent.

Estimation of flame temperature Flame temperatures on the soil surface and 1 cmbeneath were estimated using ‘Tempilaq’(temperature indicating liquid), which melts at aspecified temperature and provides estimates ofmaximum temperatures. Liquid was contained in30 cm lengths of 2 cm aluminium pipe (Saharjo 1995).The temperature sensors were placed in the vegetationat two locations.

Estimation of rate of the spread of fire Rate of the spread of fire was estimated by measuringthe average distance perpendicular of the movingflame front per minute. The monitors used astopwatch and measurement tape.

Estimation of flame height Flame length was estimated by measuring the averagedistance between the tip of the flame and the surfacefuel. Tree height equipment and a counter were used.

Fire intensity Fire intensity is the product of the available heat ofcombustion per unit area of ground and the rate of thespread of fire. It was calculated by using Byram’sequations (Chandler et al. 1983), FI: 273 (L) 2.17 , whereFI is fire intensity (kW/m) and L is flame height (m).

Results

A case study in Subanjeriji burningSouth Sumatra is famous for the many shiftingcultivation activities there. It is sometimes difficult torecognise which lands are private and whichgovernment, because there are no clearly markedborders. One thing certain is that the people seek toclaim the land and cultivate it as soon as possible.

41

Some illegally cut the reforestation trees belonging tothe government and convert the areas to rubberplantations for maximum profit (Saharjo 1997a).

Fuel preparationAll trees with both large and small diameters found inthe burning area were cut, and shrubs and grass wereincluded. A chain saw was used for large trees and amachete for small branches, shrubs and grass. Treeswith a diameter of more than 50 cm in the burningarea become sawn timber, while others with diametersof less than 40 cm, together with shrubs and grass,were left in the area as fuel for burning. These fuelswere separated in the burning area.

This activity is usually done in the dry season fromAugust to September, when rainfall is low. At thistime the air temperature is quite high, around35–36°C, and relative humidity is 70–80%. Thesehigh temperatures will hopefully reduce fuel moisturecontent, and for this the fuels are left to dry for at least2–3 weeks before burning. It is hoped that themoisture content level is low and relatively uniform,then the fire can spread easily and quickly.

Different fuel loads were found in the plot inSubanjeriji, resulting in differing smoke andenvironmental effects. Plot A has a total area of 20 hawith fuel load 15 t/ha. This fuel load was dominatedby trees with diameters of 10–20 cm, averaging10 cm. Fuel depth was 0.5 m and fuel moisture contentwas 5–25% (Table 1). In plot B, the total area forburning was 4 ha with a fuel potency of 40.3 t/ha. Thefuel load was dominated by tree diameters rangingfrom 20 to 40 cm, averaging 20–30 cm. Fuel beddepth was 0.6 m and moisture content was 7–35%. Inthe Jasinga plot, the total area burned was 1 ha with a

fuel load potency of 29.5 t/ha, mostly consisting oftrees with a diameter of less than 10 cm. Fuel beddepth was 0.3 m and fuel moisture content was 3–14%(Table 1).

BurningBefore burning, the areas selected were usuallyprotected by a firebreak 2–3 m wide. Clearingeverything within that range made this firebreak. Thisis to cut the continuing fuel load in order to preventthe spread of wildfires.

Burning was done in the afternoon, between 15.00and 17.00. The principal used in this burning was foreach head fire to meet in the middle of the burningarea, hopefully spreading directly into the centre ofthe area. With this method, the possibility of the firejumping to other regions can be reduced. It isimportant to remember that burning depends greatlyon the wind direction and slope. For this reason, theparameters should be known before burning.

Burning commenced from sharp slopes andprogressed to flat areas. For this several persons hadto be involved as burners at different burninglocations. For instance, for a 1-ha burning area atleast three burners should stand by in three differentpositions equally distanced from each other. Burningwas started under one command. Running slowly, aburner moves to an exact point clockwise or anti-clockwise, following the wind direction and makingfire points at 0.8–1 m intervals. Ideally, no gapbetween one fire point and another will allow the fireto jump (Figure 1).

With a fuel potency of 15 t/ha and dominated bytrees with small diameters, plot A with a total area of20 ha needed only 56 minutes to burn. Maximum

Table 1. Weather conditions and fires behavior when burning was conducted in South Sumatra and West Java.

Parameter South Sumatra West JavaA. B.

Weather conditionsAir temperature (°C) 33 35 36Relative humidity (%) 65 70 80Wind speed (m/sec.) 1.7 1.6 1.8Fire behaviourFuel load (t/ha) 15.0 ± 0.2 40.3 ± 6.3 29.5 ± 8.2Fuel bed depth (m) 0.5 ± 0.1 0.6 ± 0.2 0.3 ± 0.1Fuel moisture content (%) 5.2–25.2 7.1–35.3 3.1–13.9Flame height (m) 2.75 ± 0.19 3.13 ± 1.35 2.5 ± 0.3Rate of the spread of fire (m/minute) 1.30 ± 0.21 1.20 ± 0.11 1.5 ± 0.3

Flame temperature above and below soil (°C)0 m 760 1010 7201 m 302 316 932 m 159 170 76-0.01 m < 38 38–42 < 38Fire intensity (kW/m) 2468.03 ± 376.55 3962.42 ± 1753.25 2026.23 ± 648.47Slope (%) < 8 7– 30 5– 25Burning area(ha) 20 4 1

42

flame temperature reached was 760°C with a meanflame height of 2.75 m. Maximum flame height was8 m with a fire intensity of 2568 kW/m. Plot B had afuel potency of 40.3 t/ha and was dominated by treeswith larger diameter than plot A over a total area of4 ha. This needed only 25 minutes to burn. Maximumflame temperature reached was 1010°C with anaverage flame height of 3.13 m, reaching a maximumof 12 m and fire intensity of 3962 kW/m. Thisdemonstrated that high availability of fuel dominatedby large diameter trees will cause higher flametemperatures and more damaging effects.

Burning test: a case study in Jasinga, West JavaBased on the experience in Subanjeriji (South Sumatra),a burning test was applied in Jasinga (West Java). Totalarea burned was 1 ha with a slope range of 5–25%, a fuelpotency of 29.5 t/ha dominated by trees with diametersof 5–10 cm and shrubs with a fuel moisture content of3–14% and fuel bed depth of 0.35 m.

As with both plots in Subanjeriji, all the materialsfound in the plot, whether dead or alive—includingtrees, shrubs and grass—were cut down and separatedin a circular plot. This activity enabled the fuel load todry for at least 2 weeks and thus decreased the fuelmoisture content.

Before burning, the surrounding area was protectedby a 2–3 m wide firebreak by cleaning away all of thematerials within that range. Wind speed and direction,important factors in considering the starting point forthe fire, were determined with an anemometer.Measuring was done several times in order to decidethe best time for burning, ideally with a wind speed of100 m/minute and in the afternoon. Air temperatureand relative humidity was measured using dry and wetbulb thermometers.

Three burners stood by at three different pointswaiting for the command to start the fire. At13.05 p.m., burning commenced by making one firepoint, followed by two others (Figure 1). Between onefire point and another was 0.8–1 m. The head fire wasclearly observed coming from the different points andmoving toward the centre of the burning area. In only15 minutes the 1-ha plot was burned. Maximum flametemperature was 720°C with a maximum flame heightof 8–10 m and a fire intensity of 2026 kW/m (Table 1).

DiscussionThe two burning cases in land preparation inSubanjeriji and one test in Jasinga proved that highflame temperatures were caused by the high fuel

Figure 1. (I) Burning method type 1(II) Burning method type 2

Legends ➪ wind directionA, B, C burner

burning directionfirebreaks

★ burning pointfuels

(I) (II)

43

content dominated by logs with large diameters.Variations of this can be seen from the different flameheights at three different plots.

Saharjo (1996a) reported that a fuel loadcomprising shrubs 1–2 cm in diameter with a potencyof 20 t/ha can reach flame temperatures of 649°C withan average flame height of 1.5 m, while at a potencyof 10 t/ha, flame temperature will reach 550°C.Temperatures 1 cm below the soil surface can reach< 38°C, but do not disturb soil organisms (Table.1).

These facts reveal the steps that can be taken toprevent large fires during burning of land preparationareas, preventing high temperatures and the resultingpollution and smog. Essentially, trees of more than5 cm diameter should be removed, but smaller treescan be kept in the burning area together with smallbranches, shrubs and grass. In addition, observationshows that logs with diameter of more than 30 cm arequite difficult to burn, and continue burning when allother fuels have been spent. These conditions invite anincreased fire danger, especially in logged over forests.

Burning with fuel loads consisting of small diametermaterials has also been done by the people ofSampoku, Niigata, Japan (Watanabe 1995) since the17th century, preparing the land for Sugi (Cryptomeriajaponica) plantations. In this area large diameter logswere removed from the area, while small branches andleaves were left. The burning area was not so large,usually about 0.05–1.2 ha divided into blocks based onownership or rental agreement. Burning was donefrom the top to the foot of the slope. The fire was keptin horizontal lines moving slowly down the slope. Theburning was mainly done in the dry season, July orAugust, in early mornings until afternoon, as themorning wind speed is not high and can be controlled.

For burning with less smoke, the burning patternthat was done in Subanjeriji was revised, based on theburning test in Jasinga. It is recommended as onealternative for burning in land preparation, but it mustbe followed by a revision of fuel load weight andform. Only logs less than 5 cm in diameter should beused, while the larger ones should be removed. Themaximum area to be burned should be no more than1 ha. For areas of more than 1 ha, the procedure mustbe done several times in order to reduce fire jumps andsmog. Fuel drying should be conducted at least 2 or 3 weeks before burning, with shrubs and grass cutdown and separated in the plots. Relative humiditymust be around 70–80%, with a wind speed of1.6–1.8 m/second and fuel moisture content between 5and 20% in an area dominated by dry fuels. Firebreaksshould be 3 m wide and surround the fire zone. It mustbe clean in order to prevent wildfires.

For a 1-ha burning, at least three burners should beinvolved and should stand at equal distances fromeach other. They should move with the wind direction

and clockwise or anti-clockwise, but to a certain point(Figure 1). Burning should be conducted in theafternoon between 14.00 and 16.00, as hopefully atthis time the fuel load drying is relatively uniform. Ifthe burning is done in the morning, it may not burnwell because of low temperatures that affect fuelmoisture content and a high relative humidity that willblock the process and rate of the spread of fire. Thiswill invite an unpredictable fire invasion as happenedin Indonesia in 1994 (Saharjo 1996b, 1997b).Afternoon burning is relatively safe and preferredbecause when wildfires occur, fire brigades can workmore easily in locating the fires, as compared to thedifficulties and dangers of night fires.

ConclusionsFuel and fire management in land preparation forforest plantation and shifting cultivation is veryimportant in order to reduce air pollution impacts andother dangers.

For this, the burning patterns observed inSubanjeriji, South Sumatra and applied in a burningtest in Jasinga, Bogor, West Java, can be used for landpreparation. Their application is simple andinexpensive, and does not cause any environmentaldisturbances. The three cases described provide thebackground for further expanded study.

ReferencesAnonymous 1995. Sekitar 60 000 ha hutan dibakar percuma

setiap tahun. Kompas, 6 Februari 1995.Anonymous 1998. Kerugian kebakaran hutan Kal-Tim Rp 10

Trilyun. Kompas, 18 Mei 1998.Chandler, C., Cheney, P., Thomas, P., Trabaud, L. and

Williams, D. 1983. Fire in Forestry, Vol.1. Forest firebehavior and effects. New York, John Wiley, 450 p.

CIFOR. 1998. Fires in Indonesia in 1997. Center forInternational Forestry Research News, No.16:1.

Daubenmire, R. 1968. Ecology of fire in grasslands. In:Cragg, J.B., ed., Advances in Ecological Research.London, Academic Press, 5, 23–26.

Garren, K.H. 1943. Effects of fire on vegetation on theSouthern United Sates. Botanical Review 9(9), 617–1654.

Goldammer, J.G. 1993. Fire management. In: Pancel, L. ed.,Tropical Forestry Handbook. Berlin, Springer-Verlag, 2,1221–1268.

Jordan, C.F. 1985. Nutrient cycling in tropical forestecosystems. New York, John Wiley & Sons, 190 p.

Lal, R., and Cummings, D.J. 1979. Clearing tropical forest, I.Effects on soil and microclimate. Field Crop Research, 2,91–107.

Pritchett, W.L. and Fisher, R.F. 1987. Properties andManagement of Forest Soils, 2nd edition. New York, JohnWiley & Sons, 494 p.

44

Saharjo, B.H. 1995. The changes in soil chemical propertiesfollowing burning in a shifting cultivation area in SouthSumatra. Wallaceana, 75, 23–26.

Saharjo, B.H. 1996a. Pemeliharaan tanaman HTI seadanyameningkatkan bahaya kebakaran hutan: Studi kasus diHTI Acacia mangium, Sumatera Selatan. RimbaIndonesia, 31(2,3,4), 12–18.

Saharjo, B.H. 1996b. Disappearance of a newly establishAcacia mangium Willd plantation by fire in SouthSumatera. Tropical Forest Management Journal, 2 (1),65–72.

Saharjo, B.H. 1997a. Kebakaran hutan dan lahan diIndonesia. FUSII Journal, 1(2), 48–58.

Saharjo, B.H. 1997b. The effects of fire on Acacia mangiumplantation (Case study in South Sumatra, Indonesia).Journal of Biological Resource Management, 2(1), 11–14.

Saharjo, B.H. and Husaeni, E.A. 1998. East Kalimantanburns. Wildfire, 7(7), 19–21.

Schindler, L. 1998. The Indonesian fires and SE Asian haze1997/1998. Review, causes and necessary steps. Paper

presented at the Asia–Pacific regional workshop ontransboundary atmospheric pollution. Singapore, 27–28May, 1998.

Viro, P.J. 1974. Effects of forest fire on soil. In: Kozlowski,T.T. and Ahlgren, C.E., ed., Fire and Ecosystems. NewYork, Academic Press, 7–45.

Watanabe, H. 1995. Potential for agro-forestry developmentmainly on Taungya reforestation method. Paper presentedat Seminar on forest resources management, AsiaProductivity Organization. Tokyo, 17–27 October 1995.15 p.

Whitmore, T.C. 1990. An Introduction to Tropical RainForest. Oxford, Clarendon Press, 226 p.

Williams, D.E. 1960. Prescribed burning for seedbedpreparation in jack pine types. Pulp and Paper Magazine,61 (4), 194–198.

WWF 1998. New estimates place damage from Indonesia’s1997 fires at US$4.4 billion. World Wide Fund forNature press release, 2 p.

45

EAST KALIMANTAN PROVINCE covers an area of21 million hectares, comprising 17.3 million ha ofbasic forest and 3.8 million ha of non-basic forest.Based on the Forest Land Use Consensus (1983),basic forest is divided into production forest withoutrestriction (5.5 million ha), production forest withrestriction (4.8 million ha), protection forest(3.6 million ha), reserve and park forest (2 million ha),forest for research and education (17 560 ha) andconverted forest area (1.3 million ha).

After nearly 25 years of forest exploitation, thelocal government reviewed the overall condition ofthe forest areas from the environmental point of view.As a result in 1992, the local government declared TheProvincial Spatial Plan and removed 1.5 million hafrom the area of product forest (with restriction) andconverted it to protection areas, i.e. protection forest,peat area, wildlife reserve, cultural preserve, nationalpark etc., therefore the total of protection areas grewto 7 million ha.

Fires during the 1997–98 El Nino had the mostserious effects on natural forest, industrial timberestates and community forests. Much has been saidabout excessive logging, but in fact, it was not truethat logging activities contributed to deforestation inEast Kalimantan. The Department of Forestrylaunched the TPI system (Indonesian SelectiveCutting System) regulating forest exploitation in theearly 1970s and improved it by using the TPTI system

(Indonesian Selective Cutting and Planting System) in1992–1993. With those regulations, logging onlyremoved 2–7 trees per hectare. Other trees, eitherlesser known or non-marketable trees, were leftundisturbed, but there were 3–4 trees destroyed eachtime by falling trunks.

Clear cutting was done for several purposes, such asfor transmigration, mining, industry, plantations/estates (either agriculture or forest plantation estate)or slash and burn activities. There was falseinformation from several sources regardingdeforestation. Possibly they had never visited EastKalimantan Province. Several groups of internationaljournalists visited and saw the implementation of theTPI system, also members of the All Anglo-Indonesian Parliamentary Group in September 1994.Most of them were satisfied with what they saw inEast Kalimantan, including development of HTI(forest plantation estates) in several areas that areowned and handled by private timber industries. Thewood products of HTI will be used for their industriesand the use of natural forest as a source of timber willdecrease with enhancement of the development HTI.The estimated areas will be 6 million hectares inIndonesia in the year 2000.

Burning as a tool for land preparation and weedcontrol has been used for a long time in agriculture,plantation, and forestry. The indigenous farmers inEast Kalimantan use fire for land preparation of theirlots. However, they are practised in this and the firenever spreads to the forest areas. With time, more

Land and Forest Fire in East Kalimantan Province

Riyanto1

AbstractLand and forest areas in East Kalimantan Province were frequently burned in conjunction with thelong dry period of climatic deviation due to the impact of El Nino in 1997–98. Long dry periods givethe opportunity for farmers, workers of forest estates and agricultural plantations to use fire for landpreparation. However, safe burning techniques were not managed correctly and as a result wildfirespread uncontrolled to the land and forest areas. Impact of land and forest fires was serious, not onlyfor public health, but also hazy weather conditions hampered land, water and air transportation. Theworst impacts were reduction in growing stock, loss of natural regeneration, decreased biologicaldiversity, loss of crops and flooding.

The local government of East Kalimantan Province with Mulawarman University took theinitiative to conduct a workshop on land and forest fire prevention and rehabilitation and rearrange theBukit Soeharto forest tour, in which outputs covered: basic needs for rehabilitation; rehabilitationpolicy; implementation techniques for rehabilitation.

1University of Mulawarman, Samarinda, East Kalimantan,Indonesia

46

people from other islands are coming to EastKalimantan for a better life. Some of them work inagricultural sectors, and of course in doing landpreparation, most of them live by their own traditionalpractices. Under unfamiliar conditions the burning ofland is often uncontrollable and fire spreads aroundtheir lots and enters the forest areas. This is also doneby the workers of agricultural plantations and HTI,therefore during a long dry season such as an impactof El Nino, the land and forest fires could not becontrolled effectively.

It is the responsibility of local government to adopteither fire techniques or regulations, to train people inburning in the correct manner and to prohibit theestates and plantations from using fire for landpreparation. Other regulations should be implementedand pushed by law enforcement.

Based on the report from local government and bygathering data from field reports an estimated 390 000ha in East Kalimantan were affected by forest fires in1998. The Integrated Forest Fire Management (IFFM)has an office in Samarinda and I met with team leaderMr Ludwig Schindler and Ms Anja A. Hoffman onJuly 28 1999. We discussed the reported 8 millionhectares of forest fire in 1999, and I was informed thatthe 8 million hectares of forest fire encompassed Eastand South Kalimantan. By using radar imagery, theIFFM Samarinda has already produced more accurateresults, determining that the fire in East Kalimantanaffected 5.2 million hectares (3 million hectares forestarea and 2.2 million hectares non-forest area). Thesedetails have not yet been released by the IFFM andwill be discussed with the local government. Inaddition, there was an area of surface peat swamp firein East Kalimantan, but IFFM has not yet computedthe size of the area.

Land and Forest FireMost farmers in East Kalimantan prepare their land byburning weeds and shrubs. Land is prepared in theearly dry months of June or July and planting takesplace in September. During El Nino periods the windsare hot, drying and stronger than usual, and firefrequently gets out of control.

Forest concession holders also burn areas inpreparation for industrial timber estates and rubber oroil palm plantations, and these fires have a greaterimpact than those set by the farmers.

Fire is an important factor in forest and grasslandregions but since the early 1980s it has become apersistent threat in East Kalimantan. Crown fires are veryintense and cause major damage to all vegetation and soilorganic matter, whereas surface fires are far less severe.Frequent fires reduce growing stock, decrease naturalregeneration and reduce biological diversity, and thesmoke haze causes major disruptions to land, water and

air transport and affects health of humans and animals.Loss of vegetation leads to increased surface run-off andin times of heavy rain this can lead to flooding. Fires alsoaffect nearby crops such as upland rice, corn, cassava,banana and vegetables and non-wood forest productssuch as rattan and honey. Loss of soil organic matterreduces the water-holding capacity of the soil.

Fire Control MeasuresThere is great concern about the rapid rate of tropicaldeforestation, brought about by temporary orpermanent clearance of forest with fire to make way foragriculture and other purposes. Logging activities intropical rainforests do not have to cause deforestationbut selective cutting is rare. Only a few of the thousandsof species of tropical tree species are marketable andlogging usually only removes 2–7 trees per ha.

Surface coal outcrops ignited during the El Nino of1982–83. Several are still burning and they become asource of persistent fire in normal dry periods as wellas El Nino years.

Concerns about the fires of 1997–98 have led theEast Kalimantan local government to introduce newmeasures to combat the fires.

Fire fightingA local land brigade of volunteers, totalling around700 people, has been formed. They use water andchemical extinguishers, including Chinese grenadeextinguishers, and construct fire breaks. However theresult is unsatisfactory—there are too few people andthey have insufficient equipment.

Since February 1998 aircraft have bombed areas ofKutai National Park with water and Hartindo AF-31chemicals. Fifteen sorties have been undertaken, butpoor coordination has meant the results are only partlysatisfactory. Also attempts at rainmaking by seedingclouds with sodium chloride failed because of windand low humidity.

The NOAA satellite provided data that gave earlywarning of fire outbreaks and pinpointed hot spots.

TrainingA ‘Train the Trainer’ course took place in March1998. Swedish instructors taught 38 people the skillsof fire fighting and fire prevention measures so thatthey could train others. This was a joint cooperativeeffort involving the Government of East Kalimantanand Raddings Vervet UNDP.

Skilled people in East Kalimantan now number13 742, comprising 97 training instructors, 1637 forestguards, 11 447 forest concession security guards and551 people who live near the forest.

47

Firefighting equipmentThe Government of East Kalimantan has received twoconsignments of firefighting equipment from theSwedish Government. One consignment containedequipment for basic firefighting, the other equipmentto combat peat fires. There was sufficient to equipseven districts in East Kalimantan.

Prevention and RehabilitationSome years ago several seminars and workshops wereheld in Samarinda but little was done by localgovernment to implement the recommendations arisingfrom them. When the huge fires broke out in 1997–98people were affected both in and outside EastKalimantan. Now a newly established LocalEnvironmental Impact Control Board (Bapedalda) hastaken responsibility for dealing with land and forest fires.

The Board now takes steps to involve localcommunities and has worked with MulawarmanUniversity to conduct a workshop involving all partiesconcerned with land and forest fires. Participantscame from Bapedalda, Mulawarman University,national and regional departments of Forestry andHorticulture, regional Departments of Transmigrationand Agriculture, private universities, non-governmentorganisations, forest concession operators, private andstate-owned plantations, mining industries and leadersof indigenous communities. Workshop outcomes arelisted below.

Statement of basic need for rehabilitation1. East Kalimantan has rich ecosystems of great

biological diversity. Its huge natural forest areassupport human and animal life and provideeconomic benefit.

2. Land and forest fires have seriously threatened thenatural resources and environment, affected theaesthetic qualities of the forest and reducedeconomic benefits and biological diversity.

3. Smoke haze from the fires has affected publichealth, interrupted transportation over land, air andwater, and caused local and international problems.

4. Following the fires the bare land has eroded,leading to siltation of rivers and lakes and floodingfrom the increased water runoff.

5. The essence of rehabilitation is to restore theecological functions and socioeconomics of theland and forests.

Rehabilitation policy The forest ecosystem has a protective, regulative, andproductive function at the ecosystem level. Thesefunctions can acquire utility value for people andbecome a function of the cultural ecosystem.Therefore, the aim is to rehabilitate the degraded landand forest after burning by using the multipurpose

functions of the trees which should generate economicgrowth and improve quality of life.

Techniques of rehabilitation 1. To select appropriate crops/forest trees (trees that

have multipurpose functions and high economicvalue). Properties sought include fast growth,resistance to fire, ease of planting, and usefultimber, bark, leaves, fruit and seeds.

2. Develop culture techniques that can be used forlabour-intensive enterprises that will benefit peopleliving near the forest.

3. Regreening as a result of rehabilitation activities afterburning will enable land and forest areas to improvesoil and water conservation which may needed byfarmers for farming. Regreening also helps people toget a wide supply of raw materials to meet theirgrowing demands, and supplies employment.

Conclusions1. Land and forest areas in East Kalimantan Province

were frequently burned in conjunction with the longdry period of climatic deviation due to the impact ofEl Nino.

2. Long dry periods give the opportunity for farmers,workers of HTI and agricultural plantations to usefire for land preparation. However, safe burningtechniques were not managed correctly and as aresult wild fire spread uncontrolled to the land andforest areas.

3. Impact of land and forest fire was serious, not onlyfor public health, but also daily weather conditionswhich hampered land, water and air transportation.The worst impact was reduction in growing stock,loss of natural regeneration, decrease of biologicaldiversity, loss of crops and flooding.

4. The local government of East Kalimantan Provincewith Mulawarman University took initiative toconduct a workshop on land and forest fireprevention and rehabilitation and rearrange theBukit Soeharto forest tour, in which output covered:• basic needs for rehabilitation, • rehabilitation policy and • implementation techniques for rehabilitation.

Recommendations1. Since tropical rain forest is a fragile ecosystem, then

all stakeholders should shoulder the responsibilityof reducing or preventing wild fires from destroyingforest resources.

2. Local government could ask assistance ofinternational institutions that have competency andcapability in fire management, including the use ofIFFM (Integrated Forest Fire Management) whichrecently worked in East Kalimantan.

48

KOMODO NATIONAL PARK is located in the province ofNusa Tenggara Timur. It encompasses an area about173 300 ha of land and sea. One third of the area island, consisting of three main islands: Komodo(33 900 ha), Rinca (19 600 ha), Padar (2020 ha) andnumerous small islands.

This national park was established on March 61980, and listed as a World Heritage Site in December1991. The park preserves a truly unique ecosystem,including a fascinating assemblage of plants andanimals. Under the influence of the prolonged dryseason and low rainfalls, vegetation differsdramatically from other regions of Indonesia.

The park is home to active volcanoes, heartyvillagers, and many fascinating, even mysticalcreatures including the komodo dragon. Open fieldsof drought-tolerant plants dominate the landscape.

It is the savanna grassland that best characterisesthe topography of Komodo National Park. Over 70%of the islands are carpeted by sea tawny, chest highgrasses extending from sea level to 500 m. Alang-alang (Imperata cylindrica) dominated grassland,known as padang, arises from periodic burning.

Fires occur frequently in these areas, mainly associatedwith the agricultural purpose of making it easier to walkthrough the savanna and to improve grazing, but also todrive game. Fire becomes more intense during the dryseason, and many fires are difficult to suppress due tolimitations of personnel and equipment. Though fire is

important in shaping the ecology of Komodo NationalPark it needs to be managed properly to prevent itbecoming a destructive agent.

General Conditions

Geological descriptionThe Park is located in the Lesser Sunda Islands ofIndonesia. In this region massive tectonic platescollide beneath the Flores Sea; the general aspect of theislands in the park is a typical arid, uplifted region. Thepermeable ground of these areas receives and holdswater runoff from the hills, supporting an open forest

Most of the land is built on igneous core composedlargely of fine grained, slightly basic rock, of which thechief mineral is plagioclase. Flanking this volcanicmass are lenses of tuff, sandstones, and conglomerateswith intercalated limestone, sandy shoals, and clays.The eastern part of Komodo Island is composed mainlyof steeply tilted beds of corralling limestone. One of thechief volcanic belts of the world passes through theisland belts of the National Park. In many parts of thearchipelago slight earth tremors are felt every fewweeks, and as might be expected some shocks are verydestructive.

ClimatePrecipitationThe geographic range of Komodo National Parkincludes the driest part of Indonesia. Rainfall in

Forest Fire Management in Komodo National Park

Bambang Hartono and Heru Rudiharto1

AbstractKomodo National Park is designated as a World Heritage Site. Located in one of the driest parts ofIndonesia, it has a dry season lasting 9 months of the year and low rainfall (500 mm/year). Seventy percent of the area is grassland. Fire frequently occurs in the area, started for agricultural purposes toimprove grazing and drive out game. Natural fire, usually caused by lightning, is rare. While fire isimportant to maintain grassland for rusa deer and other browsers, uncontrolled fires cause greatdestruction. This paper describes the challenges faced by park management in dealing withuncontrolled fire, taking account of untrained fire fighters, insufficient tools and equipment and thedifficult terrain.

1Komodo National Park, Flores NTT, Indonesia

49

Komodo National Park is the lowest of anysurrounding area for which information is available,with a total of about 500 mm of rain per year. Thetypical rainfall pattern for this part of the world isheavy precipitation in the monsoon months ofDecember to March. Most monsoon rainstorms last afew hours or less, often in the form of cloudbursts.

There is less rainfall on Komodo than on any of thesurrounding large islands. This is partly because theJanuary storms are more to the southeast, and most ofthe rain from the northwest-moving storms arisingfrom high-pressure systems over Australia has alreadyfallen on other islands.

The rainfall pattern of Komodo is very similar tothat reported for Derby, Australia, with a similar total.Small cloudbanks commonly hide high peaks fromlate November well into May. This would suggest thatthe higher forest of Komodo probably receives smallamounts of rain year round. The pattern at thatelevation is probably similar to that of Darwin.

Ground waterFree water is available in pools at all elevations onKomodo during the rainy seasons. The largest andlongest lasting are those at high elevations in theGunung Satalibo–Ara Complex. Here naturaldepressions filled with water are enlarged by boar andwater buffalo. Some are up to 1 m deep and 25 macross, often holding water all year long. Morecommonly pools formed at medium to high elevationsalong streambeds, and rarely hold water for more thana few days after rain.

WindThe prevailing wind is southerly, from the IndianOcean. During the dry season winds may occasionallybe northerly. These increase in frequency with theapproaching monsoon, accompanied by a heavy cloudcover. By the early part of the rainy season stormwinds shift to the northwest, bringing the first hardrains. By middle February the monsoon season is at itspeak, and the rains approach Komodo from all but thesoutheast quadrant. However, the strongest windscome from the northwest and southwest.

The average wind velocities (km/hr) are 0.6 inApril–May and 5.6 in August–September. Strongwinds are infrequent and usually come from the south,accompanying rainstorms. Exposed hilltops usuallyhave only slightly higher velocities.

Relative humidityMean percent RH near the coast was highest duringthe monsoon and lowest at the end of the dry seasonsin October and November.

Solar radiationIn this part of Sunda Islands solar radiation variesfrom 13 128 cal/m3/day in June to 16 732 cal/m3/dayin October.

TemperatureMaximum air temperature during the year is 43°C,minimum 17°C. Black bulb temperature varied from23°C at night and on cloudy days to 54°C (126°F)after the short rainy season and before the beginningof the dry season. Highest temperature is recorded atnoon and slightly after. Average annual temperature atsea level on Komodo was 26.7°C.

Vegetation Many systems of tropical vegetation analysis havebeen proposed since the important work of Schimper(1903) but none has been generally accepted on aworldwide basis (Richards 1952). Of all the plantformations recognised by Schimper three major lowland types can be recognised on Komodo and adjacentislands: monsoon forest, savanna and grassland. Inaddition, most islands in this area possess smallfringing mangrove forests, as well as montane closedforest at elevations above 500 m.

Montane closed forest Most trees in the dry lowlandof Komodo are light shade producers; heavy shadeproducers are found only in the mesic quasi cloudforests found on the highest pinnacles and ridges at500–700 m. These forest are shady, cool, and moistfor most of the year.

Savanna forests Savanna forest covers most ofKomodo. The open canopy lets through a great deal ofsunlight, particularly when the leaves have fallen inthe dry season. The understorey comprises mostlymedium and tall grasses. Fires are common during thelong dry season. Grass productivity in most areas isfairly high. Fire usually occurs only once in mostareas each year and overgrazing is seldom a problemon Komodo.

Monsoon forest The other major forest association ofKomodo is the tropical deciduous monsoon forest.Like the savanna forest it is largely fire resistant andmany of the trees are almost leafless in the dry season.The surrounding savanna can be considered the broadedge or border of the forested area. It is found atelevations below approximately 500 m., an area ofalmost continually dry, hot climate known as hotground. Repeated annual burning leads to a derivedgrassland (ladang) that is an obviously deterioratedhabitat.

Grassland Tropical grassland is another importantfloristic formation in relation to fire management.Grassland formation in the Lesser Sundas isundoubtedly based on the often-repeated errors ofgrassland communities of this island that lead toestablishment of alang-alang (Imperata cylindrica).Alang-alang in Komodo is found only in the moistgardens of the main village and in a few small standsadjacent to permanent springs. The alang-alangdominated grassland that arises by periodic burning(padang) have often formed within the memory of

50

present day villagers, and the ladang to padangsequence is well know to most of them. Such padangareas may remain for a long time, even without fire.

Fire Occurrence

SourcesIn the Lesser Sundas fires, as previously mentioned,are deliberately set. Besides man, many naturalphenomena cause vegetation to ignite, includinglightning, volcanism and spontaneous combustion. Ofthese the most common is lightning. Most fires aregenerated when pulsating general storms occur after along drought, particularly if the storms pass over anarea in which the forest or grassland contains aquantity of flammable material. The pre-monsoonperiod in the Lesser Sundas meets these conditionsperfectly.

In addition, the Komodo topography favoursignition, since mountainous areas are more prone tolightning fires. The importance of lightning in theformation and maintenance of savanna and grasslandin the Lesser Sunda Islands can hardly be disputed.This does not, of course deny that man starts fires, orthat he has not been an important factor in shaping theecology of the area.

OccurrenceAlmost each year fire passes through this area, mainlyduring the dry period between March and November.Occurrences of fire in Komodo National Park duringthe last 6 years are listed in Table 1.

Fire brigadeKomodo National Park has 60 forest rangers, but onlysix have been trained in basic forest fire suppression.The rangers are posted at 10 different locations on theislands. Twenty days a month they stay at their postguards and 10 days they are at headquarters.

While at their post, guards mostly spend their timepatrolling the area and undertaking fire mitigation,prevention and suppression. There are no special firebrigades in the park. The rangers have radiocommunication at each post and report daily activitiesto park headquarters.

Once post guards report a fire, the park managercalls the duty personnel immediately and sends themto the fire location. Experience has taught most of therangers how to handle the situation and they knowwhat to do even with insufficient training.

Tools and equipmentKomodo National Park has very limited forest firehand tools and only one unit portable water pump with200 m hoses. All tools and equipment are located atpark headquarters. Whenever fire is reported, the dutyperson selects the firefighting equipment and sends itto the fire location. In many cases the fighters don’t usethe equipment, because it is inappropriate for the typeof fire. Most use traditional tools, such as palmleaves/gebang (Corypha utan) and agricultural shovels.Portable water pumps have never been used to combatthe fires, due to the lack of water resources during thefire season (dry season).

The fire fighter is not equipped with personal safetyequipment like fire-resistant boots, long sleeved shirtsand gloves, or even a mask to prevent smoke

Table 1. Fire record 1993–1998 in Komodo National Park.

No Date Location Area affected(ha)

1. Oct 1993 Loh Pinda (Komodo Island) 15002. Oct 1993 Loh Letuho (Komodo Island) 3003. Oct 1993 Loh Kima (Rinca Island) 2004. Oct 1993 Loh Ginggo (Rinca Island) 8005. Oct 1993 Muang Island 16. Oct 1993 Sok Nio (Rinca Island) 17007. Oct 1994 Padar Island 3008. Oct 1994 Gililawa (Komodo Island) 9009. Nov 1994 Loh Laju Pemali (Komodo Island) 7510. Nov 1994 Loh Keka (Komodo Island) 5011. Nov 1994 Boe Timba (Rinca Island) 412. Nov 1994 Muang Island 413. Nov 1995. Serai Island 214. Sept 1996 Loh Hung (Komodo Island) 1515. Oct 1996 Toro Somba (Rinca Island) 2016. July1997 Loh Letuho (Komodo Island) 500017. Nov 1997 Loh Laju Pemali (Komodo Island) 1018. Oct 1998 Gililawa Darat Island 1019. Dec 1998 Loh Ginggo (Rinca Island) 700

51

inhalation. Fighting fire is a high-risk job and cancause serious injury or death even with trainedpersonnel and sufficient equipment, but at KNP theycombat fires bare-handed, with only experience andfamiliarity with the area/topography to ensuresuccess.

Needs of Komodo National Park Though fire is part of the vegetation succession in thepark, to maintain rusa deer and other browser habitats,it certainly needs to be managed properly. Thefollowing basic requirements have been indicated andneed to fulfilled.• A fire brigade with well-trained personnel to mitigate,

prevent and suppress fires, and regular training.

• Sufficient personal tools and equipment as well asthe brigade.

• Establish water tanks at certain locations to providewater all year long.

• Develop and maintain fire lines/firebreaks startingin the early dry season. To date there is already a15-km firebreak.

• Provide a well equipped transportation unit, suitedto the park topography and able to deliver firefighters promptly.

ReferencesRichards, P.W. 1952. The Tropical Rain Forest. Cambridge

University Press, Cambridge U.K.Schimper, A.F.W. 1903. Plant Geography upon a

Physiological Basis. Groom and Balfour, Oxford U.K.

52

FROM AN ECONOMIC point of view, the role of forests isextremely prominent in Indonesia. The contribution offorest products to the Indonesian economy hasincreased both in terms of the national income and ofthe foreign currency earned. In addition, forests have asignificant influence on the condition of land, waterresources, housing, industrial development and theenvironment.

On the other hand, there is concern about theconservation of forests in Indonesia. Since the 1970sthere has been extremely intensive and uncontrolledexploitation of forests. In East Nusa Tenggara (NTT),uncontrolled exploitation of the land has occurred formany years. One form of traditional exploitation offorests is the burning of grass around forests, with theresult that the edges of forests have been convertedinto grasslands.

This uncontrolled exploitation of land encouragesus to find ways to manage fires and to overcome thedestruction of forests and work for their conservation.In the context of NTT, this paper will look at theproblem of fire management and reforestation from aneconomic standpoint and its implications foreconomic development in NTT. It should bementioned that this paper is hypothetical and needs tobe tested through later research.

The Basis of NTT’s EconomySince the first Pelita (Five Year Development Plan), infact ever since NTT became a separate province,NTT’s economy has been dominated by theagricultural sector in the broad sense, even thoughagriculture’s contribution to the Gross RegionalDomestic Product (GRDP) has decreased from year toyear. The agricultural sector, in its widest sense,covers food crops, plantations, livestock, forests andfisheries. Data for the GRDP for the years 1992–1997based on constant prices at the 1993 level are shown inTable 1.

The data show that the structure of NTT’s economyis still based on the agricultural sector in the widestsense. This is followed by services, trade, transport,and communication and construction. The agriculturalsector’s contribution to the GRDP is decreasing,whilst construction, trade and transport andcommunication tend to increase from year to year.Services tend to decrease although in the year 1994they increased.

In the agricultural sector, food crops made thegreatest contribution to the GRDP of NTT (see Table2). This is followed by livestock and livestockproducts, plantations and forestry.

The two tables show that the main source of incomefor residents of NTT is still agriculture. As a result, itis best to develop the NTT economy using agriculturein the broadest sense as the basis.

The Implications of Fire Management and Reforestation for Economic Development

in East Nusa TenggaraSiliwoloe Djoeroemana1, Haryono Semangun2,

Bungaran Saragih3 and Achmad Sulthoni2

AbstractThe livelihood of most of the population of Nusa Tenggara Timur depends on dryland agriculture,raising livestock and harvesting forest products. Much of the region’s forests has been removed tosatisfy the need for timber and fuelwood. This paper outlines the efforts made to prevent furtherdeforestation by replanting with trees of high economic value. Uncontrolled burning fromsurrounding grasslands and within the forest has been a major setback to these efforts, and the authorsdiscuss how the institution of more effective methods of fire management would aid the region’seconomic development.

1Wira Wacana Christian School of Economics, Sumba,Indonesia2Gadjah Mada University, Yogyakarta, Indonesia3Bogor Agricultural University, Bogor, Indonesia

Burning Off and ReforestationCurrently forests cover 12.6% of the land area of NTT.The area has been decreasing over many years due tothe uncontrolled exploitation of the forest. One way ofexploiting the forest is by burning off the grass aroundthe forest and the forest for economic purposes. Theprominent use of fire to burn off grasslands and forestsin places like Sumba led one writer to observe in 1928that in Sumba the fire from burning grasslands andforests lit up the dark evening skies.

Historically, a majority of the people of NTTdepends for their livelihood on dry land agriculture,livestock and forest products. For them, the forest is onesource of their livelihood. Therefore the forests have anextremely important role in the economic livelihood ofthe people of NTT, especially rural people.

Rural people exploit forest resources for economicpurposes. In the case of Sumba they cut down treesfrom the forest to build their homes. Initially, treeswere cut down to build their homes but later they werecut down to be sold. They also cut down trees forfirewood. Firewood is used not just in villages but alsoin the towns so that firewood becomes a source ofincome for the villagers.

The forests are also exploited for medicinalpurposes, food and a variety of products for sale. Inthe case of Sumba, the commodities sold includecinnamon, rattan, eaglewood, gnetum and colouringmaterials. All these are sources of income for them.

In an effort to offset the destruction of forests, the

government has carried out reforestation projectsusing economically valuable plants in endangeredparts of forests. Before the first Pelita, the endangeredland comprised 348 000 hectares. The plants plantedincluded teak, mahogany, cassia, gmelina, eucalyptus,dal bergia, albizzia, sandalwood, melaleuca,calliandra, candelnut, cashewnut, tamarind, javaalmond, cinnamon and jackfruit.

From the First to the Sixth Five Year DevelopmentPlan, in fact even before the First Five YearDevelopment Plan, the area of reforestation was126 000 hectares (36%) of the endangered land in theforests which totalled 348 000 hectares before theFirst Five Year Development Plan. The details ofreforestation from before the First Five YearDevelopment Plan up to the Sixth one is shown inTable 3.

As mentioned above, Table 3 shows that the landunder reforestation covers 126 000 hectares (36%). Italso shows that the remaining endangered land withinforests is 223 000 hectares (64%). Thus there is a needfor all parties, both the government and the generalpublic, to undertake reforestation.

How far has reforestation succeeded? Noinformation is yet available to show the success rate.The Forest Service for NTT has informed us that nosystematic and accurate evaluation has yet been madeabout the level of success of reforestation programs.However, community leaders and NGOs (specificallyin Sumba) have stated that in general reforestation hasnot been very successful, although in certain locations

53

Table 1. The percentage distribution of the Gross Regional Domestic Product based on 1993 constant prices according toindustry of origin/sector.

No Industry of origin 1993 1994 1995 1996 1997

Agriculture 41 40 38 38 38Mining and quarrying 1. 6 1.6 1.7 1.7 1.6Manufacturing 2.5 2.6 2.6 2.6 2.4Electricity, gas and water 0.7 0.7 0.7 0.7. 0.8Construction 8.1 8.1 8.8 8.9 8.2Trade 12.5 11.8 12.4 13.3 13.7Transport and communication 9.2 9.8 10.5 10.4 10.4Finance, rental and business services 4.4 4.4 4.3 4.6 4.7Services 20 21 21 20 20Gross Regional Domestic Product 100.00 100.00 100.00 100.00 100.00

Source: NTT in figures for 1996 and 1997.

Table 2. The percentage distribution of the GRDP based on constant prices for 1993 for the agricultural sector, 1993–1997.

No Agricultural sector 1993 1994 1995 1996 1997

Food crops 23 22 22 22 22Plantations 4.4 4.0 3.8 3.9 3.8Livestock and their products 9.6 9.7 9.0 9.0 8.7Forest 0.6 0.6 0.6 0.6 0.5Fisheries 3.7 3.5 3.1 3.0 3.2

41 40 38 38 38

Source: NTT in figures for the years 1996 and 1997.

it has been relatively successful. Comprehensiveevaluative research into the implementation ofreforestation in NTT is required.

One reason why reforestation has been ratherunsuccessful is that uncontrolled burning off hascontinued in grasslands and forests. This burning offhas consumed the trees planted in reforestationprograms. Uncontrolled burning off, carried out toopen up new fields, has resulted in neighbouring areasunder reforestation being burnt too. Also as in the caseof Sumba, burning of grasslands and forests occursbecause of cigarette butts thrown out of vehiclespassing through particular areas.

The Forestry Service has tried to control fires byusing two different methods—the yellow strips andthe green strips. The green strip is done by plantingfire-resistant plants such as lannea and glyricidiawhich have a dual function in surrounding the area ofreforestation. This approach has not been usedsystematically due to lack of funds. The yellow strip isdone by clearing the land around the area ofreforestation to a width of about 20 metres to preventfire entering the area under reforestation. This methodhas also not been used to the full due to lack of labour.

Information obtained from interviews with villagecommunity leaders and non-governmentorganisations in Sumba shows that the people ofSumba utilise certain fire-resistant plants such asramie. In the past these plants were used as fencesaround food gardens to protect them while fires wereburning off grasslands.

From the above description, it is clear that fire wastraditionally used for economic purposes. However,the uncontrolled use of fire has brought disaster in theform of destruction of the forests with all the negativeimplications for the sustainability of forest resources.This destruction will in its turn destroy the economicresources needed to raise the level of wellbeing of the

people of NTT. Reforestation has also had an economic purpose.

Economically valuable plants, in addition toimproving the structure of the soil, also help toimprove the wellbeing of the people of NTT.However, reforestation is also threatened by theburning of grass and forests.

Implications for Economic Development in NTT

The above information demonstrates that theuncontrolled use of fire is a major cause of thedestruction of forests and the failure of reforestationprograms. This has resulted in a reduction in soilfertility, elimination of forest resources and a decreasein the availability of economically valuable varietiesof wood in NTT. As a result the contribution of foreststo the economy of NTT is very small.

Starting from agriculture as the basis of NTT’seconomy, the most appropriate strategy for thedevelopment of NTT’s economy is to make agro-business the major motor. Forestry resources are veryimportant to support the development of agro-business. At the national level, manufactured goodsmade from wood, which are extractive in nature, makethe greatest contribution to agro-industrial exports.Therefore, economically valuable reforestation plants,in addition to other forest resources, have goodpotential for the development of agro-industry andagro-business in NTT.

Indonesia’s empirical experience shows that agro-industry and agro-business have significantprominence. Dasril’s analysis, as quoted by Saragihpoints to several outstanding features of agro-industryin Indonesia for the years 1971–1990. It hascontributed considerably to the aggregate value of

54

Table 3. Reforestation (in hectares) since prior to Pelita and up to Pelita VI.

No. Regency Area of Reforestation Area of land Remainingendangered Prior to Pelita Pelita Pelita Pelita Pelita Pelita reforestation endangered

land Pelita I I II III IV V VI landin forest

1. Kupang 51700 1060 310 2930 5510 1310 4400 2.0 18290 334102. TTS 19170 240 480 2125 2190 4615 6560 1.2 14180 49903. TTU 56270 270 425 2630 5540 2525 2800 1.2 12630 436404. Belu 9640 560 260 2100 3680 1090 3600 1.2 9500 1405. Alor 46930 140 180 280 3620 110 2050 946 7330 396006. FLOTIM 17560 50 120 400 2270 90 1450 1.2 7270 102907. Sikka 21250 90 280 470 2890 - 4400 1.4 9430 118108. Ende 17790 300 110 430 3940 920 2740 1.5 9950 78409. Ngada 23050 250 - 255 3770 1750 3000 961 10350 1269010 Manggarai 10720 40 - 460 3000 1730 2100 1.2 8590 212011 S. Timur 43740 260 680 210 2950 440 3230 1.3 9110 3427012 S. Barat 30970 600 140 320 3650 485 2750 1.2 9150 21820

Total 348790 3860 2985 12610 43010 15065 39080 1920.4 125780 222620

Source: Report of the NTT Forestry Service 1999.

55

non-oil and gas industrial products and industrialexports. The added value contribution of the non-oiland gas industries was 63% in 1971, 64% in 1975,66% in 1980, 67% in 1985 and 62% in 1990. Whilstthe contribution of non-oil and gas export productswas 79% in 1971, 46% in 1975, 47% in 1980, 75% in1985 and 81% in 1990.

In terms of employment opportunities, the non-oiland gas industries were also significant. In 1971 76%, in1975 63%, in 1980 71%, 1985 79% and 1990 76%. Themultiplier effect of the added value increasedcontinuously. In 1971 it was 0.9, in 1975 0.8, in 19802.2, in 1985 2.3 and in 1990 2.9. This quite highmultiplier effect is a source of growth through backwardand forward linkage as well as sideways linkages.

It is necessary to explain here that the major sourceof growth for agro-industry is private consumption.This means that the development of agro-industry todate has not been a heavy burden on the governmentbudget. In fact it has helped to encourage capitalformation and supported the principle of economicself-sufficiency. In addition agro-business has showna vertical linkage between agro-business subsystemsand a horizontal linkage with external systems orsubsystems such as the banking services, transportservices, trade services and educational services.

The potential of agro-business is also evident in itsability to survive in the development period of thenew order which undertook industrialisation with thesupport of an overvalued exchange rate policy. At themoment, agro-business in Indonesia is facing pressurefor the following reasons:1. increase in the import of agricultural products

because of a subsidy of the exchange rate;2. pressure on the export of agricultural products

because of a ‘tax’ on the exchange rate; 3. a high domestic interest rate because funds have

been absorbed into costly projects and the high

maintenance costs in banking. Nonetheless, the balance of trade on agro-business

is still in surplus and increases annually. Without thepressures mentioned above, the surplus from tradingin agro-business would be greater still.

We can therefore see that the agro-business sectorhas made an important contribution to net exports forthe 30 years in which Indonesia has been developing.The agro-business sector has also been capable ofholding out in the economic crisis. This picture showsthat with the existing agricultural basis, NTT candevelop its agro-business sector as part of thedevelopment of NTT’s economy.

Fire management and reforestation should behandled as efficiently and effectively as possible toenable the forestry sub-sector to play a positive role inthe development of agro-business in the developmentof NTT’s economy. So fire management andreforestation have positive implications for theeconomic development of NTT.

Conclusions and RecommendationsIt can be concluded that the uncontrolled use of firehad led to the destruction of forests, a thinning ofeconomic resources in forests, and the destruction ofeconomically valuable reforestation plants. Therefore,to minimise the use of uncontrolled burning, efficientand effective fire management and reforestationprograms are required for the economic developmentof NTT. In this way, the development of NTT’seconomy, which is driven by the agro-business/agro-industrial sector, can be enhanced.

Based on this conclusion, we recommend that aseries of research studies be carried out regarding themanagement of fires and reforestation in ways whichwill support agro-business as part of the economicdevelopment of NTT.

56

EAST NUSA TENGGARA (ENT) is climatically amongthe driest regions in Indonesia. This condition is dueto a combination of unevenly distributed, low annualrainfall, high wind speeds, and intense solar radiation(Monk et al. 1997). Heavy rains do occur, but theirerratic patterns—frequently after a long droughtperiod—contribute most to runoff causing erosion(Crippen International 1980a). Mountainouslandscapes further increase the adverse effects oferosion on relatively infertile soil derived from eithervolcanic or sedimentary parent materials. Savannaand grassland become the most dominant vegetationcover in this type of environment. It is to cope withthis harsh environment that most peasant farmers inthe province engage in more than one agriculturalsystem as a matter of insurance (Metzner 1982).

Fire can start either naturally or artificially insavannas or grasslands (Barbour et al. 1987; Monk etal. 1997). Frequency, intensity, and extent of naturalfire in such environments increase because grasscover provides large amounts of fuel and themicroclimate causes this fuel to dry rapidly and fire tospread readily (D’Antonio and Vitousek 1992). Thesefactors also favour burning practices for severalpurposes, notably as an integral part of agriculturalpractices evolving in savanna and grasslandecosystems (Monk et al. 1997; Crippen International1980b).

For peasant farmers lacking capital and skills, fire isthe only tool available for coping with land clearingand weeding problems. Unless more feasiblealternatives are made available, the use of fire willcontinue despite any negative impacts it may cause.Whatever are the final conclusions about the use offire as an integral component of the existingtraditional agriculture in ENT and anywhere else inthe world, opportunities do exist for developmenttowards more sustainable systems.

This paper will attempt to describe thecharacteristics of fire and agricultural systems inENT. The description will hopefully provide a soundbasis for better understanding the extent of fire inENT and a background for discussing issues related tothe need for fire management policy and practices tosupport agricultural development in the province.

