dachang, l. 2003. rehabilitation of degraded forests to improve

Rehabilitation ofDegraded Forests to Improve Livelihoods of Poor Farmers in South China

EditorLiu Dachang

Editorial BoardZhu Zhaohua, Cai Mantang, Liu Dachang and John Turnbull

CAF

Degradation of forests and forest lands is a problem in many parts of theworld and is particularly serious in south China. Chinese forest policyreforms in recent years have enabled rural households to generate incomefrom forests, to own the trees they have planted, and have offered newopportunities to manage forests sustainably. Rehabilitation of degradedforests and forest lands is one of the possible pathways to improvelivelihoods of poor farmers and others in the rural communities.

This report documents the results of four case studies in south Chinain which farmers, local officials and researchers analysed the problems ofdegraded forests and forest lands, and formulated options for theirsolution. Opportunities to improve forest management and people'slivelihoods are dependent on overcoming a range of biophysical,socioeconomic and political constraints. Action research was used toimplement and test some of the options identified. The experience andanalysis should be of value for researchers, resource managers andgovernment officials in China and elsewhere to address poverty andenvironmental concerns through a multidisciplinary, participatory andholistic approach.

ISBN 979-8764-98-6

Rehabilitation of Degraded Forests to Im

prove Livelihoods of Poor Farmers in South China

cover2.qxd 6/30/03 11:29 AM

Rehabilitation ofDegraded Foreststo Improve Livelihoodsof Poor Farmersin South China

Editor

Liu Dachang

Editorial Board

Zhu ZhaohuaCai MantangLiu DachangJohn Turnbull

2003 by Center for International Forestry ResearchAll rights reserved.Printed by SMT Grafika Desa Putera, Indonesia

Cover photos by Christian Cossalter

ISBN 979-8764-98-6

Liu, D. (ed). 2003. Rehabilitation of Degraded Forests to Improve Livelihoodsof Poor Farmers in South China.Center for International Forestry Research, Bogor, Indonesia. 97p.

Published byCenter for International Forestry ResearchBogor, IndonesiaP.O. Box 6596 JKPWB, Jakarta 10065, IndonesiaTel: + 62 (251) 622 622; Fax: + 62 (251) 622 100E-mail: [email protected] site: http://www.cifor.cgiar.org

Table of Contents

Acknowledgements vForeword vi

Chapter 1. Rehabilitation of Degraded Forests to Improvethe Livelihoods of Poor Farmers: A Synthesis ofFour Case Studies in South China 1

Introduction 1Research Methods 4Location of Research Sites 7Biophysical and Socioeconomic Settings of South China 8Extent of Forest Degradation 13Causes of Forest Degradation in South China 15Forest Rehabilitation in South China 20References 26

Chapter 2. Reclaiming Degraded Forest Lands inthe Dry, Hot Climate of Yuanmou County, Yunnan

Introduction 29Research Methods 30Background Information on Yuanmou County and Loafan Village 31Land Degradation Assessment 34Strategies and Interventions for Land Reclamation 36Conclusions 41References 43

Chapter 3. Degradation and Rehabilitation of EvergreenBroadleaf Forest in Cangwu County, Guangxi 45

Introduction 45Research Methods 46Background Information on Shanxin Village, Cangwu County 47Extent and Causes of Degradation of Evergreen Broadleaf Forest 50Disincentives to Managing Evergreen Broadleaf Forest for Commercial Timberand Other Wood Products 54Strategies for Rehabilitating Degraded Evergreen Broadleaf Forest 55Conclusions 61References 63

iv

Chapter 4. Rehabilitation of Degraded Chinese Fir Plantations andEvergreen Broadleaf Forest in Huitong County, Hunan 65

Introduction 65Research Methods 66Background Information on Dongxi and Xiangjian Villages, Huitong County 67Chinese Fir Plantations 70Evergreen Broadleaf Forests 75Disincentives to Forest Rehabilitation 77Conclusions 78References 80

Chapter 5. Rehabilitation and Sustainable Management ofDegraded Forests in Gaohong, Lin’an County, Zhejiang 81