Characteristics of Fire and Agricultural Systems

Landscape, vegetation cover and agriculturalpracticesENT landscapes are mountainous; 45% of total landarea has slopes over 40% (Dick 1991). Sedimentarylimestones are the most important rock componentson the outer arc (Sumba, Sabu, Rote, and Timor),producing a surprising variety of karst land forms andraised coral terraces. On the other hand, the inner arc

Fire and the Management of AgriculturalSystems in East Nusa Tenggara

Wayan Mudita1

AbstractEast Nusa Tenggara (ENT) boasts very diverse agricultural systems within which fire still playsimportant roles. Depending on how it is used, fire may be either beneficial or destructive. Despite theimportant roles and widespread use of fire, the current policy is more on fire suppression than on firemanagement. Fire has been blamed by the policymaker as the factor that contributes to creatingmarginal lands, while on the other hand it is praised by local communities as the most efficient toolavailable for land preparation, soil fertility improvement, and weed and pest control. However, ifproper management is not initiated, the negative impacts will prevail and unavoidably threaten thesustainability of agriculture in the province. Therefore, fire has to be incorporated into agriculturaldevelopment policy in ENT. For this purpose, the currently adopted commodity-based, productivityoriented approach must be reconsidered. An agroecosystem approach that allows trade-off betweenproductivity, stability, and equitability must be formulated on the basis of local environmentalcarrying capacity.

1Environmental and Natural Resources Research Centre, NusaCendana University Jl. Adisucipto, Penfui, Kupang 85001,NTT, Indonesia

57

(Flores and the neighbouring small islands) consists ofstrongly dissected old volcanoes as well as conicalyoung volcanoes (RePPProT 1989a). Alluvial andcolluvial plains do occur, but they occupy relativelysmall areas in comparison with other land forms.These plains, along with river terraces common inTimor and sedimentary basins in Rote, have tended tobecome important areas for certain types of permanentagriculture, notably wetland rice fields.

The rest of the areas are covered mostly bysecondary vegetation; only 10% of the total land areais currently under closed forest cover (RePPProT1989a, b). Secondary forest is not climax vegetationbut the product of disturbance. The disturbances canbe natural or anthropogenic in their origins, and fire isthe most important. Frequent burning of monsoonforest will produce savannas and grasslands, butnatural savannas and grasslands may also occur insome of the driest parts of Sumba, Timor, and othersmaller islands. Although fire may start naturally inthe climatically dry areas with abundance of standingdry biomass, fire in ENT is closely related toagricultural practices, notably shifting cultivation. Forsome reasons, this type of agricultural practice isperformed mostly in mountainous areas. First, landswith steep slopes are considered as having less weedproblems compared with flat lands where weed seedsare deposited by runoff. Second, almost all fertilecoastal alluvial plains have been used for permanentagriculture. Third, mountainous areas are relativelyfree from roaming livestock that could easily destroythe crops.

NTT boasts not only shifting cultivation but manyother agricultural systems as well (Monk et al. 1997).One of the most traditional agricultural systems, i.e.hunting and gathering, is still practised in ENT,together with other traditional mixed tree gardens andfree-grazing livestock. Some sorts of more recentpermanent agricultural systems such as permanentupland agriculture, homegardens, irrigated andrainfed rice fields, etc., also exist, although only onmore limited areas. Few people in ENT depend ononly one system but on a combination of thesesystems. These diverse agricultural systems can beseen as some sort of adaptation to the harshecosystems. Peasant farmers attempt to keep risks aslow as possible by having several fields with differentagricultural systems as dictated by theirenvironmental conditions. According to Metzner(1982), risks are also reduced by selecting droughtresistant instead of high yielding crop varieties and bybinding to traditional values.

Time and frequency of burningBurning is carried out late during the dry season,usually starting from October or November. Startingfrom the middle of the dry season, peasant farmers cut

down trees and clear shrubs either in new sites or inold fields. The cut-down trees and shrubs are left todry for 2–3 months before being burned towards theend of the dry season (Monk et al. 1997). Theavailability of dry biomass is important in shiftingcultivation since the purpose of burning is not only toclear the land but also to produce ample amount of ashrequired to fertilise the soil and to obtain heat to killweed seeds and rhizomes, soil-dwelling insect pests,and soil-borne pathogens (Rambo 1984). Whenburning is incomplete in the first occasion, secondaryburning is carried out to make sure that the requiredamounts of ash and heat have been produced.

Early burning, such as practised in Kakadu NationalPark, Australia (Russell-Smith 1995b), may also takeplace, but is certainly not intended for shiftingcultivation. Grass may set to fire accidentally early inthe dry season to promote new regrowth for cattlegrazing (Crippen International 1980b). This type offire is not intensive since there are limited fuel loadsand driving climatic factors, notably wind. Earlyburning for shifting cultivation is not performed for atleast two reasons. First, early burning is restricted inold fields because shifting cultivation involves mixedcropping of early and late crops (Friedberg 1989;Hoskins 1989). Second, peasant farmers must spendtheir times for harvesting the crops so that cuttingdown trees and clearing shrubs for new fields can onlybe done afterward (Monk et al. 1997). Burninggrasslands to promote grass regrowth and to drive outgame animals is also more common late in the dryseason since only during this period grass is scarceand time is available for hunting (CrippenInternational 1980b; Suhardjono 1988).

Purpose and scale of burningTo some extent burning is practised in all agriculturalsystems found in ENT but in some ways the reasonsare still not fully understood. Rice straw in wet ricefields is sometimes burned after harvest. In permanentsystems such as home gardens and permanent drylandfarming, dry plant materials are collected fromsomewhere else, spread under a crop canopy, andburned at the end of the dry season. According toCrippen International (1980b) burning is intended toprepare land, provide neater fields and easier access inshifting cultivation, to stimulate grass regrowth forhunting and free grazing livestock, to protect propertyfrom wild fires in permanent cultivation systems, andsimply to have fun in abandoned fields. Burning helpsto remove large quantities of mostly thorny or pricklyvegetation and thereby to deposit nutrient-rich ash onthe soil surface, naturally providing the newly plantedcrops with the necessary nutrients (Monk et al. 1997;Rambo 1984). Burning is also considered helpful forgetting rid of weeds and other pests, especially in oldfields (Barbour et al. 1987; Rambo 1984). It is the

58

increased demand for human labour in weeding ratherthan the decline of soil fertility per se that leads todecisions for cultivated fields to lie fallow for severalyears.

Use of fire in permanent systems such as inpermanent dryland farming, home gardens, irrigatedrice fields, etc. is much more restricted than that ofshifting cultivation, hunting and gathering, and free-grazing livestock systems (Monk et al. 1997). The useof fire in the later agricultural systems in ENT is mostintensive and therefore, these systems are referred tohereafter as fire-based agricultural systems.

Burning in the fire-based agricultural systems inENT is mostly small-scale. The area of most shiftingcultivation fields is less than 1 ha. Larger areas may beburned to promote grass regrowth in the grasslands,but fire in this type of vegetation is certainly lessintensive. Although small in size, the percentage ofburned areas to the total land area of ENT is muchbigger than that in a continent such as Australia. Thishas tremendous implications to the individual smallislands of the island ecosystems. As Brookfields(1990) notes, vulnerability is almost directlycorrelated to small size.

Use of Fire and the Trend of Agricultural Development

Possible impacts of fireIn reviewing impacts that burning may have onrelatively small islands of ENT, one must carefullyconsider observations made by Hess (1990). Impacts ofburning in small islands cannot be simply extrapolatedfrom those in the continents because: (1) small landmass, with coastal regions making up a higherpercentage of land, is vulnerable to disturbance, (2)water catchments are smaller and erosion levels higher,and (3) their environments may be more specialised andendemicity higher.

Shifting cultivation and other fire-basedagricultural systems are often praised for efficiency interms of energy and labour input. However, Rambo(1984) has argued that use of fire does not mean a freelunch, it is neither energetically efficient norenvironmentally benign. The difference is only on theform of energy subsidy in use, biological or fossilfuel, and the form of cost, direct or indirect.

Fire can be seen as a form of biological energysubsidy to fire-based agricultural systems, comparableto fuel energy and fertilisers to the more intensiveagricultural systems. However, the nutrient-rich ashresulting from burning for shifting cultivation in ENTis soon removed by runoff and further soil erosionsteadily depresses soil quality (Crippen International1980b). While making some nutrients biologically

more available, fire also volatilise others, notablynitrogen, sulfur and carbon (D’Antonio and Vitousek1992; Raison 1979). During the next dry season, soilwill be more exposed and its humus will breaks downmore rapidly. Finally, repeated clearing and burningrapidly encourage the growth of flammable grasslands(Freifelder et al., 1998; Mack and D’Antonio 1998).These effects are certainly less significant in larger, oldland masses such Australia. In the Kakadu NationalPark for example, in spite of the intense, highlyseasonal rainfall, the generally low slope angle slowsdown erosion (Russell-Smith et al. 1995). Even whenthere is erosion; it simply enriches the floodplains.Volatile nutrients may be deposited somewhere butthey will finally be brought back by runoff and streamsto the floodplains.

Burning for shifting cultivation may cause moredevastating off-site than on-site impacts. While someportions of the extensive deciduous scrub currentlyfound in ENT are the products of on-site impacts ofburning, most result from escape fires. Monk et al.(1997) have provided succinct empirical evidence thatfire, whether caused by human activity or natural, hasadded substantially to, if not been the cause of, thedevelopment of extensive savanna and grasslands inENT. Remnant forests are the only refuge availablefor wildlife and therefore careless burning on adjacentsites poses, contrary to that in Australia’s KakaduNational Park (Press et al. 1995; Russell-Smith1995a), a serious threat to conservation of biologicaldiversity of the area. Off-site impacts imposed byburning are not limited to habitat destruction, but theeven more far reaching and in many ways cumulativeare flooding, sedimentation, destruction of woodfences, and reduction of air quality (CrippenInternational 1980a; Metzner 1982; State Ministry ofEnvironment 1998).

Despite all debates on the positive and negativeimpacts of burning, little experimental work has beendone locally to support either side. Fire-basedagricultural systems may have a number of ecologicaladvantages under a seasonally dry tropicalenvironment, but only if they are practised at anappropriate level of population density and inaccordance to indigenous knowledge and practices(Brady 1996). As a report by the State Ministry forEnvironment (1998) has clearly pointed out, erosionof indigenous knowledge and practices on safeburning is, unfortunately, an important factorinfluencing forest and land fires in Indonesia.

Present trend of agricultural developmentShifting cultivation and free grazing of livestock havebeen practised for generations and are still extensivelypractised today. They may cause environmentaldegradation that now the province has to deal with,but they are also the only sound alternatives for a

59

region where topography, soil, and rainfall limit thedevelopment of permanent agriculture. The solution istherefore not as simple as forcing peasant farmers toadopt some sorts of intensification without clearunderstanding of the underlying ecological principles.Despite much criticism (e.g. Metzner 1982), it seemsthat the policy of agricultural development in theprovince still lacks this understanding.

The recent introduction of new, high-yieldingmaize varieties into ENT provides a good example. Asa general rule, any high-yielding variety requiresfavourable environmental conditions to produce itshigh yield. The reality is that maize in ENT is a cropof shifting cultivation. The environment of shiftingcultivation cannot naturally support the requirementsof a high-yielding variety nor can agrochemical inputbe provided by the peasant farmers who lack capitaland the necessary skills. Peasant farmers stick to localmaize because what is important to them is not toincrease production but to secure food up until thenext rainy season (Monk et al. 1997). Also this newlyintroduced maize cannot meet this purpose because itsgrains are destroyed by weevils soon after harvest.Although lower in yield, local maize has an importantadvantage that it can better withstand water shortagedue to the erratic pattern of rains.

Another example is the so-called integratedagricultural development project. Metzner (1982)considered integration of food crops, cash crops, andlivestock as being the most ideal for a region limitedin resources. To be truly integrated, the project must,to some degree, address the conflict between shiftingcultivation and free-grazing livestock. In fact,however, integration occurs only on the table of theproject manager. In the field, the crop component goessimply as an add-on to shifting cultivation in onevillage while the livestock component to anotherlocation with a large area of open land. To manygovernment officials, large open land meansgrassland, ignoring the fact that weeds have takenover most of the grassland areas (Monk et al. 1997).

Need for sound fire managementUse of fire in tropical agricultural systems has longbeen controversial (Sanchez 1976). The governmenttends to blame burning for the creation of newmarginal lands, but the policy toward fire is not clear.The Indonesian Ministry of Agriculture has noregulation related to fire and no fire management unit(State Ministry for the Environment 1998). There is aban on burning for land clearing, but that is forplantations now under the Ministry of Forestry andPlantations.

All conflicting arguments on the positive andnegative impacts of fire suggest that sound firemanagement is necessary if agricultural development inENT is to be sustainable. A complete ban on fire use is

certainly unlikely to fulfil this goal. Instead, thecurrently adopted commodity-based, productivity-oriented approach must be reconsidered and fire-basedcultivation systems be given appropriate recognitionwithin mainstream agricultural development policy. Anagroecosystem approach that allows trade-offs betweenproductivity, stability, and equitability must beformulated on the basis of local environmental carryingcapacity (Conway 1984, 1985; Mudita 1998). For thispurpose, recognition of the existing traditional systemand development of modern agroforestry systems withnitrogen-fixing and fire-resistant trees and shrubs arethe most feasible alternatives. An important advantageof agroforestry is that extensive burning to releasenutrients or kill weeds and soil-borne pests andpathogens is no longer possible and needed. However,agroforestry in itself will do little to reduce all thenegative impacts of burning until a comprehensiveimprovement has been made to the presentuncontrolled land-use practices and complicated landtenure systems leading to what Conway and Barbier(1990) call the common property problems.

Current Issues and DiscussionA great diversity of fire management practices doesexist in other regions (e.g. Russell-Smith 1995b), butintroduction of a particular practice to ENT needscareful examinations. At least, any introduced practicehas to fit to the existing agricultural systems. Morepreferably, such practice should have the capacity ofimproving the overall system performances (i.e. notonly to increase productivity) and be acceptable tolocal social and cultural values (Vivian 1995). At thesame time, locally available fire management practicesshould be identified, documented, and carefullystudied. In both approaches, training will be required,including advanced training at postgraduate level thatwill enable local university staff to carry out researchuseful for decision-making processes.

Interaction between fire and agricultural systemshas been extensively studied but in differentenvironmental settings. Understanding various factorsthat maintain this interaction locally will be helpfulfor at least two reasons. First, the effect of fire on soiland vegetation is mostly a local phenomenon.Whether fire will be beneficial or destructive dependson the available fuel loads, effects of localenvironmental factors (e.g. winds), and the practice ofburning itself. Second, disturbance to such interactionmay lead to different direction according to variationin local environmental setting. For example, invasionby an exotic plant such as Chromolaena odorata todry areas such as ENT will maintain repeated fires,but to wet areas such as in Sumatra may promotesuccession toward forest (Tjitrosoedirdjo 1999).

60

At this moment C. odorata invasion hasconsiderably changed burning practices among ruralcommunities as well as the existence of natural fire inENT. Peasant farmers consider C. odorata as beingbeneficial for shifting cultivation but at the same timedestructive for free grazing livestock. Burning doesnot control but rather promotes flush regrowth thatcreates a denser stand of this exotic plant after therainy season. This will force farmers to simplyabandon a field, leaving the drying C. odorata stemsas the flammable fuel either for intentional burning orfor natural fire during the dry season. This means thatdevelopment of an appropriate windbreak andfirebreak within an agroforestry scheme is crucial formanagement of both fire and C. odorata. The currentwork of the Environmental and Natural ResourcesResearch Centre of Nusa Cendana University is on acommunity-based project based on this approach. It ismanaged by PLAN International in IndonesiaProgramme Unit Kupang and sponsored by AusAID(Mudita 1999). However, the success of such an effortwill depend largely on the willingness of thegovernment to adequately resolve the complicatedland-use and land tenure issues within thedevelopment policy of the province. While workingreasonably well when population pressure is relativelylow, such traditional land tenure systems have seriousdrawbacks, particularly to promotion of moreadvanced types of cultivation systems necessary tocope with problems resulting from C. odoratainvasion.

ConclusionFire is used to some extent in all agricultural systemsin ENT, but it plays important roles in shiftingcultivation, free-grazing livestock, and hunting andgathering systems. Despite the important roles of fire,however, regulation on fire use for agriculturalpurposes is non-existent. Implementation of eitherintroduced or locally existing fire management needsto adequately address land-use systems and landtenure practices to achieve some degree of integrationnecessary for the development of sustainableagriculture.

AcknowledgmentThe author appreciates valuable comments providedby Dr. Bill Roscoe of Winrock International to theearlier version of this paper. Appreciation is also dueto Mr. Colin G. Wilson, Parks and WildlifeCommission of Northern Territory, for his help withreferences.

ReferencesBarbour, M.G., J.H. Burk, and W.D. Pitts 1987. Terrestrial

Plant Ecology. 2nd ed. Menlo Park, TheBenjamin/Cummings Publishing Company.

Brady, N.C. 1996. Alternative to Slash-and-Burn: A GlobalImperative. Agricultural Ecosystems and Environment58:3–11.

Brookfields, H.C. 1990. An approach to islands. In: Beller,W. d’Ayala, P. and Hein, P., ed., SustainableDevelopment and Environmental Management ofSmall Islands. Paris and New Jersey, UNESCO andParthenon Publishing Group.

Conway, G.R. 1984. Applying ecological concepts to thestudy of the intensification of the use of Indonesia’sagro-ecosystems. In: KEPAS, The Sustainability ofAgricultural Intensification in Indonesia. Jakarta, TheFord Foundation in collaboration with the Agency forAgricultural Research and Development, Ministry ofAgriculture, 13–24.

Conway, G.C. 1985. Agroecosystem analysis. AgriculturalAdministration, 20, 31–55.

Conway, G.C., and Barbier, E.B. 1990. After the GreenRevolution: Sustainable Agriculture for Development.London, Earthscan Publication.

Crippen International 1980a. Timor Island Water ResourcesDevelopment Study, Final Report, Vol. 4: Climate.Ottawa and Jakarta, Canadian InternationalDevelopment Agency, for Directorate General of WaterResources Development, Ministry of Public Works,Indonesia.

Crippen International 1980b. Timor Island Water ResourcesDevelopment Study, Final Report, Vol. 14: Forestry.Ottawa and Jakarta, Canadian InternationalDevelopment Agency, for Directorate General of WaterResources Development, Ministry of Public Works,Indonesia.

D’Antonio, C.M., and Vitousek, P.M. 1992. Biologicalinvasions by exotic grasses, the grass/fire cycle, andglobal change. Annual Review of Ecology andSystematics, 23, 63–87.

Dick, J. 1991. Forest Land Use, Forest Use Zonation andDeforestation in Indonesia: A Summary andInterpretation of Existing Information. Jakarta,Environmental Management Development in IndonesiaProject.

Freifelder, R.R., Vitousek, P.M., and D’Antonio, C.M. 1998.Microclimate change and effect on fire followingforest–grass conversion in seasonally dry tropicalwoodland. Biotropica, 30, 286–297.

Friedberg, C. 1989. Social relations of territorialmanagement in light of bunaq farming rituals.Bijdragen tot de taal-, land-en Volkenkunde 145. Part INusa Tenggara Timur, 4, 548–563.

Hess, A. 1990. Overview. In: W. Beller, P. d’Ayala, and P.Hein, ed., Sustainable Development and EnvironmentalManagement of Small Islands. Paris and New Jersey,UNESCO and Parthenon Publishing Group.

Hoskins, J. 1989. Burned paddy and lost souls. Bijdragen totde taal-, land-en Volkenkunde 145. Part I NusaTenggara Timur, 4, 430–444.

61

Mack, M.C., and D’Antonio, C.M. 1998. Impacts ofbiological invasions on disturbance regimes. Tree, 13,195–198

Metzner, J. 1977. Man and Environment in Eastern Timor: AGeoecological Analysis of the Baucau-Viqueque Areaas a Possible Basis for Regional Planning. Canberra,Australian National University, Development StudiesCentre Monograph 8.

Metzner, J. 1982. Agriculture and Population Pressure inSikka, Isle of Flores. Canberra, Australian NationalUniversity, Development Studies Centre Monograph28.

Monk, K.A., De Fretes, Y., and Reksodihardjo-Lilley, G.1997. The Ecology of Nusa Tenggara and Maluku. TheEcology of Indonesia Series Vol. V. Singapore,Periplus Editions.

Mudita, I.W. 1998. Invasi Chromolaena odorata di NTT:Menyikapi Perubahan sebagai Peluang MenujuPengelolaan Lingkungan Semi-Ringkai secaraBerkelanjutan. An oration on the XXXVI Anniversaryof Nusa Cendana University. Environmental andNatural Resources Research Centre, SpecialPublication, 1, 1–29.

Mudita, I.W. 1999. Program Aksi Lingkungan BerbasisMasyarakat untuk Mengendalikan DampakChromolaena odorata di Daerah Kering. In: I W.Mudita, ed., Prosiding Lokakarya Aksi LingkunganBerbasis Masyarakat untuk Mengendalikan DampakChromolaena odorata di Daerah Kering. Kupang,PLAN International in Indonesia Programme UnitKupang and Environmental and Natural ResourcesResearch Centre, Nusa Cendana University, 20–34.

Press, T., Brock, J., and Andersen, A. 1995. Fauna. In: Press,T.. Lea, D Webb, A. and Graham, A., ed., Kakadu,Natural and Cultural Heritage and Management.Darwin, Australian Nature Conservation Agency andNorth Australia Research Unit, Australian NationalUniversity, 167–216.

Raison, R.J. 1979. Modification of soil environment byvegetation fires, with particular reference to nitrogentransformations: a review. Plant and Soil, 51, 73–108.

Rambo, A.T. 1984. No free lunch: A re-examination ofenergetic efficiency of swidden agriculture. In: Rambo,A.T. and Sajise, P.E., ed., An Introduction to HumanEcology Research on Agricultural Systems inSoutheast Asia. College, Laguna, University of thePhilippines at Los Baños, 154–163.

RePPProT 1989a. Review of Phase I Results, Maluku andNusa Tenggara. Volume I: Main Report. Jakarta,Ministry of Transmigration, ODNRI, and ODA.

RePPProT 1989b. Review of Phase I Results, Maluku andNusa Tenggara. Volume II: Annexes. Jakarta, Ministryof Transmigration, ODNRI, and ODA.

Russell-Smith, J. 1995a. Flora. In: Press, T. Lea, D. Webb,A. and Graham, A., ed., Kakadu, Natural and CulturalHeritage and Management. Darwin, Australian NatureConservation Agency and North Australia ResearchUnit, Australian National University, 127–166.

Russell-Smith, J. 1995b. Fire management. In: Press, T. Lea,D. Webb, A. and Graham, A., ed., Kakadu, Natural andCultural Heritage and Management. Darwin, AustralianNature Conservation Agency and North AustraliaResearch Unit, Australian National University,217–237.

Russell-Smith, J., Needham, S., and Brock, J. 1995. Thephysical environment. In: Press, T. Lea, D. Webb, A.and Graham, A., ed., Kakadu, Natural and CulturalHeritage and Management. Darwin, Australian NatureConservation Agency and North Australia ResearchUnit, Australian National University, 94–126.

Sanchez, P.A. 1976. Properties and Management of Soils inthe Tropics. New York, John Wiley & Sons.

State Ministry for Environment 1998. Forest and Land Firesin Indonesia, Executive Summary. Jakarta, StateMinistry for Environment and United NationsDevelopment Programme.

Suhardjono, Y.R. 1988. Dampak Penggunaan SumberdayaLahan Kering Nusa Tenggara Timur. Paper presentedat a meeting on Dryland Management, IndonesianInstitute of Science, Serpong, Indonesia, 15 Januari1988.

Tjitrosoedirdjo, S. 1999. Strategi penerapan PHT dalampengendalian dampak Chromolaena odorata. In:Mudita, I W., ed., Prosiding Lokakarya AksiLingkungan Berbasis Masyarakat untukMengendalikan Dampak Chromolaena odorata diDaerah Kering. Kupang, PLAN International inIndonesia Programme Unit Kupang, andEnvironmental and Natural Resources ResearchCentre, Nusa Cendana University, 1–15.

Vivian, J.M. 1995. Foundations for sustainable development:participation, empowerment, and local resourcesmanagement. In: Ghai, D., and Vivian, J.M., ed.,Grassroots Environmental Action: People Participationin Sustainable Development. New York, Routledge,50–77.

62

EAST NUSA TENGGARA is a province in Indonesia withclimatic conditions distinct from other regions of thecountry. While the other regions in Indonesia mostlyhave a 6–7 month rainy season, most of the East NusaTenggara region is in drought during an 8–9 monthdry season, with heavy rainfall during the rest of theyear. The combination of climate condition andgeographical position of East Nusa Tenggara has ledto the development of a unique ecosystem dominatedby savanna and monsoon types of forest. The flora andfauna of the region are mixtures of Asian andAustralian forms.

The Province of East Nusa Tenggara has a totalland area of 4.7 million hectares, consisting of morethan 560 islands. The islands can, however, be dividedinto three major groups—Sumba Islands, the Floresgroup (Komodo, Rinca, Flores, Solor, Adonara andLembata) and the Timor group (Sabu, Rote, Semau,Timor, Alor and Pantar).

Based on the Decree of the Governor of the EastNusa Tenggara Province, Number 64 for 1996, thearea in the province declared as a public or state forestland is around 1.8 million hectares or 38% of the totalland area. However, only about 13% of the forest landis covered by primary forests, while the remaining issecondary forests, shrubs, grasslands and bare lands.Table 1 shows the classification of state forest lands.

Table 1. Classification of the state forest lands in East NusaTenggara.

Classification Area (‘000 ha)

1. Protection forests 6202. Reservation areas 102 3. National parks 1944. Other recreational parks 1655. Production forests 727Total forest lands 1808

Management and Problems of the State Forest

Institutionally, the management of state forest lands inIndonesia is under the control and responsibility ofDepartemen Kehutanan dan Perkebunan (TheMinistry of Forestry and Estate Crops or MOFEC).However, the Ministry delegates the managementcontrol over the forest land at the provincial level toKantor Wilayah Kehutanan dan Perkebunan (TheProvincial Office for Forestry and Estate Crops) in therespective provinces. The Provincial Office is mainlyresponsible for program development, coordination,monitoring and evaluation of activities, while theimplementation of the activities, at provincial orcounty levels, is carried out by Dinas Kehutanan (TheForest Service), Dinas Perkebunan (Estate CropsService), state owned enterprises and privatecompanies. It should be noted that the Provincial

Forest Land and Fire Management in East Nusa Tenggara

Slamet Riyadhi Gadas1

AbstractThe region of East Nusa Tenggara has a longer dry season compared to other regions in Indonesia.Flora and fauna in the region are mixtures of Asian and Australian types as a result of climatic andgeographic conditions. The Province of East Nusa Tenggara has more than 560 islands with a totalland area of 4.7 million ha, with about 1.8 million ha being declared as state forest land in 1996. Theforest land can be classified into conservation areas (59.8%) and production forests (40.2%). Themain problems concerning the management of forest lands in East Nusa Tenggara are claims to forestland by traditional communities, fire, shifting cultivation and grazing practices, uncontrolled treecutting, and extreme drought. With such a land composition and problems, forestry activities in EastNusa Tenggara are focused on protection and rehabilitation of conservation areas.

1Forestry Research Institute, Jl Untung Suropati, Kupang,NTT Indonesia

63

Office is an institution under the central government,while the Forest Service and the Estate Crops Servicecome under the local government.

Data in Table 1 designate almost 60% of forestedland for conservation purposes. However, most of theconservation areas are in a critical state because ofpoor agricultural practices, such as shifting cultivationin steep sites, slash-and-burn practices, over grazing,and unauthorised hunting. To prevent furtherdegradation, the forestry activities in East NusaTenggara are concentrated to protect and rehabilitateconservation areas, particularly through intensiveactivities on extension services, reforestation,breeding of endangered species, and communitydevelopment in the buffer zones.

Although the region has about 727 000 ha ofproduction forests, the yield of commercial timberfrom the forests is only 25–75 m3/ha. To preventfurther degradation of forest resources, the provincialgovernment has banned tree cutting in the stateforests. Requirements for building timber have beenfilled by bringing processed woods from otherprovinces—mainly from south and southeastSulawesi, Maluku and Irian Jaya. The annual supplyof processed woods to East Nusa Tenggara isestimated around 26 000 m3. Meanwhile, wood needsfor firewood, fencing and other purposes are expectedto be filled from private lands.

However, around 80% of the 3.5 million people inthe Province of East Nusa Tenggara are among thepoorest in the nation. Consequently, most of them areunable to purchase processed wood from otherregions. Also, the people rarely grow trees on theirown farmlands, except to supply fruit and cattlefodder. Therefore, to meet the demand on timber forbuildings, most people cut trees from the state forestswithout authorisation, causing more degradation tothe resource.

Efforts to provide timber for local needs have beenconducted by both central and provincial forestryinstitutions. In 1987 the institutions invited PerumPerhutani, a state owned forestry enterprises based inJava, to develop community forests in East NusaTenggara. Up to February 1999, using agroforestrytechniques and involving farmers from communitiessurrounding the developed forest areas, the companyhas developed 9600 hectares of plantations. Treespecies planted are teak (Tectona grandis), mahoni(Swietenia macrophylla), sandalwood (Santalumalbum), redwood (Pterocarpus sp.), and johar (Cassiasiamea). Beside trees for timber production, theproject has also developed multipurpose tree species,such as candlenut (Aleurites mollucana), cashew(Anacardium occidentalis), jackfruit (Artocarpusintegra), lamtoro (Leucaena glauca) and akasia(Acacia arabica). Between the rows of trees planted

as main crops, the farmers are allowed to grow foodcrops—mostly paddy, maize and cassava, or cashcrops, such as onion and peanut.

In 1992 PT. Fendi Hutani Lestari, a private forestcompany, obtained government approval to developtimber plantations in East Nusa Tenggara. Up toDecember 1998 the company had developed about4400 hectares of plantations with teak, mahoni,sandalwood, candlenut and cashew as main crops.

Unfortunately, the development of timberplantations in East Nusa Tenggara by the two forestcompanies is unlikely to succeed because of manysocial, economical and technical problems. BothPerum Perhutani and PT. Fendi Hutani Lestari feelthat the community does not support their presence inthe region. Many adat (traditional) community groupsclaim areas being developed by the companies. Theplantations are also growing very slowly, adding highoperational and maintenance costs, particularly forprotection of young plantations from fire, grazing, anddrought. Facing such disadvantages, the twocompanies intend to terminate their activities in EastNusa Tenggara before the end of 1999.

Fire ProblemsFire is still one of main causes of the degradation offorests in many parts of East Nusa Tenggara.According to the reports made by the ProvincialOffice, during the fiscal year of 1996–1997 about21 000 hectares of forested land caught fire, and in1997–1998 fire consumed more than 16 000 hectaresof forest area.

The most common source of fire comes from theslash-and-burn practices of farmers during thepreparation of planting sites, usually carried out fromJuly to September. The technique is chosen because itis simple and inexpensive for removing grasses,bushes and shrubs, and based on farmers’ experiencesit makes the site more productive. A simpleexperiment done by Komang Surata, a researcher ofForestry Research Institute—Kupang (FRIK),indicates that corn production from a site preparedusing a slash-and-burn technique is almost double thatof a site prepared through cuttings. Another firesource is the burning of grassland to stimulate thegrowth of cattle fodder and to make it easy forhunting.

Many efforts have made to reduce forest firesthrough various extension activities of both centraland local forestry institutions. However, probably dueto poverty and low educational conditions, theseefforts have been unsuccessful in transforming thehabits of East Nusa Tenggara’s communities in usingfire for land-clearing purposes.

64

Forestry ResearchNationally, forestry research activities funded by thegovernment are organised and controlled by BadanPenelitian dan Pengembangan Kehutanan danPerkebunan (the Forestry and Estate Crops Researchand Development Agency or FERDA), anorganisation under the MOFEC. To implementresearch activities at the regional level, FERDA isfurnished with some technical implementation units.One of the unit is Balai Penelitian Kehutanan Kupang(Forestry Research Institute—Kupang or FRIK),which is responsible for implementing forest researchactivities in the provinces of Bali, West NusaTenggara, East Nusa Tenggara, East Timor and thecounty of South East Maluku.

Considering the problems arising in its servicingregion, FRIK has established three primary researchprograms: • sustainable management of forest resources;• natural resources conservation;• forestry for sustainable rural development.

The program on sustainable management of forestresources emphasises rehabilitation techniques forconservation areas and production forests, anddevelopment of seed stand models for marketablelocal tree species improvement in both natural andman-made forests. The natural resources conservationprogram focuses on the development of varioustechniques to maintain and rehabilitate the forestecosystem and its biodiversity, for instance throughbreeding local flora and fauna, developingmanagement techniques of buffer zones, anddeveloping collections of medicinal plants.

The program on forestry for sustainable ruraldevelopment concentrates on finding methods toempower and improve the prosperity of ruralcommunities living inside and surrounding the stateforest areas, including methods to improve utilisationof non-timber forest products. To support its programactivities, FRIK has established three researchstations, on the islands of Lombok, Sumba and Timor,and several experimental gardens across the servicedregion.

ConclusionsIt can be concluded that the primary problems facedby both central and local forestry institutionsconcerning the management of forest land in EastNusa Tenggara are:• claims to forest lands by customary or traditional

communities;• forest fires, due to slash-and-burn practices;• uncontrolled shifting cultivation practices on steep

sites;• grazing practices in forest areas;• unauthorised tree cutting, in both primary and

secondary forests;• the extreme dry season, leading to low survival

rates out of planting activities.

65

FIRE HAS BEEN used by humans for tens of thousandsof years in eastern Indonesia and northern Australiafor a wide variety of purposes including to clearundergrowth, hunt game, engage in warfare andsignal presence. The landscapes we see today in theregion have undoubtedly been largely formed andmaintained by this all-pervasive use of fire. Regularfire tends to act against trees and in favour of grasses,leading to the conversion of tropical forest intosavanna and grassland (Freifelder et al. 1998). Thethreatened Imperata cylindrica (alang-alang or bladygrass) grasslands of Australia’s Cape York Peninsula(Neldner et al. 1997) and the economically destructivegrasslands featuring the same species in tropical Asia (D’Antonio and Vitousek 1992) are probablyboth the result of this long-term occurrence ofanthropogenic fires.

Against this background of fire, we have seen as aresult of human activity in recent times a sharplyescalating movement of plant species into new regionsof the globe, at rates unprecedented in geologicalhistory. This movement, both deliberate andaccidental, has resulted in many thousands of speciesbecoming permanently established in biogeographicregions that they could not otherwise have reached.

The role and behaviour of fire in tropical landscapesis rapidly changing, driven by changing land usepractices and by the increasing presence of invasivealien plant species. We will discuss the nature andimplications of these changes in this paper.

Fire as a DisturbanceIn relatively intact forest and woodland shading playsan important role in limiting the ability of grasses andshrubs to establish beneath the tree canopy (Gentleand Duggin 1997). Low-intensity fires reduce thebiomass of the natural shrubby layer, creatingopportunities for fire-tolerant alien species to exploitthe conditions of increased light availability on theforest floor (Duggin and Gentle 1998) while alsoincreasing the availability of nutrients. High-intensityfires create even more opportunities for invasive plantspecies such as Lantana camara (Duggin and Gentle1998; Fensham et al. 1994), I. cylindrica (Eussen andde Groot 1974) and Chromolaena odorata (Gautier1996) to move into forested areas by opening gaps inthe canopy layer.

While it is thought that much tropical grassland ismaintained by long-term traditional burning regimes,a subsequent reduction in the frequency or intensity offire can allow re-invasion by fire-susceptible shruband tree species. The change in land managementpractices over much of northern Australia fromtraditional aboriginal use to cattle grazing has had this

Fire and Weeds: Interactions andManagement Implications

Colin Wilson1 and Wayan Mudita2

AbstractThe landscape of eastern Indonesia and northern Australia has undoubtedly been largely shaped byfire. Frequent burning tends to convert forest into savanna and grassland by selecting against fire-sensitive trees and favouring fire-resistant grasses and shrubs. Fire can act as a disturbance, creatingopportunities for weeds to invade. A reduction in fire can enable the reversion of grassland towoodland. Northern Australia has seen the relatively recent arrival of fire-resistant, giant, perennial,African grasses (e.g: Pennisetum polystachion, Andropogon gayanus and Panicum maximum), andeastern Indonesia has been invaded by a fire-resistant, scrambling, American shrub (Chromolaenaodorata). These species interact with fire to create positive feedback loops that maintain and extendtheir dominance over the native flora. The implications of these introduced species for managementare discussed.

1Parks and Wildlife Commission of the Northern Territory,PO Box 496, Palmerston, NT 0831, Australia2Nusa Cendana University, Centre for Research on theEnvironment and Natural Resources, Jalan Adisucipto, Penfui,Kupang 85001, NTT, Indonesia

66

effect. Grazing by cattle reduces the available fuelload, pastoral land managers burn less frequently thantheir aboriginal predecessors to avoid destroyingpasture, and they tend to light fires in the early dryseason when the moisture content of grasses is stillquite high. The resultant fires are smaller and patchier,giving tree seedlings and suckers an opportunity tobecome established (Neldner et al. 1997). Thesubsequent shading further reduces the amount offlammable grasses. This reduction in burning ofsavanna has in part enabled Cryptostegia grandiflora(rubber vine), a fire-sensitive species, to become aserious weed in northern Queensland (Grice 1997).

Weeds as a DisturbanceThe invasion of fire-resistant woody species intograssland can suppress existing fire regimes(Lonsdale and Miller 1993; Mack and D’Antonio1998) by reducing the biomass of flammable grassesin the understorey. But a more common andecologically and economically far more significantphenomenon is the invasion of fire-tolerant alienweeds that leads to an increase in the frequency and/orintensity of fire in an ecosystem. These plants aresometimes dubbed ‘fire weeds’ for their promotion ofpositive feedback loops between fire and weedinvasion. As the weed becomes more abundant,increasingly intense fires become more common. Thefire-sensitive native species begin to decline, enablinga further increase in the alien invader that leads to aneven greater probability of intense fire, and so on(D’Antonio and Vitousek 1992; James 1995; Mackand D’Antonio 1998). This process transforms non-flammable, mostly native forest and woodland intohighly flammable exotic grassland with very littlebiological diversity (D’Antonio and Vitousek 1992;Freifelder et al. 1998).

Northern AustraliaThe savannas that occupy much of northern Australiahave been, and continue to be, burnt regularly. Theleaves of trees in these Eucalyptus-dominated, fire-maintained ecosystems are not highly flammable,native grass fuel does not tend to accumulate andhence the intensity of fire is generally low. Fire rarelycarries into the canopy and few mature trees are killed(Lonsdale and Braithwaite 1991).

In relatively recent times, a new phenomenon hasbeen superimposed upon this pre-existing cycle. Giant,perennial, tussock-forming grass species of mainlyAfrican origin have been deliberately introduced intonorthern Australia as fodder for cattle and many havespread beyond intensively grazed pastures (Freifelderet al. 1998; Humphries et al. 1991; Latz 1991;Lonsdale 1994; Low 1997; Macdonald and Frame

1988). Grasses in Africa have co-evolved withhominids and anthropogenic fires for millions of years(D’Antonio and Vitousek 1992). Hence they showadaptations conferring resistance to fire, such as theability to resprout from underground storage organsand to germinate and grow rapidly following fire(Freifelder et al. 1998).

This invasion of perennial African pasture grassesis beginning to alter the scale of fire in northernAustralian savannas. These grasses produce muchgreater flammable biomass than the native grassesthey displace, they cure later in the season whengeneral moisture levels are lower, and they carryflames into the tree canopy (Low 1997). Cenchrusciliaris (buffel grass), for instance, produces 2–3 timesas much flammable material as native grasses incentral Australia (Latz 1991) and has been observed tolead to progressive destruction of dry rainforestremnants as it carries successive hot fires into theedges (Fensham 1996). Pennisetum polystachion(mission grass) produces much greater fuel loads thanthe native annual grasses it replaces (Macdonald andFrame 1988; Panton 1993) and has been blamed for asignificant decline in monsoon rainforest aroundDarwin (Panton 1993). Andropogon gayanus (gambagrass) produces 3–5 times as much flammablebiomass as native grasses in the Top End of theNorthern Territory (Cook 1991) and it is feared that itcould lead to widespread conversion of native savannawoodland into exotic grass monoculture. OtherAfrican species implicated in increased intensity ofwildfires include Brachiaria [=Urochloa] mutica(para grass) on seasonally inundated wetlands, whichis blamed for the disappearance of some floodplain-fringing monsoon rainforest patches in KakaduNational Park (P. Barrow pers. comm.), and Panicummaximum (Guinea grass) in dryland savanna innorthern Queensland.

Eastern IndonesiaWhile northern Australia is experiencing an invasionof African grassy ‘fire weeds’, eastern Indonesia isfacing invasion by an American scrambling shrub,with similar consequences for fire regimes.Chromolaena odorata (Siam weed) is a fast-growingperennial shrub that forms dense tangled thickets2–3 m high in open country, but with the ability toclimb to about 6 m over other vegetation. It dies backfollowing flowering and the dry stems then burnreadily, creating a serious fire hazard. With the firstrains following fire, new shoots rapidly appear fromthe root crown or from undamaged axillary buds, andseeds freely germinate (McFadyen 1989;Tjitrosoedirdjo et al. 1991).

C. odorata was probably introduced into Asia forits ability to suppress grasses such as I. cylindrica(D’Antonio and Vitousek 1992; de Rouw 1991) and is

67

now widespread in Indonesia, having progressivelyinvaded from west to east (Sipayung et al. 1991).Eastern Indonesia has large areas of I. cylindricagrassland, maintained by fire but consideredeconomically destructive. Following fire, C. odoratais able to overgrow the grass layer and eliminate it byshading (Eussen and de Groot 1974). Some farmersconsider C. odorata desirable as it is easier to controlthan I. cylindrica and is believed to improve soilquality (Tjitrosoedirdjo et al. 1991), but it is not eatenby cattle and has greatly reduced the availability ofpasture (Sipayung et al. 1991). It is especially aproblem in plantation crops (de Rouw 1991;Tjitrosoedirdjo et al. 1991) by smothering seedlingtrees and carrying intense fire into the tree canopy.

C. odorata cannot grow in the shade under an intactclosed forest canopy, but it forms dense stands inriparian vegetation and along the edges of forestpatches. Intense fires are carried from savanna andgrassland into the forest canopy, causing forestmargins to retreat (Macdonald and Frame 1988). Inshifting slash-and-burn agriculture, C. odorata invadesthe newly abandoned fields causing a halt to thenormal succession back to secondary forest (de Forestaand Schwartz 1991; de Rouw 1991). The result isdomination of the rural landscape by dense thickets ofthe weed, and a dangerous fire hazard for villages.

Management ImplicationsThe relatively recent arrival and rapid spread of exoticfire weeds in the region has profound implications fortraditional and current fire practices. These alienspecies heighten the likelihood of intense fires, fuelledby a far greater flammable biomass than before,occurring later in the season and carrying flameshigher into the tree canopy. In the absence of activemanagement, the prospect is for conversion ofbiologically diverse and productive forest, woodlandand savanna into biologically impoverished grassland(northern Australia) or shrubland (eastern Indonesia).

A biological control program, funded by theAustralian Centre for International AgriculturalResearch (ACIAR) is attempting to reduce thedominance of C. odorata over other vegetation inIndonesia (McFadyen 1998) and may ameliorate theweed–fire feedback cycle. It is important to try toprevent the further spread of this weed in easternIndonesia and into northern Australia.

The problems caused by perennial African grassesare widespread across the world’s tropics (D’Antonioand Vitousek 1992). To remain consistent with the newparadigm of ‘ecologically sustainable development’,we should attempt to prevent their further spread, as

extensive use of such species is not ecologicallysustainable. Research is desperately needed into waysof managing existing infestations, especially in relationto fire, to minimise damage to infrastructure,agricultural production and the environment.

ReferencesCook, G.D. 1991. The effects of fire on nutrient losses from

Top End savannas. In: Moffat, I. and Webb, A., ed.,Conservation and Development Issues in NorthAustralia. Darwin, NARU, 123–129.

D’Antonio, C.M. and Vitousek, P.M. 1992. Biologicalinvasion by exotic grasses, the grass/fire cycle, andglobal change. Annual Reviews of Ecology andSystematics, 23, 63–87.

de Foresta, H. and Schwartz, D. 1991. Chromolaena odorataand disturbance of natural succession after shiftingcultivation: an example from Mayombe, Congo,Central Africa. In: Muniappan, R. and Ferrar, P., ed.,Proceedings of the Second International Workshop onBiological Control of Chromolaena odorata, Bogor,Indonesia, 4–8 February 1991. BIOTROP SpecialPublication, 44, 23–41.

de Rouw, A. 1991. The invasion of Chromolaena odorata(L.) King & Robinson (ex Eupatorium odoratum), andcompetition with the native flora, in a rainforest zone,south-west Côte d’Ivoire. Journal of Biogeography, 18,13–23.

Duggin, J.A. and Gentle, C.B. 1998. Experimental evidenceon the importance of disturbance intensity for invasionof Lantana camara L. in dry rainforest–open forestecotones in north-eastern NSW, Australia. ForestEcology and Management, 109, 279–292.

Eussen, J.H.H. and de Groot, W. 1974. Control of Imperatacylindrica (L.) Beauv. in Indonesia. MededelingenFakulteit Landbouw-wetenschappen Gent, 39, 451–464.

Fensham, R.J. 1996. Land clearance and conservation ofinland dry rainforest in north Queensland, Australia.Biological Conservation, 75, 289–298.

Fensham, R.J., Fairfax, R.J. and Connell, R.J. 1994. Theinvasion of Lantana camara L. in Forty Mile ScrubNational Park, north Queensland. Australian Journal ofEcology, 19, 297–305.

Freifelder, R.R., Vitousek, P.M. and D’Antonio, C.M. 1998.Microclimate change and effect on fire followingforest–grass conversion in seasonally dry tropicalwoodland. Biotropica, 30, 286–297.

Gautier, L. 1996. Establishment of Chromolaena odorata ina savanna protected from fire: an example from Lamto,central Côte d’Ivoire. Proceedings of the ThirdInternational Chromolaena Workshop, Abidjan, Côted’Ivoire, November 1993. University of GuamPublication No. 202, 54–67.

Gentle, C.B. and Duggin, J.A. 1997. Lantana camara L.invasions in dry rainforest–open forest ecotones: therole of disturbances associated with fire and cattlegrazing. Australian Journal of Ecology, 22, 298–306.

68

Grice, A.C. 1997. Post-fire regrowth and survival of theinvasive tropical shrubs Cryptostegia grandiflora andZiziphus mauritiana. Australian Journal of Ecology, 22,49–55.

Humphries, S.E., Groves, R.H. and Mitchell, D.S. 1991.Plant invasions: the incidence of environmental weedsin Australia. Kowari, 2, 1–134.

James, D. 1995. The threat of exotic grasses to thebiodiversity of semiarid ecosystems. AridlandsNewsletter, 37, 6–7.

Latz, P.K. 1991. Buffel and couch grass in central Australiancreeks and rivers. Newsletter of the Central AustralianConservation Council Inc., April 1991, 5.

Lonsdale, W.M. 1994. Inviting trouble: introduced pasturespecies in northern Australia. Australian Journal ofEcology, 19, 345–354.

Lonsdale, W.M. and Braithwaite, R.W. 1991. Assessing theeffects of fire on vegetation in tropical savannas.Australian Journal of Ecology, 16, 363–374.

Lonsdale, W.M. and Miller, I.L. 1993. Fire as a managementtool for a tropical woody weed: Mimosa pigra in northAustralia. Journal of Environmental Management, 33,77–87.

Low, T. 1997. Tropical pasture plants as weeds. TropicalGrasslands, 31, 337–343.

Macdonald, I.A.W. and Frame, G.W. 1988. The invasion ofintroduced species into nature reserves in tropicalsavannas and dry woodlands. Biological Conservation,44, 67–93.

Mack, M.C. and D’Antonio, C.M. 1998. Impacts ofbiological invasions on disturbance regimes. Trends inEcology and Evolution, 13, 195–198.

McFadyen, R.E. Cruttwell, 1989. Siam weed: a new threat toAustralia’s north. Plant Protection Quarterly, 4, 3–7.

McFadyen, R.E. Cruttwell, 1998. The ACIAR project for thebiological control of Chromolaena odorata: futuredevelopments. Proceedings of the Fourth InternationalWorkshop on Biological Control and Management ofChromolaena odorata, Bangalore, India, October 1996.University of Guam Publication No. 216, 115–118.

Neldner, V.J., Fensham, R.J., Clarkson, J.R. and Stanton,J.P. 1997. The natural grasslands of Cape YorkPeninsula, Australia: descriptions, distribution andconservation status. Biological Conservation, 81,121–136.

Panton, W.J. 1993. Changes in post World War IIdistribution and status of monsoon rainforests in theDarwin area. Australian Geographer, 24, 50–59.

Sipayung, A., Desmier de Chenon, R. and Sudharto, P.S.1991. Observations on Chromolaena odorata (L.) R.M.King and H. Robinson in Indonesia. In: Muniappan, R.and Ferrar, P., ed., Proceedings of the SecondInternational Workshop on Biological Control ofChromolaena odorata, Bogor, Indonesia, 4–8 February1991. BIOTROP Special Publication 44, 43–50.

Tjitrosoedirdjo, S., Tjitrosoedirdjo, Sri S. and Umaly, R.C.1991. The status of Chromolaena odorata (L.) R.M.King and H. Robinson in Indonesia. In: Muniappan, R.and Ferrar, P., ed., Proceedings of the SecondInternational Workshop on Biological Control ofChromolaena odorata, Bogor, Indonesia, 4–8 February1991. BIOTROP Special Publication 44, 57–66.

69

THE PROVINCE of East Nusa Tenggara (NTT), locatedin the southeastern part of Indonesia, experiences along dry season period and receives its annual rainfall(less than 1500 mm) from December to March. Theeconomy of NTT depends on agriculture. It isestimated that some 38% of regional gross domesticproduct (RGDP) comes from agriculture and about75–80% of the labour force is employed in agriculture.

Livestock has been raised in the area for centuries,as can be observed from the social and traditionallives of the local villagers. It was estimated in that46% of the total 550 000 farmers’ families raisedlivestock. Most of the small livestock (pigs, goats andchicken) are for local consumption, while most of thelarge livestock (cattle, buffaloes and horses) are forinter-island trade. Bali cattle are the predominantbreed in West Timor with estimated population isaround 600 000 head, whereas Ongole cattle in SumbaIsland total just 40 000 head because their repro-ductive capacity is lower than Bali cattle (Wirdahayatiand Bamualim 1990). The cattle industry is one of themain contributors to the economic structure of NTT,especially Bali cattle. Although village goats, pigs andpoultry have a valuable role in rural communities,much of the cash flow is generated from cattle sales.

Livestock are raised by the extensive system basedon communal grazing areas. Fire management on

native grasslands is not significantly apparent in thearea, except in Sumba Island during the dry season.This paper will briefly discuss some aspects of therole of livestock, the production systems and to someextent the fire management.

The Role of LivestockThree factors primarily stimulated the development ofthe livestock industry in NTT, particularly largeruminants: 1) the availability of adequate land forgrazing, 2) the availability of the export market forlive cattle and buffalo in Hong Kong in the 1970s, and3) the existence of substantial amounts of leucaena(Leucaena leucocephala) in West Timor.

A hypothetical estimation made by Ayre-Smith(1991) shows that 55% of cattle and buffalo ownersown only 1–2 head and 32.5% of owners own 3–10head of these animals, but he also estimated thatalmost 50% of cattle and buffaloes were owned byless than 5% of the livestock holders. A survey madein eight villages in two districts of West Timorshowed that 59% of farmers in the survey areas ownon average 2.4 head of cattle plus a few chickens, pigsand goats per family (Tjaong-Soka et al. 1991),indicating that livestock are an important part offarmer livelihood in NTT.

A study conducted in Kupang district showed thatfarmer income from the livestock component wassignificant but varied widely depending on the agro-ecosystems (Momuat and Bamualim 1994)—see

Livestock Production and Fire Managementin East Nusa Tenggara

Abdullah Bamualim1

AbstractMost of East Nusa Tenggara (NTT) experiences a pronounced dry season lasting 8–9 months eachyear with an annual rainfall of less than 1500 mm. Traditionally, livestock has been important in thearea for social status and to secure the income of farmers. The livestock in West Timor comprises600 000 head of Bali cattle while Sumba Island has around 40 000 head of Ongole cattle. Farmers alsoraise buffaloes and horses and keep small animals (pigs, goats and native chickens) as their mainsource of animal protein and income. Around 70% of lands in Nusa Tenggara are classified as grazingland. It is estimated that more than 90% of breeding animals in NTT are raised extensively oncommunal grazing land with little input from livestock owners. Fire management on native grasslandsis not significantly apparent in the area, except in Sumba Island during the dry season. This paperbriefly discusses some aspects of the role of livestock, the production systems and to some extent thefire management.

1Balai Pengkajian Teknologi Pertanian (BPTP) Naibonat. Jln.Tim Tim km.32, P.O. Box 1022, Kupang 85000, NTT,Indonesia.

70

Table 1. The data indicated that livestock was moreapparent in the drier regions or on marginal soil typesthan on the better soil types, as demonstrated byfarmers living on the Bobonaro clay zone. Thedifficult environment for growing productive cropsmight be the main reason why farmers rely more ontheir incomes from livestock and off-farm jobs thanon annual food crops.

It is estimated that the livestock subsectorcontributes some 12% of the RGDP, derived primarilyfrom the annual export of 70–80 thousand head ofcattle from the area, plus 25% of total cattle andbuffalo turnoff from NTT for local consumption.Buffalo and horses are used widely for draught andtransportation purposes. About 12 000 buffalo and6000 horses from NTT are sent to Java every year forslaughter.

Small animals such as poultry, goats and pigs alsoplay an important role as sources of income forfarmers and for human consumption in the villages.Most village families in NTT will own 2–3 pigs orgoats, and a few chickens. Sheep are not favoured inTimor Island, because of their role in spread ofmalignant catarrhal fever (MCF) to which Bali cattleare highly susceptible.

Livestock Production Systems

The extensive systemIt was estimated that around 70% of lands in west andeast Nusa Tenggara are classified as suitable forgrazing animals (Hasibuan and Mangunsong 1993). Ingeneral, the farmers communally use the grazing landfor large ruminants with an extensive system.However, grazing potential is constrained by lack ofwater and in some areas also by fires that may causedeterioration of native grazing land. Land degradationaround the water points and the expansion ofunwanted weeds due to post-burn overgrazing havebeen apparent in some places in West Timor.Nevertheless, at present the extent to which grazingland has deteriorated due to grazing pressure andfire–grazing interaction is unknown.

With the extensive system, grazing animalsexperience a significant body weight gain during thewet season and a body weight loss during the dryseason. Although average production per year is quite

low, it is believed that more than 90% of breedinganimals in NTT are raised extensively on communalgrazing land. Low production cost is the prime reasonthat such a system is widely practiced in the area.

An estimation of carrying capacity of differentislands in NTT is shown in Table 2. The data indicatethat carrying capacity in West Timor may havereached its climax. In this area, therefore, there is anurgent need to improve the management systems ofgrazing animals in some critical areas, toward anintensive system where sufficient feed reserves aredeveloped to guarantee the quantity, quality andcontinuity of feed supplies. In Sumba and Floresislands, on the other hand, there is still potential toincrease productivity of grazing animals throughsustainable utilisation of grazing land and by theprovision of adequate water resources.

The intensive systemFeed availability throughout the year is the keyvariable for livestock productivity and willsignificantly improve farm income. Tethered animalswith cut-and-carry feeding systems are usual in suchareas. In general, the productivity of animals raisedunder an intensive system is much better than thoseraised under the extensive system.

One of the intensive systems practiced by somefarmers in Timor Island is the Amarasi system wherefarmers intercropped their land with leucaena(Leucaena leucocephala) and maize. When leucaenawas in full production, the Amarasi system was aclassic example of how some dryland farmers wereable to fatten 4–5 cattle per family at any one time.Store animals were purchased from other parts ofTimor and fattened on a diet consisting mainly ofleucaena, producing high liveweight gains. The mostserious criticism of the Amarasi system is its totaldependence on a single species of tree legume. The

Table 1. The proportion of farmers’ income derived fromfood crops, livestock and off-farm jobs on different soiltypes in Timor (values in percentage).

Soil types Food crops Livestock Off-farm

Alluvial 53 26 21Mediterran* 79 14 7Bobonaro clay 16 50 34

*Calcareous parent material with dark reddish brown solum

Table 2. Estimated total land area, land used for livestock, number of grazing animals and carrying capacity in NTT.

Land used Number of CarryingLocations/ islands Land area (ha) for livestock (ha) animal units (AU) capacity (ha/AU)

Sumba 1085440 770 600 145 960 5.3Flores/Alor 1 909 500 406 170 129 630 3.1Timor 1 699 060 705 040 537 110 1.3Total/average 4 694 000 1 475 680 812 700 1.8

71

infestation of psyllid insects (Heteropsylla cubana) onleucaena indicated that monoculture productionsystems are high risk for small farmers. Therefore amore diverse plant community or the restriction ofcattle number, rather than the promotion of leucaena,was the key to the success of the Amarasi system.

The Sikka land-use system is a well known methodof controlling soil erosion by planting a densevegetative barrier, consisting mainly of leucaena,along the contour. The system was developed in theSikka district in eastern Flores, based on conceptsfrom the Amarasi system where intercropping annualcrops with perennial crops is dominant. It isinteresting to note that the productivity of Bali cattleraised under coconut plantations in Sikka was the bestof any survey sites in NTT (Bamualim et al. 1994).

Fire Management on Grazing AreasIt appears that grazing potential in NTT is constrainedby lack of water and in some areas also by fires thatmay cause deterioration of native grazing land,especially on Timor Island. Fire during the dry seasonhas been apparent in some grazing areas on Sumbaand Timor islands. There are two main causes of fire:1) purpose-burnt by farmers to have new grass grown,and 2) accidental burning by surrounding farmers,passing travellers or by heat from the sun.

There is not much information available on theeffect of fire on native grass production andcomposition in NTT. A study conducted in SumbaIsland indicated that fire did not change the nativegrass production and composition (Subandi et al.1998). Native grass production data gathered over a 2-year period on burnt and unburnt areas in Sumba areshown in Table 3. Although grass production did notsignificantly differ between burnt and unburnttreatments, there was a tendency for the unburnttreatment to yield slightly higher than the burnttreatment. However, the trial was conducted withoutthe presence of grazing animals, therefore the biomassproduction of unburnt grass was higher due to theinclusion of the dead material. The study alsoindicated that burning did not influence the nativegrass species in the area; most of the grass speciesreported by Hoekstra (1948) still exist on the trial site.However, the appearance of some weeds such asChromolaena odorata which colonise quite largeareas of grazing land, has limited the grasslandresource in the region. Chromolaena is easilycontrolled by fire and therefore favoured by theshifting cultivation farmers. Nevertheless, theincidence of chromolaena fire in the grazing area isuncommon.

ConclusionsThe role of livestock in the economic structure ofNTT province is significant. Because of the dryclimate, livestock production has augmented thelower income received by most farmers. Some 90% oflarge animals are raised in the extensive grazingsystem in NTT. Nevertheless, an intensive fatteningprogram for Bali cattle has been initiated in theAmarasi area for the export market since the 1970s,due to the presence of leucaena.

Fire, both intentional and unintentional, has beenapparent in the area, particularly on Sumba Island. Atthe present time, the benefit of fire for grazing area inNTT is not fully recognised. Therefore, further studiesare needed to find ways to improve grassland qualitythrough the use of fire, grazing management and theintroduction of better forage species in the area.Furthermore, a few innovative systems of land useinvolving tree legumes have been developed in theregion. Farmers in Amarasi and Sikka havedemonstrated the merits of these systems. It appearsthat there is a need to intensify the livestockproduction in the area through the improvement offeed availability and quality, livestock distributionschemes and better management systems.

ReferencesAyre-Smith, R. 1991. Livestock development in NTT. In:

Barlow, Colin, Bellis, Alex and Andrews, Kate, ed.,Nusa Tenggara Timur: The Challenges ofDevelopment, Dept. of Political and Social Change,Research School of Pacific Studies, The AustralianNational University, Canberra, 85–104.

Bamualim, A., Saleh, A., Fernandez, P.Th., dan Liem, C.1994. Produksi dan kualitas hijauan rumput alamsebagai makanan ternak sapi di Nusa Tenggara. In: TheFinal Seminar of the Cattle Health and ProductivitySurvey (CHAPS). Held at the Disease InvestigationCentre, Denpasar-Bali, on May 15–17, 1994. CHAPSBook A, 202–229.

Hasibuan, D.B. dan Mangunsong, R.D. 1993. Peluanginvestasi sub-sektor peternakan dalam pengembangandaerah lahan kering di Nusa Tenggara. Dalam‘Prosiding Lokakarya Status dan Pengembangan LahanKering di Indonesia’. Mataram, 16–18 November 1993,Buku I. (Penyunting O.E. Momuat, A.H. Taryoto,

Table 3. Native grass production with burnt* and unburnttreatments in Sumba Island (value in tonnes dry matter/ha).

Parameters First harvest Second harvest (March 1998) (March 1999)

Unburnt 3.25 3.66 Burnt 2.71 3.12

* Burnt in October 1997 and October 1998

72

Ch.J.S. Momuat dan D. Sitepu). Proyek PembangunanPenelitian Pertanian Nusa Tenggara (P3NT), 256–276.

Hoekstra, P. 1948. Paardenteelt op het Eiland Soemba.Drukkerij V/H John Kappe–Batavia–C, 190 p.

Momuat, E.O. and Bamualim, A. 1994. Crop–livestockfarming systems in the uplands of Indonesia. Paperpresented at ‘The International Seminar on UplandFarming Systems and the Role of Livestock Production’,Chiangmai, Thailand, November 21–27, 1994.

Subandi, Triastono, J., Budisantoso, E. dan Bamualim, A.1998. Pengelolaan padang rumput untuk konservasilahan dan perbaikan kualitas pakan ternak DASKambaneroe Kabupaten Sumba Timur. Prosiding

Lokakarya Nasional Pembahasan Hasil PenelitianPengelolaan Daerah Aliran Sungai. Bogor, 27–28Oktober 1998, 375–390.

Tjaong-Soka, E., Mozes, P.J., Bunga, J.M., Arianto, H. danSchottler, J.H. 1991. Perencanaan PembangunanPeternakan di Nusa Tenggara Timur. KerjasamaPropinsi NTT dengan AIDAB. Kupang, Juni 1991.

Wirdahayati R.B. dan Bamualim, A.1990. Penampilanproduksi dan struktur populasi ternak sapi Bali di NusaTenggara Timur. Proceedings Seminar Nasional SapiBali. Denpasar-Bali, Tanggal 20–22 September 1990,C 1–5.

73

THE PROVINCE of Nusa Tenggara Timur is one of the27 provinces in Indonesia at this time. It is relativelyremote and impoverished compared to the otherprovinces, but this does not necessarily mean thatdevelopment is lacking or neglected. Manydevelopment programs have been successfully carriedout since the colonial period, when resident J.A.Hazaart pioneered an integrated physical and socio-cultural development in West Timor in 1810–1832.His pioneering steps in integrated development wereunknowingly followed and spread all over theprovince by governor El Tari in 1965–1978. Despitethe efforts of all development programs, poverty stillremains the province’s most important challenge.

Geographically, Nusa Tenggara Timur is anarchipelago of semi-arid islands dominated bysavanna. Most of the more than 3.6 million people ofthe province engage in peasant agriculture. Their wayof life is influenced by the complex conditions of thephysical and the socio-cultural environment of theisland. These conditions need thorough study to gain aclear view of the hampering and constraining factorsthat must be carefully addressed in designing specialprograms for development.

The peasants have used fire as the main means andtool of their swidden agricultural activities since timeimmemorial. Probably they have acquired various

traditional wisdom and knowledge about fire fromtheir ancestors, but research into this challengingaspect is lacking. The research is needed because eachof the approximately 50 ethnolinguistic groups ofNusa Tenggara Timur may have developed their owndistinctive wisdom and knowledge in their owntraditional habitats.

Scattered and rare written materials on fire andtraditional knowledge about fire are available butinsufficient. I have collected some papers on fire fromvarious parts of Nusa Tenggara, but they are toogeneral to incorporate into this paper. This is thereason why I finally decided to rely on my personalearly learning experience and my observations towrite this paper.

The main material for the first part of this paper isderived from my personal learning experience. I haveundergone a traditional informal education andpractical training as a peasant’s son in my nuclear andextended family within my Meto ethnolinguisticgroup before joining the elementary school. Anendurable discipline to avoid and or to prevent fire hasbeen instilled in my character. I found out later thatsimilar discipline has also permeated the character ofother Meto children. The success of the Meto way ineducating and training on fire avoidance andprevention is clearly shown through a reality that thecatastrophic fires seldom destroy the Meto fire-pronehouses and villages. But its weakness is that thechildren are not educated and trained properly toavoid and or to prevent fires in the grasslands, bushes

Fire, Traditional Knowledge, and CulturalPerspectives in Nusa Tenggara Timur

Hendrik Ataupah1

AbstractThis paper is not based on scientific research. The main material for the first part is derived from myearly informal education and practical training as a son of a peasant within the Meto ethnolinguisticgroup in West Timor. This early learning in my nuclear and extended family has been enriched by mypersonal observations on the kind of informal education and practical training given to young Metogenerations. The second or the last part of this paper is based on my observation on cultural issues inmany parts of Nusa Tenggara Timur, and my personal view on the observed issues.

1Faculty of Social and Political Sciences,Senior Staff ofEnvironmental and Natural Resources Research Centre NusaCendana University

74

and forests that are considered as common accessedproperties.

In the second part of this paper I attempt tosummarise my personal observations on the use of firein agricultural activities in various parts of NusaTenggara Timur. I have had various chances to visit16 of the more than 30 inhabited islands during thelast 35 years. I have visited more than 1400 villagesinhabited by peasants, who use fire in the swiddenagricultural activities. Fire is used for huntingpurposes at the Bajak peninsular at the northeasternpart of the island of Lembata in East Flores. Fire isused at Sumba and the northwestern part of WestTimor for intensive animal husbandry purposes, andpeasants on the island of Alor have used fire over aperiod of 8 years to control rats. The fires mentionedabove may run out of control and then ruin thousandshectares of accumulated dry materials in grasslands,bushes, village forests, and protected forests in theneighbouring areas.

On the other hand, in the densely populated areassuch as in the southeastern part of West Timor, thesouthern part of the island of Adonara, at districts ofSikka in Flores and West Sumba, outbreaks ofcatastrophic fires seldom occur. The increase ofpopulation as a whole and the increase of the numberof peasants will surely have both positive and negativeimpacts on fire in the future. Thirty years ago thedanger and the risks of fire were commonplace inmany areas of the district of Ngada in Central Flores.Nowadays, fire stricken areas at Ngada are convertedinto newly irrigated rice fields and dry farming areas.Many areas in the neighbouring district of Manggaraiare converted into a kind of agro-silvipastoral system.Those agro-silvipastoral areas were unintentionallycreated in the formerly traditional swiddenagricultural areas as well as in poorly managedprotected forests.

Scientific research on fire in the agro-silvipastoralareas is worth undertaking in the future. This researchhas to be centered on the concepts of fire, grasslands,bushes, and forests. Local languages and cultures havetheir own concepts on the above-mentioned land uses.Local systems of land tenure and land rights, whichinfluence those concepts, are worth investigatingthrough such research.

Fire and Traditional Knowledge

The use of fire according to folktalesThe forefathers of the Meto ethnolinguistic group inWest Timor were gatherers and hunters. According tovarious folktales they had to live on raw food becausethey didn’t know the existence of fire and its use toprocess their food.

Once upon a time, the eldest son of the foundingforefather of the Pit-Ay clan killed a civet cat on hishunting expedition near the top of Mount Mutis. Atthe time he was ready to eat the raw meat, a youngTetun princess mysteriously appeared in front of him,and prevented him from doing so. She kindlyinstructed and assisted him to make fire for the firsttime, and let him use the fire to burn the raw meat toobtain a more delicious food. She also taught thehunter’s companion to set fire to the extendedsurrounding grasslands to sustain the availability offire. She instructed them to spread fire toward thecoastal areas of Timor, so that other Meto huntinggroups could use fire for their hunting activities andfood processing.

Several generations later, a group of technologicallyand socially more advanced newcomers defeated theMeto gatherers and hunters, and taught them to use fireto perform swidden agriculture for securing their foodsupplies. Since then, fire has been used both forhunting and agriculture, and also for various domesticuses.

The image and the characteristics of fireThe Meto ethnolinguistic group does not have a cleardefinition of fire. But the people are sure that fire is anatural phenomenon that sometimes starts accidentallyby lightning at the beginning of the rainy season.People have also kindled fires since their forefatherslearnt fire-starting techniques from foreign new-comers.

Fire is figuratively considered and handled as aliving creature that eats, licks, grasps, crawls, leapsand jumps, becomes angry, and so on. Its desired foodconsists of all kinds of dry flammable materials. Itbecomes a greedy and unfriendly creature when mancarelessly supplies its food. On the other hand, itbecomes a reliable servant or an enjoyable pet if itsfood is delivered restrictively and appropriately.

Fire is the only natural phenomenon that comprisesfour risky but beneficial characteristics—heat, light,smoke, and smell. They are distinguishable from eachother but are always closely related.

Fire’s heat is its pioneering and main characteristic,and the only characteristic with the capacity to burnall dry flammable materials. The starting of a new fireis impossible unless the heat is there, and this isimmediately followed by the other threecharacteristics. Each traditional Meto child is trainedand tirelessly reminded that there is a built-in dangerof fire at every moment after the creation of the heatfor a burning process to satisfy any needs. The risksfrom a wild fire can be detected early from afar by oneor more of the five senses of man, but the destructiveheat is practically undetectable.

75

The Avoidance and Prevention of Fire and the Concepts of Burning

Focus of traditional informal education andpractical training about firePrior to their full awareness of its dangers and risks,small children are forbidden to play with fire. As theygrow children gradually learn to avoid and/or toprevent fire. Teaching is undertaken when thechildren assist their parents and/or elder siblings inusing fire in the family furnace or are encouraged tolook after fire in the furnace or elsewhere. They aretirelessly warned and reminded to concentrate andthus avoid or prevent catastrophic fires that may betriggered by their slight carelessness, negligence orerror, especially when they are very inexperienced.

Actually, the focus of the entire processes ofinformal education and practical training on fire is todirect children to learn the right way to avoid and or toprevent fire in order to safeguard all kinds of valuableand important belongings of the family and of thecommunity. Safeguarding the fire-prone traditionalhouses and villages is the noble obligation of eachmember of a nuclear family as well as extendedfamilies.

According to the traditional social land tenure andland-use system of the Meto group, protectedgovernment-owned forests as well as communityforests are claimed as integrated parts of thetraditional areas. Cattle owners use such forests forgrazing and many other natural resources are takenfrom them. But when children receive informaleducation and practical training those forests areexcluded from their tasks of safeguarding. Thoseforests are considered common property, meaning thatthey can be accessed and used by everyone withoutany obligation, including that of extinguishing fire.The apathetic villagers will hastily try to extinguishfire in such a forest only if fires, known by the name ofsbuna in the Meto language, are blown by fast windstoward the village.

Concepts of burningThere are different contextual meanings for the verb‘burning’ in the Meto language. They are tunnu, nottu,and nollak. The verb tunnu is used when fire is useddirectly and appropriately to change the condition ofcertain kinds of raw material to produce a directlyedible food, or to obtain a desired condition of the rawmaterial for further processing to obtain various kindsof consumable food. The verb nottu is used when fireis used to burn out rubbish or to clear dry-slashedmaterial in order to clear lands for swiddenagricultural purposes. The verb nottu is also usedwhen fire is used to burn a supposedly limited area of

grassland or bushes to obtain a desired plot of land orto obtain a predetermined intentional good result. Theverb nollak is used when fire is burning out rottenfirewood, or when fire is intentionally used to burn outundesired standing dry stumps of trees or fallen drybig trunks in a second or third year swiddenagricultural field. The intention of nollak, mentionedlater, is positive. But usually the verb nollak is usedwhen the process of burning is unexpected, and/or theresult of burning and the impacts of burning arenegative. The verb nollak is used when burning isdone with bad intentions in anyone’s mind, or theburning is done by unidentified persons to destroygood grasslands, bushes, or protected forests.

Children and youngsters receive early informaleducation and practical training to avoid and/or toprevent the outbreaks of catastrophic fire during thetunnu or nottu activities. Movable dry belongingshave to be removed from where the tunnu and nottuburning activities are under way. Sakko or firebreaksare created to prevent the fire jumping towardimmovable dry flammable or readily burnt outmaterials when a nottu activity has to be undertaken ina fire-prone neighbourhood. Certain kinds oftraditional prescribed burnings have to be carefullyand practically applied when a nottu activity isunavoidable.

Children and youngsters are also trained to preventjumping fires when a nollak activity is carried out toclear away dry stumps and trunks in second and thirdyear swidden agricultural fields. But they are notexclusively and conditionally trained to prevent or toextinguish fire in the commonly accessed and usedgrasslands, bushes, and forests. When they are adultthey may need spontaneously or willingly toextinguish fire in such common properties, especiallywhen the fire threatens the houses in the village.

Cultural PerspectiveWhen the young Tetun Princess in the above-mentioned folktale instructed the hunter’s companionto set fire to the surrounding grassland or when Timorwas still sparsely populated, there was no problem offire. But nowadays everything is changing rapidly.The population of West Timor has exceeded1.1 million people and the population of NusaTenggara Timur exceeds 3.6 million. Many formergrasslands, bushlands and forests are beingtransformed into villages, small towns andtransportation facilities, swidden and otheragricultural fields. Most of the population in the ruralareas consists of poor subsistence peasants, who usefire as the main means and tool in their activities. Therelatively few rural and city elite employ thesepeasants to raise free-ranging cattle and otherlivestock. The physical and biological components of

76

the environment are degraded by the competingswidden agricultural activities and the traditionalextensive livestock raising activities.

The process of urbanisation in Kupang and otherrapidly growing small towns in Nusa Tenggara Timuris increasing. But the increasing numbers in thegeneration of the autochthonous group are reluctant tolook to the surrounding straits and seas to earn a betterliving. Fishing activities are mainly done bynewcomers from South and Southeastern Sulawesi tofeed the booming population of Kupang as well as thesmaller towns and the villages.

Although the area of grasslands, bushlands andforests is decreasing, various scattered areas are stillavailable in East Sumba, and there are many scatteredpockets of dry areas all over the province of NusaTenggara Timur. Many of these areas of hilly andmountains landscapes were declared protected forestareas by the Dutch colonial government in the 1930s.The Indonesian government expanded the declaredprotected forests in the 1950s and 1970s. But manygroups of people who live in or around the protectedforest earn their daily bread from those forests.

Unfortunately, the government accuses them ofencroaching on and destroying the protected forests orof stealing the forest’s products. Of course theincreasing population is one of the main causes of theencroachment of forests and forest products, becausesome newcomers are settling down in and around theprotected areas. Many urban people are also usingtrucks and other transportation to steal and removelogs and other forest products from the protectedforests. But many people who have for generationslived and/or earned their living in and from thecurrently protected forests are now losing their rightbecause the government declares their forests asgovernment property. The local people want to use theprotected forest as the hunting and gathering grounds,or for their swidden and other agricultural activities,or as simply their ordinary living space because theyhave inherited the forests from their forefathers.

Some parts of the forests, especially areas aroundthe water resources and fertile swidden agriculturalfields, have been converted into permanent agro-silvipastoral systems for perennial-commercial cropsand livestock. These kinds of agro-silvipastoralenterprises are valuable, inheritable possessionssimilar to those outside the protected forest. Theseactivities are termed forest encroachment by forestryofficials who are not familiar with the traditional landtenure and land use rights. Research is needed in thefuture to unveil the unknown (by outsiders) traditionalland tenure and land-use rights for the sake of thefuture sustainable agriculture and forestrydevelopment endeavours.

Foresters and other government officials frequentlyaccuse these so-called forest encroachments ofcausing forest fires. But actually there are manycauses of forest fires; some of them are still unknownor superficially known. As long as there is sufficientdry material in the grassland and bushes or on thefloor of any types of forest, fire is likely at anymoment during the long dry sessions. Fire may startfrom the hands of anyone, a careless passerby, ahunter, a peasant, or accidentally escaping from thelands of a goodhearted fire user.

Fire occurs infrequently in relatively denselypopulated areas and in relatively well managedfarmlands. Thirty years ago fire was the only reliablemeans of clearing away piles of seasonal grasses onthe northern outskirts of Bajawa, the capital of Ngadadistrict in central Flores. Nowadays the fire-pronegrasses are replaced with irrigated rice fields andcommercial perennial trees, and fire no longer occurs.

Chromolaena odorata, the introduced bushy plantis infesting overgrazed grasslands and pushing backthe decreasing forest in Sumba, Flores and Timor. It isbecoming a new kind of flammable fuel in themismanaged protected forests and the long fallowedswidden agricultural fields in some relatively sparselypopulated areas of East Sumba. Cattle owners hateand curse this new plant because their starvinganimals will not eat its abundant leaves. On the otherhand, swidden agriculturists in the densely populatedvillages welcome it because it restores the vegetationcover and the soil fertility of their followed fieldswithin 2–3 years. It is easy to slash then clear awaywith fire within 1–3 days. So, the competing andconflicting interest between peasants and cattleowners for arable lands and grazing lands is mingledwith the feeling of hate and love toward Chromolaenaodorata.

Concluding RemarksFire is being tamed and has become a servant in thefurnace within the fire-prone traditional houses. But ithas to be carefully managed in the fire-prone semi-arid savannas to become a trustful partner forlivelihood. Networking, which means cooperating toachieve a common purpose (in this case sound firemanagement), is badly needed from variousgovernment agencies, universities, non-governmentorganisations, and the community at large.

77

GEOLOGISTS COMMONLY agree that the island of Timorhad a turbulent past which impacted the island’s currenttopography, soil conditions, vegetation, land use andcultivation system. The island lies between twocontinents, one extremely wet and one dry, and theaverage number of dry months is longer than theaverage number of wet months. These factors directlyrelate to agricultural practices and the types of drought-resistant food crops cultivated by Timorese farmers.

The following discussion focuses on agriculturalpractices on the island of Timor, considering the typesof vegetation cultivated in the area, the various formsof land utilisation and traditional agriculturalactivities. In each of these three different perspectives,it will become clear how fire plays such an integralrole in the lives of Timorese. To begin with, I willdescribe the vegetation, particularly the food cropsgrown in the area by swidden farmers. (The term‘swidden’ is used in the Old English sense of ‘burnedclearing’, while the term ‘shifting cultivation’ refersmore broadly to agricultural activities where fields arecultivated for crops then left fallow.)

Food CropsDry land agriculture is still the main source of life andincome for most Timorese. Their staple crop is maize(Zea mays) (Tetun: batar malae = the foreign maize).

The name given to this crop indicates that maize is anewly introduced plant. Maize is grown throughoutthe area from the lowland to the upland. Peoplerecognise different varieties of maize by the size of thecobs, the colour of the seed and the length of thegrowing period, as follows:

Names of maize Translationbatar mutin white maizebatar mean red maizebatar ki’ik small maizebatar bot big maize

The other major crop in the area is sorghum. Indaily speech among the Tetun people in the centralregions of the island, the term for sorghum is batar ainaruk (literally, long stalked maize). In Tetun rituallanguage it is referred to as batar na’an tasi (literally,fish maize) or simply as batar tasi (maize from thesea). A well known myth in south Tetun narrates howthe first seed of sorghum was discovered in the headof a knase fish. Locally, the various kinds of sorghumare named according to characteristics of the seed, itscolor, its taste or its mythical origin, as follows:

Names of sorghum Translationbatar mean lakulot red charming sorghumbatar bua funan areca-nut blossom sorghumbatar na’i katuas respected man’s sorghumbatar laka bela flat flame sorghumbatar mean red sorghumbatar na’an tasi fish sorghum

The Role of Fire in Swidden Cultivation: ATimor Case Study

Tom Therik1

AbstractSwidden agriculture is a subsistence way of life in which fire features significantly. This paperfocuses on swidden cultivation in east and west Timor, particularly among the Meto, Tetun andBunak ethnic communities. The author describes the types of food crop grown, traditions of landallocation and the ways in which each type of land—sacred or ‘forbidden’ space, dwelling space andcultivation space—are used in a communal sense. Cultivation space is the only land allocated forswidden farming, and the means of using fire on this land and its part in the cropping cycle aredescribed.

1Centre for Regional Studies, Universitas Kristen Artha Wacana

In certain areas on Timor dry upland rice (hareleten) is planted together with maize and other rootcrops. The Timorese highly value the dry upland ricefor two reasons. Firstly, it tastes better than the wetrice. Due to its superior taste, dryland rice isconsidered appropriate for special occasions such asentertaining guests. Secondly, dryland rice has ritualimportance. In agricultural rites, only this kind of ricecan be offered to ancestors. If a person from the Tetunarea was asked, for example, why dryland rice is usedin these offerings, he or she would explain that thiswas the first rice brought by their ancestors to Timor.Following typical logic in this region, people oftenexplain: ‘Wet rice (hare we) cultivation is not the wayof our ancestors’.

People’s long acquaintance with dryland ricecultivation can be deduced from the mention of rice intheir oldest myths and the varieties of dryland riceknown in the area, such as:

Name of rice Translationhare sukabi oak (tree) ricehare kwa metan black crow ricehare Bauk morin fragrant Bauk ricehare busa ni’an dog-teeth ricehare marahuk furry ricehare ekekero taily ricehare babelik sticky rice

There are other crops which grow particularly wellin the hill regions of Timor, such as varieties of fox-tail millet (tora), which in Tetun are called ‘dog-tail’(asu ikun), sesame seed (lena) and cow pea (turis).These crops still have ritual and economic importancefor most Timorese. The hill regions are also suitablefor certain species of root crops. TheTetun distinguishbetween root crops cultivated in their garden (fehuk)and those that grow wild in forest (uhi). These twokinds of root crops can often be found at the marketduring market days. For example see below:

Cultivated root Literal translationfehuk ema people tuber fehuk samea snake tuber fehuk nona metan black woman tuber fehuk nona muti white woman tuber fehuk fafiur quail tuber

Uncultivated root Literal translationuhi rama bow (for shooting) rootuhi laku charming rootmaek (itchy) yamfia kalo raek yam (with small leaves)fehuk lambo lambo tuber

In addition to these root crops, a number ofsubsidiary plants classified as beans or legumes(Indonesian: kacang-kacangan) are also known to thearea. The Tetun use two words to describe beans.Those that grow wild in jungles are called ahan, whilecultivated beans are called fore. Under the category of

ahan two types are still important in people’s diet,ahan alas (the forest bean) and isikoma. Types of foregrown in the area include peanuts (fore rai), kidneybeans (fore tali) and mung beans (fore Wehali).

Land AllocationThe Timorese divide their physical environment intothree categories: sacred or ‘forbidden’ space, dwellingspace and cultivation space. Land utilisation isdifferent for each of these spaces. A sacred jungle, forexample, is part of forbidden space. People of thesame origin group would communally own thisjungle. For the Meto ethnic group in the western partof the island, this jungle is a free grazing zone for thecommunity. People may claim individual ownershipof wild bee hives for collecting honey, but they maynot claim the rights to the trees where the bees buildtheir hives. Formerly, elders in the community wouldgive special permission to hunt deer for consumption.But any activities related to agricultural cultivation aretotally banned in ‘forbidden spaces’.

Activities relating to agriculture are carried out inthe ‘cultivation spaces’. People grow food crops onlyin these specific spaces. This area is also communallyowned. The ‘land lord’ (kua tuaf) of a community is incharge of distributing land to every adult member ofeach house for a personal garden. In an area whereadult members are fewer than the available land, thekua tuaf can allocate a person more than one block ofland. The custodianship of a garden may be inheritedby people within the same house but they are forbiddenfrom selling the land to outsiders. When planting theirfood crops, some farmers also grow fruit trees such ascoconut and mangoes in these spaces. Fruit from thesetrees is only for personal consumption of family, andtherefore the number of such trees is limited.

The agricultural activities held in cultivation spaceare important as a source of food in the peasanteconomy. The failure of food crops cultivated in thisarea will endanger the life of the people. Therefore,the farmers spent much of their time in the fields fromthe time of field preparation, through planting,weeding and harvesting.

Dwelling space allows for domestic activities andthe location of houses. Fruit trees may be planted indwelling spaces as a source of cash income, but therisk of fire to houses and fruit trees means the practiceof slash and burn is forbidden in this area.

The Role of Fire in Swidden FarmingAs mentioned earlier, the most widespread type ofagricultural activity on Timor is swidden cultivation.However, the practice of swidden farming is carriedout only in the so-called ‘cultivation spaces’. The

78

normal sequence of events in the Timorese cultivationcycle is the clearing of fields, followed by burning andthen cropping. From the time a field is cleared untilpost-harvest fire is an essential tool in the swiddenfarming practices of Timorese farmers. A farmer’sfuture harvest depends greatly on the success of theburn. So for many Timorese, swidden farmingactivities and the use of fire cannot be separated. Fireguarantees their economic security, or in other words,people’s lives depend on the proper use of fire.

Thus in both agricultural rituals for planting seedsand in birth rituals, fire is associated with life. Manyfactors are involved in the actual firing of fields,including how well the slash dries, the thickness of theheap of slash, the wind direction, and the actual timingof the burn, i.e. whether it is done during the day ornight. In opening a new field, farmers usually slashthe undergrowth and fell most larger trees. The woodis then chopped in 0.5–2 metre lengths. The Amarasipeople (subgroup of the Meto ethnic group) called thisactivity ‘an kes’. In cases where there is not enoughwood, additional wood is brought from the forestaround the field, and piled in a single layer accross thefield (na’ nunu). To ignite the wood, dry leaves areused as fuel.

In a land dominated by Imperata cylindrica(Indonesian: alang-alang) extra care must be taken byfarmers because of the fire danger associated with thisgrass. Many of the forest fires in Timor are caused byaccidental burning of imperata by swidden cultivatorsor intentional burning by livestockers. Choosing thebest time to burn is the most important factor indetermining the success of a burn. At the same time,the success of the firing is also determined by thecompleteness of the slashing and the dryness of thematerial at the time of the burn. Because of theunpredictable weather conditions found in Timor,farmers have developed considerable local knowledgeto decide the ‘exact’ time of the burn, during a shortperiod prior to the onset of the rainy season.

The Timorese want to burn their fields before thegermination of weed seeds, to ensure that theirseedlings are not competing with weeds for food.Theweather, the exact hour for firing, and the direction ofthe wind on the day of the burn are crucial, not only forthe success of the burn but also for fire prevention. If afield is located close to fruit trees or houses with highlyflammable thatch roofs, the burn is postponed untildusk (Meto: tot fai), so that sparks can be easily spottedand extinguished. Otherwise the burn is done during thehottest midday period (Meto: tot mansin maeb) so thatthe slash is in its driest condition. The wind direction(Meto: ainne poin) and the establishment of fire breaks(Meto: nak sako) are considered in detail prior to thefiring. Figure 1 attempts to visualise the burningtechniques just mentioned:

There are a few techniques used by people to avoidan unfortunate outcome, such as:• Protective clearing• Backfiring for burning steeply sloping fields• Burn the easily ignited spots first, and later the rest

of the field.

Fire, a Life-giving ForceI have shown that ‘fire’ is not only a useful tool forswidden farming. Many rituals concerning the humanlife cycle also have something to do with fire (andwater). The importance of fire as a life-giving force isalso narrated in various kinds of myths. Some ethniccommunities such as the Sikka who live in themountain region still mark the end of dry season andthe beginning of rainy season by performing anagriculural rite where a block of imperata grass land isburned. A traditional adat official called Paka WatuApi has the ceremonial responsibility of igniting thisfire. People still strongly believe that withoutperforming this particular rite, the fertility of the landand the welfare of the people will be jeopardised.

Despite the many benefits of fire in swiddencultivation, the use of fire has been stronglycondemned. There are many allegations, particularlyfrom politicians and academics, regarding:• the ecological impact of such practices; • the carrying capacity in terms of population–land

ratios for a sustainable swidden; • the nutritive value of slash and burn.

These issues must be addressed seriously becausemisjudgment and misinformation will certainly affectthe life of countless swidden cultivators, particularlythe poor Timorese farmers who depend on fire for life.

79

nak sako

wind direction

house

tree crops

other fields

tot tahun noko ainne poin �(initial burning point)

tree crops

nak sako

Figure 1. Techniques employed for successful burning of afield.

80

IN 1997 THE world was witness to the dramatic impactof uncontrolled forest fires in Kalimantan andSumatra. The resultant thick pall of smoke and hazethat spread across much of western Indonesia,Malaysia and Singapore focused attention on majorinadequacies of forestry management in Indonesia. Inthis case the combination of extended droughtconditions and rapacious land clearing, particularlyfor industrial plantations, produced a majorenvironmental disaster which has had continuingrepercussions. Data from satellite imaging have led toestimations of up to 4 million hectares of land wereburnt in East Kalimantan alone during 1998 (Siegertand Hoffmann in press). The huge extent of the firesand the management weaknesses they have exposedhave prompted significant reforms and policy shifts atthe national level.

In eastern Indonesia the sustained use of fire formsa central aspect of indigenous land-use practice. It is a

practice that is ingrained within the agricultural cycleof farming communities throughout the region.During the 1980s I spent some years living on Timorisland in Nusa Tenggara Timur (NTT), and one of myenduring memories of that time is of the annualburning and smoke fires that ignited across thelandscape as the dry season progressed. Driving atnight through Timor, the hills and mountains wouldbe regularly aflame with multiple glowing spot firesand ragged lines of burning vegetation along ridgesand grassland savannas. Annually from June throughNovember the extended dry season and strongsoutheast trade winds provide ideal conditions for fireand the so-called cultural burning of the landscape asfarmers prepared for the coming wet season.

The pattern and rhythm of seasonal burning hasdeep historical and cultural roots in Nusa TenggaraTimur. Over 200 years ago (1770) when the Englishseafarer, Captain James Cook, sailed into Timoresewaters during his first voyage around the world, he toofelt moved to comment on the presence of fires:

Fire and Cultural Burning in Nusa TenggaraTimur: Some Implications of Fire

Management Practices for IndonesianGovernment Policy

Andrew McWilliam1

AbstractIn eastern Indonesia the sustained use of fire in agriculture has been a central factor in the conversionof extensive forest areas into permanent fire climax savanna grasslands and degraded secondarybushland. From a Government perspective the continuing practice of ‘cultural burning’ is widely seenas destructive and detrimental to the development of progressive and sustainable agriculture.

In approaching the issue of cultural burning and fire management, however, it is important torecognise the great diversity of ecological conditions and agricultural practices that exist in regionslike Nusa Tenggara Timur. The impact of this burning is highly variable across the Province andremains generally poorly understood and under-researched.

Contemporary patterns of burning practices are also highly dynamic over time, reflecting changingresponses to economic or environmental factors. For example, the successful cultivation of tree crops,particularly in southern hinterland areas of Flores, has substantially reduced the incidence of culturalburning in these regions. Slash and burn cultivation has given way to higher value permanent tree cropcultivation. Elsewhere, growing rural population densities and the emergence of semi-permanentfarming regimes have also significantly reduced the possibility of bushfires. At the same timeunregulated savanna grassland fires across the Province, and destructive burning in forest lands and re-afforestation areas continue to pose risks and challenges for longer term management of land resourcesin NTT. The paper explores some of the patterns of fire and cultural burning in NTT and suggests arange of policy implications and directions for improved management of fire in the landscape.

1Sacred Sites Authority, North Australian Research Unit 12 Sanders St Jingili NT 0810

81

‘Light land and sea breezes…we saw severalsmoakes ashore in the p.m. and fires in the night,both upon the lowland and up in the mountains.’And later: ‘We continually saw upon it (Timor)smoakes by day and fires by night and in manyplaces, homes and plantations.’

The long-term impact of this sustained use of fire inland-use practice has seen the gradual and continuingconversion of native forests into permanent fireclimax savanna grasslands, palm savannas (see Fox1977) and degraded secondary regrowth bushland.Evidence of this process can be found throughout theislands of NTT but it is particularly marked in regionssuch as the highland grasslands of eastern Sumba, theheavily populated highlands of West Timor and incentral Flores. The principal contributions to thishistorical transformation of the natural environmenthave been the widespread practice of shiftingagriculture using slash and burn clearing technologyand the regular burning of savanna lands for huntingand to regenerate livestock fodder.

Although an evidently successful adaptation to thewet–dry monsoon climate of the region and one thathas sustained the populations of the region forcenturies, this continuing practice of ‘culturalburning’ is widely seen as destructive and detrimentalto the development of progressive and sustainableagriculture. Government proscriptions on burning andswidden agriculture are frequently justified on thisbasis.

In approaching the issue of cultural burning and firemanagement, however, it is important to recognise thegreat diversity of ecological conditions andagricultural practices that exist in regions like NusaTenggara Timur, and by extension eastern Indonesiamore generally. The impact of this burning is highlyvariable across the Province and the underlying localrationale for its persistence and application remaingenerally poorly understood and under-researched.

NTT is justly renowned for the linguistic andcultural diversity of its human population, but thisdiversity is also matched by the great variety ofecological and environmental conditions foundthroughout the region. Factors such as slope, soiltypes, vegetation, frequency and intensity of burning,population history and rainfall patterns all contributeto a differentially marked impact of fire on thelandscape in the region.

In many areas the long-term practice of slash andburn agriculture, while substantially modifying thenatural environment, has nevertheless managed toretain a high level of equilibrium and ecologicalsustainability. Fox (1977) for example has written atlength on the comparative ecological advantages ofpalm economies on otherwise resource poor islandssuch as Savu and Rote. (See also Lewis 1992 centraleastern Flores, and Dove 1984 for comparison with

Sumbawa, NTB.) The fact that significant forest areasexist at all in contemporary NTT is in part testimonyto the former capacity of swidden cultivators tosustain their forest resources.

At the same time, fire-based farming can and hasbeen highly destructive to natural environments. Thecombination of steep topographies, fragile and mobilesoils with decreased fallow periods can quickly resultin deforestation and environmental degradation.During the 20th century the process of deforestationand the penetration of cultivation into increasinglymarginal areas has gone hand-in-hand with risingrural population densities. Ormeling (1956) arguedthat significantly increased cultural burning andclearing took place in the early 20th century when theimposition of the pax Nederlandica coincided withpopulation expansion (see also Dove 1984) and anincreased demand for land but with farming methodsthat remained unaltered. Ormeling conjectured that,by the 1950s, several times the estimated 60 000 ha ofcultivated bushland (‘ladang’) in West Timor wasdestroyed every year by fire.

These preliminary comments on the use and abuseof fire in contemporary land management across NTThighlight two major issues or themes that holdsignificant implications for Government policy andmanagement responses. The first of these is therelative paucity of detailed and accurate field-baseddata and analysis on which to base appropriate policyformulation. Despite the evident pressing need forconservation and fire management research in theregion, fire policy in NTT (to the extent that it existsas a formal set of strategies) operates to a significantdegree in a research data vacuum.

Very little published data on the impact or practiceof fire-based farming and cultural burning is availablefor the region (Kaho 1994 and Subandi et al. 1998 arethe most recent examples of preliminary work) andwith the current constraints imposed by financialrecession, there are limited resources for pursuing fireresearch programs. The formulation andimplementation of effective and sustainable policiestowards fire management need to be based on moreinformed studies of existing practice. Field-basedresearch data, the identification of strategic prioritiesand an iterative contribution to continuing policydevelopment are vital components in any sustainedmanagement approach.

To date, Government approaches to fire andcultural burning have been very much ideologicallydriven, viewing the practices as more or lessuniversally misguided and inherently destructive. Thecurrent Provincial policy proscribes all culturalburning in a wholly futile attempt to curb the practice.This policy position derives in large degree fromnational perspectives and guidelines, but its originscan also be in identified in historical Dutch Colonial

82

proscriptions against any local use of fire in landmanagement and clearing.

A second major issue confronting the formulationof strategies for fire policy and management is theintricate complexity of the issue in Nusa TenggaraTimur. Fire and the practice of cultural burningrepresents a multifaceted set of issues that cut across awide range of agro-ecological (Dove 1984) contextsand institutional responsibilities. The generation ofeffective management solutions and local policydirections needs to meet this practical complexity withinitiatives and coordinated programs that recogniseand incorporate local diversity. In other words it isimportant to pursue an integrated approach to firemanagement policy and programming—one thatincorporates local farming needs and managementobjectives.

Fire and Cultural Burning in ContextTo illustrate this combination of a need for more fireimpact research and the evident complexity of fireimpacts across NTT, it is instructive to explore aspectsof current use in two broad ecological zones whereunregulated burning is evidently of concern toGovernment. These include the water catchment areasin the forested highlands and the predominantlylowland savanna grasslands across the islands. I beginwith Metzner’s (1982) study of Sikka regency incentral Flores, one which provides a rare perspectiveon local fire history. In the work he notes with alarmthe speed at which Sikka’s natural vegetation andparticularly forests have disappeared.

At the time of Metzner’s research in the early1980s, all the forest reserves created during the Dutchadministration of the 1930s were reportedly underthreat. Attempts to enlarge and connect the variousforest reserves had been largely unsuccessful in theface of population pressures and what he describes as‘fiercely defended claims on land’. Related attemptsto regulate the practice of cultural burning had beensimilarly unsuccessful. Metzner comments that localfarming communities consistently ignored regulationsaimed at controlling the use and location of fire, andlimited Forestry staff resources mean thatmanagement and policing were quite ineffective. Heillustrates his argument with a list of reported illegalburnings in forestry areas of Kabupaten Sikka. Today,some 20 years after Metzner’s research there is littleevidence to suggest any reduction in forest clearing orillegal burnings in forestry zones.

The persistence of extensive swidden cultivation inforested highlands and the apparent reluctance offarmers to convert to more intensive, non-fire-basedfarming is widespread in eastern Indonesia. This isdespite long-term Government policies directedagainst fire and forest clearing often accompanied by

extension promoting more intensive agriculture,particularly irrigated rice, in lowland areas.

Suggested factorsIn the absence of any definitive study I would suggesta number of likely contributing factors to thepersistence of burning and extensive agriculture. Thefirst of these is population increase, especially overthe last century, which in NTT has been running ataround 2% per annum ( e.g. 1.87% between 1971 and1990 (Nusa Tenggara Timur Dalam Anka 1992)). Inthe absence of any modification in farmingtechnology, higher rural populations put pressure onavailable land resources and begin an inexorableexpansion into more marginal or critical lands.

The creation of forestry reserves (kawasan hutan)under the administrative management of aGovernment ministry has also contributed to pressureon available arable land resources for cultivation.With some exceptions, such as official reforestation orindustrial plantations, local farmers may not legallycultivate forestry land. Where there are officiallysanctioned projects, a local work force, in addition towages, has the chance to grow annual food cropsbetween the tree plantings until temporary permissionis withdrawn when the plantation or reforestation isestablished.

Thus the designation of extensive areas of landwithin a forestry zone effectively takes substantialareas of agricultural land out of production. At thesame time the forestry reservation system hasenclosed or appropriated forested land to which awhole range of indigenous rights and claims inhere.These rights are usually not officially recognised and Isuspect that many farming communities would justifytheir incursions and burning of forestry land in termsof asserting traditional rights in land.

A further factor in the continuing attraction ofextensive swidden agriculture over higher production-intensive alternatives is the inherent benefit in termsof labour productivity. Dove has argued that extensiveswidden agriculture yields a higher return per unit oflabour, and that this is significant because it is labournot land that is the critical limiting resource in thedryland farming regions over much of easternIndonesia (Dove 1984, 1986a). He suggests thatfarmers typically only intensify their agriculture whenpopulation densities rise with subsequent landpressure problems, or when governments compelthem to do so. Such processes are evidently at work inNTT where land pressures are increasing—theopportunities of shifting location and opening newareas of forest land are today severely curtailed forswidden farmers.

Evidence of these effects can be interpreted in anumber of areas across NTT where fire as anenvironmental issue is actually decreasing in

83

significance. In the mountainous hinterland along thecoast of southern Flores, particularly in KabupatenNgada, the successful cultivation of tree crops overthe last 20 years or so has substantially reduced theincidence of cultural burning. Slash and burn swiddencultivation has given way to higher value, more or lesspermanent tree crop cultivation. In the presenteconomic environment the transition has provedhighly rewarding to local farming communities whoare reaping the rewards for rising commodity prices.The experience points the way for an intensification oftree crop cultivation and the support of exportmarketing infrastructure in other suitable regions ofthe Province.

In other areas, including parts of Kabupaten Ende inFlores, areas in West Sumba and West Timor, greatlyincreased rural population densities in recent decadeshave led to the emergence of semi-permanent farmingregimes with significantly reduced incidence ofbushfires or cultural burning. As swidden fallowperiods are reduced and cultivation expands, there is acorresponding decrease in the availability ofcombustible dry matter and bushland. In thesecircumstances, fire as a farming tool diminishes inimportance, but the incidence of fire reduction is morea signal of declining soil fertility in these croppingareas and increasing marginalisation of agriculture

The rangelands and savanna grasslands that coverextensive areas of all the islands in NTT representanother vivid demonstration of the long-termapplication of fire. Fire-tolerant vegetation specieshave long become dominant in the savanna lands.Repeated burning has helped prevent or limitreforestation and is widely used to clear dry rankgrasses and encourage the regrowth of succulents andgreen pick for domesticated livestock, especiallycattle, goats, sheep and horses. The recent expansionof the highly flammable but fire-tolerant shrub weedChromolaena odorata, and prior to that the chokingspread of Lantana camara L. (Ormeling 1956; Fox1977) are also in part a consequence of regularsavanna grassland fires.

While fire may well be a key factor in thepersistence of anthropogenic grasslands (Dove 1984),unregulated grazing and burning can also contribute toerosion and degradation of the rangelands. Grassfiresin the height of the dry season can give rise to winderosion and later soil erosion (Metzner 1977) whenthe monsoon rains break and carry away substantialsilt loads. In some instances too it is the absence orgreatly reduced incidence of fires that signals theimpact of overgrazing or loss of topsoil.

In these circumstances research into the use andimpact of fire has significant implications for policydirections in such diverse fields as water catchmentplanning and management, soil conservation,livestock fodder production and carrying capacities,

as well as environmental conservation, weedmanagement, agricultural and settlement policies. Inother words the use of fire has multiple implicationsfor different land-use contexts and objectives. Part ofthe development of a research focus on fire practice isa need to understand the complex ecologicalinterrelationships between fire and other land-useprocesses. Ormeling’s warning some decades ago(Ormeling 1956) still has contemporary relevance.‘Reliable information concerning grass burning is,however, scarce and conflicting. Since so muchdepends on variable factors such as type of grassland,rainfall, the season when burning takes place andintensity of the burning, generalisations aredangerous.’

Ormeling might also have mentioned the variablesignificance of cultural factors here. An aspect of firepractice in places like NTT is that while at one levelthere may be a local agro-ecological explanation orrationale for burning and clearing, these practices mayalso have deeper cultural associations andsignificance which contribute to a general dispositionfor seasonal burning. An example of the kind of strongcultural role of fire in the traditional seasonal rhythmof agriculture can be seen in practices reported forBunaq speaking rural communities in central Timor.Here the passage between the dry season and the rainyseason is marked by the conduct of ritual hunts forferal pigs in the wide savanna lowlands. These huntsare sacrificial in character and enact the mythicaltraditions of the origins of the seeds of agriculture andfertility of the soil. Their conduct is accompanied ‘bythe burning of the great savannas as well as the lowerspurs of the hills’ (Friedberg 1980).

It is not clear to what extent ritual hunting for feralpig stills forms an integral precursor to cultivationrituals and the planting cycle among Bunaqcommunities. However, the point is that the idea ofburning and firing the dried grasslands is deeplyingrained as a cultural practice among thesecommunities, as it is in many other regions of theProvince. In particular the marked seasonaldichotomy of dry and wet monsoons in NTT hasencouraged a symbolic cultural elaboration whichassociates the sequence of heating and coolingproperties with the cyclical reproduction of life andfertility of life forms, including the seeds ofagriculture. Schulte Nordholt (1971) has commentedin relation to west Timorese populations, that theprocess of burning off is followed by ‘cooling’ ritualsin which ‘ the earth has to be made cool (mainikin)again and the forces of heat and fire renderedharmless’. These rituals tend to coincide with the earlyrains of the wet season and usher in a period of ‘newlife’ in agriculture.

The ecology of fire in NTT and its multi-jurisdictional nature also has implications for the

84

coordination of policy between relevant Governmentagencies and institutions. Because fires do not respectadministrative boundaries it is clear that differentGovernment agencies need to develop a cooperativeand collaborative approach to fire management. Thedegree of elaboration of any particular agency firepolicy will depend on the extent to which fire andcultural burning represents a constraint or hazardwithin discrete domains of activity. There is clearly aneed for close cooperation in terms of setting policyguidelines and recommendations between the mainresponsible Government institutions and with non-government organisations active in the field.