Introduction 81Research Methods 81Background Information on Lin’an County 83Overview of Forest Degradation in the Villages 86Degraded Timber Forests 86Degraded Bamboo Plantations and Forests 90Degraded Tea Plantations 92Development and Extension of Forest Technology 93Conclusions 95References 97

v

Acknowledgements

This publication is the result of a project initiatedby the Chinese Academy of Forestry (CAF) andfinalised with inputs by representatives of Centerfor International Forestry Research (CIFOR), andCanada’s International Development ResearchCentre (IDRC) at a meeting in Singapore in 1995.The academic insights and management skills ofProfessor Zhu Zhaohua and Mr Cai Mantang werecritical in formulating the project and coordinat-ing the implementation of the case studies in thefield. Their major contribution is gratefully ac-knowledged. They were supported by ProfessorBai Jiayu, who replaced Professor Zhu Zhaohuaas the overall Project Leader towards the end ofthe research period, Mr Jiang Chunquan and MsLi Yuzhen. Drs Neil Byron, John Turnbull andLiu Dachang of CIFOR visited field sites to pro-vide academic inputs into the research process.

This was a multidisciplinary, participatoryproject and it involved many people to a greateror lesser extent at the case study sites. Sincerethanks go to farmers of the villages of Laofan,Xiaocun, Moke and Bingyue of Yuanmou County,Yunnan; Shanxin Village in Cangwu, Guangxi;the villages of Xiangjian and Dongxi in Huitong,Hunan; and the villages of Chenjiakan, Hongqiaoand Shangfeng, Lin’an County, Zhejiang, for theirparticipation in and collaboration with this re-search project. Their active involvement and col-laboration made it effective to identify causes offorest degradation, and constraints to and strate-gies for forest rehabilitation. Their traditionalknowledge made a great contribution to the de-velopment of rehabilitation strategies.

The untiring efforts of the research teams ateach site in diagnosis, design and delivery, andthe preparation of reports is acknowledged withthanks. In Yunnan, Professor Lai Yongqi led ateam from the CAF Research Institute of ResourceInsects; in Guangxi, Professor Zheng Haishui led

a team from the CAF Research Institute of Tropi-cal Forestry; in Hunan, Professor Xu Guozhenled a team from the Central South Forestry Uni-versity; and in Zhejiang, Mr Zhou Guomo led ateam from Zhejiang Forestry College and Lin’anCounty Forestry Bureau.

We also thank the following individualsand institutions for their participation in, and valu-able support in many ways for our research: inYuanmou County: staff of Bureaus of Forestry,Meteorology and Statistics, and the governmentsof Zuolin Township, Nengyi Township andLaocheng Township; in. Cangwu: Mr ShaoJingxian, Mr Huang Yicai, Mr Xie Shangao, MrLiang Shusheng, Mr Lu Shi’an and XieHongsheng of Cangwu County Forestry Bureau;and staff of Guangxi Academy of Forestry,Cangwu County Government, Cangwu CountyCommittee of Science and Technology, CangwuAgriculture Bureau; and staff of Wangfu Town-ship Government, Wangfu Forest Station andWangfu Agriculture Station of Cangwu County;in Huitong: staff of Huaihua Prefecture ForestryBureau, Huitong County Government, HuitongCounty Forestry Bureau, and Ma’an TownshipGovernment of Huitong County; and in Lin’anCounty: staff of Lin’an County Forestry Bureau,Gaohong Township Government, and GaohongForest Station.

Dr Brian Belcher provided a critical review.Mr Christian Cossalter and Dr John Turnbullmade a valuable contribution to the final editingand lay out. Dr John Graham provided continu-ous encouragement to those preparing this publi-cation.

Finally, sincere thanks to IDRC and CIFORfor their financial support for the research, andthe preparation and publication of the researchreport.

vi

Foreword

Degradation of forests and forest lands is aproblem in many parts of the world and isparticularly serious in south China where 70 percent of the population lives in rural areas. Manynatural disasters, such as catastrophic flooding inthe Yangtze River basin, have been attributed todeforestation and there is real poverty in areasthat have severe soil erosion and degraded forests.Forest policy reforms in the last 20 years haveenabled rural households to generate income fromforests, to own the trees they have planted, andhave offered new opportunities to manage forestssustainably. Rehabilitation of degraded forests andforest lands is one of the possible pathways toimprovement of livelihoods of poor farmers andothers in the rural communities.

In China, where individual farmers ownmuch of the forest, management of the forest isjust one component of the farming system andactivities must be integrated with otheragricultural tasks. Opportunities to improvemanagement and livelihoods are dependent onovercoming biophysical, socioeconomic andpolitical constraints. There has been increasingrecognition among foresters and other naturalresource managers that to address the problemsof rehabilitating degraded forests and landsrequires a holistic approach. A multidisciplinaryeffort involving all stakeholders, in particular thelandholders, is required to analyse problems andformulate options for their solution.