ConclusionsThis preliminary overview of fire practices and theirsignificance in NTT has indicated a number of policyimplications, which I represent below in summaryform.• There are limited current field based data on which

to base policy formulation and priorities forGovernment strategic planning and programming.There is a need to increase research capability andthe role of research in policy formulation. There isalso a need to bring together and utilise in aconcerted way, the variety of existing informationand relevant studies that can support policyformulation. This includes both field-based dataresearch and comparative research data fromrelevant project experience such as the WorldWildlife Fund integrated studies in southeastSumba, Riung in Flores and the Mutis MountainsForestry Zone in West Timor. Relevantorganisations beyond NTT include the IntegratedForest Fires Management Project Centre and theSustainable Forest Management Project (both basedin Samarinda, East Kalimantan), and the Centre forInternational Forestry Research (CIFOR) in Bogor.

• The practice of fire and cultural burning in NusaTenggara Timur cannot be considered in isolationfrom inter-related processes of land-use manage-ment and cultural practice. This understanding hasimplications both for the design and gathering ofresearch data on fire impact, but also for thesubsequent development of improved managementstrategies. Successful small-scale pilot anddemonstration projects provide a phased approachfor wider applications. The often uniquecharacteristics of local interventions however,present challenges to scaling-up managementsolutions for broader application.

• Government has limited skilled resource capacity todirectly manage the two critical areas of savannagrasslands and forests. Indeed it can be argued thatthere is effectively no Government organisationwith the responsibility for managing grasslands as

grassland. (I am indebted to Jefferson Fox forraising this issue of which much more might besaid. To some extent it reflects a longstandingproblem in eastern Indonesia where landmanagement policy and programming is developedand promulgated on the basis of agriculture practicefrom wet tropical environments of westernIndonesia.) Central agencies such as the Ministriesof Forestry, Animal Husbandry and Agriculture alltend to view grasslands as degraded forms ofotherwise more productive land that need to berevegetated or rehabilitated into something else.Improving Government resource managementcapacity, therefore, also requires significantinstitutional development and policy shifts (see alsoDove 1986b).

• In addition to Government, it is vital to involvelocal farmers and local non-Governmentorganisations (Lembaga Swadaya Masyarakat) inthe process of improved land management,knowledge transfer and exchange. This may requirea combination of strategic conservation andpolicing priorities with pragmatic partnerships andequity positions encouraged between ruralcommunities and management institutions.

• Fire management is a public education issue.Dissemination of knowledge and improved practicemust be promoted through appropriate publicinstitutions such as radio, village government,church and school programs.

ReferencesDove, M. R. 1984. Man, land and game in Sumbawa: some

observations on agrarian ecology and developmentpolicy in eastern Indonesia. Singapore Journal ofTropical Geography, Vol 5, No 2, 119.

Dove, M.R. 1986a. The ideology of agriculturaldevelopment in Indonesia. In: MacAndrews, C., ed.,Central Government and Local Development inIndonesia. Singapore, Oxford: Oxford UniversityPress.

Dove, M.R. 1986b. The practical reason of weeds inIndonesia: peasant vs state views of Imperata andChromolaena. Human Ecology, Vol. 14(2), 163–190.

Friedberg, Claudine 1980. Boiled woman and broiled man:myths and agricultural rituals of the Bunaq of CentralTimor [Trans from French by: E. Traube]. In: Fox,James J., ed., The Flow of Life: Essays on EasternIndonesia. Cambridge Mass: Harvard University Press,266–289.

Fox, James J. 1977. The Harvest of the Palm: EcologicalChange in Eastern Indonesia. Cambridge Mass:Harvard University Press.

Kaho L.M.R 1994. Tinjauan ekologis tentang pembakaranpadang rumput danperladangan serta kemungkinanpengelolaannya [Ecological survey on the burning ofgrasslands and fallow field gardens and the possibilities

85

for management]. Seminar paper presented at theFalkultas Peternakan (Undana) workshop ‘ Pengelolaanpembakaran padang rumput dan ladang berpindah’[The management of fire in grasslands and shiftingcultivation land] . Kupang, NTT.

Lewis E. D. 1992. The relations of the social order,agricultural practice and environment in Tana’Ai,Flores. In: Fox, James J., ed, The Heritage ofTraditional Agriculture among the WesternIndonesians. Canberra: The Australian NationalUniversity, 47–66.

Metzner, Joachim K. 1977. Man and Environment in EasternTimor: a geoecological analysis of the Baucau-Viqueque Area as a possible basis for regionalplanning. Canberra: The Australian NationalUniversity.

Metzner, Joachim K. 1982. Agriculture and populationpressure in Sikka, Isle of Flores: A contribution to thestudy of the stability of agricultural systems in the wetand dry tropics. Canberra: Development StudiesCentre, The Australian National University,Monograph No. 28.

Nusa Tenggara Timur Dalam Angka 1992. Propinsi NTT:Kantor Statistik.

Ormeling, F.J. 1956. The Timor Problem: A geographicalinterpretation of an underdeveloped island. Groningen,Djakarta: J B Wolters.

Siegert F., and Hoffmann, A.A. in press. Evaluation of ForestFires 1998 in East Kalimantan (Indonesia) using multi-temporal ERS-2 SAR images and NOAA-AVHRRdata. International Conference on Data Managementand Modelling using Remote Sensing and GIS forTropical Forest Land Inventory, Jakarta 26–29 October1998.

Schulte Nordholt, H. G. 1971. The Political System of theAtoni of West Timor. The Hague: Martinus Nijhoff.

Subandi, J. Triastono, E., Budisantoso dan T Bamualin 1998.Pengelolaan Padang Rumput untuk Konservasi lahan danPerbaikan Kualitas Pakan Ternak di DAS Kambaneroe,Kab. Sumba Timur’ [Grassland management for soilconservation and pasture quality improvement inKambaneroe, Eastern Sumba]. In: Pengembangan SistemUsahatani Konservasi Lahan Kering di DAS Kambaniru.Kab. Sumba Timur-NTT. Pokja Penelitian danPengembangan Sekretariat Tim Pengendali BantuanPenghijauan dan Reboisasi dan Balai PengkajianTeknologi Pertanian. Naibonat-Kupang. Bogor: PusatPenelitian Tanah dan Agro-Klimat.

86

EAST SUMBA REGENCY AREA LEVEL II has an areaaround 7000 km2 and is located in the southern regionof the Republic of Indonesia. It directly faces theIndian Ocean and the Australian continent and is oftenaffected by the presence of dry trade winds. It has awet season for four months (Dec.–March) and an 8-month dry season (April–Nov.).

The conditions mentioned above are made worse bythe mountainous and hilly topography and relativelylow land fertility. Based on the facts available,192 000 ha have slopes of more than 40%, while amongthese there are 176 000 hectares on very steep hillsides.The level of erosion is quite high and is marked by thevisible presence of rocks on the land surface, as well asthe existence of gullies.

The population of East Sumba in 1997 was morethan 170 000 people, with a distribution of 23 peopleper km2. It is estimated that in the year 2000, thepopulation will reach more that 200 000.

The main livelihood of the population in general isfarming, with farm operations of dryland farming aswell as livestock. Farming for drylands is still mostlydominated by traditional slash-and-burn methods.This method is thought by the community to be thebest because it doesn’t require costs or much hardlabor, and the labor time is relatively short. But if it islooked at from an environmental conservation point ofview, this has a great effect on the ecosystem andlowers the land productivity.

In accordance with the Method of Forest UseAgreement, the area of forest in East Sumba is237 600 hectares or 34% of the region’s total area. Iflinked with Ordinance No. 5 1967 concerning basicforestry stipulations, then the area mentioned hasfulfilled the requirements, i.e. more than 30%. But iflinked with vegetation/plant conditions, it is a greaterconcern, because dense forest consists of only31 250 hectares (13%), and swamp or marshlandconsisting of 6000 hectares (2.5%).

Based on its function, the forest in East Sumba canbe classified into six sections, which are:

Area (ha)• Protected forest : 105000 • Stable production forest : 29700 • Limited production forest : 20000• Convertible forest production : 60 • Forest nature reserves : 15600 • Forest wildlife reserves : 8100

In the Regency of East Sumba are seven watercatchment areas (DAS), they are, DAS Kambaniru,Kadhang, Watumbaka, Melolo, Kaliongga, Nggongiand Tidas. The condition of these areas is very criticalbecause during the wet and dry seasons there is quite ahigh fluctuation of the rate of water flow as well asmud content, and a very erratic river flow during thewet season.

Forest potential is relatively low. The kinds offorest products that stand out are, tamarind(Tamarindus indicus), Aleurites mollucana, Schleora

Forest Fragmentation and Water CatchmentHydrology — Case Study on Sumba

Juspan1

AbstractIn East Sumba dryland farming is still mostly dominated by traditional slash-and-burn methods.Forestry potential is relatively low. Natural forest resources have many functions and many benefits,and need to be guarded from all types of intrusions that destroy. One such element is fire. This paperexamines the cause and impact of forest fires in East Sumba, detailing their causes and the impact onvegetation, soils and hydrology as well as communities. The paper looks at prevention through betterinformation, silvicultural practices, employment of local knowledge and tradition, communityinvolvement and control through legislation.

1Department of Forestry and Land Conservation RegencyArea Level II, East Sumba, Indonesia

87

oleosa, Melaleuca cinamomun, cane and various otherforest woods. While the fauna potential of endemicbirds, among others, are the Sumba button quail(Burung Puyuh Sumba), Sumba green pigeon(Merpati Higau), red-naped fruit dove (Merpati bulukuduk merah), Sumba cicada bird (Burung penangkaplalat coklat), and Cacatua sulphurea citrinocristata(Kakatua Cempaka).

Dividing Factors of the Causes of Forest Fires

GeneralForest resources that have such a large function andmany benefits must be guarded from all types ofintrusions that destroy. One such element is fire. Outof control fires or burning can cause massivedisturbances to forests, this is the destruction offorests and the shattered product of many generations.

Fire has many bad impacts. Damage can range fromdisturbances to the trunk of trees to the destruction ofthe whole tree. Burning weakens resistance to pestsand diseases. Burns on tree trunks create chances forpests and diseases to break through. The destruction ofone generation of forests by fire means a loss of timethat is needed to reach the standard of coveragementioned previously.

Soil characteristics are also destroyed because ofthe loss of humus and organic soil matter, and in timethe soil is exposed to the effects of heat and wind. Theland becomes infertile and erodes, becomes porousand water levels drop. The loss of manure from thefire causes the fertility of the land to drop, and alsodestroys the priceless habitat of the wild animals.

Basics about forest firesFactors supporting the existence of forest fires areheat, fuel and atmosphere. These three elements areusually drawn in a fire triangle. The spreading of firedepends on fuel and climate (wind, moisture andtemperature). Strong winds speed up the spreading offire by increasing the air supply. Temperature is alsoan important factor, because not only does it affect thefuel and the movement of air or wind, but also has aneffect on the fire brigade itself. In hot weathersituations the capacity of the fire brigade to work andit’s abilities decrease.

Causes of fireForest fires in East Sumba are generally created byhumans. Besides this, brush fires on fields result fromvery dry conditions, which occur periodically every year.

Forest fires can be intentionally or unintentionallycreated. Some reasons for unintentional causes are

from smokers, campfires/camping, forest workers,shepherding and hunting. In many cases forest firesstart from the intentional use of fire by those who clearland, workers preparing fields, sheep herders whowish to stimulate grass growth, hunters herding preyand others.

Forest and field fires actually represent anoccurrence that is linked to the lifestyle of theSumbanese community which places a high value onlivestock and hunting activities and also overcomingfood crises in the dry season with crop rotation andweak farming management. The root of the problem isthe social prestige connected with owning livestock.

During the famine or food shortage seasons,marginal farmers in general make efforts in huntingwild boar and deer, which are viewed by thecommunity as nutritious foods, also by searching theforests for tubers/roots to satisfy their carbohydraterequirements as substitutes for rice and corn.

To make the catching of wild boar and deer easier,and to restimulate the growth of tubers/roots, theforest and scrub is burnt. Because this burning is notcontrolled it can spread to areas where it was notintended.

Large proportions of forest fires originate from themethod of burning for dry land use at the village level.The clearing of forests by the community is usuallydone according to customs, to provide land foragricultural products that continue every year (rice,corn, various kinds of tubers and various kinds ofbeans). The average amount of land cleared by thecommunity is restricted to their own plantations(approx. 0.5 hectares per family), but because thisburning is not controlled fire then spreadseverywhere. This is what causes forest areas and otherpastures to burn as well.

The planting of new forests can also result in theexpansion of brush fires. Fire is used for the clearingand preparation of land for planting by theintercropping method. Often fire spreads tosurrounding forest areas and causes forest fires.

Impact of Forest FiresForest fires are not only considered the main obstaclein the production of forests, but also the main factorthat causes a lowering in quality of life. A studyconcerning plant dynamics has shown that species thatare inclined to regrow from seeds or stumps tend toappear again in burnt-out locations. With this, speciesthat easily burn will appear again and create adangerous type of fuel in that particular area.

Fires that happen every year have resulted in theflourishing of Chromolaena odorata (Tai Kabala) inparts of East Sumba. The flourishing of this plant, forthe moment, is viewed by farmers as a weed that

88

greatly disrupts agricultural efforts as well as being ahiding place for forest pigs (pests). In accordance withthe culture of the area, Chromolaena on agriculturalland is overcome by burning. This activity is also seenas the only exact way of clearing weeds.Unknowingly, this condition has hastened the spreadof Chromolaena, which at this moment has penetratedrare forest areas, resulting in suppression of growth ofthe natural succession of seedlings.

Other dangers from forest fires are the destructionof land fertility and the increase of erosion. Areaswhere the tops of slopes are burnt tend to decrease thecapacity of water catchment areas below. There havebeen many observations stating that the decrease inquality of land areas arises from repeated burning thatresults in soil erosion and floods, which subsequentlyimpact on water ways, rivers, lakes and weirs.

The impact of fire which is felt the most is the risingof air temperature (heat) because the production ofoxygen through tree leaves becomes restricted.

The burning of forests and fields results in thedestruction of microorganisms that fertilise the land.This burning also decreases land productivity, whichin turn causes community agricultural yields todecrease. To increase land productivity then requiresfertilisation, which of course requires a high cost. Theburning of forests and fields also causes the loss oftopsoil. The water table level drops, which results inthe decrease in the amount and rate of water flow insprings, rivers and lakes.

Fires, which occur at the peak of dry season destroythe fertilisers and topsoil microorganisms that areimportant in the decaying process.

Efforts in Forest Fire PreventionPrevention is better than cure. This statement is alsorelevant to forest fires, where with a good preventionprogram, fires don’t need to occur, and with that thecost of fighting them along with the damage causedcan be avoided. This can be achieved througheducation, siliviculture practices and preventativelaws. A number of efforts in forest fire preventionhave already been undertaken. Among these are:

InformationBecause humans cause most forest fires, whether fromcarelessness or intentionally, with support andcooperation the community becomes very importantfor protection programs to work.

This requires repetitive information given to changebehaviour as well as the encouragement of communityinterest towards forests and making them care aboutforests. Information can be conveyed through:• the use of community figures who are organised to

work for forest fire prevention;

• mass media publications in the area;.• audiovisual publications;• circulated letters;• the installing of notice boards.

Silvicultural practicesGood silvicultural practice involves clearing of landfrom time to time and the removal of dead trees/deadvegetation. It is important to note the role of certaintrees in lowering the risk of fires. It is then possible tochoose species for planting that are fire-resistant/enduring of fire (examples are teak, Gmelinaarborea, Schleora oleosa, tamarind).

Local knowledgeUnderstanding local culture that has been handeddown the generations—for example the presence ofsacred forests, which are considered places wheregods reside. This is belief of marapu communities.The forest represents a place of prayer for marapucommunities to praise the gods. This belief that isfollowed by the marapu communities states thatwhoever destroys the place/forest mentioned willreceive the punishment of the gods.

Cultural oaths‘Rotu Pandang’ is one such oath that still exists in thecommunity. It means the community is not allowed toburn forests or fields. Because the whole communityof a village pledges rotu pandang, then whoever breaksthis pledge will receive punishment/misfortunes fromthe gods.

Formation of groups for safeguarding forests These groups for safeguarding forests are formed bycommunity groups themselves and are facilitated by anumber of community non-profit organisations, theForestry Department and the Land ConservationCouncil of Regional Area Level II East Sumba.

The aim of forming these groups is to assist thegovernment in safeguarding forests in its area fromvarious threats like forest fires, theft of forestproducts, clearing of forests and unauthorisedshepherding. For example, in Wanggameti village theKMPH, ‘Paberingu Mahanu’, has succeeded inpreventing forest fires in their village for the past6 years, so that now secondary forests have grown,especially in the areas around hills surrounding thevillage.

Attention of community non-profit organisationsA number of local and outside non-profitorganisations care greatly about forest conservationon the Island of Sumba. Among them are TananuaFoundation, Mbaha Eti Foundation, Kehati

89

Foundation, World Neighbors (WN), Bird Life,WWF, Cornell University and others. The presence ofthese institutions has greatly helped prevention offorest fire activities on the Island of Sumba.

Preventative lawsThe regulation of legislation has a very important rolein the prevention of fires. Pushing disciplined fireusage is much needed, especially for those who tendto keep violating the laws. The community needs to begiven information and educated about these laws.Because of this preventative laws still represent a wayto guarantee prevention and they are aimed at thosewho don’t care.

If the laws are upheld and the punishments for thosewho violate them are announced or published, thiscould have an impact of lessening the occurrence offires. Even though the enforcement of laws representsone important part of preventing forest fires, thisshould be viewed as an educational tool that must beput to use skilfully and wisely.

Fire breaks/green stripsFire breaks/green strips are only applied torehabilitated forest plantations. Fire breaks/greenstrips approximately 10 metres wide are made aroundthe plantation location with the intent of preventingthe spreading of fire into rehabilitated locations.

ConclusionForest fires represent one factor that causes loss offorest potential. Reality shows that if fire spreads it ishard to control. Forest fires have a great impact on thesocial economy, ecosystem and hydrology. Becauseof this, prevention/control of fires must continue to beexecuted to create forests and ecological reservations.This valuable lesson represents a challenge and achance to care for the social economy and maintainthe water and nutrient cycles in certain areas to ensurethe continuing function and benefit of the forests.

90

THE TRANS-FLY IS a biogeographic entity thatcomprises the area of monsoonal climate which liesbetween the Aramia and lower Digul Rivers insouthern New Guinea. Annual rainfall varies between800 and 2000 mm, most of which falls in the wetseason from December to May (Paijmans et al. 1971a;McAlpine et al. 1983). However, there is a gradient inwhich annual rainfall decreases and the dry seasonbecomes more pronounced southwards. The coastalplain and floodplains of the major rivers are mostlyless than 3 m above sea level. Inland the flat to slightlyundulating Oriomo Plateau is largely below 30 mabove sea level (Paijmans et al. 1971b) .

The vegetation consists of grassland, permanentswamp, savanna woodland and monsoon forest,distributed in a mosaic which depends on rainfall, soiltype and duration of seasonal inundation, and thehistory of fire and other human activities (Paijmans1971,1976). This complex of vegetation typescorresponds to Gillison’s (1983) bioclimatic ProvinceNo. 4-AIVS. The alternation of dry and wet seasonsand the dominance of grass in the herb layer ensuresthat fire is an annual event (Henty 1982). Manycomponents of the Trans-Fly savanna biota are sharedwith those of northern Australia, but there is also astrong contribution from New Guinea (Walker 1972;Axelrod and Raven 1982; Gressitt 1982).

The following account is necessarily anecdotalsince, apart from the descriptive works referred to

above, there has been little empirical research onvegetation ecology and the behaviour and effects offire in the savannas of southern New Guinea. Unlessexplicitly stated, the observations were made by theauthor during 1979–1982 and 1995, and have notpreviously been published.

Fuel TypesGrass is the main fuel for fires in the Trans-Fly. Fuelloads vary from sparse low swards in the shade of thedenser types of woodland or under heavy grazing to talldense reedbeds of Phragmites karka in seasonalswamps. There are no figures for grass standing cropfor southern New Guinea. Extensive areas of woodlandand grassland are seasonally inundated by flood water.Extensive open grasslands exist in two main types:those on black soils subject to prolonged inundation;and sedge–grasslands on impermeable soils.

There are no measurements of fire behaviour inTrans-Fly savannas but the following generaldescription is possible. Fallen dry leaves and bark ofMelaleuca spp. are important in the spread of firewhere the grass layer is sparse. During early burningoperations in Wasur National Park, Irian Jaya, thesefuels supported more intense fires than those in grassswards. A litter of dry leaves allows late dry seasonfires to enter monsoon forest and thicket. These firesare usually slow-moving and with low flame height.Frequent grass fires in woodlands ensure that dead

Fire in the Trans-Fly Savanna, Irian Jaya/PNGNeil Stronach1

AbstractThe vegetation of the Trans-Fly in southern New Guinea is a complex mosaic of grassland, savannawoodland and monsoon forest, the distributions of which are determined partly edaphically and partlyby anthropogenic fire. Human population density has been low historically and remains low in mostareas. A strong traditional use of fire includes fires set for hunting, to protect sensitive resources, andto maintain grassland and woodland at the expense of denser woodland and monsoon forest. Failureof the annual rains may lead to drying out of swamps, followed by severe fires in swamp grassland,woodland and forests and the combustion of peaty soils. Social and cultural changes have introduceda suite of non-traditional uses of fire, mostly for reasons of access to remote areas for commercialpurposes, and indifference to traditional management goals. In conjunction with heavy grazing byrusa deer, burning has eliminated dense Phragmites reedbeds, setting in train a number of majorecological changes which lead to the widespread encroachment of seasonal swamp grassland bymelaleuca woodland and forest. Rare severe burning of melaleuca swamp woodland may reverse thistrend, but fuel loads in sedge-grassland are insufficient to prevent the growth of even small melaleucaregeneration.

1Fota Wildlife Park, Carrigtohill, Co. Cork, Ireland

91

wood does not accumulate. However, severe fires inmelaleuca swamp woodland and forest may kill manytrees, providing large quantities of fuel in subsequentyears.

Although grass standing crop on higher ground mayincrease as annual rainfall increases (Fitzpatrick andNix 1970; Nix 1975), the lower the rainfall in any oneyear, the greater the area of swamp that dries outsufficiently to burn. Not only does herbage becomeavailable as a fuel for fires, but so also do woodyvegetation and the organic deposits in peaty soils. Theseverity of fires increases markedly when the wetseason rains fail, as in 1978–79. As grass standingcrop increases so also does the likelihood of burning.The rate of fuel accumulation in Trans-Fly grasslandshas not been measured but may approximate toWalker’s (1981) model in which fuel accumulates upto a maximum after four years, and to the situationdescribed by Williams et al. (1999) for the similarsavannas in northern Australia.

Several factors affect fuel loads independently ofprimary production and the progressive seasonaldrying. On the coastal plain there is, locally, intensivegrazing by agile wallabies (Macropus agilis) andexotic rusa deer (Cervus timorensis), reducing fuelloads. Elsewhere, heavy grazing by cattle induces thereplacement of grass by a sparser forb community.This effect can be seen in parts of Wasur NationalPark east of Tomerau. Cattle rearing is promoted inthe Trans-Fly of Irian Jaya, but in Papua New Guineathere has been a policy of keeping livestock awayfrom the international boundary for the purposes ofdisease control.

Grassland and swamps may be heavily dug over byferal pigs (Sus scrofa x Sus celebensi), particularly inareas remote from human habitation. Pig-rootedground supports less grass, forming fire breaks andseed beds for the regeneration of Melaleuca spp.,promoting the spread of woodland and forest at theexpense of grassland and open swamps.

Anthropogenic FireTraditionally, the indigenous people of the Trans-Flycombine subsistence hunter-gatherer and small-scalecultivator lifestyles. Fire has been used extensively tomanage their resources. Early burning is used toprotect key fire-sensitive resources such as gardens,and coconut and sago palm groves from damaging latefires. Most fire in the Trans-Fly is anthropogenic.However, the traditional use of fire in the Trans-Flyhas not been systematically described.

Fire is used opportunistically in hunting. Stands oftaller grasses are burned to concentrate and flush outbandicoots, wallabies and other species. This activitytends to produce a progressively burnt landscape,beginning each dry season with small-scale early

burns and ending with small-scale hot burns in relictpatches of fuel. Post-burn sprouting of perennialgrasses may attract grazers, enhancing opportunitiesfor hunting. No attempt is made to control huntingfires once lit. This suggests a high degree ofconfidence that fires lit for one purpose will be usefulin other ways or will not be damaging.

In discussing the use and effects of fire, indigenouspeople of the Trans-Fly express satisfaction at the‘cleaning’ of the land by fires, a sense of purpose alsoobserved among the aboriginal people of northernAustralia (Russell-Smith et al. 1997) and theWangindo people of southern Tanzania (Stronachunpubl.). This cleaning relates to the removal of litter,the suppression of shrubs and tree regeneration, andthe subsequent regrowth of grass species that supportgrazers such as the agile wallaby. Thus fire is used tomaintain grassland, open woodland and a mosaic ofthese with monsoon forest and thicket, a mixture ofhabitats which supports a higher diversity and densityof prey and food plant species. The removal of thegrass sward makes walking easier and hazards such assnakes and feral pigs more visible.

Of particular interest in the area of traditionalburning practices is the attitude towards severe fires.Wet season rainfall has decreased and become lessregular since the virtual failure of the rains in the1978–1979 wet season (Kitchener 1997). The severefires of the 1979 dry season in the Tonda WildlifeManagement Area caused considerable alarm amongthe local people, suggesting that there had been littleexperience of these conditions in living memory. Thesounds of exploding melaleuca trees weremisidentified, serious injuries were sustained whenpeople walked into smoldering swamp soils, andpeople cultivating remote gardens were alarmed at thesevere grass fires.

The present situation in the Trans-Fly is one inwhich traditional fire management is being replacedby less structured practices. Recent administrationshave encouraged the concentration of a formerlydispersed population, reducing traditional earlyburning in remote areas. In Irian Jaya there has beenan influx of settlers and urbanisation around Merauke.The younger people are less clear about the aims oftraditional burning and, with the intrusion of outsideinfluences and the cash economy, the need to continuethem has declined. If the traditional burning practicesare to be understood the older people must beinterviewed and asked to demonstrate theirtechniques.

The new burning practices have much to do withease of access. Hunters who supply the town ofMerauke with deer meat use motor bikes for huntingand transporting meat, burning extensively whereverthey operate, particularly in the now relict Phragmiteskarka reedbeds. Cattle herders burn wherever the

92

grass sward is not overgrazed and can carry a fire.Increased access to the area by way of the Trans-IrianHighway and other new roads and tracks ensures thatthe new practices are applied widely. My impressionis that the new practices are less discriminate and,while progressive, are unlikely to protect somehabitats. The management of Wasur National Park hascarried out some early burning in the woodlands andto protect Asteromyrtus symphyocarpa regenerationfor the commercial extraction of rubbing oils.

Effects of Fire on VegetationThe most obvious effect of fire in the Trans-Fly hasbeen to prevent the ascendancy of woody vegetationat the expense of grassland. The absence of fire inTrans-Fly woodlands allows the development of fire-sensitive monsoon forest, a change whichsubstantially alters the resource base of people. Thedevelopment of monsoon forest in woodland can beobserved widely in Wasur National Park despitecontinued access by people, and between Dimisisi andthe Fly River where people no longer live. In openwoodland maintained by fire, monsoon forest initialsare nevertheless conspicuous. For example, forestinitials frequently take the form of small thickets ofDillenia alata, often growing on eroded termitaria.There are no estimates of the rate of invasion of Trans-Fly savannas by monsoon forest, but interviewingresidents may provide useful historical information onthis aspect.

Fire enters monsoon forest and thicket duringexceptionally dry years, when leaf litter provides asubstantial fuel. There was considerable tree deathand consequent opening of the canopy following firesin these habitats in Tonda Wildlife Management Areaduring the dry season of 1979. The subsequent growthof grass (mostly Imperata cylindrica) provided thefuel for further fires the following dry season, leavingalive only those species of tree that would be expectedto survive in woodland (e.g. Acacia mangium, Alstonascholaris). Woodland replaced monsoon forest withintwo years under the influence of fire.

Fire is also implicated in changes in grassland,notably in conjunction with grazing by rusa deer. Thepeople of the coastal plains indicate that in livingmemory, and coinciding with the establishment of therusa deer, there have been major changes in theswamps and grasslands. Robust swamp grasses,notably Phragmites karka, have declined. Seasonalinundation has been neither so deep nor so long. Largeareas of grassland and seasonal swamp have beenencroached mostly by Melaleuca spp. (Stronach 1995;Kitchener 1997).

Phragmites is eliminated by deer grazing inconjunction with burning, being replaced by a variety

of swamp grassland types. Melaleuca does notregenerate in dense phragmites but does on the shortgrasslands that replace it, particularly following heavygrazing which reduces fire intensity. These extensivevegetation changes appear to be reinforced by rainfallchanges, and to have set in train significant changes inhydrology which in turn accelerate the vegetationchanges (Stronach 1995; Kitchener 1997). Burning inconjunction with deer grazing eliminates Imperatacylindrica which is often replaced by short swards ofChrysopogon aciculatus (Henty 1982; Stronach 1995;Kitchener 1997). Traditional burning regimes inphragmites and imperata before the colonisation ofdeer probably did not have these effects since, in theabsence of grazing both phragmites and imperatarecover to their former density following burning.

More research is required to provide empirical dataon all aspects of vegetation fires. We lack estimates ofgrass standing crop and its variations in space andtime. Similarly, fire behaviour, distribution and timingremain unmeasured. The greatest need is for a surveyof past and present fire management objectives asseen by the traditional land managers and by thosenewcomers who also burn grass. This is a particularlyurgent task since the most reliable information on thetraditional use of fire rests with the oldest people.Without an analysis of traditional burning practicesand their social and resource management contexts, itwill be difficult to interpret the changes in vegetationcomposition and structure and in hydrology that haveoccurred in the Trans-Fly during the past few decades.The role of fire in vegetation dynamics, particularly itseffects on woody vegetation growth and its interactionwith grazing in causing vegetation changes, alsoneeds considerable study.

AcknowledgementsMy field experience was gained while working withthe Livestock Division, Department of PrimaryIndustry, Papua New Guinea, and with the WasurNational Park Project (WWF Project No. ID 0105),Irian Jaya. The Governors of Fota Wildlife Park,Ireland, and WWF Australia supported myparticipation in this workshop.

ReferencesAxelrod, D.I. and Raven, P.H. 1982. Paleobiogeography and

origin of the New Guinea flora. In: Gressitt, J.L., ed.,Biogeography and Ecology of New Guinea. TheHague, Junk, 919–941.

Fitzpatrick, E.A. and Nix, H.A. 1970. The climatic factor inAustralian grassland ecology. In: Moore, R.M., ed.,Australian Grasslands. Canberra, Australian NationalUniversity Press, 3–26.

93

Gillison, A.N. 1983. Tropical savannas of Australia and thesouthwest Pacific. In: Bourliere, F., ed., TropicalSavannas (Ecosystems of the World, 13). Amsterdam,Elsevier, 183–243.

Gressitt, J.L. 1982. Zoogeographical summary. In: Gressitt,J.L., ed., Biogeography and Ecology of New Guinea.The Hague, Junk, 897–

Henty, E.E. 1982. Grasslands and grassland succession inNew Guinea. In: Gressitt, J.L., ed., Biogeography andEcology of New Guinea Vol. I. The Hague, Junk,459–473.

Kitchener, D. 1997. Wasur National Park — Tonda WildlifeManagement Area. Biodiversity Research PlanningWorkshop, Jakarta, WWF Indonesia Programme.

McAlpine, J.R., King, G. and Falls, R. 1983. Climate ofPapua New Guinea. Canberra, Australian NationalUniversity/CSIRO.

Nix, H.A. 1975. Environmental control of breeding, post-breeding dispersal and migration of birds in theAustralian region. Proceedings of the 16th InternationalOrnithological Congress 1974, 272–305.

Paijmans, K. 1971. Vegetation, forest resources and ecologyof the Morehead-Kiunga area. In: CSIRO LandResearch Series No. 29, Land Resources of theMorehead-Kiunga Area, Papua and New Guinea.Melbourne, CSIRO, 88–113.

Paijmans, K. 1976. Vegetation. In: Paijmans, K., ed., NewGuinea Vegetation. Amsterdam, Elsevier, 23–105.

Paijmans, K., Blake, D.H. and Bleeker, P. 1971a. Summarydescription. In: CSIRO Land Research Series No. 29,Land Resources of the Morehead-Kiunga Area, Papuaand New Guinea. Melbourne, CSIRO, 12–17.

Paijmans, K., Blake, D.H. and Bleeker, P. 1971b. Landsystems of the Morehead-Kiunga area. In: CSIRO LandResearch Series No. 29, Land Resources of theMorehead-Kiunga Area, Papua and New Guinea.Melbourne, CSIRO, 18–45.

Russell-Smith, J., Lucas, D., Gapindi, M., Gunbunuka, B.,Kaparigia, N., Namingum, G., Lucas, K., Giulani, P.and Chaloupka, G. 1997. Aboriginal resourceutilisation and fire management practices in westernArnhem Land, monsoonal northern Australia: notesfrom prehistory, lessons for the future. HumanEcology, 25, 159–195.

Stronach, N. 1995. The rusa deer of Wasur National Park,Irian Jaya. Typescript, WWF Project No. ID 0105, 3 p.

Walker, D. 1972. Bridge and barrier. In: Walker, D. ed.,Bridge and Barrier: the Natural and Cultural History ofTorres Strait. Research School of Pacific Studies,Publication BG3, Australian National University,Canberra, 399–405.

Walker, J. 1981. Fuel dynamics in Australian vegetation. In:Gill, A.M., Groves, R.H. and Noble, I.R., ed., Fire inthe Australian Biota. Canberra, Australian Academy ofScience.

Williams, R.J., Cook, G.D., Gill, A.M. and Moore, P.H.R.1999. Fire regime, fire intensity and tree survival in atropical savanna in northern Australia. AustralianJournal of Ecology, 24, 50–59.

94

Fire Management Issuesin Northern Australia

95

VAST TRACTS of the sparsely inhabited savannalandscape of northern Australian are burnt each yearby pastoralists, Aboriginal landholders, andconservation (national park) managers, particularly inthe Top End of the Northern Territory and thenorthern Kimberley region (e.g. Press 1988; Head etal.1992; Allan and Willson 1995; McMillan et al.1997; Russell-Smith et al. 1997). Elsewhere, butparticularly on pastoral tenures, fire is little used aspart of property management; rather, it is often seen asa threat to infrastructure and pasture. As a resultsignificant concerns have been expressed with respectto the cultural, economic, and ecologicalsustainability of contemporary fire regimes in thatregion (e.g. McDonald and Batt 1994; Rose 1995;

Grice and Slatter 1997; Jacklyn and Russell-Smith1998; Saint and Russell-Smith 1998).

In parts of northern Australia dominated byfrequent, extensive fires occurring typically in thelatter half of the dry season under severe fire-weatherconditions, various studies have demonstrated thatsuch contemporary fire regimes incur significantimpacts on fire-sensitive native species (e.g. cypresspine (Callitris intratropica)—Bowman and Panton1993) and communities (e.g. small rainforestpatches—McKenzie and Belbin 1992; Russell-Smithand Bowman 1992). While no broad consensus iscurrently available as to the long-term impacts/effectsof different fire regimes on eucalypt-dominated forestand woodland savanna (e.g. Bowman et al. 1988;Lonsdale and Braithwaite 1991), it is generallyrecognised that biodiversity is most likely to beconserved in those landscapes where fire regimesthemselves are highly patchy both in space and time(e.g. Woinarski 1990; Braithwaite 1995). Recent, ifcurrently localised, invasion of high biomass,flammable agricultural plants (e.g. the grassesAndropogon gayanus, Urochloa mutica andPennisetum polystachyon) into natural systems, has

Fire Management and Savanna Landscapesin Northern Australia

Jeremy Russell-Smith1,2, Grant Allan1,2,Richard Thackway3, Tony Rosling3 and Richard Smith4

AbstractThe vegetation cover of northern Australia ranges from open forest or woodland savanna (dominatedby eucalypts over a range of highly flammable annual and perennial grasses) to hummock and tussockgrasslands occupying sandy and fertile fine-textured soils. Like monsoonal eastern Indonesia, themajor fire period occurs over the long dry season, typically between April/May–October/November.People light fires for a range of land management purposes; lightning strikes cause relatively few firesat the start of the annual wet season. Based on regional mapping of fires from satellite imagery(mostly NOAA-AVHRR and LANDSAT) from the 1980s, we can identify two broad patternsconcerning the application of fire in northern Australia. In northwestern and northern Australia, andpossibly also on parts of Cape York in the northeast, intense wild fires typically late in the dry seasonburn vast tracts annually. Ecological studies indicate that such fire regimes are having catastrophicimpacts on native fire-sensitive species, communities, and habitats. Conversely, elsewhere acrossnorthern Australia but especially on more productive pastoral lands, the restricted application/absenceof burning is in some cases leading to native and exotic woody species thickening/invading, likewisewith profound ecologic and economic consequences. Growing recognition of these issues has led tothe development of collaborative fire management programs in various parts of northern Australia.Similar cooperative approaches involving practitioners from northern Australia working withrelevant parties in eastern Indonesia and Papua New Guinea would bring significant benefit to thestudy of regional landscape management issues.

1 Cooperative Research Centre for Tropical Savannas,Darwin, Northern Territory, Australia2 Bushfires Council of the Northern Territory, PO Box 37346,Winnellie NT 0821, Australia3 Environment Australia, GPO Box 636, Canberra ACT 2601,Australia 4 Remote Sensing Services, Department of LandAdministration, PO Box 471, Wembley WA 6014, Australia

96

the capacity to significantly increase fire managementproblems in the region over the longer term throughmarkedly increased fire intensities (Lonsdale 1994;Panton 1993).

By contrast, relatively little burning is undertakenin many other regions of northern Australia,especially on more arable lands under pastoralmanagement (Craig 1997; Grice and Slatter 1997). Insome pastoral areas there is growing appreciation thatdensities of woody plants, both native (e.g. Acacia,Eucalyptus) and exotic species (e.g. Acacia nilotica(prickly acacia) and Parkinsonia aculeata), haveincreased markedly in recent decades with majorconsequences both for loss of productive pasture (e.g.Burrows et al. 1990; Hodgkinson 1991; Scanlan et al.1991), and loss of extensive areas of native habitat andits associated species (e.g. golden-shoulderedparrot—Garnett and Crowley 1995). Increase inwoody plants, combined with grazing pressure,reduced burning frequency and the use of intense,shrub-killing fires, have all contributed to habitat loss.

To illustrate the contemporary spatial and seasonalpatterning of burning across northern Australia, in thispaper we provide a landscape-scale assessment of theextent and seasonality of fires during 1997 and 1998.For both years, mapping of fires was undertaken every9 days through the dry season derived from relativelycoarse-resolution (1.1 × 1.1 km pixels at orbital nadir)

NOAA–AVHRR imagery (see McMillan et al. 1997for details). Fires occurring up until the middle of Julyare described here as Early Dry Season (EDS) fires;from late July onwards they are described as occurringin the Late Dry Season (LDS). GeographicInformation System (GIS) analysis of the patterningof EDS and LDS fires is undertaken here with respectto broad-scale coverages describing rainfalldistribution in the preceding wet season, vegetation,and land-use.

The definition of the tropical savanna region usedhere originates from the Tropical SavannasCooperative Research Centre and is based on theInterim Biogeographic Regionalisation of Australia(IBRA) (Thackway and Cresswell 1995). The map inFigure 1d is derived from this provincialbiogeographic regionalisation.

Climate, Vegetation, Land Use, andBiogeographic Regionalisation

Broad-scale northern Australian rainfall distribution,vegetation, land-use and biogeographic features aredescribed in Figure 1 (a-d), respectively. Rainfall,which occurs over the savanna region mostly betweenOctober and March under the influence of the Asianmonsoon, declines rapidly southwards, from over

PastoralAboriginalConservationDefenceMarine ConservationVacant Crown Land

Kimberley/VRDTop End/GulfCape YorkArid ZoneMitchell GrasslandsCairns CoastSoutheast QueenslandNSW Coast

2400–3200mm2000–2400mm1600–2000mm1000–1200mm800–1000mm600–800mm400–600mm300–400mm200–300mm<200mm

Closed ForestForest - EucalyptForest - Acaia (lancewood/brigalow)WoodlandOpen WoodlandShrublands - AcaciaShrublands - Chenoped/OtherSpinifex - sandy soilsSpinifex - rocky soilsTussock Grasslands - FloodplainTussock GrasslandsLittoral Complex & Lakes

Figure 1. Maps of annual rainfall distribution, vegetation, landuse class, and biogeographic regionalisation, for the Australiantropical savannas region (indicated north of black line on all maps), where: (a) mean rainfall isohyets, 1969–1996(source: Bureau of Meteorology, quarter degree grid-cell digital database); (b) major vegetation communities, derivedfrom Australian Surveying and Land Information Group (1990); (c) land-use classes derived from Australian Surveyingand Land Information Group (1993); and (d) Interim Provincial Regions, from Environment Australia (1996).

a. c.

b. d.

97

2000 mm per annum in a few coastal areas to around500 mm per annum inland (Figure 1a). Although theamount of rainfall received in any one area is highlyvariable from year to year, the wet season is a highlyreliable event (Taylor and Tulloch 1985). Suchclimatic conditions are conducive to the production ofgrassy fuels sufficient for carrying ground fires on anannual basis in higher rainfall areas, to once every fewyears under lower rainfall conditions (Walker 1981).

In northern parts of the tropical savannas,vegetation cover is mostly eucalypt-dominatedwoodland developed on a range of typically nutrient-poor soils, becoming increasingly open-canopied andlower in stature with declining rainfall (Figures 1b and2). In the south, woodland savannas give way to thevast hummock grasslands of the central Australiandune fields. Other regionally significant vegetationtypes include: (1) pastorally productive tussock (or‘Mitchell’) grassland communities predominantly inwestern Queensland, with restricted areas in theNorthern Territory and Western Australia; (2)scattered hummock grassland communities developedon rocky infertile substrates (e.g. sandstone), withsignificant components of spinifex (Triodia spp.) anda range of typically fire-sensitive shrubby species.Rainforest communities are confined mostly to thehumid tropics of northeastern Queensland, elsewhere

occurring as small patches scattered within thesavanna mosaic.

Generalised land use of the savanna lands ofnorthern Australia is given in Figures 1c and 3. Asindicated below it is useful to distinguish between‘land-use class’ and the more problematic issue oftenure, especially given recent (if overdue) legalrecognition of the rights of aboriginal people to sharetitle (‘native title’) in a range of tenure situations. Thegreat majority of land is used ostensibly for pastoralproduction, especially the grazing of cattle (Bostaurus and Bos indicus), and also sheep in parts ofwestern Queensland. Most of this land is leasehold(i.e. leased from respective State and NorthernTerritory governments), although in the NorthernTerritory leasehold pastoral properties whenpurchased by aboriginal people have, from the mid-1970s until recently, been claimable as freeholdfollowing successful demonstration of unbrokentraditional relationships to the land in questionthrough the highly legalistic ‘land claim’ process.

Also, in Western Australia, currently around 25%of pastoral properties in the northern savannas areleased by aboriginal-owned enterprises. Landsallocated to freehold aboriginal tenure in recognitionof traditional/cultural usages (i.e. for no specificeconomic purpose) constitute the next most common

0

50000

100000

150000

200000

250000

300000

Are

a(k

m2 )

Vegetation community

Western Australia 91148 93631 16105 5341 13666 103379 656Northern Territory 43631 210789 115301 9960 83602 7055 34957 98365 18177 8807 3118Queensland 55492 271932 213524 44500 250101 8262 1175 84996 7832 11625 4129 15970

Eucalypt forest Woodland Open

woodlandAcacia shrub- lands

Tussock grasslands

Littoral & lakes

Spinifex : sandy soil

Spinifex : rocky soil

Acacia forest

Floodplain tussock

grasslands

Chenopod / mixed

shrublandsClosed forest

Figure 2. Extent of major vegetation communities occurring in the tropical savannas region of Western Australia, NorthernTerritory and Queensland.

98

land-use type, especially in the Northern Territory.Finally there are unallocated lands—Governmentlands (i.e. ‘vacant crown land’) especially in WesternAustralia—claimable through the ‘native title’process. Also, in the Northern Territory, lands setaside for conservation purposes include significantareas under aboriginal freehold tenure (e.g. Kakaduand Nitmiluk National Parks), managed jointly withFederal and Territory Government agencies,respectively. Despite the small area used for miningpurposes, such land use constitutes by far the greatesteconomic return to the regional economy (Gray1996).

Patterning of Fires in Northern Australia,1997–1998

The mapped distribution of fires occurring in northernAustralia in the EDS and LDS of 1997 and 1998 ispresented in Figures 4 a, b respectively, and thebreakdown by State and Territory jurisdictions isgiven in Figure 5. In sum, the data indicate that pixelsdepicted 244 000 km2 and 242 000 km2 respectivelyburnt (at least partly) in consecutive years, with thegreatest extent, and largest fires, occurring in the LDSin both years. These data have a number ofqualifications, including difficulties with mappingfires under cloudy conditions at the start of the wetseason (Oct/Nov), and estimating the amount actuallyburnt in any one 1.1 × 1.1 km2 pixel. Nevertheless, the

data illustrate that relatively little burning occurred inQueensland in both years, especially in the moreproductive pastoral areas of the tropical savannas.Other data available for the Northern Territory andWestern Australia indicate that such observations arebroadly typical of other years in the 1990s, for thesejurisdictions at least (Allan and Willson 1995;McMillan et al. 1997).

The extent of burning by rainfall class for thepreceding wet season shows that both in 1997 and1998 the majority of burning consistently occurred inthe LDS across the region (Figure 6 a, b). Consideringonly the more extensive vegetation types (see Figure2) it is apparent that, in both years, most burningoccurred in eucalypt-dominated open woodlandhabitats under relatively high rainfall conditions, andleast burning occurred on pastorally productivetussock grasslands (Figure 7). The relative lack ofburning undertaken on pastoral lands is evident alsowhen considered by land-use class (Figure 8).

DiscussionFire is an active partner in northern Australiansavannas, as intrinsic as climate, substrates/soils, waterrelations, vegetation/habitats, people/cultures/politics,land uses. Burning, in some regions of northernAustralia, takes place annually/biennially; in otherareas, particularly on pastoral lands, it is currentlylittle used. Both situations have implications for

0

10

20

30

40

50

60

70

80

90

100 Pr

opor

tion

ofno

rther

nsa

vann

are

gion

(%)

Landuse class

WA 1.807062476 7.96709352 66.7507955 4.434613892 19.04043461 NT 0.559102006 28.57328922 63.83848941 5.434725635 1.594393732 QLD 0.105487315 2.253271406 92.43765076 2.942524483 2.261066035

Defence Aboriginal Pastoral Conservation VCL

Figure 3. Proportion of five major land-use classes occurring in tropical savannas region of Western Australia, NorthernTerritory and Queensland; VCL = Vacant Crown Land.

99

ecologic and economic resource sustainability. Firecan be managed, usefully and sustainably, for variouspurposes.

Effective fire management requires appropriateresearch, education, sound policy and infrastructure. Itrequires also cooperation between adjoininglandholders/managers; in northern Australia at least,these are relatively few in number. And yet,substantial cultural differences apply concerning themanagement of land. In this paper we have simplifiedthese to a number of classes: pastoral, aboriginal,unallocated, conservation, defence, mining.Ecotourism is already a major industry across mosttenures, as is potentially education. The issue oftenure (‘security of tenure’) is intrinsic also—hencethe current legal debate in Australia concerningindigenous (aboriginal) and non-indigenous people.Also the management of lands at scales ranging fromFederal/State/Territory to regions, sectors, properties,habitats, etc.

The above issues apply differently, but equally, ineastern Indonesia and PNG. In northern Australia wehave commenced a number of regional firemanagement projects that involve the participation oflocal communities, across a range of land useactivities/tenures. In the Northern Territory, forexample, we are undertaking two such projects: onefocusing on aboriginal lands with high conservation,tourism, mining, and lesser pastoral, values (seeCooke, these proceedings); the other on landsprimarily pastoral, also involving aboriginal,

< 600600 - 800

800 - 10001000 - 1200

1200 - 16001600 - 2000

> 2000

EDS

LDS

0

5

10

15

20

25

30

35

Prop

ortio

nof

Are

aof

Rai

nfal

lCla

ss(%

)

1996/1997 Rainyear Rainfall Class ( mm )

< 600600 - 800

800 - 10001000 - 1200

1200 - 16001600 - 2000

> 2000

EDS

LDS

0

5

10

15

20

25

30

35

Prop

ortio

nof

Are

aof

Rai

nfal

lCla

ss(%

)

1997/1998 Rainyear Rainfall Class ( mm )

Figure 4. Extent of mapped fires in the tropical savannasregion for the Early Dry Season (EDS—green) andLate Dry Season (LDS—red) of (a) 1997, (b) 1998.

0

20000

40000

60000

80000

100000

120000

Are

aof

Stat

eB

urnt

(km

2)

EDS 13706 17325 44151 60761 8757 8819

LDS 36572 54290 113649 69480 27540 31607

WA 1997

WA 1998

NT 1997

NT 1998

QLD 1997

QLD 1998

Figure 5. Extent of Early Dry Season (EDS) and Late DrySeason (LDS) burning in 1997 and 1998 for thetropical savannas region, Western Australia,Northern Territory, Queensland.

Figure 6. Proportion of tropical savannas region burnt inrelation to rainfall for the preceding wet seasonperiod, in (a) 1997, (b) 1998.

a

b

100

conservation, defence, tourism, but limited miningvalues (see Dyer, these proceedings). These projectsaim to develop options for better fire managementpractice in these regions through engagement,community participation and, where appropriate,applied research. Managing fire is of course only oneland management issue; nevertheless itconcerns/involves everyone.

ReferencesAllan, G.E. and Willson, A. 1995. Learning more about fire

in the Top End: processing and interpreting spatial andtemporal fire history data derived from NOAAAVHRR satellite images. NARGIS 95: 2nd NorthAustralian Remote Sensing and GeographicInformation Systems Forum. Supervising Scientist,Commonwealth of Australia, Canberra, 21–26.

Australian Surveying and Land Information Group 1990.Vegetation—Atlas of Australian Resources, DigitalDataset, AUSMAP, Department of AdministrativeServices, Canberra.

0

5

10

15

20

25Pr

opor

tion

ofV

eget

atio

nC

omm

unity

Vegetation Community

Burnt 1997 23.830998 19.470471 11.566175 2.1024605 1.6208892 9.3184239 13.725451 14.024552 2.5375831 14.889985 6.5268387 0.5886036

Burnt 1998 15.648235 20.671442 14.308472 1.0429673 1.1414833 7.9678575 6.285393 11.534491 4.2023915 18.370637 2.2768042 0.9643081

Burnt 97+98 4.7910172 6.6410627 2.3652167 0.183621 0.1017701 1.2005034 0.1044219 1.1180861 0.0153793 7.1699545 0.6347454

Eucalypt Forest Woodland Open

Woodland Acacia

Shrublands Tussock

Grasslands Littoral &

Lakes Spinifex : sandy soil

Spinifex : rocky soil

Acacia Forest

Floodplain Tussock

Grasslands

Chenopod / Mixed

Shrublands

Closed Forest

Figure 7. Proportions of tropical savannas major vegetation communities burnt in 1997 and 1998.

0

5

10

15

20

25

Prop

ortio

nof

Veg

etat

ion

Com

mun

ity

Vegetation Community

Burnt 1997 23.830998 19.470471 11.566175 2.1024605 1.6208892 9.3184239 13.725451 14.024552 2.5375831 14.889985 6.5268387 0.5886036

Burnt 1998 15.648235 20.671442 14.308472 1.0429673 1.1414833 7.9678575 6.285393 11.534491 4.2023915 18.370637 2.2768042 0.9643081

Burnt 97+98 4.7910172 6.6410627 2.3652167 0.183621 0.1017701 1.2005034 0.1044219 1.1180861 0.0153793 7.1699545 0.6347454

Eucalypt Forest Woodland Open

Woodland Acacia

Shrublands Tussock

Grasslands Littoral &

Lakes Spinifex : sandy soil

Spinifex : rocky soil

Acacia Forest

Floodplain Tussock

Grasslands

Chenopod / Mixed

Shrublands

Closed Forest

Figure 8. Proportions of tropical savannas major land-use classes burnt in 1997 and 1998.

101

Australian Surveying and Land Information Group 1993.State and Northern Territory Boundary. DigitalDataset. Australian Surveying and Land InformationGroup, Canberra.

Bowman, D.M.J.S., Wilson, B.A. and Hooper, R.J. 1988.Response of Eucalyptus forest and woodland to fourfire regimes at Munmarlary, Northern Territory,Australia. Journal of Ecology, 76, 215–232.

Bowman, D.M.J.S. and Panton, W.J. 1993. Decline ofCallitris intratropica in the Northern Territory:implications for pre- and post-colonisation fire regimes.Journal of Biogeography, 20, 373–381.

Braithwaite, R.W. 1995. Biodiversity and fire in the savannalandscape. In: Solbrig, O.T., Medina, E. and Silva, J.,ed., Biodiversity and Savanna Ecosystem Processes.Springer-Verlag, Berlin, 121–140.

Burrows, W.H., Carter, J.O., Scanlan J.C., and Anderson E.R.1990. Management of savannas for livestock productionin north-east Australia: contrasts across the tree–grasscontinuum. Journal of Biogeography, 17, 503–512.

Craig, A.B. 1997. A review of information on the effects offire in relation to the management of rangelands in theKimberley high-rainfall zone. Tropical Grasslands, 31,161–187.

Environment Australia 1996. Interim Provincial Regions forAustralia (IBRA regions grouped by Agro-EcologicalRegions). Digital Dataset, Environment Australia,Biodiversity Group, Canberra.

Garnett, S.T. and Crowley, G.M. 1995. Ecology andConservation of the Golden-shouldered Parrot. CapeYork Peninsula Land Use Strategy. Office of theCoordinator General of Queensland (Brisbane),Department of Environment, Sport and Territories(Canberra), Queensland Department of Environmentand Heritage (Brisbane).

Gray, W. 1996. What lies ahead for the tropical savanna?Industries and management regimes. In: Ash, A., ed.,The Future of Tropical Savannas: an AustralianPerspective. CSIRO, Melbourne, 149–158.

Grice, A.C. and Slatter, S.M. 1997. Fire for the Managementof Northern Australian Pastoral Lands. Proceedings ofa Workshop, Townsville, April 31–May 2 1996. MeatResearch Corporation, Sydney.

Head, L., O’Neill, A., Marthick, J. and Fullager, R. 1992. Acomparison of Aboriginal and pastoral fires in thenorth-west Northern Territory. In: Moffatt, I. andWebb, A., ed., Conservation and Development Issuesin Northern Australia. North Australia Research Unit,Australian National University, Darwin 130–144.

Hodgkinson, K.C. 1991. Shrub recruitment response tointensity and season of fire in a semi-arid woodland.Journal of Applied Ecology, 28, 60–70.

Jacklyn, P. and Russell-Smith, J. 1998. Proceedings from theNorth Australia Fire Management Workshop, 24–25March 1998, Darwin. Tropical Savannas CooperativeResearch Centre, Darwin.

Lonsdale, W.M. 1994. Inviting trouble: introduced pasturespecies in northern Australia. Australian Journal ofEcology, 19, 345–354.

Lonsdale, W.M. and Braithwaite, R.W. 1991. Assessing theeffects of fire on vegetation in tropical savannas.Australian Journal of Ecology, 16, 363–374.

McDonald, K.R. and Batt, D. 1994. Fire management onconservation reserves in tropical Australia. Proceedingsof a Workshop, Malanda, Queensland, 16–30 July1993. Queensland Department of Environment andHeritage, Brisbane.

McKenzie, N.L. and Belbin L. 1991. Kimberley rainforestcommunities: reserve recommendations and managementconsiderations. In: McKenzie, N.L., Johnston, R.B., andKendrick, P.G., ed., Kimberley Rainforests Australia.Surrey Beatty & Sons, Sydney, 453–468.

McMillan, C., Craig, R., Steber, M., Adams, J., Evans, F. andSmith, R. 1997. Fire history mapping in the Kimberleyregion of Western Australia 1993 to 1996—methodologyand results. Proceedings of the Australian NationalBushfire Conference (Bushfire 97), Darwin. CSIROTropical Ecosystems Research Centre, Darwin, 257–262.

Panton, W. J. 1993. Changes in post World War IIdistribution and status of monsoon rainforests in theDarwin area. Australian Geographer, 24, 50–59.

Press, A.J. 1988. Comparisons of the extent of fire indifferent land management systems in the Top End ofthe Northern Territory. Proceedings of the EcologicalSociety of Australia, 15, 167–175.

Rose, D.B. 1995. Country in Flames. Biodiversity Unit,Department of the Environment, Sport and Territories(Canberra), North Australia Research Unit, AustralianNational University (Darwin).

Russell-Smith, J. and Bowman, D.M.J.S. 1992. Conservationof monsoon rainforest isolates in the Northern Territory,Australia. Biological Conservation, 59, 51-63.

Russell-Smith, J., Ryan, P.G. and DuRieu, R. 1997. ALANDSAT MSS-derived fire history of KakaduNational Park, monsoonal northern Australia,1980–1994: seasonal extent, frequency and patchiness.Journal of Applied Ecology, 34, 748–766.

Saint, P. and Russell-Smith, J. 1998. Malgarra: Burning theBush. Proceedings of the Fourth North Australian FireManagement Workshop, June 1997, Kalumburu,Western Australia. Gordon Reid Foundation (Perth),Bush Fires Board of WA (Perth), and TropicalSavannas Cooperative Research Centre (Darwin).

Scanlan, J.C., Wilson, B.J. and Anderson, E.R. 1991.Sustaining productive pasture in the tropics. 2.Managing woody vegetation in grazing lands. TropicalGrasslands, 25, 85–90.

Taylor, J. and Tulloch, D. 1985. Rainfall in the wet–drytropics: extreme events at Darwin and similaritiesbetween years during the period 1870–1973. AustralianJournal of Ecology, 10, 281–295.

Thackway, R. and Creswell, I. 1995. An interimbiogeographic regionalisation for Australia: aframework for establishing the national system ofreserves, Version 4.0. Australian Nature ConservationAgency, Canberra.

Walker, J. 1981. Fuel dynamics in Australian vegetation. In:Gill, A.M., Groves, R.M. and Noble, I.R., ed., Fire andthe Australian Biota. Australian Academy of Science,Canberra, 101–127.

Woinarski, J.C.Z. 1990. Effects of fire on the bird communitiesof tropical woodlands and open forests in northernAustralia. Australian Journal of Ecology, 15, 1–22.

102

THE FEDERAL Aboriginal Land Rights (NorthernTerritory) Act 1976 established land councils asstatutory authorities to assist aboriginal people in theNorthern Territory (NT) to claim back crown landsand to manage former aboriginal reserves that becameinalienable freehold aboriginal land under that act.

Aboriginal people constituted about 23% of thetotal Northern Territory population of 195 000 in 1996(ABS 1999) and own a little more than 40% of theland area of 1.35 million km2. There are land claims toa further 10% of the NT (Department of ForeignAffairs and Trade 1995). Aboriginal people make upabout 2.1% of the Australian population of 18 million(ABS 1999) and own about 13% of the 7.7 million km2

land mass (Fourmile 1996). There are approximately 28 000 aboriginal people

in the Northern Land Council (NLC) area and about170 000 km2 of aboriginal-owned land (ABS 1998).Probably only about half those aboriginal people havetitle to their traditional lands. For the rest most of theirtraditional lands have been appropriated by thepastoral industry since the mid-19th century.

Aboriginal people’s relationship with land differsfrom that of the majority of settler Australians. Mostimportantly land is not a commodity to be sold ortraded away but rather it is owned as a sacred trustinvolving discretely defined groups of landowners, their

ancestors and descendants. These ancestors may still beseen in the landscape in particular natural features, oftenreferred to as sacred sites. Aboriginal people who stillhave that deep spiritual attachment to country are ingeneral not interested in moving somewhere else tofollow work or other opportunities—their hearts remainattached to the country wherein lie the bones of theirancestors.

Although most aboriginal groups from the Top Endlands suited to pastoralism lost control of their landsin the mid-19th century, some have regained about40 000 km2 through purchase of pastoral stations.These are mostly economically marginal properties onwhich continuing pastoral enterprise rarely generatessufficient income for appropriate management. Theyprovide little of economic benefit to their owners aspastoral enterprises but are significant in regionaleconomies as well as the social and spiritual lives oftheir aboriginal owners.

The remaining 130 000 km2 of aboriginal land inthe NLC areas is mostly made up of a number ofsmaller aboriginal lands and the largest, formeraboriginal reserve Arnhem Land, which wasproclaimed in 1931 after more than a century of failedEuropean attempts at settlement.

Fire Management on Aboriginal Lands in theTop End of the Northern Territory, Australia

Peter Cooke1

AbstractIn the Top End of the Northern Territory many aboriginal groups continue to practise traditionalmanagement of fire despite a variety of impediments. There is a growing recognition of the positiveeffects of traditional burning regimes but many non-aboriginal settlers still regard aboriginal landscapeburning as anarchic pyromania. This antipathy colours aboriginal perceptions of Government burningregulations and has led some communities to cease traditional burning, believing they may be jailed forlighting fires. The Northern Land Council’s Caring for Country Unit helps landowners address fireproblems and develop sustainable means of appropriate fire management through participatoryplanning processes, collaboration between aboriginal ecological experts and non-aboriginal scientists.The Unit also encourages utilisation of economic projects such mining, tourism and buffalo harvesting,to support traditional fire management and resettlement of unpopulated areas and to build capacitywithin and between aboriginal resource agencies concerned with land management. Future effectivemanagement will depend on finding an appropriate mix of aboriginal and non-aboriginal knowledgeand technology and the resources to sustain application of that mix.

1Northern Land Council Caring for Country Unit PO Box 42921, Casuarina NT 0810

103

Concentration of PopulationsAs Christian churches established missions from thebeginning of the 20th century aboriginal people weredrawn away from traditional lands where they livedindependent lives and were brought into the Christianfold at least partly through attraction to exotic goodslike tobacco and rice. Communal feeding and issuingof rations reinforced the intent of taming the bushAborigines and leading them away from their ownculture towards assimilation into the Euro-Australianculture (Abbott 1950).

Where people weren’t drawn into missions theyoften moved to other sources of trade goods, placeslike the buffalo camps of what is now Kakadu, andcattle stations near to reserves. The effect over timewas progressive depopulation of aboriginal lands anda falling away of aboriginal land managementpractice, including fire, across the broad landscape aspopulations became more concentrated.

Western Arnhem Land people, for example, weredrawn progressively down from the Arnhem Plateaufrom the mid to late nineteenth century, joining thebuffalo camps and mines from the Alligator Riverswestward (Christian and Aldrick 1977).

In other areas on the larger reserves considerablenumbers of groups remained predominantly ontraditional lands, living hunter/gatherer lifestyles untilthe 1950s and early 1960s. By the late 1960s therewere few aboriginal bands living on traditional estateson aboriginal land away from missions andsettlements, leaving many large areas essentiallyunmanaged.

Change of Rules, for a WhileAt that time there was scant recognition that traditionalaboriginal land management was ecologically soundpractice, and little acknowledgment of the broadlyrecognised rules and customary practices ofmanagement amongst Aborigines. Anthropologist andnaturalist Donald Thomson (Thomson 1949) notedsuch customary rules associated with firemanagement. He observed that responsibility fordirecting burning of grass was shared by the old men ofan estate-owning clan (members of which were relatedto that estate and clan through patrifiliation—father–son) and by men of other clans with hereditaryrights. The latter individuals are related throughmatrifiliation (mother–son) to the land-owning clan,but related patrifilially to their own respective clansand estates.

The differentiation of the roles of these groups,referred to in aboriginal English as owners andmanagers, continues to be a core concept of aboriginalrelationships to land and of management of land. Thereciprocity of responsibility between these groups

extends fire management beyond the boundary ofsingle estates and involves neighbours, who are alsokin, collectively in fire management.

But neither aboriginal owners nor managers wereinvolved to any degree when the federal governmentattempted to impose a western-style forestry landmanagement regime over a large area of centralArnhem Land around Maningrida in the 1960s. Theproject’s government experts planned to increase thepercentage of naturally occurring cypress pine(Callitris intratropica) by a process of ‘enrichmentplanting’ in open forest savannas dominated bystringybark (Eucalyptus tetrodonta) and woolybutt(Eucalyptus miniata). Total fire exclusion was a majormanagement objective.

Despite the depopulation of the bush generally,individuals still looked for opportunities to hunt ontheir country, often at weekends and in the rightsituations using fire as a hunting tool. In October 1968a number of fires were lit within the forestry‘protected area’. Fuel that had accumulated afterseveral years of fire exclusion contributed to the firstrecorded ‘crowning’ in northern forests. About twothirds of the protected area was razed: all regenerationand many of the mature trees of cypress pine weredestroyed (Haynes 1978).

At Maningrida, aboriginal resentment of exoticmanagement methods reached a peak in 1972 afterforestry managers sent in heavy machinery toextinguish a small fire lit within a sacred ceremonialground during preparations for a ceremony. At theinsistence of landowners the forestry project wasclosed down and non-local staff withdrawn.

Reflecting on Maningrida’s experience withforestry-style management, government forestryofficial Chris Haynes (Haynes 1978) concluded:‘Everything to do with fire management by anessentially alien body interfered with the customsdeveloped over millennia and the resentment wasexpressed very actively with the lighting of many firesinside the managed area’.

Back to CountryFrom about the time of the Forestry Branch’sdeparture, a revival of customary management wasunder way in many areas as people began to exercisethe power of self determination offered by the federalWhitlam Government which came to power in 1972.

Dissatisfied with the life at missions andsettlements, and fearful that Government would givetheir country to mining companies or otherdevelopers, many aboriginal people had begun tomove back to traditional estates to establish smallcommunities. These small communities, referred to as‘homeland centres’ or ‘outstations’, often have muchthe same population as the basic traditional land-using

104

unit, sometimes referred to by anthropologists as aband, and numbering around 25–30 people.

This decentralisation, or outstation movement, hasgrown and consolidated since the early 1970s untilthere are now hundreds of such small communities inthe Northern Territory. In most cases the aboriginalpeople in these small communities have some accessto health and education services and have beenprovided with water supplies and housing in thelocations of their choice. But there are still few formaleconomic enterprises in these areas. People make andsell arts and crafts, some are considering tourism orother small enterprise, but the mainstay of theoutstation economy remains hunting and gathering.These activities are augmented through theCommunity Development Employment Program(CDEP), a Commonwealth program under whichaboriginal people work for unemployment benefits,also referred to as ‘the dole’. Others activities such asmining, buffalo harvesting and tourism are proceedingunder various arrangements but predicated on theinformed consent of landowners under the AboriginalLand Rights Act.

The return to being on traditional land and usingthat land has brought back traditional management,including traditional fire management.

Since Chris Haynes first began to describe some ofthe systematic elements of aboriginal burning andrelate it to ecological effect (e.g. Haynes 1985),various researchers have directed their attention to anunderstanding of the ecological basis for andecological outcomes from traditional burning.

In recent years satellite imagery allows us tocontrast the burning patterns from occupied andtraditionally managed country with large areas thatremain unpopulated and unmanaged. Although theideal sized aboriginal fire is too small to show up onAVHRR imagery, the observable patterns reflect acommon practice of burning early in the dry season,when fires are unlikely to travel long distances, toachieve a mosaic of burnt and unburnt country. Thisprevents to a large degree the extensive, uncontrolledwildfires that are commonly observed in unpopulatedareas in the late dry season and allows for containmentof deliberately lit hot fires for hunting later in the year.Aboriginal people burn to hunt, to promote new grasswhich attracts game, to make country easier to travelthrough, to clear country of spiritual pollution after adeath, to create firebreaks for later in the dry seasonand a variety of other reasons which overall ‘bring theland alive again’ (Yibarbuk 1998).

There is a now an increasing, but not universal,recognition of the ecological soundness of Aboriginalburning practice. A recent collaborative projectinvolving landowners and researchers from the Parksand Wildlife Commission of the Northern Territory(PWCNT) and the Cooperative Research Centre for

the Tropical Savannas looked in some detail at theeffect of traditional burning on some estates borderingthe Arnhem Land escarpment on the Cadell River. Inthis area one family had refused to move to thesettlement of Maningrida and had maintainedunbroken occupation and unbroken practice oftraditional fire management. The conclusions of theresearchers, using a range of accepted indicators, wasthat these clan estates were being well managed tomaintain ecological integrity (Yibarbuk et al.unpubl.).

Urge to EmulateAs non-aboriginal fire managers gain someunderstanding of the basic principles of aboriginal firemanagement, there is an increasing tendency for themto seek to emulate what they perceive as the effects oftraditional burning. In an attempt to achieve greatereconomic efficiency and minimise labour requirementsthey have introduced technological wizardry such ashelicopters and aerial burning from fixed wing aircraft.

But there is a danger therein, as the research teamfrom the Cadell River project wisely cautioned in theconcluding remarks to their report on the Cadellresearch:

‘Premature or clumsy attempts to ‘transplant’ thetechnology without supporting the ongoingparticipation and recognition of its most skilledpractitioners is not only likely to produce inferiorresults, but also squander an important opportunity todevelop genuine partnerships among Australia’sindigenous and non-indigenous natural resourcemanagers (Yibarbuk et al. unpubl.).’

In Kakadu National Park, jointly managed byaboriginal landowners and Parks North Australia, wehave heard landowners recently voicingdissatisfaction with the hybridised burning regimeoperating there. Some landowners complain aboutaerial burning by park staff, saying that it marginaliseslandowners from continuing fire management as acultural practice and also suggesting that it lacks theecological finesse and effectiveness delivered by aperson walking through the landscape and lightingfires. Parks staff and landowners are now entering aperiod of discussion and dialogue to work through theissues and seek practicable solutions.

Some landowners complain about the increasingextent of wet season burning which has been highlyeffective in removing or radically reducing theamount of speargrass (annual Sorghum spp.) availableas fuel the next dry season. Kakadu landowners arealso debating appropriate floodplain fire managementwith Kakadu staff now that buffalo, which denudedthe flood plains of vegetation for over 100 years, havebeen almost eliminated and the plains again now carrylarge fuel loads each dry season.

105

But while there is debate about the best methods offire management in Kakadu, fire managers there haveone advantage over landowners for much of the rest ofthe Top End—there are significant human andfinancial resources to put whatever system is agreedon into effect.

From Farm to Feral LandscapesAway from such generously funded areas the situationis mixed. From the point of view of the NLC and itsCaring for Country Unit, fire management is notconsidered a critical issue in areas that landownershave been able to resettle and where the principles oftraditional fire management continue to beincorporated into daily life. Satellite imagery shows ussome continuation of early and mosaic burningpatterns across the more closely settled parts ofaboriginal land in the Top End. To a pleasing degreefire is being used a management tool for particularends and all is well down on the firestick farm—or isit? That situation could easily change for the worse ifexotic species like gamba grass (Andropogon gayanus)in woodlands and para grass (Urochloa mutica) onflood plains are allowed to invade and radicallyescalate fuel loads. The usual rules and parameters fortraditional burning won’t apply as they did in the past.

The Caring for Country Unit encourages localcommunity-based agencies to establish landmanagement programs under the CommunityDevelopment Employment Program (CDEP) schemesthat operate in most areas. Through these landmanagement programs awareness can be spread aboutthe danger posed by exotic grasses and the agenciesmobilised to find and eliminate infestations as theyoccur.

The Problem without PeopleIn the Northern Land Council’s area the most difficultlands for fire management are unpopulated areas andareas where landowners have no access to fundsthrough the CDEP program to support landmanagement programs. Although there are severallarge areas with very few settlements, westernArnhem Land provides a case study where there hasbeen some detailed investigation of the problem andsome possible solutions proposed.

The Western Arnhem Land Plateau is one of severallarge areas where depopulation is a major obstacle toeffective fire management. The escarpment is veryimportant ecologically as well as culturally and inparts includes many important fire sensitive speciesand communities. There is a record of humanhabitation for 60 000 years (Russell-Smith et al. 1993;Bridgewater et al. 1998). Landowners from the

plateau are now scattered in communities around theplateau and for a number of reasons they have beenunable to resettle the plateau country and bring it backunder traditional management. Fire maps derivedfrom AVHRR imagery over the past few yearsillustrate the fire problem; to the north and northeastmore traditional patterns of fire are observable inresettled areas and in the south and southeast there isalmost no early burning in some years and in othershuge, very hot late dry season fires burn for weeks at atime.

The Caring for Country Unit of the NLC, incollaboration with the Jawoyn Association and theBushfires Council of the NT, is working withlandowners to develop a sustainable fire managementregime in the area over a 3-year period. Thecollaborators are trying to develop effective methodsof participatory planning—the land is privateaboriginal land and for solutions to be sustainablethere must be a sense of ownership and commitmentto them by aboriginal landowners. The Maningridaforestry experience should have taught us that much!

Focusing Varied ResourcesGiven the lack of personal financial resources of thelandowners there is also a need to develop a strongpartnership of collaboration between landowners anda variety of other stakeholders. These includeaboriginal organisations, business interests andscientists. A meeting of landowners at Manmoyihoutstation in September 1998 looked at ways ofgetting better fire management by posing a series ofquestions for group problem solving andbrainstorming (Northern Land Council et al. 1998).

At that meeting landowners discussed:• how can buffalo harvesting help fire management?• how can mining exploration help fire management?• how can tourism help fire management?• how can the undertaking of ceremonial and other

cultural practices help fire management?• how can aboriginal resource agencies help fire

management? Some tentative responses are listed below. Buffalo There is aboriginal-owned buffalo

harvesting in the southeast and likelihood of otherareas again being mustered west of there. Earlyburning can be incorporated into early reconnaissancefor buffalo harvesting later in the dry and elders withdetailed knowledge of traditional burning can havemore input into burning that can be carried duringmustering operations.

Mining Mineral exploration can create tracks whichenable people to get into previously inaccessiblecountry for early burning. Landowners working withexploration companies on the ground can carry outburning. Mining companies may contribute to costs of

106

getting people into country for early burning becauseit will protect their camps and equipment from thepossibility of being burned out late in the year. But ifnot properly controlled the tracks can result inunauthorised access, which may bring irresponsibleburning and import weeds.

Tourism There is potential to find non-Governmentfunds to pay landowners to walk through the countryon traditional walking tracks carrying out earlyburning. Bushwalkers can pay for the experience ofbeing guided across this environmentally andculturally spectacular landscape by aboriginal guides.The knowledge of these traditional walking tracks isin danger of being lost with the death of the last peoplewho grew up on the plateau and walked extensivelyover it. Helicopters will be used map the tracks withold people who can’t walk long distances and youngpeople will come along to learn the routes and to beintroduced the spirits of the country by their elders.Then the younger people can ‘test drive’ theecotourism possibilities with some parties ofbushwalkers before deciding whether to develop theidea into a full fledged business.

Cultural traditions There are important culturaltraditions associated with using fire to hunt kangaroosin particular places on the escarpment. Olderlandowners will supervise a number of these hunts topass on knowledge to younger people and as a focusfor bringing people back to country. The NLC andJawoyn Association will try to assist as many familiesas possible to get back onto the plateau during themid-year school break in 1999 with the longer termhope that this could lead to more permanentoccupation or more regular visits.

Modern as well as older cultural practice is alsoimportant in creating awareness of the importance offire management. A band of musicians fromManmoyih outstation, the Nabarlek band, recordedand released a CD called Manwurrk or bushfire with anumber of songs about the cultural importance of fireafter our first workshops. It has done a lot to getpeople thinking more about fire in particular andlandscape management in general.

Landowners from the corners of the plateau havebeen enthusiastic about re-establishing cultural andsocial links with groups of the plateau diaspora. Laterthis year a meeting of the local resource agencies andlandowners from around the plateau will determinewhat resources the agencies can commit to thecollective task of managing fire on the plateau andkeeping its management well in the hands oflandowners.

Developing CommitmentThere seems to be great potential in pursuing firemanagement strategies that integrate and involve thebroad range of land use activities and land users in thebusiness of fire management. There are many usersand potential users of aboriginal land–miners, buffalocatchers and tourism operators as well as landownersto name a few. It seems essential to develop thegeneral understanding that participating in landscapemanagement of fire and other key areas such as weed,erosion and feral animal control must be a basiccondition of land use.

Effective communication is a key element inworking towards this goal. In our Western ArnhemLand Project we ensure that wherever possiblelinguists or aboriginal interpreters are available to liftthe quality of communication.

We should not underestimate the inadequacy of ourcurrent communication strategies and the negativeeffect they can have on management. A BushfiresCouncil sign that has stood for some years on the roadfrom Bulman to Katherine can serve as an example. Itreads: No lighting of fires during fire danger periods.Penalty $1000 or 6 months imprisonment. I would nothave thought aboriginal people took much notice of itall until a senior landowner from Bulman told me thatpeople took it very seriously and interpreted it to meanthat people should stop following traditional burningpractice or face jail.

CommunicationCommunication is without doubt the key.

Better communication amongst aboriginal people isneeded to ensure that knowledge of traditionalburning practice is passed to new generations.

Better communication is needed to get groupsseparated by historical circumstance talking togetherabout mutual interest in country and assisting eachother to resettle country or develop effectivemanagement.

Better communication is needed to involvedisparate non-aboriginal land users in recognition thatfire management is a responsibility of all land users.

Finally, better communication is needed to focus allthese groups on developing locally designed strategiesthat make best use of western science and oftraditional knowledge to increase the capacity oflandowners to preserve the quality of their country forfuture generations.

107

ReferencesAbbott, C.L.A. 1950. Australia’s Frontier Province. Sydney.

Angus and Robertson, 144–145. ABS 1998. 1996 Census of Population and Housing: 1998c

Aboriginal and Torres Strait Islander People (2034.0-8), 1998d Indigenous Profile (2020.0). Canberra,Australian Bureau of Statistics.

ABS 1999. Special Article: Aboriginal and Torres StraitIslander Australians: a statistical profile from the 1996Census. Canberra, Australian Bureau of Statistics.

Bridgewater, P.B., Russell-Smith, J., Cresswell, I.D. 1998.Vegetation science in a cultural landscape.Phytocoenologia, 28, 1–17.

Christian, C.S. and Aldrick, J.M. 1977. Alligator RiversStudy: A review of the Alligator Rivers Regionenvironmental fact-finding study. Canberra, AGPS.

Department of Foreign Affairs and Trade (DFAT) 1995.Northern Territory Facts. International Public AffairsBranch, DFAT, Canberra.

Fourmile, H. 1996. Making things work: Aboriginal andTorres Strait Islander involvement in bioregionalplanning. In: Approaches to Bioregional Planning. Part2. Background papers to the Conference, 30 Oct–1 Nov1995, Melbourne. Department of the Environment,Sport and Territories, 145, 253.

Haynes, C.D. 1978. Land, trees and man (gunret, gundulg,dja bining). In: Commonwealth Forestry Review, 57,99–106.

Haynes, C.D. 1985. The pattern and ecology of munwag:traditional aboriginal fire regimes in north-centralArnhemland. Proceedings of the Ecological Society ofAustralia, 13, 203–214.

Northern Land Council, Jawoyn Association, BushfiresCouncil NT, and Aboriginal landowners 1998. Lookingafter the Arnhem Land Plateau: Manmoyih MeetingSeptember 1998. Darwin, Northern LandCouncil,1–10.

Russell-Smith, J., Lucas, D., Brock, J. and Bowman,D.M.J.S. 1993. Allosyncarpia-dominated rain forest inmonsoonal northern Australia. Journal of VegetationScience, 4, 67–82.

Thomson, D. 1949. Economic Structure and the CeremonialExchange Cycle in Arnhem Land. Melbourne,Macmillan & Co Ltd, 17–18.

Yibarbuk, D. M.1998. Notes of traditional use of fire onupper Cadell River. In: Langton, M., ed., BurningQuestions. Darwin, Centre for Indigenous and NaturalCultural Management, Northern Territory University.

108

CONCERN WITH the role of fire in land management innorthern Australia arises from two important issues thatthreaten the sustainable development of savannarangelands (Jacklyn and Russell-Smith 1998). The firstis an apparent increase in the incidence and extent ofuncontrolled, late dry season wildfires across largeareas of northern Australia, particularly in high-rainfall,eucalypt woodlands and forests. Such fire regimes haveadverse impacts on pastoral management and thestability of rangeland ecosystems. The second andcontrasting issue is a reduction in the incidence, extentand intensity of burning, mostly throughout the drier,more productive pastoral lands. Apart from othereffects, reduced fire is suggested as being a major causeof significant increases in tree and shrub populations.

Fire may play an important role in managing woodyplant populations, maintaining pasture condition,managing grazing distribution, reducing the hazard ofuncontrolled wildfires and increasing the availability ofnutritious herbage to cattle. Fire also has a role in aidingthe establishment and management of improvedpastures, contributing to the control of exotic weeds andmaintaining biodiversity.

The Pastoral Industry in Northern AustraliaPastoralism is the most extensive land use in northernAustralia and is focused on extensive beef production.

Significant changes to market conditions haveoccurred in recent years. During this time fire andother land management issues have become importantto the sustained productivity of the industry.

Cattle are raised within vegetation and climaticzones that range from woodlands in the seasonally wetmonsoon tropics in northern or coastal areas to semi-arid natural grasslands and arid shrublands in southerninland areas. Northern cattle herds are dominated byBos indicus pure or crossbred cattle that are wellsuited to tropical conditions and the typical lowquality dry season forage of native pastures.Production systems are simple, require few inputs andrely almost entirely on native grass pastures. Grazingmanagement options are limited mainly to adjustingstocking rates and total grazing pressure, burning,strategic spelling and controlling the number andlocation of watering points and fences.

In recent years rapid growth of the live cattleindustry has been a key development. In 1997 livecattle exports from northern Australia totalled 680 000head, a massive ninefold increase from levels in 1990.This represents a total value of $322 million (MRC1997). The major destination for live cattle fromnorthern Australia in 1997 was Indonesia, whichaccounted for around 60% of all exports. As a result ofthe Asian currency crisis in August 1997 live cattleexports from northern Australia to Southeast Asiaplummeted to 346 000 for 1998. This represents only7 and 25% respectively of the 1997 aggregate exportsto Indonesia and the Philippines.

The Role of Fire on Pastoral Lands inNorthern Australia

Rodd Dyer1

AbstractIn northern Australia fire has been regarded as an essential component of pastoral land management,but its use has not always been appropriate or beneficial. This paper examines the interaction betweenlivestock grazing and fire, the effects of grazing intensity on the frequency, extent and intensity ofnatural fire regimes, and the need for intentional burning. The author describes the role of fire inmanaging woody plant populations, maintaining pasture conditions, managing grazing distribution,increasing the availability of nutritious herbage to cattle, and reducing the hazard of uncontrolledwildfires. The author also emphasises the need for land managers to understand fire ecology and theinteractions between season, burning and grazing on a landscape scale in order to use fire as asustainable management tool.

1Northern Territory Department of Primary Industry andFisheries, Katherine Research Station, PO Box 1346,Katherine NT 0851

109

Fire Regimes on Pastoral LandThe landscape that faced early pastoralists wasfashioned by thousands of years of burning. Fire was anatural and frequent part of the landscape (Pyne 1991)and an integral part of aboriginal culture and daily life.They used fire frequently and skilfully for hunting,ceremonies, protecting food reserves and ‘cleaningthe country’ (Haynes 1985; Braithwaite and Roberts1995). Accounts of early explorers suggest thataboriginal burning occurred mainly between the latewet and the late dry seasons but was most frequentduring the early dry season (Braithwaite 1991). Firesduring the late dry season would be ignited byfrequent lightning during dry storms that preceded thesummer wet season (Braithwaite and Estbergs 1985).

There is strong evidence that contemporary fireregimes throughout northern Australia differsignificantly from those that prevailed prior to theestablishment of the pastoral industry (Braithwaiteand Estbergs 1985). Early pastoralists burnt country topromote fresh growth for green herbage that wouldattract cattle and make mustering easier (Smith 1960).Today however, the frequency, extent and intensity ofburning on semi-arid pastoral land generally havedecreased. In many areas, managers are hesitant toburn grass that could be used as feed, particularlywhere rainfall is low and highly variable betweenseasons. Mineral supplementation with Bos indicuscattle breeds enables the utilisation of poor qualitycarry over pasture. In some areas this has had theeffect of increasing grazing pressure, severelyreducing fuel and opportunities for burning. A recentworkshop noted that many pastoralists are hesitant toburn for a variety of reasons (Grice and Slatter 1997).

In contrast throughout large areas of the wet tropics,frequent uncontrolled wildfires dominate thelandscape particularly during the late dry season whenfire weather conditions are extreme (Jacklyn andRussell-Smith 1998). These fires are prevalent in lessproductive pastoral lands, in areas with high seasonalrainfall, tall grass pastures and low total grazingpressure. Wildfires are generally more destructive thanless intense, patchy early dry season fires. They are aserious threat to the survival of fire sensitive plant andanimal communities as well as pastoral productivity.

Role of Fire in Pastoral Land ManagementFire has a range of potential uses in the managementof pastoral land. It is possible that one, several or noneof these issues may exist on a given piece of land.Pastoral land managers must therefore identify whatgoals may require the use of fire, then implement afire regime to suit these goals. The components of afire regime that must be considered are fire type,frequency, extent, season and intensity.

The interaction between fire and grazing makes therole of fire on pastoral land unique, though littlequantitative information exists on fire-grazinginteractions. A trade-off exists between utilising grassas forage for livestock or fuel for fires (Ludwig et al.1999). In the post-burn phase, grazing managementaffects the response of both the grass and woodylayers to burning. Obviously managing grazingpressure both prior to and following burning iscritical. Some of the uses of fire in northern Australiaare discussed below.

Management of tree and shrubsFire has an important role to play in the managementof the tree–grass balance in northern Australianrangelands (Burrows et al. 1990; Scanlan et al. 1991).A reduction in frequency, extent and intensity ofburning has contributed to increased densities ofnative trees and shrubs into once open grasslands andwoodlands throughout Australia and the world(Scholes and Archer 1997; Hodgkinson 1990). Astrees and shrubs increase in size and density, increasedcompetition with the pasture layer for moisture andnutrients can lead to significant decreases in pasturegrowth. The impacts of competition are morepronounced in areas with lower and sporadic rainfalland on shallow soils (Scanlan and McKeon 1993).This may have the effect of reducing livestockcarrying capacities, burning opportunities andtolerance to drought.

The role of fire is one of managing the tree–grassbalance rather than controlling woody plants. An‘appropriate’ balance for particular pasture lands isdifficult to assess, however evidence from historicalphotographs in the Victoria River District in theNorthern Territory (Daryl Lewis pers. comm.)suggests that significant increases in tree and shrubdensity have occurred on the more productive pastorallands. Most native woody species are well adapted tofire and persist by vegetative sprouting or seedlingestablishment (Gill 1981). Rather than a single fire,effective management requires the implementation ofappropriate fire regimes. Frequency of burning isdetermined by the rates of woody regrowth and fuelaccumulation. Fires however can also be usedopportunistically to control the seedling establishmentthat occurs episodically after favourable seasons.

The season and intensity of burning will bedetermined by management objectives. Where fire isalready a frequent disturbance, lower intensity, earlydry season fires can suppress a shrubby mid-storey,maintaining open eucalypt woodlands in the wet-tropics. Conversely, a mature population structure of arosewood (Terminalia volucris) on black soilcommunities in the Victoria River District mayrequire intense, late dry season fires every 4–6 years

110

(Dyer et al. 1996). Intense, late dry season fires arealso effective in maximising canopy scorch and plantmortality in country heavily invaded by Acacialysiphlioa and Acacia cowleana around Daly Waters,NT. Seedling establishment is variable after burning.

Reduction of wildfire hazard Prescribed early dry season fires can play a role instrategically reducing fuel loads and the risk ofwidespread and destructive late dry season wildfires.Apart from being disruptive to station managementand consuming large areas of pasture for grazing,these fires are potentially damaging to fire-sensitiveplant communities.

Where uncontrolled fires prevail, prescribed use ofearly dry season fire along with greater controlmeasures during the late dry are required to create adeliberate shift in the frequency, season, extent andintensity of burning to one that is more ecologicallyand economically appropriate. Early dry season firesare patchy, less intense and more easily managed(Williams et al. 1995). As well as reducing fuel levels,they create a mosaic of burnt and unburnt areas thatimpede the progress of wildfires later in the dryseason.

Successful manipulation of fire regimes to reducefrequent and extensive late dry season fires has beenachieved to some extent in the management ofKakadu National Park in northern Australia. Thefrequency and extent of late dry season fires inAustralia has been reduced by increasing strategicburning during the early dry season (Press 1987,Russell-Smith et al. 1997). This provides a usefulmodel that could be adapted to pastoral lands whereregular wildfire is a problem.

Managing grazing distributionFire can be used strategically to more evenly spreadgrazing pressure across pastures. Normally cattleactively select palatable pasture species andpreferentially graze certain parts of the landscape(Wilson et al. 1984). Under extensive conditions,cattle return to previously grazed patches, continuallyignoring areas of rank mature pasture. Repeateddefoliation of these preferred-patch areas can rapidlylead to a reduction in plant yield, death of desirablespecies, increases in undesirable species anddevelopment of bare and scalded areas (Bridge et al.1983). As selective grazing continues heavily grazedpatches expand and coalesce.

Burning rank, ungrazed pasture can break thepattern of selective patch grazing and promote a moreuniform grazing distribution (Andrew 1986a). Cattlewill preferentially graze regrowth from recently burnt

areas (Ash et al. 1982) and display little selectionbetween palatable and normally unpalatable speciesduring the early part of the growing season (Andrew1986b). To avoid overgrazing the same patches,burning must be rotated to new areas within paddocks,primarily to areas that are less heavily utilised.

Maintaining pasture conditionFire and grazing can be used strategically to managepasture condition and species composition. Burningseason, frequency and grazing management canbe manipulated to determine the impact fire has on thepasture layer. Burning during the dry season,when grasses have seeded and become dormant,generally does not affect the composition of perennialgrass pastures that are in good condition (Mott andAndrew 1985a,b).

Wet season burning can result in changes in yieldand species composition (Smith 1960). In the wet-tropics of northern Australia strategic wet seasonburning can be used reduce the proportion of nativeannual sorghum that dominates the graminoid layer ofeucalypt woodlands (Stocker and Mott 1981; Lane andWilliams 1997). Each year sorghum pastures producemassive fuel loads (5–8 t/ha) that promote frequent anddamaging dry season fires and provide little nutritionalvalue to cattle. Burning these pastures prior to thedevelopment of mature seed can result in a significantreduction in sorghum, as the seed bank is transient andmost seed germinates or dies during the following wetseason (Mott 1978; Andrew and Mott 1983).

When pastures are unburnt and lightly grazed forseveral seasons dry matter accumulates, tying upnutrients and smothering new growth. This results inpastures becoming moribund and unproductive.Burning can rejuvenate pasture without long-termchanges in yield, cover and species composition(Norman 1969). There is evidence however thatburning of pastures in poor condition should beavoided. Heavily utilised arid short-grass pastures inthe Victoria River District are often dominated byBrachyachne convergens (summer couch), an annual‘increaser’ species. Burning these pastures,particularly with intense fires, appears to destroy seedexposed on the soil surface, so that the soil remainsbare and unprotected after the next wet season. (Dyerand Cobiac unpubl.).

Control of exotic weeds The establishment and spread of exotic weeds presentsa serious threat to the stability and productivity ofpastoral lands (Grice and Brown 1996). Fire has a rolein the containment and control of established exoticweed infestations as well as the protection of pastoral

111

land against further invasion (Grice 1997a). Weedssuch as rubbervine (Cryptostegia grandiflora), chineeapple (Ziziphus mauritiana), mesquite (Prosopis spp.)and prickly acacia (Acacia nilotica) occupy significantareas throughout Queensland, result in serious lossesin pastoral production and incur major economic costs.Most of these species are already established or havethe potential to invade pastoral lands within theNorthern Territory and Western Australia. Fire haspotential for cost-effective treatment of large areas. Itis likely however that fire will be most effective as acomponent of an integrated approach to weedmanagement (Grice 1997a), being used in combinationwith either biological, chemical and mechanicalagents.

Research indicates that exotic weed species varygreatly in their sensitivity to fire, and may differ intheir vulnerability to fire at different life stages. Fireshows potential for the control of rubbervine andmesquite (Prosopis pallida). High mortality rates ofboth juvenile and adult plants occur following burning.By contrast juvenile and adult stages of chinee appleresist fire and sprout after burning from the base ofcanopy (Grice 1997b; Campbell pers. comm.).

The susceptibility to burning exhibited by smallerplants of most exotic weed species suggests that theuse of fire prior to maturation and seed set will reduceestablishment and help contain their spread (Grice1997a). Even if fire does not cause high mortality ofadult plants, it may reduce further spread bydestroying seedlings or ungerminated seed. Althoughburning may not be effective against the adult phaseof weeds such as chinee apple, research indicates thatfire may help minimise spread by destroying seed(Campbell pers. comm.).

Establishment and management of improved pastures Mott (1982) has reviewed the impact of fire onimproved pastures. Fire can aid the establishment andmanagement of Stylosanthes (stylo) pastures (Stockerand Sturtz 1966). Various species and cultivars ofstylo have successfully augmented native pastures inthe open eucalypt woodlands of northern Australia,leading to improved livestock growth. Establishmentwithout cultivation is possible by sowing seed aftereffective rains into ash left by burning. Improvedgermination and seedling survival result from removalof accumulated litter, reduced competition fromnative pasture and provision of a favourable seedbed.

Prescribed fire can also be used to manage thegrass–legume balance in pastures improved withstylos. Dominance of stylo and decline of native grasscan occur after several years of grazing. Whenlegumes exceed 40% of pasture yield, fire can be used

to thin the stylo population and regain grassdominance. Perennial stylos will re-shoot followingburning while annual species will persist viagermination of seed stored in the soil.

ConclusionFire clearly has an important role in the managementof pastoral land in northern Australia. Burning doesoccur throughout pastoral lands, however there areconstraints to the widespread use of fire in normalmanagement. Fire must be a carefully managed tomeet specific sustainable objectives. In pastoral landsthe interaction between fire and grazing is critical. Thechallenge for land managers is to manipulate grazingand fire to achieve specific objectives. Work on theimpacts and management of fire is under way atspecific locations in northern Australia. This workaims to gain a better knowledge of fire ecology andmanagement and to provide solutions that ensure theuseful integration of fire into land management.

ReferencesAndrew, M.H. 1986a. Use of fire for spelling monsoon

tallgrass pasture grazed by cattle. Tropical Grasslands,20(2), 69–78.

Andrew, M.H. 1986b. Selection of plant species by cattlegrazing native monsoon tallgrass pasture at KatherineNT. Tropical Grasslands, 20(3), 120–127.

Andrew, M.H. and Mott, J.J 1983. Annuals with transientseed banks: the population biology of indigenousSorghum species of tropical northwest Australia.Australian Journal Ecology. 8, 265–276.

Ash, A.J., Prinsen, J.H., Myles, D.J. and Hendricksen, R.E.1982. Short-term effects of burning native pasture inspring on herbage and animal production in southeastQueensland. Australian Society of Animal Production.14, 377–380.

Braithwaite, R.W. 1991. Aboriginal fire regimes of monsoonalAustralia in the 19th century. Search, 22 (7), 247–249.

Braithwaite, R.W. and Estbergs, J.A. 1985. Fire patterns andwoody vegetation trends in the Alligator Rivers regionof northern Australia. In: Ecology and Management ofthe World’s Savannas. IX International SavannaSymposium 1984, 359–364.

Braithwaite, R.W. and Roberts, S 1995. Aboriginal burningand conservation management in Kakadu NationalPark, tropical Australia. Wildfire September, 29–38.

Bridge B.J., Mott, J.J and Hartigan, R.J. 1983. The formationof degraded areas in the dry savanna woodlands ofnorthern Australia. Australian Journal of Soil Research,21(1), 91–104

Burrows,W.H., Carter, J.O., Scanlan, J.C. and Anderson,E.R. 1990. Management of savannas for livestockproduction in northeast Australia: contrasts across thetree–grass continuum. Journal of Biogeography, 17,503–512.

112

Dyer, R.M., Cobiac, M.D., Café, L.M. and Stockwell, T. 1996.Developing sustainable pasture management practicesfor the semi-arid tropics of the Northern Territory. FinalReport NTA 022 submitted to the Meat ResearchCorporation by Northern Territory Department ofPrimary Industries and Fisheries, Katherine, Australia.

Gill, A.M. 1981. Adaptive responses of Australian vascularplant species to fires. In: Gill A.M., Groves, R.H. andNoble, I.R., ed., Fire in the Australian Biota. Canberra,Australian Academy of Science, 243–271

Grice, A.J. and Brown J.R.1996. Fire and the populationecology of invasive shrubs in the tropical woodlands.Proceeding of the Nicholson Centenary Meeting,Frontiers of Population Ecology, 1995 Canberra.589–597.

Grice, A.C. 1997a. Fire for the strategic management ofrubbervine (Cryptostegia grandiflora), an invasiveshrub. In: McKaige, B.J., Williams, R.J. and Waggitt,W.M., compilers, Bushfire ’97 Proceedings. AustralianBushfire Conference, 1997, Darwin. CSIRO TropicalEcosystems Research Centre, 230–244.

Grice, A.C. 1997. Post-fire regrowth and survival of theinvasive tropical shrubs Cryptostegia grandiflora andZiziphus mauritiana. Australian Journal of Ecology, 22,49–55.

Grice A.J., and Slatter, S.M. 1997. Fire in the Managementof Northern Australian Pastoral Lands. Proceedings ofa Workshop held in Townsville, Queensland, Australia,April 1996. Tropical Grassland Society of Australia,Occasional Publication No. 8.

Haynes, C.D. 1985. The pattern and ecology of munwag:traditional Aboriginal fire regimes in north centralArnhem Land. Proceedings of the Ecological Society ofAustralia, 13, 203–214.

Jacklyn, P. and Russell-Smith, J. 1998. Proceedings from theNorth Australia Fire Management Workshop. Darwin.March 1998. Tropical Savannas CRC, NorthernTerritory University, Darwin, Northern Territory.

Lane, A.M. and Williams, R.J. 1997. The effect of wetseason burning on understorey vegetation in a humidtropical savanna in northern Australia. In: McKaige,B.J., Williams, R.J. and Waggitt, W.M., compilers,Bushfire ’97 Proceedings. Australian BushfireConference, Darwin, CSIRO Tropical EcosystemsResearch Centre, 25–30.

Ludwig, J.A., Cougenhour, M.B. and Dyer, R.M. 1999.Modelling the impacts of fire and grazing on theproductivity and sustainability of tropical savannas,northern Australia. Proceedings of the InternationalRangeland Congress. Townsville, Australia (in press).

Mott, J.J. 1978. Dormancy and germination in five nativegrass species from savanna woodland communities ofthe Northern Territory. Australian Journal of Botany,26, 621–631.

Mott, J.J. 1982. Fire in improved pastures of NorthernAustralia. Tropical Grasslands, 16(3), 97–100.

Mott, J.J. and Andrew, M.H. 1985a. Effects of fire on thegrass understorey of Australian savannas. In: Walker,J., Davies J.R. and Gill, A.M., ed., Towards an ExpertSystem for Fire Management in Kakadu National Park.Tech. Memo. 85/2 CSIRO Division of Land and WaterResources, 77–91.

Mott, J.J. and Andrew, M.H. 1985b. The effect of fire on thepopulation dynamics of native grasses in tropicalsavannas of northwest Australia. Proceedings of theEcological Society of Australia, 13, 231–239.

MRC 1997. Profile of North Australian Beef IndustryChange, 1985–1996. Meat Research Corporation,Sydney South, New South Wales.

Norman, M.J.T. 1969. The effect of burning and seasonalrainfall on native pasture at Katherine, NT. AustralianJournal of Experimental Agriculture and AnimalHusbandry, 9, 295–298.

Press, A.J. 1987. Fire management in Kakadu National Park:the ecological basis for the active use of fire. Search,18(5) Sept/Oct, 244–248.

Pyne, S.J. 1991. Burning Bush: A Fire History of Australia.New York, Henry Holt and Co.

Russell-Smith, J, Ryan, P.G. and DuRieu, R. 1997. ALANDSAT MSS-derived fire history of KakaduNational Park, monsoonal northern Australia,1980–1994: seasonal extent, frequency and patchiness.Journal of Applied Ecology, 34, 748–766.

Scanlan, J.C., Wilson, B.J. and Anderson, E.R. 1991.Sustaining productive pasture in the tropics 2:Managing woody vegetation in grazing lands. TropicalGrasslands 25:85–90.

Scanlan, J.C. and McKeon, G.M. 1993. Competitive effectsof trees on pasture are a function of rainfall distributionand soil depth. XVII International Grassland Congress,Paper No.7:19*1 1993.

Scholes, R.J. and Archer S.R. 1997. Tree–grass interactionsin savannas. Annual Review of Ecology andSystematics, 28, 517–544.

Smith, E.L. 1960. Effects of burning and clipping at varioustimes during the wet season on tropical tallgrass rangein northern Australia. Journal of RangelandManagement, 13, 197–203.

Stocker, G.C. and Mott, J.J. 1981. Fire in the tropical forests andwoodlands of northern Australia. In: Gill, A.M., Groves,R.H. and Noble, I.R., ed., Fire in the Australian Biota.Canberra, Australian Academy of Science, 425–439.

Stocker, G.C. and Sturtz, J.D. 1966. The use of fire toestablish Townsville lucerne in the Northern Territory.Australian Journal of Experimental Agriculture andAnimal Husbandry, 6, 277–279.

Williams, R.J., Gill, A.M. and Moore, P.H.R. 1995. Intensityof experimental fires in a tropical savanna at Kapalga,Kakadu National Park. In: Bushfire ’95 Proceedings.Australian Bushfire Conference, Hobart, Tasmania.

Wilson, A.D., Harrington, G.N. and Beale, I.F. 1984.Grazing Management. In: Harrington, G.N., Wilson,A.D. and Young, M.D., ed., Management of Australia’sRangelands. Melbourne, CSIRO, 129–139.

113

IN AUSTRALIA, where grasslands are the predominantsource of fuel for fires that burn over large areas,satellites orbiting about 800 km above the earth areused to determine the location of bushfires, map thearea burnt and also estimate the fuel load andflammability conditions that predispose towards fires(Smith et al. 1998b). This is achieved at a scale andaccuracy not possible from human observations on theground.

Use of this technology for bushfire managementwithin grasslands in Australia has resulted from thedevelopment of an integrated system by SatelliteRemote Sensing Services, Department of LandAdministration, Western Australia (Smith et al.

1998a). This system ingests and processes NOAAdata received in Perth, Darwin and Melbourne anddistributes the resulting satellite information tomanagers in the field and strategic planners in theoffice. The following products are now routinelyproduced from the near real-time processing ofAVHRR data:1)Fire hot spots—to locate active fires and fire fronts

in remote areas (Perth and Melbourne data only);2)Fire-affected areas—to map area burnt, establish

fire history and evaluate the effectiveness ofcontrolled burning (Perth, Melbourne and Darwindata);

3)Curing Index—to estimate the flammability ofgrassland fuels at the onset of the dry season forplanning controlled burning strategies;

4)Normalised Difference Vegetation Index—tomonitor seasonal grass growth to estimate fuel loadbuild up and future fire risk.

Role of Remote Sensing in Bushfire Management

R.C.G. Smith, R.L. Craig, M.T. Steber, A.J. Marsden,N.A. Raisbeck-Brown and J. Adams1

AbstractTo assist in the management of bushfires occurring over extensive areas of Australia a FIREWATCHservice based on the Advanced Very High Resolution Radiometer (AVHRR) on the United StatesNOAA polar orbiting satellites has been developed. For this service AVHRR data are acquiredautomatically during the day from receiving stations in Perth and Melbourne, using the Internet forprocessing at the Leeuwin Centre in Western Australia. Additional AVHRR data are acquired weeklyon tape from Darwin. Following processing, FIREWATCH information products are disseminated toclients in regional areas of Australia using the Internet and fax.

From the the visible, near infrared and thermal infrared spectral bands of AVHRR active fires, firescars, fuel load build up and fuel flammability are mapped under cloud-free conditions. To assist intactical management, the locations of hot spots of active fires are posted on the Internet and faxed toregional offices within 2 hours of the AVHRR overpass. To provide a fire history for management ofcontrolled burning and to document the fire regime, the area burn by fire is mapped every 9 days. Fuelload build up and curing of the grasslands are estimated using the Normalised Difference VegetationIndex.

During October 1998, hot spots of fires in southeastern Indonesia were successfully detected fromAVHRR data received in Perth. This trial indicated that southeastern Indonesia, with a dry season andcloud-free skies similar to northern Australia, would enable AVHRR optical sensors to be used tomap fires, fire regimes, fuel loads and fuel flammability in near real-time. Therefore using AVHRRdata from Darwin and Perth, the FIREWATCH project could be extended as part of anIndonesian–Australian collaborative research program into the management of fire in tropicalsavannas. This extension of FIREWATCH would provide a common data set on fire activity andseasonal vegetation conditions on both sides of the Timor Sea.