Between 1990 and 1994 the ChineseAcademy of Forestry (CAF), with financialsupport from the International DevelopmentResearch Centre (IDRC) of Canada, designed andimplemented a Farm Forestry Program that tooka multidisciplinary and integrated approach tosolving on-farm forestry problems. The resultsof this work were reported in a book - IntegratedResearch in Farm Forestry (ed. Cai Mantang and

Hu Shaofang, Chinese Science and TechnologyPress, Beijing, 1995). Building on thisexperience, a new project was initiated in 1995to develop strategies to assist in the rehabilitationof the degraded forests and forest lands insouthern China, to improve local livelihoods andto provide inputs into policy deliberations. Amultidisciplinary, participatory approach withaction-orientated research was designed andcarried out at selected sites in four provinces. Thenew project was managed by the ChineseAcademy of Forestry and implemented inpartnership with the Center for InternationalForestry Research (CIFOR), CAF ResearchInstitute of Resource Insects, CAF ResearchInstitute of Tropical Forestry; Central SouthForestry University, Zhejiang Forestry College,Lin’an County Forestry Bureau with cooperationof local and regional institutions andcommunities. IDRC again provided financialsupport.

The analysis of the CAF/CIFOR/IDRCproject reported here should enable a betterunderstanding of the constraints andopportunities for improving the livelihoods ofpoor rural communities and rehabilitatingdegraded forests and forest lands. Some of theparticipants have already contributed todeliberations at a symposium entitled “Policyreform and forestry in China: lessons for Chinaand the world” held at Dujiangyan, Sichuan inJune 2001. I hope that this report will provideinspiration for researchers, resource managersand government officials, both in China andelsewhere, to address poverty and environmentalconcerns through a multidisciplinary,participatory and holistic approach.

David KaimowitzDirector General, CIFOR.

Chapter One

IntroductionDeforestation in the tropics is high with an annualloss of 14.2 million ha converted to other landuses (FAO 2001). Deforestation and forestdegradation are closely linked and are sometimesreferred to as ‘forest decline’ (Contreras-Hermosilla 2000) which can be taken to includelosses of forest productivity in terms of wood andnon-wood products and environmental services.It directly threatens the livelihoods of millionsof forest-dependent people.

The underlying causes of forest declineinclude: market failures, mistaken policyinterventions, governance weaknesses, andbroader socioeconomic and political causes suchas population growth and distribution ofeconomic and political power (Kaimowitz andAngelsen 1998, Contreras-Hermosilla 2000).Clearly forest decline is a complex phenomenonwith many socioeconomic, political and culturalcauses. Pressures to exploit forests, especially

Rehabilitation of Degraded Forests to Improvethe Livelihoods of Poor Farmers: A Synthesis ofFour Case Studies in South ChinaCai Mantang1, Liu Dachang2 and J.W. Turnbull3

natural forests, have not stopped and severalfactors suggest forests will continue to declineand have less capacity to provide both timberand non-timber forest products. It follows thatmore efforts will have to be made to rehabilitateand sustainably manage degraded natural forestsand develop productive plantations if thepredicted increases in consumption of forestproducts are correct.

Degradation of forests and lands is a commonphenomenon all over the world and is a veryserious problem in China. Much of China’sextensive area to the south of Yangtze River,referred to as ‘south China’ in this account, hasa subtropical climate in which the typicalvegetation is evergreen broadleaf forest. Duelargely to human actions, nearly all primaryevergreen broadleaf forest has disappeared andonly a few areas exist, usually in isolatedmountainous locations. Secondary forests andman-made plantations have replaced the original

Authors are listed in alphabetical order.1 Chinese Academy of Forestry, Beijing. Currently at Beijing Development Institute, Peking University, Beijing, P.R. China.2 Southwest Forestry University, Kunming P.R. China and Center for International Forestry Research, Bogor, Indonesia.Currently at The Mekong Institute, Khon Kaen, Thailand.3 Center for International Forestry Research, Bogor, Indonesia. Currently at Forestry and Forest Products, CSIRO, Canberra,Australia.

isi Christian 6/30/03, 9:25 PM1

2 Rehabilitation of Degraded Forests to Improve the Livelihoods of Poor Farmers

evergreen broadleaf forest but most of theseforests are degraded to some extent. There hasbeen a significant loss of biodiversity, with areduced number of species and a standingbiomass much lower than in primary forest. Insome cases, the vegetative cover of forest landhas been reduced to sparse shrubs or even bareland.