1Satellite Remote Sensing Services (SRSS), Department ofLand Administration (DOLA), Leeuwin Centre, 65 BrockwayRoad, Floreat, Western Australia 6014http://www.rss.dola.wa.gov.au

114

The success of the system depends on being able toaccess AVHRR data in near real-time from locationsaround Australia. The Internet is used to bring thePerth and Melbourne data to SRSS at the LeeuwinCentre for processing. AVHRR data from Darwin areacquired on tape. Fire information derived fromAVHRR is distributed rapidly using the Internet andwhere the speed is inadequate fax is used. As a resultthe impact of distance and remoteness across theAustralian continent is contracted and all persons haveaccess to the same information. The same servicecould be readily extended to southeastern Indonesiausing AVHRR data collected in Australia or Indonesiaitself, providing a basis for improve interactionbetween fire managers on either side of the Timor Sea.

Mapping Fire Affected Areas and Fire Hot Spots in Rangelands

Land surface albedo in the visible, near infraredand shortwave infrared wavebands From one satellite orbit to the next during the daytime,changes in surface reflectance resulting fromreduction of vegetation cover by fire can be detected.An increase in reflectance is the common response onbright coloured soils in the spinifex areas ofAustralian rangelands. However, where significantcarbon residue is left, reflectance is decreased and thefire scar appears dark, a response common in thewetter tropical savannas of northern Australia. Thissignature reversal makes automatic fire scar detectiondifficult. Decreased reflectance following a bushfirecan also occur on dark soils, which have a lowerreflectance than the surface vegetation cover. Thusfrom changes in the reflectance of the surface in thevisible, near infrared and short wave infrared regionsthe fire-affected area (FAA) can be mapped fromsatellite. In agricultural situations the areas burnt areoften too small and in forest the tree canopy too densefor the change in surface reflectance to be detected bythe satellite sensor. From AVHRR only FAAs greaterthan 400 ha can be effectively mapped, but fire hotspots (FHSs) of areas of about 1 ha can be mapped.

Landsat-TM with 30-m resolution and a 16-dayrevisit cycle is the main low-earth polar-orbitingsatellite measuring in the visible, near infrared andshortwave infrared regions, but unlike AVHRR it isnot available in near real-time, at no cost or with mid-thermal spectral coverage. AVHRR with 1-kmresolution and a daily revisit cycle measures in thevisible, near infrared and thermal infrared spectralregions (Table 1). A single Landsat-TM image coversan area 180 × 180 km and costs $1500, while NOAA-

AVHRR data cover a 2500-km swath × 5000 kmalong track and is free-to-ground.

The more frequent revisit times of AVHRR resultsin a high probability of detecting FAAs greater than400 ha. With Landsat-TM, FAAs as small as 0.1 hacan be detected but, if vegetation regrows rapidly andcloud extends the 16-day revisit time to severalmonths, some FAAs may disappear before they can bemapped.

Land surface temperature in the thermal infraredIn the mid thermal infrared (3.55–3.93 µm) wavebandpeak radiance emitted occurs at temperatures around800oK. Therefore the heat generated by active firescan be detected from areas less than the 1-kmresolution of AVHRR during the day and night.Hence many small fires in agricultural fields andunder forest canopies are detected when the FAA istoo small be ‘seen’ by AVHRR. However at 3.55-3.93 µm emissivity is below 0.9 and the sensor can besaturated by reflected solar irradiance, which causes‘false’ hot spots to be detected, particularly in desertregions. Therefore FHS detection from AVHRR isrestricted to night time, but geometric accuracy islimited by a lack of visible ground features for precisenavigation. Also fires that start and end during the daymay not be hot enough to be detected by night-timeimagery.

Fires change surface emissivity as well as albedo,therefore the thermal infrared (10.5-12.5 µm) issuitable for measuring land surface temperatures andis also used in combination with other spectral bandsfor discriminating FAAs and FHSs.

Land surface roughnessFires, by removing the vegetation cover, change theroughness of the surface and therefore FAAs can intheory be detected using radar in the microwaveregion (λ=1mm-1m). An advantage of using radar isthe possibility in tropical regions of measuring FAAsunder cloud. The main commercial satellite availablefor making such measurements is RADARSAT(λ=5.6 cm) with HH polarisation and a repeat cycle of24 days. The length of the cycle can be reduced bypointing the sensor. The single wavelength andpolarisation of RADARSAT limits the ability to mapFAAs, but if the change in surface roughness issubstantial then the FAA can be detected. A singleimage covering 100 x 100 km costs about $6000. Themain application of RADARSAT will be where thereis persistent cloud such as in tropical rainforests orduring the tropical savanna wet season. We have notyet adapted RADARSAT to bush fire monitoring inAustralia because of the cost and undevelopedbenefits.

115

Mapping Grassland Fuel Load and Flammability

Fuel load and its flammability, two additionalvariables contributing to the risk of bushfire, can beestimated from the visible and near infraredreflectances. From AVHRR the seasonal flush ofgreen vegetation growth is measured using theNormalised Difference Vegetation Index (NDVI)calculated using the visible and near infraredreflectances as

NDVI = (Rnir-Rvis)/(Rnir+Rvis)

where Rvis is the land surface reflectance in thevisible waveband (0.58-0.68 µm) and Rnir is the landsurface reflectance in the near infrared waveband(0.725–1.1 mm) (Smith 1994).

The NDVI in grasslands has a typical seasonalpattern in response to the growth and senescence ofthe annual grasses (Figure 1). The increase in NDVIduring the wet season is related to the amount of fueladded by new growth and the decrease in NDVI withthe onset of the dry season is related to theflammability of this fuel (Paltridge and Barber 1988).Total fuel load can be estimated in the absence ofintensive grazing from the biomass accumulated in thecurrent season and the time since the last fire.Similarly the NDVI during senescence estimates thelevel of curing or flammability of the grassland. Thecuring index is estimated from an inverse function ofthe NDVI to give minimum curing (CI=0) at NDVI=1( Paltridge and Barber 1988).

Figure 1. Change in NDVI with seasonal growth ofgrasslands in relation to fire fuel load and fuelflammability.

Satellite SelectionAs indicated previously, choice of the satellite sensorselected for bushfire monitoring depends on the costof real-time accessibility to the data, vegetation type,risk of cloud, revisit frequency, spectral bands andspatial resolution required. Candidate satellites aregenerally in low earth, polar and sun synchronousorbit. As satellite parameters are interrelated, thehigher the spatial resolution the higher the cost of dataand the less frequent the revisit times. Our focus is onuse of the NOAA-AVHRR sensor (Table 1) becauseof its direct broadcast, low cost, frequent revisit andwide spectral coverage, which all compensate for itslow spatial resolution.

Table 1. Characteristics of the NOAA-AVHRR satellitesensor.

Band Wavelength Waveband Applications(µm)

1 0.58–0.68 Visible Surface features, cloud,vegetation, albedo

2 0.73–1.10 Near infrared Water, vegetation,firescars, albedo

3 3.55–3.93 Thermal Fires and volcanoes4 10.5–11.3 Thermal Sea, land, cloud

temperature andevaporation

5 11.5–12.5 Thermal Sea, land, cloudtemperature andevaporation

Swath 2800 km; Scan angle +/- 55.4°; Resolution at nadir 1.1 km; Revisit 12 hours; Repeat cycle 9.2 days.

Other satellite sensors are used when differencerequirements exist. Hence Landsat-TM is used to mapthe extent of wild fires in southern areas for disasterassessment and for the management of fire in theKarijini National Park in the Pilbara, WesternAustralia.

Western Australia’s Fire and Emergency Serviceconcept for the integrated use of AVHRR to provideinformation on bush fires, fire history, fuel load andflammability for tactical and strategic management ispresented in Figure 2. This model requires ultimatelythe incorporation of the fire information into aGeographic Information System with othergeographic data for querying to address the morespecific questions of fire managers and researchers.This system is evolving as these questions are beingasked of the data.

Growth

Time of year

New fuel load at peak NDVI

ND

VI

Curing andincreasingflammability

116

Application in Operational Bush Fire Management

The application of the four sets of information ispresented in Figure 2.

Figure 2. The application and frequency of informationderived from the NOAA-AVHRR satellite sensorfor tactical and strategic fire management by landmanagers (adapted from Mal Cronstedt, Fire andEmergency Services, WA).

Fire detectionThe use of the four sets of satellite information forbush fire management by the Fire ManagementServices in northern Australia is illustrated in Figure 2.The most frequently used information is the FHS fromthe automatic processing of nighttime NOAA-12, -14and -15 data, posted on the World Wide Web (WWW)

at http://www.rss.dola.wa.gov.au/noaafd/NOAAfd.htmlwithin two hours of the overpass (Figure 3). Locationaccuracy at night is limited to the orbital model usedfor satellite navigation. Therefore to provide greateraccuracy and interpretation of fire fronts, the earlymorning overpasses from Monday to Friday areprocessed manually by panning through the image andlooking for groups of dark (hot) pixels in band 3 ofAVHRR. This information is then faxed toappropriate regional managers and posted on theWWW. All the computer processing takes placewithin 3 hours of the overpass to provide landmanagers with a near real-time update of the locationof currently burning fires.

Figure 3. Example of a hot spot image on 3 April 1999 fromNOAA-15 at 20.23 hrs. Red denotes hot spots thatmet all four criteria and green all hot spots that metthree of the criteria used in the contextualalgorithm employed.

The image (Figure 3) indicates a number of smallfires from the burning of agricultural stubble prior toseeding in southwestern Australia. An initial studyusing AVHRR data during October collected in Perthindicated that this technique of fire detection is alsoapplicable in eastern Indonesia (Figure 4).

The hot spots in Figure 4 indicate that hot spots offires in southeastern Indonesia can be similarlydetected. Using the NOAA-AVHRR data fromDarwin to complement that from Perth the Australian

GREEN VEG COVER • Fuel load & fire risk assessment • Prevention Planning • Fortnightly

CURING – Flammability • Fire weather forecasts • Fire control readiness • Timing of controlled burns • Weekly

FIRE SPOTTING • Remote fire detection • Response priorisation • Resource allocation • Daily

FIRE SCAR MAPPING • Planning • Historical overview • Efficiency of controlled burns • Annual and historical

117

FIREWATCH project could be extended to coversoutheastern Indonesia to provide similar products forfire management and fire research.

Fire historyThe next most important information is the FAAwhich, when repeatedly mapped over time, is used tocreate a fire history (Allen and Russell-Smith 1999—see Figure 5). This information is also used to assessthe efficiency of controlled burns in the currentseason, the area affected by specific fires and fuel loadbuild up, to undertake strategic planning of controlledburns and to access the ecological impact of thefrequency and season of the fire. Early-season coolfires in general cause less ecological damage towoody species than later-season hot fires. The FAAhas been shown to have adequate accuracy for thispurpose (Langaas et al. 1999).

Fuel load build up and curing of grasslands orannual cropsFuel load build up is estimated from the NDVI imagesthat are produced twice a month from the combinationof a sequence of overpasses to get a cloud ‘free’ view(Figure 6). Level of curing (Figures 7 and 8) isestimated from the NDVI of individual overpasses

Fire Hotspots detected on the Islands of Sumba, Flores and Timor, Indonesia, October 1997

Produced by: Satellite Remote Sensing ServicesDepartment of Land Administration, Leeuwin Centre, Perth, WAFires mapped using NOAA-12 satellite imagery during October 1997.The red crosses indicate the position of fires during this period.

Figure 4. Hot spots detected in Indonesia from 14 fourteenNOAA-AVHRR overpasses collected in Perthduring October 1997.

Figure 5. History of areas burnt 1993 to 1996 in theKimberley region of north western Australia.

1:burnt 1 out of 4 years

2: burnt 2 out of 4 years

3: burnt 3 out of 4 years

4: burnt all four years

118

during the senescence phase of the seasonal NDVIcycle (Figure 1).

The curing index in Figure 7 of the agricultural areaof southwestern Australia was produced for the Fireand Emergency Services, WA. It is used incombination with meteorological information toforecast the fire risk and to decide whether a total fireban should be imposed. This same index is producedfor northern Australia for the purpose of planning thetiming of controlled burns using aerial incendiarydevices to get optimum results across large areas(Figure 8).

ConclusionThe success of FIREWATCH depends on theprocessing of large amounts of AVHRR data in nearreal time into relevant information products. Data

from nighttime orbits of three satellites—NOAA-12, -14 and –15—received in Perth and Melboure andtransferred to the Leeuwin Centre by Internet arecurrently automatically processed for FHS. Thisresults in up to 20 overpasses being processed everyday for FHS information. Automation of this processhas enabled provision of this service 24 hours a day 7days a week to Western Australia, Northern Territoryand South Australia.

Provision of this satellite information as a routineservice has received favourable response frommanagers of bush fires in Western Australia andNorthern Territory. The Bush Fires Service in WAreports that negative attitudes of land managers to firemanagement, typified by the feeling that little can beachieved, have been changed positively when reliablenear real-time information has been made available.The set of AVHRR information products representedin Figure 2 provides new tools for the management ofbush fires in remote areas where information is sparse.Figure 2 depicts the future development of these toolswithin a GIS that will give land and fire managersbetter ability to query this information. The AVHRRsensor has limitations but these need to be weighedagainst the requirement for low-cost near real-timeinformation, which only AVHRR’s daily, regionalcoverage can achieve.

The current FIREWATCH service could beextended to southeastern Indonesia where the climateis similar to northern Australia. AVHRR datacollected in Darwin would provide coverage 3000 kmto the north and east of Australia. The mainrequirement would be the establishment of an Internetlink between Darwin and Perth of sufficient speed toget the AVHRR data to the Leeuwin Centre for nearreal-time processing and delivery to fire managers insoutheast Indonesia by Internet, fax and ordinary mail.Such a service would provide a useful basis forresearch into the management of fire in tropicalsavannas. It would provide a comparable set of datafor research and management on both sides of theTimor Sea.

ReferencesAllen, G. and Russell-Smith, J. 1999. Patterns of fire in north

Australia from AVHRR images. Abstract, 4th NorthAustralian Remote Sensing and GIS Conference 28–30June, 1999 Darwin, 63.

Langaas, S., Ryan P.G. and Russell-Smtih, J. 1999.Assessing DOLA/Firewatch AVHRR fire scar maps ofthe Kimberleys, Western Australia 1997–98, usingairborne transect data. Abstract, 4th North AustralianRemote Sensing and GIS Conference 28-30 June, 1999Darwin, 67.

Paltridge G. W. and Barber, J. 1988. Monitoring grasslanddryness and fire potential in Australia using NOAA-

Composite NDVIimage, October1997 showing thecured grasslandsin north westernAustralia and thegreen pasturesand crops in southwestern Australia

GreenVegetationCoverBlue = high Green = moderateBrown = low

Composite NDVIimage, April 1998showing the greengrassland growthin northwesternAustralia and thecured crops andpastures in southwestern Australia.

The perennialforests and densewoodlands remaingreen.

Data in whiteareas were cloudaffected.

Figure 6. Contrast between the NDVI in October 1997 andApril 1998 showing the changes in greenvegetation cover of grasslands due to changes inseasonal growing conditions.

119

Figure 7. Example of a curing index image from 13December 1998 from a single afternoon NOAA-14overpass for use in bush fire risk forecasts.

Figure 8. Curing index of 26 March 1997 produced for theBush Fires Service to estimate fuel dryness toassist in timing controlled burns.

120

AVHRR data. Remote Sensing of Environment, 25,381–394.

Smith R.C.G. 1994. Final Report Australian VegetationWatch. Rural Industries and Development CorporationPublication, DOL-1A.

Smith R.C.G., Craig, R.L., Steber, M.T., Marsden, A.J.,McMillan, C.E. and Adams, J. 1998a. Use of NOAA-AVHRR for fire management in Australia’s tropicalsavannas. Proceedings 9th Australasian RemoteSensing and Photogrammetry Conference, July 20–24,Sydney NSW, Australia, Paper 123, Volume 1.

Smith R.C.G., McMillan, C.E., Craig, R.L., Adams, J. andSteber, M.T. 1998b. Satellite monitoring of bush firesin Western Australia. Journal of Arid Land Studies, 7S,27–30.

121

EFFECTIVE COMMUNICATION is an important part ofeffective fire management in northern Australia. Thisis a message that consistently emerges fromworkshops and fora on fire management involving abroad range of fire managers across the region(Jacklyn and Russell-Smith 1998; Russell-Smith andSaint 1998). Numerous techniques and skills ofcommunication are needed for effective firemanagement—many more than are found in mostprofessional communicators. And reliable expertise isespecially vital when one considers the nature ofbushfires in northern Australia.

Effective Communication for Effective Fire Management

Some of the characteristics of northern Australianbushfires that are relevant to communication are listedbelow.

Large size of northern fires Figure 1 shows satellite imagery of the fire history fornorthern Australia in 1998. Some of the fires over thistime were very large, spreading over land tenureboundaries across different pastoral properties butalso across land managed by quite different groups.This can be seen in the land-use map of northern

Australia (Figure 2). It shows that pastoralists andaboriginal communities manage the largest areas andthat significant chunks of land are managed forconservation and by the defence forces. Added tothese groups are tourist operators who influenceattitudes, and mine operators who manage fire onmine sites.

Given the size of northern fires, effective firemanagement requires that all these groups in a givenregion may need to communicate with each other todevise a coordinated approach. An example of this isseen in the Kakadu region in which a national park isbordered on the west by pastoral properties anddefence forces land (Mt Bundy) and on the east byaboriginal land (western Arnhem Land). Wildfiresfanned by strong southeasterly late dry season windscan spread from western Arnhem Land into the parkor from the park onto Mt Bundy. To prevent suchsituations park fire managers, aboriginal firemanagers and defence force fire managers need tocooperate.

Wide variety of fire management knowledgeLocal and traditional knowledge gathered over manyyears or generations can be needed for good firemanagement because of the fine-grained and long-term nature of fire’s impact on the landscape andbecause much of the landscape has been shaped byaboriginal use of fire (Russell-Smith 1997; Jones1995). However, when a combination of these factorsimpacts on a landscape that is changing through

Communication and Fire Management in Northern Australia

Peter Jacklyn1

AbstractHow do we communicate effectively in the area of fire management? There has been little detailedresearch on this issue. This paper draws on the little information available, primarily from recentworkshops on fire management in northern Australia, to draw some rather broad conclusions aboutwhere we should be heading. It starts by listing the reasons why communication is important for firemanagement in northern Australia. Some broad rules about how such resources might best be used ina communication strategy for fire management are then discussed. The paper concludes with a briefoverview of the activities of the Tropical Savannas Cooperative Research Centre in fire managementcommunication.

1Tropical Savannas CRC Northern Territory University,Darwin NT 0909

122

human intervention, and this occurs in remote areaswith sparser populations, scientific research andtechnology can also be useful for effective firemanagement (Andersen et al. 1999). And here the fireresearchers need to communicate well with firemanagers who have local and traditional knowledge.

This is no simple task. Research scientists have adifferent cultural outlook from fire managers innational parks, and this has led to communicationdifficulties over fire management in Kakadu (Andersenand McKaige 1997). The even greater culturaldifferences that can exist between aboriginal andpastoral fire managers and some fire researchers hindercommunication (Andersen and McKaige 1997).Aboriginal and pastoral communities will often regardresearchers as ‘blow-ins’ who have little respect for,

and get little respect from the communities they havefleeting contact with. Equally, some fire researchers arefrustrated by what they see as an unsystematic approachtaken to fire management by the fire managers.

Remoteness of many fire managersOnce people decide that they want to manage firemore effectively they need access to relevant firemanagement information. At present this is difficult:most scientific papers on fire management are tuckedaway in university libraries, yet most fire managersare in remote locations across the tropical savannas.Furthermore much of the local and traditionalknowledge is not in written form. Intellectual propertyissues are crucial when considering the widerdissemination of this knowledge.

E125° E130° E135° E140° E145°

E125° E130° E135°

0 200 400

kilometres

E140° E145°

S20°

S18°

S16°

S14°

S12°

S20°

S18°

S16°

S14°

S12°

Mapped from NOAA-14 Satellite Imagery by: Satellite Remote Wensing Services Department of Land Administration Leeuwin Centre, Perth, W.A.

Figure 1. Western Australia, Northern Territory andQueensland fire history — 1998.

Pastoral land

Aboriginal land

Vacant crown land

Defence land

Nature reserve

Marine reserve

Figure 2. Tenure in north Australia.

123

Emotive nature of fireFire management can be an emotive and politicallypotent issue and public attitudes, regardless of theirfoundation, can shape political responses to firemanagement and legislation. In Western Australia, forexample, fire management policy is largelydetermined by attitudes prevailing in the populoussouthern–western corner, with a consequent emphasison fuel reduction burns and fire suppression to reducethe danger to life and property from catastrophic fires.This policy can be inappropriate when applied in theKimberley region. To get more appropriate firemanagement policies, greater public awareness of firemanagement is useful (Anderson 1997; Jacklyn andRussell-Smith 1998). For northern Australia the‘public’ here are largely the residents of the largeurban centres in Western Australia (WA), Queensland(Qld) and the NT, who influence the fire managementpolicymakers in those jurisdictions.

Importance of legislation Because of the danger to life and property, firemanagement is bound by laws in each state and insome cases these laws may need to be changed toaccommodate more effective fire management innorthern Australia. Hence, there is also a need for moretargeted awareness communication—the lobbying ofpoliticians and other key decision-makers on firemanagement (Jacklyn and Russell-Smith 1998).

Improving Communication in Fire Management

Given the importance to fire management of the formsof communication identified above, what follows aresome ‘rules of thumb’ about how such communicationmay be encouraged.

Use meetings—especially workshops and hands-on demonstrations Creating better communication between different firemanagers and between fire researchers and firemanagers requires encouraging conversation betweenpeople who often have quite different cultures andperceptions concerning fire. Here a valuabletechnique is to use meetings or workshops wherepeople talk informally and on a one-to-one basis.

It can also be important that at such meetings theinformation provided by the different participants isgiven equal value. For example a workshop that is runafter the major decisions have already been made forthe purposes of publicising research and/ormanagement may be resented by those not party to thedecisions. Consultation needs to take place at allstages of research planning (Andersen and McKaige

1997). Workshops aiming to bridge cultural gaps havea greater chance of success if participants feel theyhave a reasonable chance of influencing workshopoutcomes.

An example here is the North Australian FireManagers Workshop run by the Tropical SavannasCRC in Darwin in March 1998, where participantsincluded fire researchers and fire managers from acrossthe savannas representing aboriginal communities, thepastoral industry and the defence forces. The workshopwas designed to identify fire management issues in thevarious regions and sectors of the tropical savannas.There were opportunities for break-away discussiongroups involving a mix of different fire managers witha common interest in a particular region or sector—which encouraged a useful exchange of diverse viewson a particular problem. This workshop received goodfeedback from almost all participants. It was, however,an expensive undertaking.

An emphasis on ‘hands-on’ tackling of concreteproblems was seen in the successful Kalumburu FireWorkshop, which involved mainly pastoral andaboriginal fire managers getting together with fireresearchers near Kalumburu in the northernKimberley for a demonstration of remote sensingtechniques in mapping fires. An area was set alight—asatellite passed overhead and the consequent imagesof the fire scars were recorded. Getting people tocooperate in practical demonstrations of firemanagement techniques like this can be a veryeffective way of communicating the value of thetechniques as well as getting different fire managersand researchers to better understand each other’s pointof view. Again this workshop was very well receivedby its participants.

Establish good community connections The skills needed to conduct such workshops arerelatively straightforward organisational abilities.However, in order to get useful attendance from oftentight-knit aboriginal and pastoral communities, goodrelationships with those communities are also needed.Establishing mutual respect and trust with acommunity—not just through workshops but throughmany different avenues—builds the basis for effectivecommunication with a group. These relationshipscan’t be bought or learned or built quickly.

Another important asset is good connection withkey organisations in different sectors and fire managergroups, e.g. with the fire services, pastoral groups,tourist operators, aboriginal organisations etc. Eventssuch as the North Australian Fire ManagersWorkshop, which saw over 100 people attend, aregood at building such links. The Tropical SavannasCRC is also well placed here, having associationswith 16 partner agencies, most of them involved infire management.

124

Seek professional advice when appropriate More conventional public relations and media skillsare needed to run public awareness campaigns. Theproduction of brochures and the planning of mediacampaigns targeted at large urban audiences are oftenbest handled by professionals with the design andgraphic skills and the media experience. Urbanaudiences are saturated with slick messages and ittakes a well-packaged publication or media campaignto succeed here (Andersen and McKaige 1997;Macnamara 1984). It can also take a deal of luck to getyour message across to an urban audience—particularly if the message attempts to conveycomplex environmental issues. Mazur (1998)describes the convoluted and unpredictable process ofinfluencing public opinion on major environmentalissues through mass media coverage.

Considerable resources can be wasted trying toreach a mass audience through the more glamorousmedia outlets, whereas smaller, regional audiences areeasier to reach, particularly if the fire issue is locallyrelevant. Small local newsletters or newspapers willoften publish press releases verbatim if the subject isconsidered of interest.

Use smart lobbyingSimilarly, when it comes to lobbying politicians,experience in the field helps. CSIRO has recently beenrunning national science briefings in ParliamentHouse in Canberra, featuring key scientists givingshort, punchy lunchtime presentations to politicians.The organisers found that the presentations must betailored to politicians—not more than seven minuteslong, for example, or the politicians tend to startlooking at their watches (Parsons pers. comm). It alsohelps to be aware of the priorities of key politicians.

Another important audience comprises keydecisionmakers who aren’t politicians, such asopinion leaders in the fire manager communities.Lobbying here can be very effective but often requiresa good relationship with the group concerned.

Establish an accessible database of firemanagement information All fire managers do not need direct access to such adatabase. But if land management agency and fireservice officers have access to such information, andif it can be spread by various other media, then it willbe indirectly accessible by most fire managers and abroader audience.

This strategy has a low probability of completefailure. Public awareness campaigns can misfire or bemisinterpreted, lobbying can run into brick walls andworkshops can attract the wrong people, yet aninformation database will always be useful to

somebody. And such databases once created are therefor the future when circumstances may be morefavourable for successful fire management.

It may also be important to make such a databasenot just an instrument for disseminating the ‘gospel’on fire management—but to make it also a way ofgathering fire information from users. This allows it tobetter incorporate useful local and traditionalknowledge (assuming intellectual propertyconsiderations are accommodated) and allows it to bemore effectively ‘owned’ by user groups.

Continuously evaluate the effectiveness of your communicationGiven the lack of tried and tested methods in this areathere is a need to continually test the communicationtechniques employed. This is particularly the case infire management in north Australia because thedifferent cultures involved mean that one particulartype of communication will rarely work for all groupsinvolved in fire management—and what does anddoesn’t work may not be intuitively obvious.

Fire Management Communication in the Tropical Savannas CRC

The Tropical Savannas Cooperative Research Centre(TS-CRC) is an unincorporated joint venture between16 partner agencies covering the major research andland management agencies servicing the tropicalsavannas, together with the region’s universities. Italso has close links with stakeholder organisationssuch as the bushfire services, the Aboriginal LandCouncils, pastoral research agencies and touristbodies in north Australia. Thus it has a ready-madenetwork suitable for fire management communicationinvolving researchers from a range of disciplines andfire managers from a range of stakeholders. For moreinformation on the Tropical Savannas CRC and firemanagement refer to Hill in these proceedings.

The North Australian Rural Fire Managers Forum This forum involves the three bushfire services innorthern Australia (The Fire and Emergency ServicesAuthority of WA, the Bushfires Council of the NT,the Queensland Fire and Rescue Service) and the TS-CRC. Regular meetings allow whole-of-northAustralia communication strategies to be formulatedthat exploit the links and strengths of the agenciesinvolved. One of the first initiatives of the forum is afire awareness brochure for north Australia, targetedat visitors.

125

‘Clearing House’ of savanna information This will be a computer-based web-accessibledatabase of land management information that will belaunched in July 1999. One of the main areas coveredwill be fire management. The idea is that users will beable to visit our Clearinghouse website and then clickon the region of north Australia where they areseeking fire management information. They will thenget a summary of fire information for that region andcan choose the particular fire topic that interests them(fire and weed control, fire and biodiversity, fire andgrazing etc.). Alternatively they could get an overviewof fire management information for the whole of thetropical savannas.

The database is designed both to disseminateinformation and to incorporate information fromusers, and it will be made more accessible to firemanagers on the ground by exploiting the TS-CRC’slinks with the extension and outreach services of itspartner agencies (see ‘Extension and training’ below).

‘Savanna Links’ newsletterThis quarterly newsletter has a circulation of around3000, encompassing a broad distribution of tropicalsavanna land managers and land managementagencies. It frequently features articles on firemanagement and associated land management issues.It can also be used to publicise information on firemanagement available in the Clearinghouse. Allissues are available on-line and are integrated into theClearinghouse.

Regional management studies These focus on three particular regions in the tropicalsavannas: the Victoria River District in the NT, theDesert Uplands and the Burdekin catchment, both inQld. The studies integrate different researchdisciplines and industry sectors and try to getmanagers on the ground working with researchers. Allthe studies involve fire management to some extent,and it is a focus of the VRD study. The ManagementStudies will provide important communicationadvantages such as getting fire researchers to workwith fire managers; there will be ample opportunityfor practical demonstrations and workshops.

Extension and training The TS-CRC’s extension and training projects willhave major components incorporating firemanagement. These projects will exploit the researchlinks of the TS-CRC, the fire managementinformation in the Clearinghouse and the links to firemanagers on the ground of the Management Studies.Some publications on fire management aimed at abroad audience have already been released, e.g.Schulz (1998).

ReferencesAnderson R. 1997. The Northern Territory rural/urban

interface: ‘never the two shall meet’. In: McKaige, B.J.,Williams, R.J. and Waggitt, W.M., ed., Bushfire ’97Proceedings, Darwin, CSIRO, 31–35.

Andersen, A.N., Braithwaite, R.W., Cook, G.D., Corbett,L.K., Williams, R.J., Douglas, M.M., Gill, A.M.,Setterfield, S.A. and Muller, W.J. 1999. Fire researchfor conservation management in tropical savannas:introducing the Kapalga fire experiment. AustralianJournal of Ecology, 23 (in press).

Andersen, A.N. and McKaige, B.J. 1997. Burning issues:communicating fire research in northern Australia. In:Wills, R.T. and Hobbs R.J., ed., Ecology For Everyone:Communicating Ecology to Scientists, the Public andthe Politicians. Sydney, Surrey Beatty & Sons, 88–96.

Jacklyn, P. and Russell-Smith, J., ed. 1998. Proceedingsfrom the North Australia Fire Management Workshop.Darwin, Tropical Savannas CRC, 24–25 March 1998.

Jones, R 1995. Mindjongork–legacy of the firestick. In:Rose, D., ed., Country in Flames, Proceedings of the1994 symposium on biodiversity and fire in northAustralia. Commonwealth of Australia and NorthAustralian Research Unit, 11–17.

Macnamara, J.R. 1984. Public Relations Handbook. Sydney,Margaret Gee Media Group.

Mazur, A. 1998. Global environmental change in the news.International Sociology, 13(4), 457–472.

Russell-Smith, J. 1997. The relevance of indigenous firepractice for contemporary land management practice:an example from western Arnhem Land. In: McKaige,B.J., Williams, R.J. and Waggitt, W.M., ed., Bushfire’97 Proceedings, Darwin, CSIRO, 81–85.

Russell-Smith, J. and Saint, P. 1998. Malgara: burning thebush. Fourth North Australian Fire ManagementWorkshop. Kalumburu, North Kimberley WesternAustralia, June 1997.

Schulz, D., ed. 1998. Fire on the Savannas. TropicalSavannas CRC, Darwin.

Tropical Savannas CRC. http://savanna.ntu.edu.au

126

Institutional Linkages and Cooperative Approaches

127

THE EMPHASIS of this paper is on the Northern Territorylinks established with Indonesian universities,especially from the perspective of the NorthernTerritory University (NTU). My comments are basedon firsthand experience in collaborative research witheastern Indonesian universities, as a member of NTUdelegations to annual meetings with Indonesian partnerinstitutions, and as a former Dean of the Faculty of Artsat NTU, which has been particularly active indeveloping institutional links in Southeast Asia.

NTU is the largest research and teachingorganisation in the Northern Territory. The Universitymaintains close links with other research organisationsoperating in the NT in a range of fields, but especiallyin the environmental sciences. NTU was established in1989, through a merger of the Darwin Institute ofTechnology (DIT, formerly Darwin CommunityCollege) and the University College of the NorthernTerritory (UCNT). From its inception, NTU has had astrong orientation towards Southeast Asia in bothteaching and research. Indonesian studies have been animportant part of this orientation.

Southeast Asian and Indonesian Studies in NTAs a small university contending with an increasinglycompetitive national environment of tertiaryeducation and research institutions, NTU has sought

to focus its activities regionally. NTU has identifiedthe northern Australian and the Southeast Asianregions as of special importance in its teaching,research and community service mission. WithinSoutheast Asia the BIMP-EAGA region (Brunei,Indonesia, Malaysia, Philippines–East Asia GrowthArea) is considered of particular relevance to NTU’sresearch agenda. This is also consistent with theNorthern Territory government’s promotion ofDarwin as Australia’s ‘gateway to Asia’.

The combination of northern Australia andadjacent, primarily insular, parts of Southeast Asia is anatural and logical orientation. Despite culturaldifferences, northern Australia and eastern Indonesiain particular share some striking similarities in naturalenvironment as well as experiences of social andeconomic ‘development’ in the late 20th Century.

There is considerable potential for research acrossthe environmental and social sciences thatsystematically explores commonalities and contrastsacross insular Southeast Asia and northern Australia.The present project on ‘Fire Management in EasternIndonesia and Northern Australia’, involving theCooperative Research Centre for SustainableDevelopment of Tropical Savannas, is an importantstep in this direction. The project is complemented byother research initiatives at NTU. The activities of theCentre for Indigenous Natural and Cultural ResourceManagement are highly relevant for an understandingof Australian aboriginal relations with the environment.A complementary research project titled ‘Arafura

Eastern Indonesian and Northern TerritoryTertiary Institutional Linkages

Chris Healey1

AbstractFrom its inception the Northern Territory University (NTU) has had a strong orientation towardsSoutheast Asia in both teaching and research. Indonesian studies have been an important part of thisorientation. This paper provides an overview of the development of formal linkages between thetertiary institutions of the Northern Territory and their neighbours in Eastern Indonesia, and offerssome comments on the potential for these links to contribute to a better understanding of firemanagement and sustainable development in our region. It also contains observations about thesignificance of the current waves of unrest in Indonesia for reaching longer-term research objectiveson fire management, sustainable agriculture and forestry development.

1School of Southern Asian and Australian Studies, NTU,Darwin NT, Australia

128

Linkages’ has recently been initiated in anthropologyat NTU (Healey 1999). This project embraces thestudy of change and development over the last 2500years of the peoples and cultures of the lands in andbordering the Arafura Sea, and includes attention tosociocultural dimensions of resource management,particularly in eastern Indonesia.

A Centre for Southeast Asian Studies (CSEAS) wasestablished at the UCNT to coordinate researchinterests in the region. The CSEAS is located withinthe Faculty of Arts, but with a brief that encompassesinterests across the full spectrum of disciplinesdealing with Southeast Asian issues. (Note that sincethe presentation of this paper at the Workshop, theformer Faculty of Arts has been regrouped into alarger Faculty of Law, Business and Arts, which nowhosts the CSEAS.)

The CSEAS actively fosters research andscholarship on Southeast Asian topics, especially inthe humanities, social sciences and environmentalsciences, and serves as a point of contact for a numberof informal and formal institutional linkages. (Thefoundation Director of the CSEAS was Dr Paul Webb.Dr Ian Walters succeeded him in 1997. In the sameyear the Centre went through a review of itsoperations by the NTU Research Committee, whichendorsed continuing support for the activities of theCentre for a further 5 years. In line withrecommendations of the review panel, the CSEAS is

becoming more directly involved in research anddelivery of award and non-award teaching programs.)

Table 1 outlines some of the major areas of researchand teaching at NTU dealing specifically withIndonesia. The majority of research projectsconducted by staff and postgraduate students that arefocused locally or regionally, rather than at nationallevel, are conducted in the eastern Indonesianprovinces of Maluku, Nusa Tenggara Timur, TimorTimur and Irian Jaya.

Several other Northern Territory institutions haveestablished research interests in Southeast Asia, andespecially eastern Indonesia. These include the NorthAustralia Research Unit (NARU) of the ResearchSchool of Asian and Pacific Studies at the AustralianNational University, the Museums and Arts Galleriesof the Northern Territory, and a number of NorthernTerritory government departments, including AsianRelations and Trade, Education, Health andCommunity Services among others. Governmentdepartments and instrumentalities have establishedcontacts with Indonesian counterpart departments.NARU recently entered into an MOU with theIndonesian Academy of Sciences (LIPI: LembagaIlmu Pengetahuan Indonesia).

NTU has a considerable concentration of expertiseacross a range of fields in those areas of easternIndonesia most relevant to the concerns of thisWorkshop. To the potential benefits of this expertise

Table 1. Indonesian studies at NTU.A summary of major areas of activity by broad discipline area within the last 5 years in both teaching and research is outlinedbelow. This list is intended to indicate the range of interests, rather than be exhaustive.

Discipline Research & teaching interestsAnthropology and archaeology Early hominids, human ecology, ethnobiology, migration and urbanisation, natural resource

management, maritime adaptations, ritual and religion, material culture

Economics National economy, tourism

Engineering Power generation and conservation

Environmental sciences Ecology and natural resource management, botanical studies, ornithological studies,conservation biology

Fine Arts Traditional and contemporary arts practice

Health sciences Nursing education

History History in eastern Indonesia, the role of the church in development in Indonesia, regionalhistories in Timor, Flores

Indonesian language, Old Javanese, Balinese languages and literature, modern Indonesian literature, European literature and cultural studies representations of Indonesia, modern Indonesian theatre

Law Business law in Indonesia, Islamic law

Linguistics Teaching Indonesian as a foreign language, studies of regional languages in Bali, Timor,Irian Jaya

Marine sciences Aquaculture, fisheries

Music Traditional and contemporary music practice

Political science Indonesian government, media in political system, Australian foreign policy

Multidisiplinary Several research projects involving input from: Anthropology, Environmental Sciences &Linguistics, Anthropology, History, Indonesian Studies & Political Science

129

are added the international network of institutionallinks built up over recent years. This networkcomplements local and national links, of which thepartners in the Cooperative Research Centre for theSustainable Development of Tropical Savannas are aprime example.

International Institutional LinksAt present NTU has entered into formal agreementswith 24 institutions in nine countries in the Asia-Pacific region. Table 2 summarises these variousinternational agreements, all of which are supportedby government-to-government agreements. Thirteenof the agreements specifically include reference tocollaboration in research, and 11 of these agreementsare with institutions in Southeast Asia. Six of theseagreements are with universities in eastern Indonesia(see Figure 1). NTU is thus relatively well preparedthrough formal institutional links to undertakecollaborative research in eastern Indonesia.

The Indonesian universities with which NTUmaintains links are the leading state universities in theProvinces of Bali (Universitas Udayana), Nusa TenggaraBarat (U. Mataram), Maluku (U. Pattimura), NusaTenggara Timur (U. Nusa Cendana), Irian Jaya (U.

Cenderawasih), and the private university servicing EastTimor (U. Timor Timur). A brief history of thedevelopment of the memorandum of cooperation(MOC) with eastern Indonesian universities, up to 1994,is contained in an NTU newsletter on the MOC (MOC1994). Details of annual meetings are also published inthe same newsletter series.

These six eastern Indonesian universities alsomaintain a close association with each other, and meetcollectively with NTU on an annual basis (Table 3).Meetings are important as an effective means ofmaintaining cooperative relations, and provide anopportunity to review the progress on current projectsand to discuss future directions.

Evaluation and Prospects for the FutureI concentrate here on the MOC between NTU and thesix eastern Indonesian universities. While manyspecific projects are discussed at annual meetings ofMOC members, they can be grouped into severalgeneral areas for the sake of this outline:1. Capacity strengthening of university operations:

involving library development, universityadministration, technical training, power and watersupply

Table 2. NTU agreements with overseas institutions.

Country Institution TYPE* R/T#Brunei Universiti Brunei Darusalam MOC RT

Further Education & Management Consultants A T

China Tsinghua University, Beijing MOU RTChinese Management Association Hong Kong MOA TShanghai Chinese Relief Foundation LOI THainan University MOC RT

India Myrnd School of Catering, Hotel & Computer Management MOA T

Indonesia Akademi Ilmu Pelayaran Negeri LOU TBalai Pendidikan & Latihan Pelayaran Dasar LOU TUniversitas Cenderawasih (Jayapura) MOC RTUniversitas Mataram (Lombok) MOC RTUniversitas Nusa Cendana (Kupang) MOC RTUniversitas Pattimura (Ambon) MOC RTUniversitas Timor Timur (Dili) MOC RTUniversitas Udayana (Denpasar) MOC RT

Malaysia Kolej TAFE Malaysia MOA TPenang National College LOI TUniversiti Malaysia Sarawak MOC RT

Papua New Guinea PNG Institute of Public Administration MOC T

Philippines University of the Philippines MOU TMindanao State University MOC RT

Thailand Sripatum University MOU T

Vietnam Nha Trang University of Fisheries LOI RTMinistry of Fisheries LOI RT

* Type of Agreement: A = Agreement; LOI = Letter of Intent; LOU = letter of Understanding; MOA = Memorandum of Agreement; MOC =memorandum of Cooperation; MOU = Memorandum of Understanding.# R = Agreement includes research; T = Agreement includes teaching.

130

2. Language training: in-country courses inIndonesian language

3. Social sciences and cultural studies: anthropology,customary and business law, linguistics, tourismstudies, demography

4. Environmental and health sciences, agriculture andhorticulture, aquaculture, marine sciences, animalhusbandry, forestry, nursing training

5. Student and cultural exchange, sporting linksSome proposed cooperative projects in teaching

and research require external funding, and lapse forlack of adequate resources. However, a number ofprojects have been completed successfully or arecontinuing. NTU continues to provide to MOCpartners training and support in library developmentand general university administration. NTU offers ahighly successful intensive in-country Indonesianlanguage program through cooperative teachingarrangements with Nusa Cendana University inKupang and Mataram University in Lombok. Inaddition, the Faculty of Arts and its associated Centrefor Asia-Pacific Arts runs intensive courses in Bali oncontemporary and traditional visual arts and music, in

Figure 1. Eastern Indonesian MOC Partners

(Source: Memorandum of Co-operation News, International Division, NTU, No.1/97)1. UNUD–Universitas Udayana, Denpasar, Bali2. UNRAM–Universitas Mataram, Mataram, Nusa Tenggara Barat3. UNDANA–Universitas Nusa Cendana, Kupang, Nusa Tenggara Timur4. UNTIM–Universitas Timor Timur, Dili, Timur Timor5. UNPATTI–Universitas Pattimura, Ambon, Maluku6. UNCEN–Universitas Cenderawasih, Jayapura, Irian Jaya7. NTU–Northern Territory University, Darwin, Northern Territory

Wallace’s Line

Table 3. MOC meetings with Indonesian partners and NTU.

Meeting Host Institution1. April 1991 U Pattimura, Ambon, Province of

Maluku

2. June 1992 NTU, Darwin

3. May–June 1993 U Pattimura

4. June–July 1994 U Mataram, Mataram, Provinceof Nusa Tenggara Barat

6. June 1995 NTU

7. May 1996 U Cenderawasih, Jayapura,Province of Irian Jaya

8. June 1997 U Udayana, Denpasar, Provinceof Bali

NB:1. At the 4th meeting, the numbering of the next meeting wasincorrectly recorded in the Minutes as number 6. There was no 5thmeeting.2. Because of the economic crisis that hit Indonesia in 1997, andserious civil unrest, particularly in Maluku and Timor Timur,meetings scheduled for 1998 and 1999 were postponed.

131

collaboration with other Balinese tertiary institutionsand private organisations.

Turning to cooperative research, the mostsuccessful projects have been in anthropology andmarine sciences. Both areas attracted substantialfunding under the Targeted Institutional LinksProgram through the Australian Vice-ChancellorsCommittee, as well as funding from other sources. Anongoing anthropology project on change anddevelopment in Maluku Province involvesanthropologists from NTU and Pattimura Universityin Ambon. This project received $120 000 in 1991,the first externally funded project to involve any MOCpartners. A project on the development of aquacultureindustries for trochus shell and other valuable marineresources has involved NTU and a number of MOCpartners, but especially the Nusa Cendana andPattimura Universities. It received $250 000 in 1993.

NTU has provided a limited number of specialscholarships to enable staff from Indonesian MOCpartners to undertake further studies at NTU. To date,several staff members have obtained laboratorytechnician qualifications, while others havecompleted or are currently enrolled in postgraduateresearch degrees in anthropology and linguistics.

I think it is fair to say that cooperative researchactivities undertaken under the terms of MOC linkswith eastern Indonesian universities have so farproved highly successful. This evaluation takes intoaccount the success of research teams to attractexternal grant support, and the outcomes of researchin terms of successful graduations of postgraduatesfrom MOC partner universities, publications, anddevelopment of related research projects.

Indonesian MOC partners have demonstratedconsiderable enthusiasm for collaboration with NTU.A major impediment to achieving greatercollaboration is the difficulty of obtaining funds forresearch projects. There is a tendency for IndonesianMOC partners to expect NTU to take the lead in thedevelopment and management of collaborativearrangements, and to secure the necessary funding.This is unfortunate for two reasons. First, it tends toinhibit our MOC partners taking the initiative forprojects, including the development of closercooperative research links between Indonesianinstitutions. Second, Indonesian institutionspotentially have access to national and internationalsources of funding that are not available to NTU as aninstitution from the so-called ‘developed’ world.

Five of the six MOC partners are located inprovinces that lie east of the Wallace Line (seelocation in Figure 1). Among biogeographic featuresshared with northern Australia are savanna andwoodland habitats with marked wet–dry monsoonalclimates. These are widespread in southern Irian Jayaand the island of Timor, but are present in all

provinces of MOC partners, including in western andsouthern Bali, and the south and southeast islands ofMaluku. In addition, there are broad similarities oftraditional patterns of exploitation and management ofenvironments across much of eastern Indonesia,including the place of fire in subsistence strategies.

While broad environmental parameters and patternsof human interaction with the landscape are wellknown, details of variations at local level remainpatchy at best. There is already considerable localexpertise in the social and environmental sciencesamong staff of all MOC partners to fill in these gaps inknowledge relevant to fire management issues.

The research capabilities of MOC partners havebeen improving in recent years, as more staff inIndonesian universities complete research degrees—many from overseas institutions. This is important, asit means MOC partners will gain enhanced capacity toprovide solid training for future generations ofresearchers. Unfortunately, relatively few universitiesin Indonesia offer research degrees, and as aconsequence a research culture is not as highlydeveloped as it might be. Nonetheless, there isundoubtedly great potential for a more coordinateduse of the human and intellectual resources of theMOC partners.

There is also potential for building closer linksbetween NTU, our MOC partners, and othergovernment and various regional, national andinternational nongovernment organisations (NGOs)working in the areas of environmental resourcemanagement, social and economic development in theregion. The Nusa Tenggara Community DevelopmentConsortium is an example of a well-establishedregional organisation in southeast Indonesia, but thereare numerous other, often small, local NGOs (usuallylabelled Yayasan in Indonesian). MOC partners in theregion are particularly well placed to mediate andadvise in the development of these links.

The major threats to the research potential ofeastern Indonesian universities come from theunstable economic, political and social environmentof contemporary Indonesia. The sharp and continuingdownturn in the Indonesian economy since 1997 hasleft universities seriously under-funded, andundermined their ability to provide researchinfrastructure. The recurrent waves of social andpolitical violence are a major threat to nationalsecurity. The human and material costs are high,particularly in Ambon and East Timor, wherereligious, inter-ethnic and political violencecontinues. Staff of Pattimura University have beenkilled, and there has been severe disruption touniversity operations and destruction of universityfacilities, including the new marine scienceslaboratory on the north coast of Ambon near Hila. It isdifficult for Indonesian nationals to maintain an active

132

involvement in research in an environment of suchturmoil, and virtually impossible in Maluku and EastTimor (Brouwer and Soselisa 1999).

Assuming that the central government andprovincial authorities are ultimately able to restorelaw and order in the worst effected areas, the currentcrisis is unlikely to affect researchers in the long term.However, it would be unwise to assume that MOCpartners would simply be able to engage incollaborative research with the capacity theydisplayed prior to the current crisis without a period ofrehabilitation.

ConclusionThe MOC between NTU and eastern Indonesianuniversities is now well established with a good track-record of successful collaboration in research in boththe social and environmental sciences. This should beencouraging for future cooperation building oncurrent networks. Subject to securing externalfunding, targeted projects should be sustainable andable to achieve goals, to the benefit of all participants.

Economic and political crises are likely to impacton resource use. Increasing poverty will drive manypeople to look for alternative or supplementary meansof livelihood. The ability and will of national andprovincial governments to implement and enforceregulations on environmental resource managementmust also be questioned as economic and politicaltensions continue. These largely social and culturalfactors may have profound environmentalconsequences, including, for example, shiftingpatterns of subsistence-oriented agriculture, andchanging emphases in the intensity and extent ofanimal husbandry, plantation agriculture, fisheries

and aquaculture, forestry and forest clearance. Thesefactors and their environmental impacts will need tobe monitored. Researchers in MOC partnerinstitutions are well placed to undertake such aprocess, provided that material resources, training andcollegial support are made available and consolidateand enhance current research capabilities.

AcknowledgmentsI thank Professor Greg Hill for his invitation tocontribute to this Workshop, and to the Office ofInternational Relations at NTU for access to files andinformation. This paper was prepared during thetenure of a Visiting Fellowship with the NorthAustralia Research Unit of ANU, and I thankNARU/ANU for their support. I am also grateful to ananonymous reviewer for helpful comments on a draftof this paper.

ReferencesBrouwer, Arie and Hermien Soselisa 1999. Histories,

Malukan identities, violence and roads toreconciliation. Paper presented to 5th InternationalMaluku Research Conference, Darwin, NorthernTerritory University.

Healey, Chris 1999. Arafura linkages: a new programme ofresearch at Northern Territory University, Australia,International Institute of Asian Studies NewsletterLeiden. Leiden University (in press).

MOC 1994. ‘The Memorandum of Cooperation—a shorthistory’, MOC: Memorandum of Co-operation News,Darwin, External Relations Division, NorthernTerritory University, August 1994, 5–7.

133

NUSA TENGGARA is a region of Eastern Indonesiawhich is situated to the east of the Island of Bali andsouth westerly from the Australian continent (seeFigure 1). The region encompasses three provinces,namely West Nusa Tenggara (NTB), East NusaTenggara (NTT) and East Timor (Tim-Tim). Thepopulation of the three provinces is 7.4 million. Therace of Nusa Tenggara represents a transition betweenthe Malaysian and Papuan races, and possesses verycomplex and various ethnic characteristics. Per capitaincome is approximately one third of the national percapita income. Almost 60% of the population has onlyreached primary school level of education.

Most of the Nusa Tenggara community relies ondry farming, which is very dependent on rainfall, astheir main source of income. The agricultural systemwas until recently simple, with rotation and slash andburn methods. However in the last decade, themanagement of plantations has intensified.

Characteristics of the Nusa TenggaraCommunity Development Consortium

Based on the natural, economic and social conditionsmentioned above, along with experience learnt whileassisting the community, in 1992 a few developers

took the initiative of developing a cooperativenetwork. The network that was created began with 12institutions (eight non-profit organisations, oneuniversity and three donor institutes), with theaspiration of increasing the effectiveness ofagroforestry development as an economic basis forhighland and dryland farmers.This network, known asthe Nusa Tenggara Community DevelopmentConsortium (Konsorsium Pengembangan MasyarakatDaratan Nusa Tenggara—KPMDNT), has grown inparticipation, programs, and activities and also ininstitutional management.

The Consortium represents an umbrella institutionfor communication, coordination and cooperation. Asa network organisation delivered by developers, theexistence of this Consortium is a medium forinformation and experience exchange in technology,methodology, approach, and policies betweendifferent groups. Groups which are involved in thenetwork, amongst others, are non-profit organisations,government institutions, universities, donorinstitutions, local communities and other groups thathave an interest in, and commitment towards, thecommunity problems and conditions in themanagement of natural resources in NTT, NTB andEast Timor (see Figure 2).

The Consortium has formulated a mission that canbecome an operational reference in networkdevelopment. It encompasses a program to raisecommunity prosperity and continuity in themanagement of natural resources through:

Nusa Tenggara Community Development Consortium

Putra Suardika1 and Suhardi Suryadi2

AbstractThe Nusa Tenggara Community Development Consortium is a network aimed at lifting the wellbeingof farming communities in the Nusa Tenggara region in Eastern Indonesia. It does so throughinformation exchange and hands-on experience in methodology, technology and policy, linking upfarming communities with people from government and non-government organisations, and fromeducational institutions. This paper describes the network’s origins and how it has developed throughconsultation with all parties involved. Its focuses on developing activities that answer real needs offarming communities. Success of the network can be gauged from evidence of increasingparticipation of communities in program development using the techniques of particpatory ruralappraisal (PRA), and greater awareness of conservation issues.

1World Neighbours Insular Southeast Asia, Bali, Indonesia2 Institute for Social and Economic Research Education and

Information, Mataram, NTB, Indonesia

134

Priority conservation sites

Figure 1. Nusa Tenggara, Eastern Indonesia (shaded islands).

Figure 2. Provincial divisions of Nusa Tgenggara, April 1999, showing the eight conservation locations.

135

• umbrella institutes for coordination andcommunication between development elementsconcerned with the care of natural resources;

• empowerment of the community and its institutionsalong with human resource development in themanagement of natural resources;

• the provision of information and knowledge on allaspects such as traditional wisdom, social economy,ecology, technology and methodology;

• integration of gender concepts in developingprograms and institutions;

• attempts at perfecting appropriate developmentknowledge that is acceptable to all parties.

Organisation structure and mechanismsThe sixth annual meeting in Mataram NTB in 1997became the milestone for beginning to nurture a moreformal consortium organisation. Although without alegal framework, the organisational structure will bean indication that the consortium intends to find itsplace as a development institution in Nusa Tenggara.The structure of the consortium is shown in Figure 3.

The ultimate decisions of the network are madethrough a consortium general meeting which is heldonce every two years. The network publishes theJALUR Bulletin every 3 months and shares workgroup activities (seminars, workshops, exercises etc.).

Programs and activitiesIn the beginning the network gave attention to thetechnical issues of dryland agriculture through itsagroforestry program. But in its development, theprograms and activities carried out by the consortiumhave grown. After 8 years five main issues have becomethe focus for the consortium, and these have developedinto program areas and work groups. They are:

• Community economy development that is based ondryland farming management—agroforestry,integrated farming, postharvest and marketing. Theagro-economy work group oversees the programsand activities.

• Participative approaches and management programsthrough participatory rural appraisal (PRA)methodology and conflict management. The PRAwork group developed the programs and activities.

• The development of programs and institutions witha gender perspective. The gender work groupdeveloped these programs and activities.

• Community development in continuing managementof forest resource conservation regions. Programsand activities managed by the KSDA working group.

• Development of media information, illumination,awareness and campaigns that support the fourprograms mentioned above. These programs andactivities were managed by the media work group.Community development programs in forest

resource conservation area management focus theirattention on the eight conservation locations shown inFigure 2. The program of strengthening communityeconomy and management of dryland agriculturegives attention to communities and land that is inproximity to or within conservation areas. Thedirection of the programs is shown in Figure 4.

Generally the mentioned programs are carried outthrough activities, which are, amongst others:• Research and documentation • Training• Cross visits or study exchanges• Courses or comparative studies• Meetings• Workshops• Seminars• Media publications

Consortium General Meeting

PRA Work Group

Gender Work Group

Community Development

Work Group

Media Work Group

Agro- economy

Work Group

Secretary

Coordination Council

Figure 3. Consortium structure.

136

Impacts of the ConsortiumImpacts of the programs that are seen now do not justconstitute the consortium work results directly, butrepresent the signs and accumulation of cooperativeprocesses and coordination between stakeholders aswell as work from various participating institutions.Impacts of the consortium can been seen at variouslevels and they are, community, programs and policies.

Community level• The formation of a number of local institutions in a

few conservation areas that have multiple functions,these are to strengthen community unity, increase theeconomic wellbeing of members and conserve forestresources.

• Communities surrounding the forest preservationareas of Sesoat and Taliwang have the trust of theforestry department to manage the forest preservationareas with community forest methods.

• Insight, knowledge and community abilities inmanaging plantations that are stable and intensiveincrease through training study exchange, mediainformation and assistance.

• Community participation and abilities grow inmanaging programs and fighting for their rights withparticipative approaches through PRA methodology.

• The formation of communication forums held bycommunities in conservation areas who have aninterest in the management of natural resources, forexample the communication forum of Wanggameticommunity.

Program level• The formations of nuclear teams that coordinate

activities at the local level in developingconservation area management.

• Research approach at regional level can bring allinterested parties together to find a way out ofnatural resource management problems together.

• There are staff resource development programs inPRA methodology, conflict management, programmanagement and networks. There are even enoughstaff institutions that get access to PRA methodologydevelopments in and outside of Nusa Tenggara.

• There is an awareness and understanding towardsthe importance of gender issues in developingprograms and institutions.

• There is a growth in partnership enthusiasm betweengovernment institutions, non-profit organisations,and the community in developing programs.

Policy level• There has been a change in government approach

from the ‘top down’ becoming a mixture between‘top down’ and ‘bottom up’ and the use of PRAmethodology in program development especiallyby the Forestry Department.

• There is an admission of the success of non-profitorganisations in a number of areas especially fromthe community point of view.

• There is a new approach in conservation areamanagement that previously was only orientatedtowards preservation but is now directed to increasecommunity prosperity and preserve the area.

• A number of participating consortium institutionshave active roles in developing participatory ruralappraisal methodology, developing communityforests through communication forums andparticipative meetings in the network at national level.

Analysis of the Consortium’s DevelopmentAfter being active for 8 years, the consortium representsone of a few communication network organisationsbetween institutions that exist in Indonesia. What isinteresting about the consortium is that the work areadoes not end at the border of one province but workswithin three provinces. Many network communicationinstitutes and cooperatives like the consortium havedeveloped in Indonesia but there are a limited few thatare still active. If there are any, they are only specificinstitutes and work areas, for example the non-profitorganisation forum of Central Java.

To recite the existence of the consortium in carryingout programs and activities, a number of factors thatsupported its development are:• Gradual planning and development through the

organisations and activities that are in accordance

Media Development

Gender Issues

PRA Methodology

Management of Natural Resources based on

communities (8 locations)

Agroforestry & community

economy

Figure 4. Program unification.

137

with the demands and problems of communitydevelopments.

• Democratic management marked by transparencyof activities and acknowledgment of the rights andresponsibilities of all sides as well as cooperation indecision-making processes.

• Facilitation in shaping a concrete activity thatbecame required by participant institutions; whichin turn became the glue between efforts incoordinating perceptions and commoncommitments.

• Substantial number of institutions with a commit-ment and dedication to the role and activities of theconsortium.

• Support from a number of donor institutes, bothpolitical support and funding for the realisation ofits various activities.

Consortium development has been hampered by anumber of problems and obstacles, amongst others:• Many of the consortium’s elite members from non-

profit organisations are lost to the consortium whenthey leave their institution to find new experiencesin other institutions or regions.

• Participants are spread through three provinces,thus increasing costs in carrying out each program.The consortium needs to cultivate new donors suchas commercial businesses and seek more fundingcommitments from participating institutions.

• The involvement of government institutions in theconsortium is still based upon personal approach,not an institutional one. The high frequency ofpersonnel movement in the government means thatconstant effort is needed to develop newunderstanding towards new participatinginstitutions and new government officials.

View of the FutureThe consortium can be said to constitute the onlynetwork institution by a number of groups inpromoting the process of participatory highlandcommunity development. The consortium representsa strength in building alliances, especially betweenregional governments that are not totally in agreementwith practical political ways of influencingdevelopment policies. With the present condition ofeconomic and monetary crisis, this consortium canoffer enterprising highland village communitydevelopment concepts.

Aside from that, Indonesia’s changing politicalsystem that has become more open and democraticcreates the possibility for the community to havecontrol over government development initiatives thatdeviate from and are not in accordance with

community aspirations. The problem is how far is theconsortium able to anticipate a number of futureIndonesian developments. The consortium willprevail, if it is suitable to the needs of the participantsand it can face the next few challenges:• The main well of funds of the consortium network

is still dependent on one donor institution, whichmeans there is a need to find diversified funding fornetwork development at regional and local levels.

• The components and areas of participants in theconsortium are not in proportion. This matter isvisible in conservation management program,where location priority programs and participantsare mostly from NTT, and so are other programs.

• There are demands from a number of networkparticipants concerning how the consortium canhave a higher profile role while working with someissues that are strategic and connected with publicpolicy.

• The campaigning of development policies is stillnot carried out at optimal levels by the consortiumwhether at regional level or national level. The fewconsortium successes need to be communicatedeffectively and thus influence government policyon highland community development.

• There are no processes in place to measure theconsortium’s level of success in managing naturalresources. Each side has its own indicators and theconsortium has been challenged to formulateindicators and develop an evaluation andmonitoring system that is participatory.

• Natural resource management polices are currentlyorientated towards community prosperity andenvironmental preservation. This matter representsa challenge for the consortium to develop conceptsand models of natural resource management basedon community and continuity.

• With regional autonomy and the need for objectiveinformation concerning natural resourcemanagement, the consortium is challenged todevelop human resources in many aspects, throughparticipative meetings, ecological research, thedistribution of legal knowledge and promotion ofagricultural technologies.The consortium has had impacts at all levels—

community, program, institute, and policy making.These are cumulative and provide the drive/motivationfor all participants to extend and maintain theirnetworks, and continue to work cooperatively at alllevels. An evaluation of the consortium and its impactsat all levels is currently under way and it is hoped tohave the results of the evaluation ready for publishingtowards the end of 1999.

138

THE TRI-NATIONAL WETLANDS (TNW) Programfocuses on three conservation areas. Two of these—Wasur National Park, in the Indonesian Province ofIrian Jaya and Tonda Wildlife Management Area inPapua New Guinea’s Western Province, are contiguousand protect a large proportion of a biogeographical areain southern New Guinea known as the Trans-Fly. Thethird reserve involved in the program is KakaduNational Park in Australia’s Northern Territory.

The three reserves contain extensive lowlandwetlands and lie within the monsoon climate zone ofsouthern New Guinea and northern Australia. Eachreserve includes large expanses of seasonallyinundated floodplains, an immensely rich system ofcritical importance for large numbers of migratory andresident waterfowl and waders. All three areas alsocontain large areas of savanna grasslands andwoodlands, small rainforest patches and monsoongallery forests. In addition are large tracts ofundisturbed intertidal mudflats and mangrove in thecoastal regions of the reserves. Indigenous people

reside within each of the areas and have practisedtraditional forms of land management for centuries.The most significant management tool employed isthat of dry season burning and this has had a profoundimpact on landscape and vegetation types that we seein the region today.

All three reserves share similar managementproblems related to exotic weed infestations (Salviniaand Mimosa pigra in Kakadu, Eichornia crassipes,Stachytarpeta spp., Sida acuta, Chromolaena odorataand Mimosa pigra in Wasur and Tonda). These weedspose a critical threat to the integrity of the reserve’snatural ecosystems. Additional grave threats are theimpacts of feral and introduced animals includingcattle, horses, dogs, cats and wild pig in all reserves,as well as the presence of vast numbers of introducedrusa deer (Rusa timorensis) and numerous introducedfish species in Wasur and Tonda. Other managementconcerns common to all three reserves include non-indigenous fire regimes and large-scale vegetationchanges relating to invasion of woody scrub onto thefloodplains grasslands.

Each reserve is currently at a different stage ofmanagement planning and implementation.

Development of Regional Synergies betweenNorthern Australia and the Trans-Fly Region

of Indonesia and Papua New GuineaMichele Bowe1

AbstractThe Trans-Fly savannas of southern New Guinea (including Papua New Guinea and the Indonesianprovince of Irian Jaya) are broadly similar in the nature of their habitats, wildlife and indigenouslandscape management practices, to areas of Cape York and the Northern Territory in Australia.These areas also share similar threats to the integrity of their ecosystems and native wildlife, such asrecent changes in fire management, the impacts of feral animals and exotic weeds. World Wide Fundfor Nature (WWF), through its offices in Australia, Papua New Guinea and Indonesia, isimplementing a program called Tropical Wetlands of Oceania (TWO). The program area covers thesavannas and coastal and freshwater wetlands of northern Australia and southern New Guinea. Animportant component of TWO is the Tri-National Wetlands Program (TNW), which aims to promotesynergies between three protected areas in the region. These are Kakadu National Park in Australia’sNorthern Territory, Wasur National Park in Irian Jaya and Tonda Wildlife Management Area which iscontiguous with Wasur and is located in Papua New Guinea’s Western Province.

This paper focuses on the Tri-National Wetlands Program, describing how synergies between theprotected areas are evolving. Particular emphasis is given to how the program was developed, theinteraction of the various levels at which it operates, and especially the numerous lessons that havebeen learnt over several years of planning and implementation. These lessons have been distilled intoa list of general pointers to guide other agencies and organisations intending to set up complex andmultisectoral programs relating to the research and management of natural resources.

1Tropical Wetlands of Oceania Program World Wide Fund forNature, Australia Darwin, NT Australia

139

• Kakadu is a World Heritage Site, and is managedthrough a joint management agreement between theAboriginal traditional owners and the Director ofNational Parks and Wildlife. The Kakadu Board ofManagement was established in 1989 and has amajority of aboriginal members, representing thetraditional owners of land in the park. The Boarddetermines policy for the park and is responsible forpreparing park management plans. The park is nowoperating to its Fourth Plan of Management and,compared with either Wasur or Tonda, has anadvanced management structure and infrastructure,plus a federal budget to aid implementation of itsmanagement plan.

• Wasur National Park is the managementresponsibility of the Indonesian Ministry ofForestry’s Department of Plantations and NatureConservation (PKA). From 1991 up to the presentthe World Wide Fund for Nature Indonesia andPKA (formerly PHPA) have operated a projectentitled the Conservation and Development ofWasur National Park. This project aims to developand implement solutions to the most pressingproblems faced by Wasur. A draft managementplan has been produced for the park that is currentlyawaiting ratification. PKA have, in the last 18months or so, resourced the park to include amanager and some 30 plus rangers, as well as asmall amount of office infrastructure. The parkhowever is not well resourced from an operationalbudget perspective.

• Tonda Wildlife Management Area (WMA) ismanaged by a committee of traditional landowners,which determines the rules for protection,management and exploitation of the fauna. Theinitial rules were established in 1975 and have beenamended at least four times. In recent years thecommittee has not been very effective inmanagement of the area. Lack of funds precluderegular committee meetings from taking place andthere is a general lack of understanding about therole of the committee and a lack of awareness ofsome of the more pressing ecological problems.The amount of government support for the WMA issmall. A small-scale WWF project is being run inthe area, which focuses on awareness raising andcapacity building for the committee. A WWFproject officer also assists in the facilitation oflinkages with the relevant groups in Irian Jaya.All three sites are important under the Convention

on Wetlands of International Importance Especially asWaterfowl Habitat (Ramsar). The wetlands of Kakaduwere designated a Ramsar site in 1991, Tonda WMAis PNG’s only Ramsar site and Wasur has beennominated to become Indonesia’s fourth Ramsar site.

The aim of the TNW program is to provide aplanning framework which will enable the three

governments to cooperate in the field of wetlandsmanagement. Specifically at a reserve level the TNWwill facilitate a variety of collaborative activitieswhich will further the management and conservationof the three reserves. Activities include:1 The facilitation of a tri-partite memorandum of

understanding (MOU) between the threegovernments that outlines a formal sister-parkrelationship between Kakadu, Wasur and Tondareserves.

2 Using the Ramsar convention as a technical andconvening forum, a tripartite coordinatingcommittee will be formed which will determine thescope of work at a field level under the program.Such activities could include:• wetlands research programs and cooperative

information exchange especially in relation tomigratory birds, habitat change, indigenousmanagement of the reserves, control ofthreatening introduced flora and fauna and thedevelopment of a regional database;

• cross-visits between the reserves to facilitatetraining for wetland managers and rangers;

• management planning assistance for the lessdeveloped reserves through regular workshopsand progress meetings.

At this stage it is suggested that the tripartitecoordinating committee via its secretariat wouldreport activities and developments to the RamsarBureau.

Ultimately is envisaged that a trans-boundaryconservation area arrangement would be establishedbetween Wasur and Tonda. This would enable cross-border collaboration for effective, cooperativemanagement planning between the two reserves.

Reasons for Program DevelopmentThe program grew from the activities of WWFIndonesia, which initiated its project in WasurNational Park in 1991. Together with the IndonesianMinistry of Forestry’s Department of ForestProtection and Nature Conservation (PHPA), a jointproject focused on the production and implementationof a National Park Management Plan. The plan wasdeveloped following several years of research intomanagement issues. Attempts to manage the mostserious threats to the park including introduced andferal animal species, exotic plants and changed firemanagement regimes, quickly revealed the need towork in conjunction with neighbouring Tonda.

Many of the threats that Wasur and Tonda face,such as fire, feral animals and weeds, do not ‘respect’the international border. There is an obvious need tolink management actions and strategies across theborder, for example weed control work that takesplace in Wasur will not be effective unless

140

communities in Tonda have a similar awareness of theproblem and have the resources and capacity to act aswell. Wild fires are also an issue that threatens bothreserves. Customary owners of both Wasur and Tondapracticed fire management, with an awareness of theimpacts on landscape, wildlife, important foodresource areas and gardens. In more recent years, hotfires burn large tracts of land later in the season, withthe result that people’s gardens are destroyed,important areas of vine forest are damaged, and firesfrequently burn uncontrolled until the next rains, oftenspreading across the international border.

Finally an obvious example of the need to worktogether to achieve successful management of Wasurand Tonda is the impact of large numbers of feralanimals, especially the rusa deer. Some initialresearch has shown that the deer are causing large-scale disturbance to natural habitats, especially thegrasslands within the reserves. The deer is perceiveddifferently on either side of the border. In Indonesiaalthough it is protected, there has in the past beenwidespread poaching of the animal to provide meatfor the nearby township of Merauke. This has had theunintentional benefit of keeping the populationsuppressed to some extent. In PNG, the deer are notutilised heavily by the local population, although thereis some trophy hunting from a wildlife lodge inTonda. In both Wasur and Tonda there is not as yet awidespread perception that the deer are a threat. Thereis no traditional knowledge about the deer since it isan introduced species and in both sites there is a lackof expertise in wildlife and landscape ecology. InWasur, although WWF had been raising the issues ofweed control, feral animals and fire for several years,there had been reluctance (perhaps born out of a lackof real understanding) on the part of the parkmanagers to act on any of these issues.

In 1993 after 3 years work in Wasur, WWFIndonesia brought a group of park managers andtraditional owners from Wasur to northern Australiato visit Kakadu National Park. The visit was anunqualified success. Managers and communitiescould see for the first time a well resourced andmanaged national park. They could also see forthemselves that active fire management was anintegral part of the management plan, that other placesalso controlled feral animal populations and that weedcontrol and prevention was an exceedingly highpriority. Interest in tackling similar issues in Wasurwas high and the possibility of further technicalassistance, especially from another park managementauthority, meant that the managers in Wasur did notfeel isolated and could promote these seemingly alienconcepts to head offices in Jakarta with greaterconfidence.

It was within this framework of similarity ofenvironments—threats, management needs and the

need to promote information exchange—that the Tri-National Wetland program began to take shape. It wasobvious from the experience of this initial visit thatthe customary owners and managers of the threereserves could learn much from each other, and ateven a very basic level, experience and informationsharing could do much to promote effective protectionof all three reserves. In 1995 WWF Indonesia soughtfunding for a program that would further facilitatecollaboration between the three reserves in the form ofcooperative research programs, cross-visits andtraining of wetland managers, joint managementplanning for Wasur and Tonda and regular meetingsof staff.

Forming and Maintaining SynergyAt the start of the program there were largedifferences between the level of management activityin each reserve. In addition the managementmechanisms of the three reserves are very different.Both Wasur and Tonda needed funding support aswell as technical and institutional support to assist inthe management of the reserves. In the case of Wasurthis was already provided to some extent by the WWFproject on site, and there was already a good workingrelationship between the government agency, thecommunities and WWF. This meant that there was aprevious relationship on which to build furthercomponents of work related to the TNW program. InTonda, where management of the WMA is decided bya management committee comprising locallandowners, the case was very different. Anyinvolvement in the TNW program would need the fullinformed consent of the WMA committee. In 1996WWF began a small-scale project in Tonda, whichaimed to initiate discussion and agreement in PNGconcerning a trans-border conservation project linkingTonda with Wasur. There had always been some localtension across the border area between PNG andIndonesia and so gaining the good will and consent ofthe local landowners was a critical move.

In the early stages of the development of theprogram it was evident that because of itsinternational nature, there would be a need for someformal mechanism at a government level under whichthe collaboration could take place. The linkagesbetween Wasur and Tonda required special attentionas there was the added dimension of a sharedinternational border, across which it was hoped thatactivities could take place. Therefore before lookingat mechanisms for collaboration between the threegovernments, an initial focus was identifying optionsfor Wasur and Tonda. As part of the field project inTonda, WWF commissioned a report that identifiedmechanisms for legal and institutional coordinationfor a proposed cross-border conservation area.

141

The report concluded that any cross-bordercollaboration necessitated the inclusion of the projectwithin the framework of an existing bilateral treatybetween the PNG and Indonesian governments. The‘Basic Agreement between the Government of PapuaNew Guinea and the Government of the Republic ofIndonesia on Border Arrangements’ includes articlesrelevant to the proposed TNW program, includingarticle 18 on protection of flora and fauna in thevicinity of the border. Pursuing collaboration througha high-level forum such as this would be timeconsuming and complex. However, the BasicAgreement on Border Arrangements would providethe framework in which to address a broader range ofissues relevant to the cross-border conservation area,such as environment, security, border crossings andtrade. Both the Indonesian and PNG governmentagencies have tabled the TNW at border liaisonmeetings that are held regularly under the BasicAgreement, to seek initial endorsement of theprogram.

In addition to the endorsement of the programbilaterally between Indonesia and PNG, it wasconsidered necessary by all parties to formulate anagreement between the three countries involved.There are four levels of agreement deemed necessaryto achieve good cooperation in the TNW. • The first, as described above, was the agreement of

landowning communities, especially in PNG. Theprocess of consultation with communities in Wasurand Kakadu was less structured than the process inTonda, involving informal consultation anddiscussion in Wasur, and discussion at a meeting ofthe Board of Management in Kakadu prior to aproposed visit of traditional owners to Wasur.

• The second level of agreement was at the ‘parkmanagement’ level, in other words the manager ofKakadu National Park, the head of Wasur NationalPark and in the case of Tonda, the WildlifeManagement Area Committee.

• The third level of agreement was at an agency levelin the three countries. This is the Department ofEnvironment and Conservation in PNG, based inPort Moresby; PHPA (now PKA) in Indonesia, basedin Jakarta; and the Australian Nature ConservationAgency (ANCA now Environment Australia) basedin Darwin and Canberra. In Australia the confusionover which office should handle the cooperativearrangement has caused problems.

• An additional level of agreement that ultimatelyproved necessary to describe was that of WWFitself since three WWF agencies were operating inthe program. WWF Indonesia was based in Jakarta,WWF Australia was based in Sydney, Australia andWWF South Pacific Programme ran its operationsfrom Suva, Fiji and had a national office based inPort Moresby.

Liaison between all these partners has proved to bea sometimes difficult and lengthy process.Misunderstandings about who is doing what havesometimes arisen, but of more concern was the factthat it had been difficult to secure coherent planningand implementation of activities related to the TNWprogram. In the early stages of the program there wasa tendency for a great deal of enthusiasm and energybut little synergy between partners.

It was also evident that a program requiring somuch consultation and meeting to maintain clearplanning processes would be expensive to maintain.Initially WWF acted as an informal facilitator toconvene meetings and promote discussion. Asresponses from all three governments were positive itwas decided that the program should be launchedofficially. Thinking ultimately about the sustainabilityof the program, WWF proposed that the TNWprogram be linked to the Ramsar Convention as aconvening forum. All three countries are signatory tothis treaty and so would normally meet every 3 yearsat the Conference of Parties. The Ramsar Conventionalso promotes international cooperation betweenmember states. Therefore as both a cost-savingmeasure and also to enable the three countries toreport on progress with the TNW program as part oftheir obligations as signatories to Ramsar, it wasdecided that the TNW program should be operatedwithin the broader framework of Ramsar. Ramsarwould also be able to provide technical assistance inthe long term.

The TNW program was officially launched in 1996at the 6th Conference of Parties (Ramsar) in Brisbanewith the financial assistance of WWF andEnvironment Australia. At this first meeting ofparticipants at an agency level, an in-principleagreement was reached to seek endorsement of theprogram at a ministerial level in the three countries. Itwas further agreed that on-ground work includingworkshops and initial exchange visits should proceedto allow formulation of training requirements anddrafting of an implementation plan. This work wouldform the basis from which formal agreements couldbe drawn up for Government approvals. Followingthis meeting, WWF initiated its field project in Tondaand received additional funding for the work fromWWF Netherlands and Environment Australia. Thisfunding enabled further training and planning work toestablish research priorities in the Trans-Fly. Severalpublications were ultimately produced—importantrecords of discussions and opportunities that wereidentified and could subsequently be incorporated intoa more formal memorandum.

A second meeting of TNW program participantsoccurred in 1997, where discussions focused on thedevelopment of formal mechanisms for collaborationbetween the three countries. It was decided that a

142

formal Memorandum of Understanding should bedeveloped. This MOU would formalise the existinginformal relationships that had developed between thethree countries as a result of field-level activities.Shortly afterwards in 1998. Indonesia developed thefirst draft of an MOU for discussion and amendmentby the other two governments. The MOU wasoriginally sent to Environment Australia for theirinitial comment and then was to be passed to PNG.Unfortunately changes in staff and departmentrestructuring in Environment Australia have resultedin a bottleneck and very little progress has been madetowards getting the MOU finalised. To date theDepartment of Environment and Conservation inPNG has not seen the draft MOU.

Although progress with the MOU has been slowthere has been a great deal of activity happening in thefield over the last 18 months. Several cross-visits havetaken place, several training courses run, research andmanagement begun in Wasur and a second phase ofawareness and capacity-building activity begun inTonda. Enthusiasm for the program at all levelsremains high and there is every indication thatalthough government bureaucracies move slowly, amultilateral MOU will ultimately support the wealthof field activity relating to the program.

Lessons and Conclusions• Things take time to grow, consolidate and bring

results. The Wasur work started in 1991 and WWFis still pursuing a formal government collaborationmechanism, although field-level activities areproceeding strongly. Policy change does not happenquickly and this has to be taken into considerationat the outset of a program because it has financialand other resource implications.

• Focusing at a formal institutional level isinsufficient if this is not backed up by field-levelactivity. Similarly field-level activity will notmagnify into something greater if there is no policybackup.

• The best programs grow out of existing work orprojects and the agenda have to be developed andsupported by local agencies. Outside agenda rarelywork if there has been no ‘buy in’ or sense ofownership from local groups, agencies andindividuals. There must be community and localagency (both government and non-government)input into the design phase of the program.

• Always build on existing local institutions and takeadvantage of existing processes rather than createnew ones. This ultimately ensures continuity andsustainability since the program will not depend onoutside resources to maintain momentum.

• Complement existing and related programs wherepossible and from the outset find out who is doingwhat and where and how this can link in.

• Determine common objectives for the programfollowing consultation and set these down clearlyso that they are understood and remembered(especially institutionally when key individualsmove away—see below).

• Maintain a constant dialogue between stakeholders,even when there has been temporarily nosubstantial progress, to keep people’s interest aliveand maintain clarity of process.

• The support of key individuals is extremelyimportant for rapid short-term progress, but long-term progress and success will be achieved ifconcepts are institutionalised and result in a changein policy or legal frameworks that ultimatelychange the attitudes of people.

• Formal agreements at a decision-maker or policylevel provide confidence to field-level personnelthat their actions will be approved.

• It can be extremely difficult to maintain continuity.With the example of the development of the MOU,WWF has had to deal with as many as sevenindividuals in one government agency alone in theperiod of 8 months. Drawbacks will inevitablyoccur as departments restructure or are absorbedand your key people move on.

• NGOs often have greater staffing continuity thangovernment departments and are good brokers orneutral facilitators of complex programs. Theyoften have a better field presence than somegovernment agencies, as has been the case in bothPNG and Irian Jaya in past years.

• Often there are relatively few good people to workwith and it is tempting to overload the same few keypeople with all the new programs that start up in anarea.

• Practical training works better on site than off aspeople often have difficulties about extrapolatingnew situations and information to their own place.In a program like the TNW, practical on the jobtraining is more important than formal educationalprograms.

143

THE FIRE management workshop incorporated twobreakaway sessions, with participants allocated to oneof five working groups. Session 1: General discussionof regional fire management issues both in thesavanna lands of eastern Indonesia and northernAustralia. Session 2: Based on Session 1, thesediscussions focused on specific identified regionalissues: development of broad-scale fire monitoringapproaches; cultural issues, tenure, communityawareness and education; ecological research issues;agricultural and forestry research issues; and sectorallinkages and collaboration.

The main findings of the two sessions aresummarised below.

Session 1: Defining the IssuesThe first session aimed to broadly define key firemanagement issues in eastern Indonesia and northernAustralia, and identify research and training needsthat could help address gaps and assist with thedevelopment of improved fire management practicesand policies.

IssuesDocumenting the current extent and patterning ofburning There has been significant development in recentyears in documenting the extent, seasonality andpatterning of burning (e.g. frequency of fire with

respect to different regions and land-use sectors)across northern Australia, but information isunavailable for eastern Indonesia. Fire map data havevarious uses, including locating the daily occurrencesof fires, and informing and monitoring thedevelopment of appropriate regional fire managementpractices and policies.

Documenting fire management practices A considerable (albeit far from complete) body ofinformation is available, documenting past and currentfire management practices and approaches of differentland-use sectors in northern Australia, yet few data areavailable for eastern Indonesia. An understanding oftraditional fire management practices in localcommunity settings is essential, for example, forinforming the development and application in bothregions of appropriate fire management systems incontemporary land-use contexts.

Applied research Some information is publicly available (particularly

for northern Australia) concerning the use of fire inpastoral management (e.g. manipulating stock tograze recently burnt areas; conservation of perennialgrasses; establishment of improved pasture species;woody plant control), but scant data are available forother primary industry and natural resourcemanagement applications (e.g. management of weedssuch as Chromolaena in agricultural and agroforestrysettings; nutrient and soil conservation underincreasingly shorter swidden cycles; water catchment

Workshop Breakaway Sessions

management). It was generally felt that, to involvelocal communities and to encourage eventualimplementation of practical solutions, an interactiveapproach through demonstration or case studies wasthe best means of addressing such land managementresearch needs.

Tenure Land tenure issues in both eastern Indonesia andnorthern Australia are fundamental to the rights (andresponsibilities) of individuals and communities onthe one hand and government agencies on the otherfor undertaking fire management. In easternIndonesian settings both central Government andunofficial, traditional tenure systems may applyindependently to the same tracts of land. Tenure issuesin northern Australia include the recently recognisedlegal rights of indigenous (Aboriginal and TorresStrait Islander) custodians to joint ‘Native Title’ onmuch Government and pastoral leasehold land,including rights and responsibilities for undertakingtraditional fire management. As such, anunderstanding of tenure frameworks is fundamental todeveloping appropriate fire management practices andpolicies in both jurisdictions.

Policy and legislationConsiderable fire management policy and associatedlegislative development is required both in easternIndonesia and northern Australia. This appliesparticularly to recognising the peculiar requirementsof living in, and managing, fire-prone savannaenvironments. In the eastern Indonesian context, thecurrent official total exclusion fire policy is set by thecentral Government, largely as a response to theinternational ramifications of biomass burning in thelarge tropical forest blocs of western Indonesia andnorthern Irian Jaya. In Queensland and WesternAustralia, state-wide policies are determined in thesoutheast and southwest respectively, under vastlydifferent southern Australian, fire-weather conditionsand seasonality. Thus, greater understanding of thefire management requirements of savanna systemsboth in eastern Indonesia and northern Australia will,in time, serve the development of more appropriateregional fire management policies and legislation.

Training, extension and communication The above issues can only be addressed through targetedtraining and education programs, from the localcommunity grass-roots level through to the policymakers. Experience in both regions, but in easternIndonesia especially, has shown that hands-ondemonstration activities are practical, effective meansfor achieving this. As well, there is a need for thedevelopment of regional ‘information clearing houses’,where end-users can readily access information on arange of topics pertaining to land management andrelated issues, or at least be guided in the right direction.

Collaboration, cooperation and linkages As with above information issues, workshopparticipants fully recognised the need for, andadvantages of, developing closer institutional tiesbetween eastern Indonesia and northern Australia.Whereas northern Australian agencies and institutionscan offer various levels of technical and educationalexpertise of relevance to fire management in savannasystems, eastern Indonesia provides excellentexamples of cooperative, information-sharingnetworks involving large numbers of NGOs,educational institutions and agencies. Both regionshave much to offer each other.

Identified research and training needsFollowing is a succinct list of research and trainingneeds identified by the five groups from the firstbreakaway session:• Develop the capacity to undertake a comparative

assessment of the current extent and patterning ofburning in eastern Indonesia and northern Australia,using remote sensing (RS) and GeographicInformation System (GIS) approaches;

• Such an assessment would require the concomitantdevelopment of appropriate land-use, vegetationand cadastral map coverages, as a basis fordeveloping a fire management database—in easternIndonesia this might involve considerable workundertaken either as student projects and/or throughtechnical assistance;

• Undertake an assessment of contemporary firemanagement practices as used in land management,particularly in eastern Indonesia;

• Using above information describe the biophysical,economical, social, cultural, and political effectsand implications of current fire regimes;

• Identify key economic and natural resource issues offire management, and potential locations where suchissues could be addressed through demonstrationactivities and community interaction;

• Implement above demonstration activities andprovide practical training to local communities andrepresentatives from provincial communityorganisations (NGOs) and relevant landmanagement agencies;

• Develop productive land management institutionallinkages between eastern Indonesia and northernAustralia through: (1) the undertaking of aboveresearch and training needs as collaborativeprograms; and (2) encouraging student exchanges;

• Assist with the establishment of a network of landmanagement information dissemination centres ineastern Indonesia and northern Australia;

• Develop a phased implementation plan to educateland administrators and policy makers of the practicalneeds for undertaking sound fire managementpractice in fire-prone savanna environments.

144

Session 2: Describing Targeted Research RequirementsEach participant was allocated to one of five groups,keeping in mind respective areas of expertise. Eachgroup was asked to consider and identify specificresearch, training, information and institutionalrequirements in a broad eastern Indonesian–northernAustralian context.

Discussion group participantsGroup 1: Satellite imagery/GIS/remotely sensedmonitoring: Dr Richard Smith, Dr Jennifer Robinson, Mr PaulRyan, Dr Jeff Fox, Prof Nigel Tapper, Mr CameronYates, Mr Grant Allan, Mr Marthen Mallo, Dr NeilStronach, Dr Richard Noske

Group 2: Cultural factors and land tenure,education/ community awareness: Dr Peter Jacklyn, Prof Chris Healey, Mr Peter Cooke,Dr Andrew McWilliam, Ir Esthon Foenay, Mr GlenWightman, Dr Hendrik Ataupah, Ms Kath Thorburn,Dr Sukwong, Ms Carol Palmer, Prof Sulthoni

Group 3: Fire in land management—ecologicalresearch requirements: Dr Jeremy Russell-Smith, Dr Dick Williams, Mr PiersBarrow, Pak Maraden Purba, Dr Phil Cheney, MrChris Done, Pak Tri Agung, Mr Bambang Hartono,Mr Robert Muller

Group 4: Fire in land management—agriculture andforestry land: Mr Rodd Dyer, Mr Colin Wilson, Prof HaryonoSemangun, Prof Sulthoni, Mr Michael Riwu Kaho, DrTony Grice, Mr Wayan Mudita, Dr AbdullahBamualim, Dr Subandi, Dr Frank McKinnell, IrSlamet Gadas

Group 5: Sectoral links and collaborations: Prof Greg Hill, Mrs Heather Crompton, Dr SiliwoloeDjoeroemana, Prof Saragih, Mr Saharjo, Mr BambangHartono, Prof Riyanto, Mr Suhardi Suryadi,Mr Juspan, Dr Mark Johnston, Mr Putra Suardika,Ir Retno Nuningsih, Ms Michele Bowe

Given the outcomes from the first breakawaysession, each group was asked to consider the use ofcase study areas and practical demonstrationapplications. As well, each group was asked to focuson what sort of research actions would enhanceactivities (e.g. agricultural, agroforestry, conservation,economic, training) already being undertaken in thesecase study areas.

Group 1: Satellite imagery/GIS/remotely sensedmonitoring This group discussed the use of remote sensing andGIS in two areas, the Trans-Fly region of Irian Jaya

and the island of Sumba. The practical experience ofNeil Stronach and Jeff Fox in these two respectiveregions served to guide this focus. Other membersincluded those with relevant Australian technicalexperience. The group considered that, in order tounderstand the distribution of fire in the landscape, itwas necessary also to describe and understand theroles and distributions of climate, population, land-use, vegetation/habitat.

Most discussion focused on the role, andappropriate application of, satellite sensors, in relationto specific research questions, for example, • Where are fires burning on any given day?• What is the seasonal distribution and extent of

burning over an annual cycle in eastern Indonesia? • What is the seasonal distribution and extent of

burning over an annual cycle in northern Australia? • How can such information be put to use (e.g. for

monitoring trends, as bases for economic andenvironmental assessments, as practical landmanagement tools)?To be useful the program would require

participation with, and assessment of, localcommunities.

The research issues defined by this group are listedin Table 1. To interpret the table, consider the firstquestion ‘When?’. It needs to be understood that thesatellite sensors (AVHRR, MODIS) are of relativelycoarse resolution (especially AVHRR, with individualpixels of maximum 1.1 × 1.1 km2) or as-yet-unlaunched (MODIS). Other satellite sensors listed inthe table comprise the high resolution (Spot,LANDSAT—maximum 30 x 30 m pixels) and radar(sees through clouds; wide range of pixel sizes). Thedifferent sensors have various, useful researchapplications, as abbreviated in the table. As the tableprogresses from ‘when’ through to ‘why’, satellitesensing assumes diminishing importance and othersocio-geographic data are of increasing importance. Afirst issue, however, is the current location of relevantreceiving stations (Indonesia, Philippines, Australia),and what historical satellite data are available.

Group 2: Cultural factors and land tenure,education/community awareness The conclusions of this group are summarised below.

Key issues• Establishing cooperative linkages—between land

councils, NGOs, regional Indonesian universities,and government agencies;

• Communication of traditional/local fire manage-ment knowledge:– between older and younger generations

(connected with loss of knowledge)– between local community and school curricula– between local community and general public

145

• Making fire management knowledge locallyrelevant;

• Better communication, coordination and courtesybetween neighbours;

• Communication of scientific/technical firemanagement knowledge;

• Improving general access to fire managementknowledge:– traditional/local– scientific/technical through general dissemination

and through daily access to tools and images;• Land tenure issues;• Raising awareness about fire issues:

– among general public and among policy makers;• Development of appropriate training and extension

in fire management;• Lobbying for a review of legislation and policy.

Projects • Scoping parameters to identify further more

specific projects and study area (e.g. Sumba andFlores and Timor);

• Comparison between NT and Indonesia of specificissues relating to fire management, range ofstakeholder attitudes/regulations;

• An examination of the practical relevance ofcustomary agricultural and forestry practices formanagement of fire regimes;

• Improving cattle pastures (Sumba) through

146

appropriate fire regimes;• Fire histories in study areas (Wasur, Wanggameti)

at different scales including local ethno-historicalassessment of satellite imagery;

• Can PRA in Indonesia be adapted for Australianneeds?

• Networking agency for information sharing,including ongoing process for communication.

Group 3: Fire in land management/ecologicalresearch This group considered areas within eastern Indonesiawhere fire management in savanna landscapes was asignificant resource management issue. Membersdiscussed four locations, describing a range of land uses.The group found that developing such a project, whichultimately might involve demonstration activities beingundertaken at a number of discrete locations, wouldrequire considerable linkage, coordination, training,education, and extension.

Case Study 1—Sumba: Forest Research Centre nearWaingapuSiteArea of 500 ha, close to town, mostly grassland, flatterrain, limestone/rocky soil, mainly weeds, 8 monthsdry, ~1000 mm rainfall annually, used by farmers onan annual lease

PrioritiesHow to improve cattle productionUse of fire to manage grassland areas for weedmanagement, grazing

Research issuesImproving land and pastoral management

Demonstration Location convenient for government officials/community/training Land currently managed by forestry department anduniversity

Case Study 2—Sumba: WanggametiSiteCatchment of 70 000 ha, of which 40 000 ha areNational ParkHilly area with rainforest

ProblemsFire in surrounding grassland burns up through hills

ResearchFire management practices for conserving forest areasand for maintaining grassland grazing. Requires goodunderstanding of traditional and contemporary burningpractices.

DemonstrationFire management practices as part of broader study ofsustainable land use.Based on existing studies being undertaken through

Table 1. Research issues potentially addressed by theapplication of different remote sensors, particularly inrelation to the Trans-Fly region of Irian Jaya and the islandof Sumba in Nusa Tenggara Timur.

Question Sensor Research issue

When AVHRR Diurnal, Persistence

MODIS Cloud, Orbits, Proportionof fires perceived

Where AVHRR, Rate of regrowth

MODIS, radar, Curing

How much AVHRR, Rate of regrowthMODIS, radar, Size of fires

–small scale landsat,radar,spot;

–cloud radar

What GIS, radar Vegetation types, landuse, soili.e. existing data

Who GIS Existing data—tenure,admin., settlement

Consequences Change analysis, what changes?

Why GIS + field Capture local knowledge interviews + capture imperatives.

Ethno-meteorology

the Nusa Tenggara Community DevelopmentConsortium.

Case Study 3—Trans-Fly region of Irian JayaProblemsFeral animals (Rusa deer);Habitat loss, (e.g. invasion of grasslands by Melaleucaand weeds, e.g. Mimosa, with effects on native species,e.g. migratory birds);Outside burns passing through/Trans-Fly regionalissues;Poor official awareness of fire management issues, orhow fire might be integrated with other managementobjectives

ResearchHabitat management;Economic resource management issues (e.g. ecotourism,meat supply);Asteromyrtus oil production;Sustainable harvesting of Rusa deer;Fire ecology and management of floodplain systems;Understanding traditional and contemporary burningpractices;Education/training;Putting training into practice.

Case Study 4—West Timor: Mutis MountainsProblemsAppropriate vs current fire regimes in Eucalyptusurophylla forests;Regeneration requirements of E. urophylla stands

ResearchProductivity and availability of pastures for grazingoutside vs inside the E. urophylla forests;Understanding traditional usage of fires andassessment of changes in fire regimes;Awareness/education/training of wise, sustainableburning practices;Fire regimes and regeneration of Eucalyptus urophylla(an economically important regional timber).

Group 4: Fire in land management—agricultureand forestry land This group found that the general aim of fire use inagriculture and forestry should be to improve theproductivity of the land. However, a number of problemsrelated to the use of fire were identified, including:• Conflicting demands between land uses in relation

to fire and a lack of knowledge of long-term fireimpacts;

• Side effects of changed land-use practices,especially shorter rotations in slash and burningpractices leading to declining land condition;

• Effects of fire on tree establishment and growth,and the need for fire protection in forestry;

• Forage quality deficits and the impacts of fire onland condition from a livestock perspective.

Issues related to land useLivestock grazingForage quality/production;Timing and intensity of fire;Fire resistant species/fences (protection from uncon-trolled fire);Grazing trials.

Slash and burn cropsErosion prevention / control;Soil fertility;Crop combinations;Resistant tree species;Systems of permanent agriculture;Timing / intensity of fires;Use of fallows;Management of Chromolaena and other weeds;Economic means of making fire breaks;Effects on biodiversity;Inventory of traditional wisdom.

PlantationsAgroforestry crop combinations, and alternatives;Fire-resistant species;Timing of burn-offs, and other control measures;Design and integration of agroforestry plantations incommunity agriculture.

ForestryEarly burning to prevent wild fire;Cash crops within forests;Biodiversity issues;Generally, as for plantations above.

Approach: a comprehensive land use managementstudy in West Timor Both research and training are required in thefollowing areas:– Land resource mapping—GIS;– Current use of fire—who, when and why;– Impacts of current use of fire on current land use

components;– Recommendations and management guidelines.

There is an obvious link between northern Australiaand eastern Indonesia in terms of similarity ofenvironmental and fire issues which suggestspotential for useful exchange of information.

Group 5: Sectoral linkages and collaborationThis group first listed the relevant organisations whichmight be involved in the undertaking of a fire and landmanagement project:

Indonesia• Ministries of Agriculture, Forestry, Education and

Environment.;• Perun Perhutani;• ICRAF / CIFOR / BIOTROP;• Provincial Planning Board;• Information Centre on Semi-Arid Agriculture in

Kupang;

147

– future agricultural uses/patterns, i.e.

herbicides use,transmigration, Effects on fire management.horticulture

• Spatial and temporal distribution of fire– seasonal patterns of burning and distribution– development of drought indices e.g. ENSO – El Nino;– curing patterns of grasses, and related applications

of satellite imagery e.g. NDVI from NOAA –AVHRR, radar;

– distribution of information to policy makers, landmanagers;

– use and application/training.

Final Workshop DiscussionIn the final open forum of the workshop participantswere invited to summarise the background, aims, anda framework for a future project(s) that might bedeveloped as a result of issues identified during theworkshop.

General points of agreement– The role of fire is not well understood by all

savanna community stakeholders, and there is arequirement for research and training in appropriatefire management practices in eastern Indonesia andnorthern Australia;

– A knowledge base concerning fire exists in theregion but is not well documented;

– Appropriate fire management practices can befound for different land-use purposes;

– Differentiation should be made betweencontrolled/prescribed burning and uncontrolled/wildfire. This distinction already exists in theIndonesian language—‘api’ refers to burning as atool and ‘kebakaran’ refers to uncontrolled/unintended fire;

– That expertise and suitable networks exist toachieve the research and training required for theabove undertaking;

– That ‘action research’ should be the guidingprinciple for undertaking demonstration or casestudy activities. Action research was defined as aprocess of framing questions, and then undertakingresearch, in association with clients. The processthus incorporates existing knowledge and allstakeholder interests/views.

– That the current workshop has established arationale for developing an ongoing project;

– That the summary of resolutions should becirculated to a group of key workshop participantsfor further development.

148

• FRIK (Forestry Research Institute, Kupang);• Nusa Tenggara Community Development

Consortium;• NGOs from World Vision, CARE, WWF, World

Food Program.

Australia• Tropical Savannas Cooperative Research Centre

(University, CSIRO, Government);• Northern Territory Government Department of

Asian Relations, Trade and Industry (DARTI);• Cattle industry;• Aboriginal Land Councils;• Key Centre for Sustainable Wildlife Management

at NTU, especially in relation to Rusa deer in IrianJaya;

• NGOs;• Additional researchers, especially from southern

Australia.

Linking mechanisms• Information Centre Kupang—linked to Ministry for

Education and Ministry for Forestry or Agriculture;• Tropical Savannas CRC—via Northern Territory

University.

Project linkagesNeed for establishing a working group withrepresentatives from a range of established agencies.

ProblemsPoverty/standard of living (note also that one in fourpeople in the NorthernTerritory live in poverty,especially Aborigines). Improved fire management, aspart of developing better land management systems,would assist with alleviating chronic environmental,social, and economic problems

Solutions‘Fire Management Package’- create one or more fire management models- research and test in various case studies

Thematic emphasisA two-way means of communication needed,especially in these three areas:

ForestsLivestock northern Australian–eastern Indonesia linkage.Weeds

Information collation and review—coordinatedthrough the Information Centre on Semi-aridAgriculture (Kupang), and Tropical SavannasCooperative Research Centre (Darwin)• Fire and ecological consequences– land rehabilitation;– different land uses/situations;– develop principles/research topics/overview (needed

for use by policy makers, government agencies,educational institutions and community groups);

}

}

Purpose/background for developing acollaborative fire management programThere was consensus amongst workshop participantsfor:– the development of a project involving fire

management to improve living standards andsustainable ecological practice, involving botheastern Indonesia and northern Australia;

– the need for more information to assist withsustainable land management, including different fireregimes tailored for different land-use requirements.

Possible titles of collaborative project 1. Improving regional land use in relation to fire2. Constructive use of fire in sustainable land

management, agriculture and forestry development3. Role of controlled (i.e. prescribed)/uncontrolled fire …

Possible locations for undertaking future casestudies For eastern Indonesia, the site selection was similar tothat already established by the Nusa TengarraCommunity Development Consortium, plus theTrans-Fly region of Irian Jaya (Indonesia).Stakeholder interests were also listed for nominatednorthern Australian sites.

Eastern IndonesiaTimor Mutis

Kupang

Sumba WanggametiWaingapuPaberiwai

Flores RiungKomodo Komodo Island

(World Heritage area)

Irian Jaya Wasur

Northern AustraliaVictoria River District pastoral; conservation

Western Arnhem Land indigenous; conservation

Kimberley pastoral; indigenous;conservation

Cape York pastoral; indigenous;conservation

Burdekin (Qld) pastoral

Ongoing developmentA core group was nominated for further developmentof the project:Pak Tri Agung WWF Indonesia / NGO

Consortium, KupangDr Hendrik Ataupah UNDANA University,

Kupang

Dr Abdullah Bamualim Balai PengkajianTeknologi Pertanian(BPTP) (Research Institutefor Farming Technology)

Ms Michele Bowe WWF Oceania, DarwinDr Siliwoloe Djoeroemana Sekolah Tinggi Ilmu

Ekonomi, Kristen WiraWacana, (Christian Schoolof Economics)

Professor Greg Hill NTU and TropicalSavannas CRC, Darwin

Dr Jeremy Russell-Smith BFC and TropicalSavannas CRC, Darwin

Professor Bungaran Saragih Bogor AgriculturalUniversity, Bogor

Mr Putra Suardika Nusa TenggaraCommunity Development Consortium

Post-Workshop Discussion with Indonesianand Australian Project Representatives

Group: Pak Tri Agung, Dr Hendrik Ataupah, DrAbdullah Bamualim, Dr Siliwoloe Djoeroemana, MrJulian Gorman, Prof Greg Hill, Dr Peter Jacklyn, PakMichael Riwu Kaho, Dr Bronwyn Myers, Dr JeremyRussell-Smith, Prof Bungaran Sarigih

The above group met on the day after the workshopto discuss possible arrangements for developing afurther collaboration in relation to Nusa Tenggara in theareas of fire research, management, training, andinstitutional linkages. Indonesian participants indicatedthat, based on the findings and direction of theworkshop undertaken over the preceding three days,they had discussed the development of a broad projectoutline which they wished to present for consideration.

In short, it was suggested that the part of theongoing project in Nusa Tenggara should:• Focus on three demonstration study areas in eastern

Indonesia—at Mbay on Flores, at Paberiwai onSumba, and Fatulew (near Kupang) on Timor;

• At each site, incorporate an active, participatoryresearch approach, involving local community,members of the Nusa Tenggara CommunityDevelopment Consortium, Government agencies,researchers and academics;

• Be integrated across all three sites under a commonset of land management objectives, with thecoordinating and information dissemination rolesundertaken through the Information Centre forSemi-arid Agriculture, based in Kupang;

• Develop close applied research, management,training and institutional linkages between easternIndonesian and northern Australian partners.

149

150

In order to progress the development of the project,the following actions were agreed upon:• Indonesian colleagues will prepare detailed

background profiles of each of the three study sites.This work would be undertaken under the auspicesof the Information Centre for Semi-aridAgriculture.

• This material will be presented and discussedfurther at a workshop in Kupang in early 2000, withpossible visits to each of the three study areas priorto the workshop. A formal presentation will bemade subsequently, in Bogor, to the IndonesianGovernment.

• Australian colleagues will commence preparationof the next-round submission to ACIAR, seekingongoing funding as a Large Project.

• Uncommitted Darwin workshop funds will be usedto facilitate development of project componentsoutlined above.Participants agreed that the workshop had provided

a sound platform for ongoing cooperation andcollaboration.

Lokakarya pengelolaan kebakaran dibagi dalam duasesi terpisah dan peserta diberi kesempatan memilihsatu dari lima kelompok kerja. Sesi 1: diskusi umummengenai persoalan-persoalan pengelolaankebakaran regional baik di lahan savana di Indonesiabagian timur maupun di Australia bagian utara. Sesi 2: Berdasarkan pada Sesi 1, diskusi sesi iniberfokus pada persoalan regional spesifik yangteridentifikasi: pengembangan pendekatan-pendekatan pengawasan kebakaran secara luas;persoalan budaya, hunian & kepemilikan, kesadaranmasyarakat dan pendidikan; persoalan penelitianekologi, persoalan penelitian pertanian dankehutanan; dan hubungan-hubungan sektoral dankerjasama.

Temuan-temuan utama dari dua sesi ini dapatdirangkum sebagai berikut:

Sesi 1: Pendefinisian PersoalanSesi pertama bertujuan untuk mendefinisikan secaraluas persoalan utama pengelolaan kebakaran diIndonesia bagian timur dan Australia bagian utara,dan mengidentifikasikan kebutuhan-kebutuhanpenelitian dan pelatihan yang dapat membantumenjembatani kesenjangan serta membantumengembangkan praktek dan kebijakan pengelolaankebakaran yang telah dikembangkan.

Persoalan

Pendokumentasian luasan dan pola pembakaransekarang iniTelah banyak kemajuan yang dicapai dalam tahun-tahun terakhir ini dalam pendokumentasian luasan,musim dan pola pembakaran (contoh: frekuensikebakaran di berbagai daerah dan sektor pemanfaatanlahan) di seluruh Australia bagian utara, tetapiinformasi yang serupa belum tersedia untukIndonesia bagian timur. Data peta kebakaranmemiliki berbagai manfaat, termasuk menentukanlokasi terjadinya kebakaran harian, danmenginformasikan serta mengawasi perkembanganpraktek dan kebijakan pengelolaan kebakaranregional yang tepat.

Pendokumentasian praktek-praktek pengelolaankebakaran Satu lembaga informasi yang besar telah ada(sekalipun belum lengkap) dan telahmendokumentasikan praktek dan pendekatanpengelolaan kebakaran masa lalu dan sekarang atasberbagai sektor pemanfaatan lahandi Australia bagian utara, namun hanya ada sedikit

data yang tersedia untuk Indonesia bagian timur.Pemahaman akan praktek-praktek pengelolaankebakaran tradisional pada latar masyarakat setempatmerupakan hal yang penting, sebagai contoh dapatdigunakan untuk menginformasikan perkembangandan penerapan sistem-sistem pengelolaan kebakarandengan tepat di kedua daerah tersebut dalam kontekspemanfaatan lahan kontemporer.

Penelitian terapanBeberapa informasi telah tersedia untuk umum(khususnya untuk Australia bagian utara) mengenaipenggunaan api dalam pengelolaan pengembalaan(contoh menggiring ternak untuk makan rumput didaerah-daerah yang baru saja terbakar; pelestarianrumput hijau tahunan; penetapan satu jenis padangrumput yang lebih baik; pengendalian tumbuhanyang berkayu), tetapi hanya ada sedikit data yangtersedia untuk penerapan dalam industri primer danpengelolaan sumber daya alam lainnya (contohpengelolaan rumput liar seperti Chromolaena dilahan pertanian dan agro-kehutanan; pelestarian haradan tanah dengan tingkat perputaran sistim tebas-bakar yang semakin pendek; pengelolaan daerahserapan air). Untuk melibatkan masyarakat setempatdan memotivasi mereka menerapkan pemecahan-pemecahan praktis, pendekatan interaktif melaluidemonstrasi atau studi kasus dianggap sebagai caraterbaik dalam upaya memenuhi kebutuhan-kebutuhanpenelitian pengelolaan lahan.

Hunian & kepemilikanPersoalan hunian & kepemilikan lahan baik diIndonesia bagian timur maupun Australia bagianutara merupakan hal mendasar terhadap hak (dantanggung jawab) individu dan masyarakat pada satusisi dan badan-badan pemerintahan pada sisi lainyang berupaya menerapkan pengelolaan kebakaran .Di daerah Indonesia bagian timur baik Pemerintahpusat maupun yang tidak resmi, sistem hunian &kepemilikan tradisional dapat berlaku secara bebasterhadap lahan yang sama. Persoalan hunian &kepemilikan di Australia bagian utara meliputipengakuan hak-hak hukum akhir-akhir ini terhadapperwalian penduduk asli (Aborigin dan Torres StraitIslander) untuk bergabung dalam ‘Native Title’ (HakMilik Penduduk Asli) pada banyak tanah pemerintahdan pemegang sewa lahan pengembalaan, yangmeliputi hak dan tanggung jawab untukmelaksanakan pengelolaan kebakaran tradisional.Oleh karena itu pemahaman kerangka kerja hunian &kepemilikan merupakan hal mendasar untukmengembangkan praktek dan kebijakan pengelolaankebakaran yang tepat di kedua wilayah hukumtersebut.

151

Sesi Lokakarya yang Terpisah

Kebijakan dan perundanganPengembangan kebijakan pengelolaan kebakaranyang sungguh-sungguh dan perundangan yangberkaitan diperlukan baik di Indonesia bagian timurdan Australia bagian utara. Pengembangan inikhususnya bertujuan untuk mengenali persyaratan-persyaratan tertentu untuk hidup dan mengaturlingkungan savana yang mudah terbakar. Dalamkonteks Indonesia bagian timur, kebijakan resmitentang kebakaran yang betul-betul terpisahditetapkan oleh Pemerintah pusat, sebagai tanggapanterhadap ramifikasi internasional atas pembakaranbiomassa di blok-blok hutan tropis yang luas diIndonesia barat dan Irian Jaya utara. Di Queenslanddan Australia Barat, kebijakan-kebijakan negara yangluas ditentukan berturut-turut di bagian tenggara danbarat daya, dengan kondisi cuaca-api dan musimAustralia selatan yang sangat berbeda. Karena itu,pemahaman yang lebih luas terhadap persyaratanpengelolaan kebakaran atas sistem-sistem savanabaik di Indonesia bagian timur dan Australia bagianutara akan membantu pengembangan kebijakan-kebijakan dan perundangan yang tepat mengenaipengelolaan kebakaran regional.

Pelatihan, pembinaan dan komunikasiPersoalan di atas hanya dapat dibahas melaluiprogram pelatihan dan pendidikan yang ditargetkanmulai dari masyarakat yang paling bawah sampaikepada para pembuat kebijakan. Pengalaman dikedua daerah ini, khususnya di Indonesia bagiantimur, telah membuktikan bahwa kegiatan-kegiatandemonstrasi yang praktis merupakan cara praktis danefektif untuk mencapai tujuan ini. Disamping itu,perlu adanya pengembangan ‘situs Web’ regionaldimana para pemakai akhir dapat langsungmengakses informasi mengenai serangkaian topikyang berhubungan dengan pengelolaan lahan danpersoalan terkait atau paling tidak dapatmengarahkan mereka ke arah yang benar.

Kolaborasi, kerjasama dan hubunganSebagaimana persoalan-persoalan informasi di atas,para peserta lokakarya telah benar-benar menyadarikebutuhan dan manfaat untuk meningkatkan ikatan-ikatan hubungan kelembagaan antara Indonesiabagian timur dan Australia bagian utara. Sementarabadan-badan dan lembaga-lembaga Australia bagianutara dapat menawarkan berbagai tingkatan keahlianteknis dan pendidikan yang relevan denganpengelolaan kebakaran pada sistem savana, Indonesiabagian timur menyediakan contoh kerjasama danjaringan penggunaan informasi bersama dankerjasama yang sangat baik, yang melibatkansejumlah besar LSM, lembaga dan badan pendidikan.Kedua wilayah ini memiliki banyak hal yang dapatditawarkan kepada satu sama lainnya.

Kebutuhan penelitian dan pelatihan yangteridentifikasiBerikut ini adalah daftar ringkas mengenai kebutuhanpenelitian dan pelatihan yang diidentifikasikan olehke lima kelompok dari sesi terpisah pertama:• Mengembangkan kapasitas untuk melakukan

penilaian komparatif atas luasan dan polapembakaran sekarang ini di Indonesia bagiantimur dan Australia bagian utara, denganmenggunakan pendekatan-pendekatanpendeteksian jarak jauh (RS) dan SistemInformasi Geografis (GIS);

• Penilaian seperti itu akan memerlukanpengembangan yang mengiringi pemanfaatanlahan, tumbuhan dan cakupan peta batas lahanyang tepat, sebagai dasar untuk mengembangkandatabase pengelolaan kebakaran — dalam kasusIndonesia bagian timur hal ini mungkinmelibatkan banyak pekerjaan yang dapatdilaksanakan sebagai proyek siswa dan/ataumelalui bantuan teknis;

• Melakukan penilaian terhadap praktek-praktekpengelolaan kebakaran kontemporersebagaimana yang digunakan dalam pengelolaanlahan, khususnya di Indonesia bagian timur ;

• Menggunakan informasi di atas untukmenggambarkan pengaruh-pengaruh biofisika,ekonomi, sosial, budaya dan politik dan implikasirezim kebakaran sekarang ini;

• Mengidentifikasikan pengelolaan ekonomi dansumber daya alam yang utama dari persoalanpengelolaan kebakaran, dan lokasi-lokasipotensial dimana persoalan-persoalan demikiandapat diatasi melalui kegiatan demonstrasi daninteraksi dengan masyarakat;

• Mengimplementasikan kegiatan-kegiatandemonstrasi di atas dan menyediakan pelatihanpraktis bagi masyarakat setempat dan perwakilandari organisasi-organisasi masyarakat di daerah(LSM) dan badan-badan pengelolaan lahanterkait;

• Meningkatkan hubungan kelembagaanpengelolaan lahan produktif antara Indonesiabagian timur dan Australia bagian utara melalui:(1) pelaksanaan penelitian di atas dan pemenuhankebutuhan pelatihan sebagai program bersama;dan (2) peningkatan pertukaran siswa;

• Membantu pendirian jaringan kerja pusat-pusatpenyebaran informasi pengelolaan lahan diIndonesia bagian timur dan Australia bagianutara;

• Mengembangkan rencana penerapan bertahapuntuk mendidik para pejabat pertanahan danpembuat kebijakan mengenai kebutuhan-kebutuhan praktis untuk melaksanakan praktekpengelolaan kebakaran yang baik di lingkungansavana yang mudah terbakar.

152

Sesi 2: Menjelaskan Persyaratan-persyaratanPenelitian yang Ditargetkan

Setiap peserta dialokasikan pada satu dari limakelompok dengan mempertimbangkan bidangkeahlian mereka masing-masing. Setiap kelompokdiminta untuk mempertimbangkan danmengindentifikasikan persyaratan-persyaratanspesifik untuk penelitian, pelatihan, informasi danlembaga dalam konteks Indonesia bagian timur –Australia bagian utara yang luas.

Peserta Kelompok Diskusi

Kelompok 1: Pengawasan Imaginer Satelit/GIS/pendeteksian jarak jauh

Dr Richard Smith, Dr Jennifer Robinson, Dr PaulRyan, Dr Jeff Fox, Prof. Nigel Tapper, Cameron Yates, Grant Allan,Marthen Mallo, Dr. Neil Stronach, Dr. Richard Noske

Kelompok 2: Faktor-faktor budaya danhunian & kepemilikan lahan,pendidikan/ kesadaranmasyarakat

Dr. Peter Jacklyn, Prof. Chris Healey, Mr. PeterCooke, Dr Andrew McWilliam, Ir. Esthon Foenay, Mr Glen Wightman, Dr HendrikAtaupah, Ms Kath Thorburn, Dr Sukwong, Ms Carol Palmer, Prof. Sulthoni

Kelompok 3: Kebakaran dalam pengelolaanlahan—persyaratan-persyaratan penelitian ekologi

Dr Jeremy Russell-Smith, Dr Dick Williams, Mr.Piers Barrow, Pak Maraden Purba, Dr Phil Cheney, Mr Chris Done,Pak Tri Agung, Mr Bambang Hartono, Mr Robert Muller

Kelompok 4: Kebakaran dalam pengelolaanlahan—lahan pertanian dankehutanan

Mr. Rodd Dyer, Mr. Colin Wilson, Prof. HaryonoSemangun, Prof. Sulthoni, Mr.Michael Riwu Kaho, Dr TonyGrice, Mr Wayan Mudita, Dr Abdullah Bamualim, Dr.Subandi, Dr Frank McKinnell, Ir. Slamet Gadas

Kelompok 5: Hubungan antar sektoral dankerjasama:

Prof. Greg Hill, Mrs Heather Crompton, Dr SiliwoloeDjoeroemana,Prof. Saragih, Mr. Saharjo, Mr. Bambang Hartono,Prof. Riyanto,

Mr. Suhardi Suryadi, Mr. Juspan, Dr Mark Johnston,Mr. Putra Suardika, Ir. Retno Nuningsih, Ms Michele Bowe

Melihat hasil-hasil dari sesi terpisah pertama, setiapkelompok diminta untuk mempertimbangkanmanfaat bidang-bidang studi kasus dan penerapan-penerapan praktek demonstrasi. Setiap kelompokjuga diminta untuk menitik-beratkan perhatian padajenis tindakan penelitian apa yang dapatmeningkatkan kegiatan-kegiatan yang telahdilakukan dalam bidang-bidang studi kasus ini(contoh: pertanian, agro-kehutanan, pelestarian,ekonomi, pelatihan).

Kelompok 1: Pengawasan Imaginer Satelit/GIS/pendeteksian jarak jauh Kelompok ini membahas manfaat pendeteksian jarakjauh dan GIS di dua daerah, daerah antar lintas udaradi Irian Jaya dan pulau Sumba. Pengalaman praktisdari Neil Stronach dan Jeff Fox di dua daerah yangberbeda ini membantu mengarahkan fokus ini. Paraanggota lain termasuk mereka yang memilikipengalaman teknis Australia yang relevan. Kelompokini mempertimbangkan bahwa untuk memahamipenyebaran kebakaran pada satu bentang lahan, kitaperlu juga menggambarkan dan memahami peran danpenyebaran iklim, populasi, pemanfaatan lahan,tumbuhan/habitat.

Sebagian besar diskusi bertitik-tolak pada peran danpenerapan sensor-sensor satelit yang pantassehubungan dengan pertanyaan-pertanyaan penelitianyang spesifik, sebagai contoh; Dimana terjadi kebakaran pada hari-hari tertentu?, Seperti apa distribusi musim dan luasan pembakaranpada satu putaran tahunan di Indonesia bagian timur?Seperti apa distribusi musim dan luasan pembakaranpada satu putaran tahunan di Australia bagian utara? Bagaimana memanfaatkan informasi tersebut (contohuntuk memonitor kecenderungan, yang merupakandasar penilaian ekonomi dan lingkungan, sebagaialat-alat praktis pengelolaan lahan)?

Agar bermanfaat program tersebut memerlukanpartisipasi dari masyarakat setempat dan penilaianterhadap mereka.

Persoalan-persoalan penelitian yang didefinisikanoleh kelompok ini ada di dalam Tabel 1. Untukmengartikan tabel tersebut, pertimbangkanpertanyaan pertama ‘Kapan?’. Kita perlu memahamibahwa sensor-sensor satelit (AVHRR, MODIS)merupakan resolusi kasar relatif (khususnyaAVHRR, dengan piksel individu maksimal 1.1 x 1.1km2) atau seperti yang belum diluncurkan (MODIS).Sensor-sensor satelit lain yang ada dalam tabel terdiriatas yang beresolusi tinggi (Spot, LANDSAT—

153

piksel maksimal 30 x 30 m) dan radar (melihatmelalui awan; ukuran piksel yang berbeda-beda).Masing-masing sensor memiliki berbagai manfaatdalam terapan-terapan penelitian, sebagaimana yangdisingkat dalam tabel. Karena tabel mulai dari‘Kapan’ sampai ‘Mengapa’, peran pendeteksiansatelit menjadi berkurang, namun peran data sosio-geografi lain meningkat. Akan tetapi, persoalanpertama adalah lokasi stasiun penerima terkait yangada sekarang ini (Indonesia, Filipina, Australia), dandata historis satelit apa saja yang telah tersedia.

Kelompok 2: Faktor-faktor budaya dan hunian &kepemilikan lahan, pendidikan/kesadaranmasyarakat

Rangkuman kesimpulan hasil diskusi kelompokadalah sebagai berikut:

Persoalan utama• Menetapkan hubungan kerjasama — antara

badan-badan pertanahan, LSM, universitas didaerah-daerah di Indonesia dan badan-badanpemerintahan;

• Komunikasi pengetahuan pengelolaan kebakaransetempat/tradisional:– antara generasi yang lebih tua dan lebih muda

(dihubungkan dengan pengetahuan yanghilang);

– antara masyarakat setempat dan kurikulumsekolah;

– antara masyarakat setempat dan masyarakatumum;

• Membuat agar pengetahuan pengelolaankebakaran relevan bagi masyarakat setempat;

• Komunikasi, koordinasi dan sopan-santun yanglebih baik antar tetangga;

• Komunikasi pengetahuan ilmiah/teknispengelolaan kebakaran ;

• Meningkatkan akses secara umum terhadappengetahuan pengelolaan kebakaran :– tradisional/setempat;– ilmiah/teknis melalui penyebaran umum dan

melalui akses sehari-hari terhadap peralatandan gambar;

• Persoalan hunian & kepemilikan lahan;• Meningkatkan kesadaran mengenai persoalan

kebakaran:– antara masyarakat umum dan para pembuat

keputusan;• Pengembangan pelatihan dan perluasan yang

tepat dalam bidang pengelolaan kebakaran ;• Melobi untuk meninjau perundangan dan

kebijakan.

Proyek-proyek• Cakupan parameter untuk mengidentifikasikan

secara lebih jauh proyek-proyek dan bidang-bidang studi yang lebih spesifik (contoh: Sumba,Flores dan Timor);

• Perbandingan antara NT dan Indonesia mengenaipersoalan tertentu yang berhubungan dengan

154

Pertanyaan Sensor Persoalan Penelitian

Kapan AVHRR Harian, TetapMODIS Awan, Orbit, Proporsi kebakaran yang diterima.

Dimana AVHRR, Tingkat pertumbuhan ulangMODIS, Radar, Pemulihan

Berapa besar AVHRR, Tingkat pertumbuhan ulangMODIS, Radar, Luasan kebakaran

– skala kecil → landsat, radar, spot;

– awan → radarApa GIS, Radar Jenis tumbuhan, pemanfaatan lahan, tanah yakni:

data yang tersediaSiapa GIS Data yang tersedia—hunian & kepemilikan,

administrasi, penyelesaianKonsekuensi Merubah analisis, perubahan apa?Mengapa GIS + Wawancara Mengumpulkan pengetahuan setempat +

lapangan mengumpulkan hal-hal penting. Etno-meteorologi

Tabel 1. Persoalan penelitian yang mungkin dapat diatasi dengan penggunaan berbagai pendeteksi jarak jauh,khususnya yang berhubungan dengan daerah antar lintas udara di Irian Jaya dan pulau Sumba di Nusa TenggaraTimur.

pengelolaan kebakaran, perbedaan sikap/peraturan-peraturan dari para pemilik lahan

• Pemeriksaan kebiasan praktek-praktek tradisionaldalam bidang pertanian dan kehutanan yangterkait untuk rezim pengelolaan kebakaran;

• Mengembangkan pengembalaan ternak (Sumba)dengan rezim pembakaran yang tepat;

• Sejarah-sejarah kebakaran di daerah-daerah studi(Wasur, Wanggameti) pada skala-skala yangberbeda termasuk penilaian etno-historissetempat dari imajiner satelit;

• Dapatkah Penilaian terhadap Keikut-sertaanDaerah Pedalaman (diketahui dengan istilahPRA) di Indonesia diadaptasikan untukkeperluan-keperluan di Australia?

• Jaringan kerja lembaga untuk berbagi informasitermasuk komunikasi yang berkelanjutan.

Kelompok 3: Kebakaran dalam pengelolaanlahan/penelitian ekologi Kelompok ini mempertimbangkan daerah-daerahyang ada di dalam kawasan Indonesia bagian timurdimana pengelolaan kebakaran di lahan savanamerupakan persoalan pengelolaan sumber daya yangpenting. Para anggota membahas empat lokasi, yangmenggambarkan rentang pemanfaatan lahan.Kelompok ini mendapatkan bahwa pengembanganproyek seperti ini, yang mungkin melibatkankegiatan-kegiatan demonstrasi diadakan di sejumlahdaerah yang terpisah, akan memerlukan hubungan,koordinasi, pelatihan, pendidikan dan pembinaanyang sungguh-sungguh.

Studi Kasus 1—Sumba: Pusat PenelitianKehutanan di dekat Waingapu

LokasiLahan seluas 500 ha, dekat kota, sebagian besarpadang rumput, berlahan datar, berbatu kapur/tanahberbatu, terutama rumput liar, 8 bulan kering, curahhujan tahunan ~1000mm, dimanfaatkan olehpeternak dengan sistem sewa tahunan.

PrioritasBagaimana meningkatkan produksi ternakMenggunakan api untuk mengelola lahan padangrumput dalam upaya pengelolaan rumput liar,pengembalaan

Persoalan penelitianMeningkatkan pengelolaan lahan dan pengembalaan

DemonstrasiLokasi yang cocok untuk aparatpemerintah/masyarakat/pelatihan

Lahan yang sekarang ini dikelola oleh departemenkehutanan dan universitas

Studi Kasus 2—Sumba: Wanggameti

LokasiDaerah serapan air seluas 70 000 ha, yang mana 40 000 ha merupakan Taman NasionalLahan berbukit dengan hutan hujan.Masalah-masalahKebakaran di padang rumput sekitar merambatsampai ke daerah yang berbukitPenelitianPraktek-praktek pengelolaan kebakaran untukmelestarikan lahan-lahan hutan dan untukmempertahankan pengembalaan di padang rumputmemerlukan pemahaman yang baik terhadap praktekpembakaran tradisional dan kontemporer.DemonstrasiPraktek-praktek pengelolaan kebakaran merupakanbagian studi yang lebih luas dari pemanfaatan lahanyang berkelanjutan.Berdasarkan studi yang telah dilakukan olehKonsorsium Pengembangan Masyarakat NusaTenggara

Studi Kasus 3—Daerah antar lintas udara di IrianJaya

MasalahBinatang liar (Rusa);Kehilangan habitat, (contoh penyerbuan padangrumput oleh Melaleuca dan rumput liar, contohMimosa, yang berpengaruh terhadap spesies asli,contoh: migrasi burung);Persoalan tentang kebakaran di luar yang melewatidaerah antar lintas udara;Kesadaran aparat yang rendah mengenai persoalanpengelolaan kebakaran , atau bagaimanamenggabungkan tujuan-tujuan pengelolaankebakaran dengan tujuan-tujuan pengelolaan lain.

PenelitianPengelolaan habitat;Persoalan pengelolaan sumber daya ekonomi (contohwisata alam, penyediaan daging);Produksi minyak Asteromyrtus;Panen Rusa yang berkelanjutan;Ekologi dan pengelolaan kebakaran terhadap sistem-sistem dataran di sekitar sungai yang terkena banjirpada saat airnya meluap (floodplain);Memahami praktek-praktek pembakaran tradisionaldan kontemporer;Pendidikan/pelatihan;Melaksanakan pelatihan

155

Studi Kasus 4— Timor Barat: Pegunungan Mutis

MasalahRezim pembakaran yang tepat vs. rezim pembakaransekarang ini di hutan-hutan Eucalyptus urophylla ;Persyaratan-persyaratan re-generasi hutan E. urophylla

PenelitianProduktivitas dan ketersediaan padang rumput di luarvs. di dalam hutan-hutan E. urophylla ;Memahami penggunaan api secara tradisional danpenilaian terhadap perubahan rezim api;Kesadaran/pendidikan/pelatihan untuk praktek-praktek pembakaran yang bijak dan berkelanjutan;Rezim api dan re-generasi Eucalyptus urophylla(kayu regional yang penting secara ekonomis)

Kelompok 4: Kebakaran dalam pengelolaanLahan —Lahan Pertanian dan KehutananKelompok ini menemukan bahwa tujuan umumpenggunaan api dalam pertanian dan kehutananseharusnya dapat meningkatkan produktivitas lahan.Akan tetapi, sejumlah masalah yang berhubungandengan penggunaan api telah dikenali, hal inimencakup:Konflik tuntutan antara pemanfaatan lahansehubungan dengan kebakaran dan kurangnyapengetahuan mengenai dampak-dampak penggunaanapi dalam jangka panjang;Pengaruh-pengaruh sampingan dari praktek-praktekperubahan pemanfaatan lahan, khususnya rotasipraktek tebas-bakar yang lebih pendek yangmenyebabkan penurunan kondisi lahan;Pengaruh-pengaruh kebakaran pada pertumbuhandan perkembangan pohon, dan perlunyaperlindungan terhadap kebakaran di kawasan hutan;Penurunan kualitas makanan ternak dan pengaruh-pengaruh kebakaran pada kondisi lahan dariperspektif peternakan

Persoalan yang berhubungan dengan pemanfaatanlahan

Pengembalaan ternakKualitas/produksi makanan ternak;Waktu dan intensitas kebakaran;Spesies /pagar tahan api (perlindungan darikebakaran yang tidak terkendali);Percobaan-percobaan pengembalaan

Hasil panen dari praktek tebas-bakarPencegahan/pengendalian erosi;Kesuburan tanah;Kombinasi hasil panen;Spesies pohon yang tahan;

Sistem pertanian yang permanen;Waktu/intensitas kebakaran;Pemanfaatan lahan-lahan kosong;Pengelolaan Chromolaena dan rumput liar lainnya;Cara yang ekonomis untuk membuat sekat api;Pengaruh-pengaruh terhadap aneka hayati;Inventaris kebijaksanaan tradisional;

PerkebunanKombinasi hasil panen agro-kehutanan dan alternatif;Spesies yang tahan api;Waktu pembakaran dan tindakan pengendalianlainnya;Perancangan dan integrasi perkebunan agro-kehutanan dalam masyarakat agraris

KehutananPembakaran dini untuk mencegah kebakaran yangtidak terkendali;Hasil bumi tunai di dalam hutan;Persoalan aneka hayati;Biasanya, mengenai perkebunan-perkebunan di atas

Pendekatan: studi pengelolaan pemanfaatan lahanyang komprehensif di Timor Barat Baik penelitian maupun pelatihan diperlukan dalambidang-bidang berikut ini:– Pemetaan sumber daya lahan—GIS;– Penggunaan api sekarang ini—siapa, kapan dan

mengapa;– Pengaruh pemakaian api pada komponen

pemanfaatan lahan sekarang ini– Rekomendasi dan tuntunan pengelolaan

Ada hubungan yang jelas antara Australia bagianutara dan Indonesia bagian timur dalam halkesamaan persoalan lingkungan dan kebakaran yangmenyarankan adanya potensi untuk bertukarinformasi yang bermanfaat.

Kelompok 5: Hubungan sektoral dan kerjasamaKelompok ini mulai dengan membuat daftarorganisasi terkait yang mungkin terlibat dalampelaksanaan proyek pengelolaan kebakaran danlahan:

Indonesia• Departemen Pertanian, Kehutanan, Pendidikan

dan Lingkungan;• Perum Perhutani;• ICRAF / CIFOR / BIOTROP;• Badan Perencanaan Daerah (Bappeda);• Pusat Informasi Pertanian Lahan Semi-Kering di

Kupang;

156

• FRIK (Badan Penelitian Kehutanan, Kupang);• Konsorsium Pengembangan Masyarakat Nusa

Tenggara;• LSM dari World Vision, CARE, WWF, World

Food Program;

Australia• Pusat Penelitian Kerjasama Savana Tropis

(Universitas, CSIRO, Pemerintah);• Departemen Hubungan Asia, Perdagangan dan

Industri Pemerintah Negara Bagian NorthernTerritory (DARTI);

• Industri peternakan;• Badan Pertahanan Aborigin;• Pusat Utama untuk pengelolaan Margasatwa

yang Berkelanjutan di NTU, khususnya yangberhubungan dengan Rusa di Irian Jaya;

• LSM-LSM;• Para peneliti tambahan, khususnya dari Australia

bagian selatan;

Mekanisme hubungan• Pusat Informasi Kupang—berhubungan dengan

Departemen Pendidikan dan DepartemenKehutanan atau Pertanian;

• Savana Tropis CRC—melalui Northern TerritoryUniversity;

Hubungan ProyekPerlu untuk mendirikan satu kelompok kerja denganperwakilan dari berbagai badan yang telah didirikan:

Masalah-masalahKemiskinan/standar hidup (perlu dicatat bahwa satudari empat orang di NorthernTerritory hidup dalamkemiskinan, khususnya orang Aborigin). Pengelolaankebakaran yang lebih baik, sebagai upaya daripengembangan sistem pengelolaan lahan yang lebihbaik, akan membantu mengurangi masalah-masalahlingkungan, sosial dan ekonomi yang kronis.

Pemecahan‘Paket Pengelolaan kebakaran ’– menciptakan model-model pengelolaan

kebakaran – meneliti dan menguji model ini dalam berbagai

studi kasus

Penekanan temaKomunikasi dua arah diperlukan, khususnya dalamtiga bidang berikut ini:

HutanTernak hubungan Australia bagian utaraRumput liar – Indonesia bagian timur

Pengumpulan dan peninjauan informasi—dikoordinasikan melalui Pusat Informasi PertanianLahan Semi-Kering (Kupang), dan Pusat PenelitianKerjasama Savana Tropis (Darwin)

• Konsekuensi kebakaran dan ekologi– rehabilitasi lahan;– pemanfaatan lahan/situasi yang berbeda;– mengembangkan prinsip/topik penelitian/tinjauan

(diperlukan oleh para pembuat keputusan,badan–badan pemerintah, lembaga pendidikan dankelompok masyarakat);

– pemanfaatan/pola pertanian di masa mendatang,yakni:

herbisida,transmigrasi, Pengaruh-pengaruh kebakaran terhadap pengelolaan hortikultura,

• Penyebaran tempat dan waktu kebakaran– pola-pola pembakaran musiman dan penyebaran – pengembangan indeks musim kering, contoh:

ENSO - El Nino;– Pola-pola pemulihan rumput dan penerapan yang

berhubungan dengan imaginer satelit, contoh:NDVI dari NOAA – AVHRR, Radar;

– penyebaran informasi kepada para pembuatkebijakan dan manajer-manajer lahan;

– pemanfaatan dan penerapan/pelatihan

Diskusi Akhir Lokakarya

Dalam akhir forum terbuka lokakarya, para pesertadiajak bersama-sama meringkas latar belakang,tujuan dan kerangka kerja untuk proyek-proyek dimasa mendatang yang mungkin dikembangkansebagai hasil dari persoalan-persoalan yangteridentifikasi selama lokakarya.

Poin-poin umum yang disepakati:– Tidak semua masyarakat pemilik lahan savana

memahami dengan baik peran api, dan karena ituperlu diadakan penelitian dan pelatihan mengenaipraktek-praktek pengelolaan kebakaran yangtepat di Indonesia bagian timur dan Australiautara;

– Peristiwa kebakaran hanya merupakan satupengetahuan saja di satu daerah, tetapi tidakdidokumentasikan dengan baik

– Praktek-praktek pengelolaan kebakaran dapatditemukan untuk tujuan pemanfaatan lahan yangberbeda-beda;

– Pembedaan harus dibuat antara pembakaran yangterkendali /ditentukan dan kebakaran yang tidakterkendali/menyebar dengan cepat. Perbedaantersebut telah ada dalam bahasa Indonesia – ‘api’

157

}

}

yang mengacu ke pembakaran sebagai alat dan‘kebakaran’ mengacu kepada api yang tidakterkendali/yang tidak disengaja;

– Bahwa keahlian dan jaringan kerja yang cocoktelah tersedia untuk mencapai penelitian danpelatihan yang diperlukan untuk usaha-usaha diatas;

– Bahwa ‘penelitian-tindakan’ seharusnyamerupakan prinsip bimbingan untukmelaksanakan demonstrasi atau kegiatan-kegiatan studi kasus. Penelitian-tindakandidefinisikan sebagai satu proses pembentukpertanyaan dan kemudian melaksanakanpenelitian, bekerja sama dengan klien. Dengandemikian, proses ini menggabungkanpengetahuan yang sudah ada dengan semuaminat/pandangan pemilik lahan;

– Bahwa lokakarya sekarang ini telah menetapkansebuah rasionalisasi untuk mengembangkansebuah proyek yang berkelanjutan;

– Bahwa ringkasan resolusi harus diedarkankepada satu kelompok yang terdiri dari pesertautama lokakarya untuk pengembangan lebihlanjut.

Tujuan/latarbelakang untuk pengembangankerjasama program pengelolaan kebakaran Diantara para peserta lokakarya telah adakesepakatan untuk :– mengembangkan sebuah proyek yang

melibatkan pengelolaan kebakaran untukmeningkatkan standar hidup dan praktekekologi berkelanjutan yang melibatkanIndonesia bagian timur maupun Australiabagian utara;

– kebutuhan akan informasi yang lebih banyakuntuk membantu pengelolaan lahan yangberkelanjutan yang mencakup berbagai rezimkebakaran yang dibuat untuk memenuhipersyaratan pemanfaatan lahan yang berbeda.

Judul-judul proyek kerjasama yang mungkindipertimbangkan:Meningkatkan pemanfaatan lahan regional danhubungannya dengan kebakaranPenggunaan api yang konstruktif dalam pengelolaanlahan, pengembangan pertanian dan kehutanan yangberkelanjutan.Peran api yang terkendali (yakni: yang ditentukan)/tak terkendali …

Lokasi-lokasi yang mungkin untuk melaksanakanstudi kasus di masa mendatangUntuk Indonesia bagian timur, pemilihan lokasinyamirip dengan yang telah dikembangkan oleh

Konsorsium Pengembangan Masyarakat NusaTenggara, ditambah dengan daerah antar lintas udaradi Irian Jaya (Indonesia). Minat-minat pemilik lahanjuga dimuat dalam daftar lokasi-lokasi yang ditunjukdi Australia bagian utara.

Indonesia bagian timurTimor Mutis

KupangSumba Wanggameti

WaingapuPaberiwai

Flores RiungKomodo Pulau Komodo (daerah Warisan

Dunia)Irian Jaya Wasur

Australia bagian utaraVictoria River District pengembalaan;

pelestarianWestern Arnhem Land penduduk asli;

pelestarian Kimberley pengembalaan;

penduduk asli;pelestarian

Cape York pengembalaan;penduduk asli;pelestarian

Burdekin (Qld) pengembalaan

Pengembangan yang berkelanjutan Satu kelompok inti telah ditunjuk untukpengembangan proyek lebih lanjut:Pak Tri Agung WWF Indonesia /

Konsorsium LSM,Kupang

Dr Hendrik Ataupah UniversitasUNDANA, Kupang

Dr Abdullah Bamualim Balai Pengkajian Teknologi Pertanian (BPTP)

Ms Michele Bowe WWF Oceania, Darwin

Dr Siliwoloe Djoeroemana Sekolah Tinggi IlmuEkonomi, KristenWira Wacana

Professor Greg Hill NTU dan Savana Tropis CRC, Darwin

Dr Jeremy Russell-Smith BFC dan Savana Tropis CRC, Darwin

Professor Bungaran Saragih Institut PertanianBogor, Bogor

Mr. Putra Suardika KonsorsiumPengembanganMasyarakat NusaTenggara

158

Diskusi Pasca Lokakarya antar PerwakilanProyek dari Indonesia dan Australia

Kelompok : Pak Tri Agung, Dr Hendrik Ataupah, DrAbdullah Bamualim, Dr Siliwoloe Djoeroemana, ,Mr Julian Gorman, Prof. Greg Hill, Dr Peter Jacklyn,Pak Michael Riwu Kaho, Dr Bronwyn Myers, DrJeremy Russell-Smith, Prof. Bungaran Sarigih

Kelompok di atas mengadakan pertemuan seharisetelah lokakarya untuk mendiskusikan kemungkinanpengembangan kerjasama lebih lanjut sehubungandengan Nusa Tenggara, di bidang penelitiankebakaran, pengelolaan, pelatihan dan hubunganantar-lembaga. Peserta dari Indonesia menyatakanbahwa berdasarkan pada penemuan dan tujuan darilokakarya yang dilakukan selama 3 hari sebelumnya,mereka telah mendiskusikan pengembangan garis-garis besar proyek secara luas yang ingin merekaingin sajikan untuk bahan pertimbangan.

Secara ringkas, mereka menyarankan bahwa bagiandari proyek berkelanjutan di Nusa Tenggara harus:• berfokus pada tiga daerah studi demonstrasi di

Indonesia bagian timur: —di Mbay -Flores, diPaberiwai - Sumba, dan di Fatulew (dekatKupang) - Timor;

• pada setiap lokasi, menggabungkan pendekatanpenelitian yang aktif dan bersifatmengikutsertakan yang melibatkan masyarakatsetempat, anggota Konsorsium PengembanganMasyarakat Nusa Tenggara, badan-badanpemerintahan, para peneliti dan para akademisi;

• menyatu dengan 3 lokasi penelitian tersebutdengan seperangkat tujuan pengelolaan lahanyang sama, dimana peran koordinasi danpenyebaran informasi dilaksanakan melalui PusatInformasi Pertanian Lahan Semi-Kering yangberpusat di Kupang;

• mengembangkan penelitian terapan, pengelolaan,pelatihan dan hubungan antar-lembaga yangdekat antara mitra-mitra kerja di Indonesiabagian timur dan Australia bagian utara.

Dalam rangka meningkatkan pengembangan proyek,tindakan-tindakan berikut telah disetujui:

• Rekan kerja Indonesia akan menyiapkan profillatar belakang yang terperinci dari tiga lokasistudi masing-masing. Pekerjaan ini akandilakukan di bawah pengawasan Pusat InformasiPertanian Lahan Semi-Kering.

Bahan ini akan disajikan dan didiskusikan lebihlanjut pada lokakarya di Kupang pada bulanSeptember 1999*, dengan kemungkinan mengadakankunjungan ke ketiga lokasi studi sebelum lokakaryadimulai.Penyajian formal mengenai lokakarya ini akandiberikan kepada Pemerintah Indonesia di Bogor. Rekan kerja Australia akan mulai menyiapkanpengajuan berikutnya kepada ACIAR dan mencaridana yang berkesinambungan sebagai satu ProyekBesar. Dana lepas dari lokakarya Darwin akan digunakanuntuk membiayai pengembangan komponen proyekyang digariskan di atas.Para peserta setuju bahwa lokakarya ini telahmenyediakan dasar yang mantap bagi kerjasamayang berkelanjutan.

* Catatan: Karena adanya ketidakstabilan politik diKupang, rencana kunjungan pada bulan Septembertelah dialihkan ke Bali pada bulan November 1999,Kunjungan ke lokasi studi di Wasur (Irian Jaya)direncanakan pada bulanDesember 1999.

159

160

IndonesiaPak Tri AgungWorld Wide Fund for Nature - Kupang, Nusa Tenggara Project,PO Box 1123, Pos Oebobo, Kupang 85000, NTT, IndonesiaPh: 62 380 821472 Fax: 62 380 821472triagung@kupang.wasantara.net.id

Ir Bambang HartonoHead, Balai Taman Nasional KomodoPelabuhan Bajo, Flores, NTT, IndonesiaPh: 62 385 41004 Fax: 62 385 41006

Mr Luigi Michael Riwu KahoFakultas Peternakan, UNDANA PenfuiJl.. Motoain 21, Naikoten I, Kupang, IndonesiaPh/ fax: 62 380 824 259P2SLPT@kupang.wasantara.net.id

Dr Hendrik AtaupahLecturer, Faculty of Social and Political Sciences,Nusa Cendana University (FISIP UNDANA)Jl. Adisucipto, Penfui, Kupang 85001, NTT, IndonesiaPh: 62 380 822625/825589 Fax: 62 380825627

Ir JuspanDinas PKT, Sumba, Timur, IndonesiaPh: 62 387 61333

Prof Ir RiyantoVice Rector for Academic Affairs, University MulawarmanRektorat Kampus Gunung Kelua, Jalan Kuaro Kotak Pos1068 Samarinda 75119, East Kalimantan, IndonesiaPh/ fax: 62 541 39890csf@smd.mega.net.id

Dr Abdullah BamualimBalai Pengkajian Teknologi Pertanian (BPTP) (ResearchInstitute for Farming Technology)Departemen Pertanian, Badan Penelitanian Dan EngebanganPertanian, PO Box 23, Kupang 85000, NTT, IndonesiaPh: 62 380 833766/826893 Fax: 62 380 833766

Mr Marthen MalloWorld Wide Fund for Nature - Kupang, Nusa Tenggara Project,PO Box 1123, Pos Oebobo, Kupang 85000, NTT, IndonesiaPh: 62 380 821472 Fax: 62 380 821472triagung@kupang.wasantara.net.id

Dr Bambang Hero SaharjoFaculty of Forestry, Bogor Agricultural UniversityPO Box 168 Bogor 16001, West Java, IndonesiaPh: 62 251 627750 Fax: 62 251 621244manhut@bogor.wasantara.net.id

Participants

Mr Grant AllanDr Hendrik AtaupahDr Abdullah BamualimMr Piers BarrowMs Michele BoweDr David BowmanDr Phil CheneyMr John ChildsDr Garry CookMr Peter CookeDr Siliwoloe DjoeroemanaMr Rodd DyerDr Jeff FoxProf Jim FoxIr Slamet GadasProf Johann GoldammerDr Tony GriceProf Chris HealeyProf Greg HillDr Peter JacklynMs Barbie McKaige

Dr Andrew McWilliamMr Wayan Mudita Dr Bronwyn MyersIr Retno NuningsihIr Maraden PurbaMr Noel PreeceMr Luigi Michael Riwu KahoProf Ir RiyantoDr Jeremy Russell-SmithDr Bambang Hero SaharjoProf Dr Bungaran SaragihProf Dr Haryono SemangunDr Richard SmithDr Neil StronachDr Somsak SukwongProf Dr Ir A. SulthoniProf Nigel TapperDr Tom TherikDr Dick Williams

Referees

161

Dr Siliwoloe DjoeroemanaSekolah Tinggi Ilmu Ekonomi, Kristen Wira Wacana, Christian School of EconomicsJalan Prof Dr W.Z. Yohanes 12 A, Waingapu 87113, NTT, IndonesiaPh: 62 387 62337 Fax: 62 387 62279kriswima@indo.net.id

Mr Wayan MuditaCentre for Research of the Environment & NaturalResources, Nusa Cendana University (UNDANA)Jl. Adisucipto, Penfui, Kupang 85001, NTT, IndonesiaPh/ fax: 62 380 825627puskom@kupang.wasantara.net.id

Prof Dr Bungaran SaragihDirector of Centre for Development Studies,Bogor Agricultural UniversityJalan Pajajaran (Kampus IPB Gunung Gede) Bogor 16153, IndonesiaPh: 62 251 328105 Fax: 62 251 344113psp-ipb@bogor.wasantara.net.id

Ir Esthon L. FoenayProvincial Development Planning Board for East NusaTenggara (BAPPEDA),Ketua Bappeda Tk. I NTT, Jl Polisi Militer No. 2, Kupang, NTT, IndonesiaPh: 62 380 831712/833462 Fax: 62 380 821310/833664

Ir Retno NuningsihDirector of Agricultural PolytechnicJl.. Adisucipto, Penfui, Kupang 85001, NTT, IndonesiaPh: 62 380 831827 Fax: 62 380 826865jotic@kupang.wasantara.net.id

Prof Dr Haryono SemangunFaculty of Agriculture, Gadjah Mada UniversitySekip, Yogyakarta 55281, IndonesiaPh: 62 274 563 062 Fax: 62 274 563 062swcu@server.indo.net.id

Ir Slamet GadasForestry Research Institute, KupangJl. Untung Suropati, No. 7 (Belakong) Kupang, NTT, IndonesiaPh: 62 380 823 357 Fax: 62 380 833 102bpkk@kupang.wasantara.net.id

Ir Maraden PurbaKepala Taman Nasional WasurJl. Raya Mandala, Spadem No.2, Kotak Pos 109, Merauke,Irian Jaya 99611, IndonesiaPh/fax: 62 971 22495wwfwasur@jayapura.wasantara.net.id

Mr Putra SuardikaNTCDC, World Neighbors Indonesia, PO Box 71, Ubud, Bali 80571, IndonesiaPh: 62 361 975 707 Fax: 62 361 976 487wnisea@denpasar.wasantara.net.id

Dr SubandiAgency for Agricultural Research and Development,Research Institute for Maize and other CerealsPO Box 1173, Ujung Pandang, IndonesiaPh: 62 411 371529/371016 Fax: 62 411 371961balitjas@upandang.wasantara.net.id

Prof Dr Ir A. SulthoniDepartment of Forestry Protection, Fakultas Kehutanan,Univ Gadjah Mada (UGM) Bulaksumur, Yogyakarta 55281, IndonesiaPh: 62 274 561947 Fax: 62 274 901420

Mr Suhardi SuryadiLP3ES NTBJl.. Bung Hatta 11/6, Mataram, NTT, IndonesiaPh: 62 370 27386lp3es@mataram.wasantara.net.id

Dr Tom Therik University Kristen Artha Wacana, Kampus UkawCentre for Regional Studies, Jl.. Adisucipto, Desapa,PO Box 147, Kupang, NTT, IndonesiaPh/fax: 62 380 832584 ttherik@kupang.wasantara.net.id

AustraliaMr Steve BalchMember for JingiliPO Box 42988, Casuarina, NT 0811, AustraliaPh: 61 8 8927 5433 Fax: 61 8 8999 5333

Mrs Heather CromptonACIARGPO Box 1571, Canberra ACT 2601, AustraliaPh: 61 2 6217 0548 Fax: 61 2 6217 0501crompton@aciar.gov.au

Prof Greg HillNorthern Territory UniversityDarwin NT 0909 AustraliaPh: 61 8 8946 6550 Fax: 61 8 8946 6712g_hill@banks.ntu.edu.au

Mr Piers BarrowProject Officer, Natural Resource Management,Kakadu National ParkPO Box 71, Jabiru NT 0886, AustraliaPh: 61 8 8938 1100 Fax: 61 8 8938 1115Piers.Barrow@ea.gov.au

Mr Chris DoneConservation and Land ManagementP.O. Box 942 Kununurra W.A. 6743 AustraliaPh: 61 8 9168 4200 Fax: 61 8 9168 2179chrisd@calm.wa.gov.au

Dr Peter JacklynNorthern Territory UniversityDarwin NT 0909 AustraliaPh: 61 8 8946 6285 Fax: 61 8 8946 7107peter.jacklyn@ntu.edu.au

162

Ms Michele BoweWorld Wide Fund for NatureTropical Wetlands of Oceania Program, GPO Box 1268, Darwin NT 0801 AustraliaPh: 61 8 8941 7554 Fax: 61 8 8941 6494mbowe@wwf.org.au

Mr Rodd DyerDepartment of Primary Industry and FisheriesPO Box 1346, Katherine 0851, NT, AustraliaPh: 61 8 8973 9750 Fax: 61 8 8973 9777rodd.dyer@dpif.nt.gov.au

Mrs Janet LawrenceACIARGPO Box 1571, Canberra ACT 2601 AustraliaPh: 61 2 6217 0537 Fax: 61 2 6217 0501lawrence@aciar.gov.au

Dr David BowmanCentre for Indigenous Natural and Cultural ResourceManagement, Northern Territory University, Darwin NT 0909 AustraliaPh: 61 8 8946 7762 Fax: 61 8 8946 7755d_bowman@banks.ntu.edu.au

Mr Andrew EdwardsBushfires Council NTPO Box 37346, Winnellie NT 0813, AustraliaPh: 61 8 8944 8464 Fax. 61 8 8944 8455andrew.edwards@pwcnt.nt.gov.au

Dr Frank McKinnellConservation and Land Management – PerthLocked Bag 104, Bentley Delivery Centre, Perth WA 6948, AustraliaPh: 61 8 9442 0319 Fax: 61 8 9386 6399frankm@calm.wa.gov.au

Dr Phil CheneyCSIRO Forestry and Forest ProductsPO Box E4008, Kingston ACT 2604 AustraliaPh: 61 2 6281 3879 Fax: 61 2 6281 8348Phil.Cheney@ffp.csiro.au

Dr Tony GriceCSIRO Tropical AgricultureDavies Laboratory, PO Box Aitkenvale, Townsville, Qld 4815 AustraliaPh: 61 77 538 543 Fax: 61 77 538 600Tony.grice@tag.csiro.au

Dr Andrew McWilliamSacred Sites Authority, NARU12 Sanders StJingili, NT 0810, AustraliaPh: 61 8 89853446

Mr Peter CookeExecutive Officer, Caring for Country Unit, Northern Lands CouncilPO Box 42921, Casuarina NT 0811 AustraliaPh: 61 8 8920 51 09 Fax: 61 8 8945 2633peter.cooke@nlc.org.au

Prof Chris HealeySchool of Southern Asian and Australian Studies, NTUDarwin NT 0909 AustraliaPh: 61 8 8946 6725 Fax: 61 8 8946 6955c_healey@banks.ntu.edu.au

Mr Robert MullerProject Officer, Kakadu National ParkPO Box 71, Jabiru NT 0886, AustraliaPh: 61 8 8938 1100 Fax: 61 8 8938 1115Rob.Muller@ea.gov.au

Dr Richard NoskeSenior Lecturer in Ecology, School of Biological andEnvironmental SciencesScience Faculty, Northern Territory University, Darwin NT 0909 AustraliaPh: 61 8 89 466 359 Fax: 61 8 89 466 847richard.noske@darwin.ntu.edu.au

Prof Nigel TapperMonash University, School of Geography andEnvironmental ScienceClayton Campus, Wellington Road, Clayton, Victoria 3168, AustraliaPh: 61 3 9905 2931/2929 Fax: 61 3 9905 2948Nigel.Tapper@arts.monash.edu.au

Mr Frank ZichNT Herbarium, PWCNTPO Box 496, Palmerston NT 0831, AustraliaPh: 61 8 8999 4516 Fax: 61 8 8999 4527f.zich@ais.net.au

Ms Carol PalmerWorldwide Fund for Nature. Tropical Wetlands of OceaniaProgram, PO Box 1268, Darwin NT 0801, AustraliaPh: 61 8 8941 7554 Fax: 61 8 8941 6494c.palmer@wwf.org.au

Ms Kathryn Thorburn3/5 Bremer Street, Ludmilla, NT 0820, AustraliaPh: 61 8 8948 0541kethorburn@hotmail.com

Mr Noel PreeceEcoz ConsultancyGPO Box 381, Darwin NT 0801, AustraliaPh: 61 8 8981 1100 Fax: 61 8 8981 1102ecoz@ozemail.com.au

Mr Glen WightmanParks and Wildlife Commission of the Northern TerritoryPO Box 496, Palmerston NT 0831, AustraliaPh: 61 8 8999 4513 Fax: 61 8 8999 4527glen.wightman@pwcnt.nt.gov.au

Dr Jennifer RobinsonMurdoch UniversityEnv. Science, Murdoch University, Perth 6150, AustraliaPh: 61 8 9360 6102robinson@essun1.murdoch.edu.au

Other countriesDr Jeff FoxDirector, Program on Environment, East-West Center1601 East-West Road, Honolulu, Hawaii 96848Ph: 1 808 944 7248 Fax: 1 808 944 7298foxj@ewc.hawaii.edu

Dr Mark JohnstonNational Agricultural Research InstituteP.O. Box 4415 Lae 411, Morobe Province, PNGPh: 675 472 1751 Fax: 675 472 2242nari@datec.com.pg

Dr Neil StronachFota Wildlife Park LtdCarrigtwohill, Co. Cork, IrelandPh:353 21 633007/812678 Fax: 353 21 812744fota@indigo.ie

Dr Somsak SukwongRegional Community Forestry Training CentreKasetsart University, PO Box 111, Bangkok 10903, ThailandPh: 662 942 5700 Fax: 662 561 4880ftcsss@nontri.ku.ac.th

163

Dr Dick WilliamsCSIRO Wildlife and EcologyPMB 44, Winnellie, DarwinNT AustraliaPh: 61 8 89448426 Fax: 61 8 8944 8444dick.williams@terc.csiro.au

Dr Jeremy Russell-SmithBushfires Council NTPO Box 37346, Winnellie NT 0821, AustraliaPh: 61 8 8984 4000 Fax: 61 8 8947 2263jeremy.russell_smith@nt.gov.au

Mr Colin WilsonParks and Wildlife Commission of the Northern TerritoryPO Box 496, Palmerston NT 0831, AustraliaPh: 61 8 8999 4698 Fax: 61 8 8999 4793colin.wilson@PWCNT.nt.gov.au

Dr Richard SmithWestern Australian Department of Lands Administration,Satellite Remote Sensing ServicesPO Box 471, Wembley WA 6014 AustraliaPh: 61 8 9383 9342 Fax: 61 8 9383 7142richard_smith@notes.dola.wa.gov.au

Mr Cameron YatesBushfires Council NTPO Box 37346, Winnellie NT 0821, AustraliaPh: 61 8 8984 4000 Fax: 61 8 8947 2263