While there has been an overall improvementin China’s living standards, the gap between thesupply and demand for forest products has beenwidening partly due to low productivity ofdegraded forests. Numerous natural disastershave been attributed to deforestation and the lossof environmental functions provided by forests(Yang 2001). At the local level, the lowerproductivity of degraded forests and forest landshas a direct impact on rural communities whichare often partially dependent of forests for theirlivelihoods.

In the early 1990s, about 70% of populationin south China lived in rural areas (ChinaStatistical Bureau 1993). These people rely onforests for livelihood support in many ways.They depend on forests for energy (fuelwood),for timber used on their farms, and for timberand non-wood forest products (NWFPs) to sellto generate cash income. As a consequence offorest degradation, they are now able to collectless variety and quantity of products from forestsand so generate less income than formerly. Thismeans a decline in their livelihoods unless theycan generate income in other ways.

Forest degradation has on-site and off-siteecological and environmental effects, includingdeteriorating agro-ecological conditions,increased flooding, lower water quality andsiltation of dams. The deterioration of agro-ecological conditions results in decline inagricultural productivity, which in turn hasconsiderable impact on the welfare of farmersand their families. Moreover, if the local supplyof timber and other forest products is curtailed,people will either suffer from the shortage orhave to rely more on expensive imports.

Some of the principles set out in the RioDeclaration of the United Nations Conference

on Environment and Development in 1992 toguide states and peoples in global environmentand development matters relate to forestry, e.g.for environmental protection to be an integralpart of development. Agenda 21 provided anaction plan and there was an appeal for countriesto formulate and implement national forestryprogrammes. The link between integratedforestry development in mountainous areas,where the majority of the 65 million poverty-stricken people in China reside, and the processof poverty alleviation is recognized (Jiang1997). A forestry action plan has been preparedfor China’s Agenda 21 which includescomprehensive forestry development andpoverty alleviation in mountainous areas as aspecific goal (Zhao 1997).

Effective implementation of such a planrequires a good understanding of the extent andcauses of forest degradation and the actionsrequired to rehabilitate and manage degradedforests sustainably. There has been research oncauses and prevention of forest degradation insouth China since the late 1960s (Institute ofForestry and Pedology 1980, Research Team ofIntensive Cultivation of Chinese Fir 1992, Chen1992, Xu 1992). Zhu (1991) examined humanimpacts on changes in biodiversity and soilfertility in secondary broadleaf forest. Researchhas also been carried out to identify tree speciesand develop tree-planting technologies for thoseareas with harsh ecological conditions (Yu 1997,Yang et al. 1999). However, experience showsthat forest degradation largely is a result ofhuman intervention and so understanding itscauses and rehabilitating it needs more than abiological perspective.

Political change and social transition arefeatures of modern China and numerous changesin government policies have impacted onforestry. Several changes of policies related toland tenure and management of non-state forestshave sometimes resulted in severe deforestation(Liu 2001). After 1950 there was a situation ofcollective ownership and compulsory low pricewood delivery system under the totallygovernment-controlled planned economy


3Cai Mantang, Liu Dachang and J.W. Turnbull

(Table 1.1). Subsequently the governmentintroduced market-orientated reforms changingproperty rights and marketing arrangements. Inrural areas, poverty makes many farmers adverseto taking even small economic risks and theyhave become very cautious about acceptingpolicy changes (Yin 1994). In thesecircumstances, technology-oriented options andinterventions alone are not sufficient tosuccessfully rehabilitate the degraded forest andlands and improve livelihoods. Zhu (1997)describes several case studies showing the needfor the participation and cooperative interactionof leaders, scientists, extension technicians andfarmers. It is clear that factors that cause forestdegradation or inhibit rehabilitation need to beexamined from multidisciplinary perspective,including socioeconomics and governmentpolicies, and with participation of allstakeholders. Only this way will it be possibleto develop effective options and strategies.

In 1995 our research project was initiated todevelop strategies to assist in rehabilitation of

the degraded forests and forest lands, improvelocal livelihoods, and provide input to policydeliberations. It was implemented by theChinese Academy of Forestry and the Centerfor International Forestry Research, with supportfrom Canada’s International DevelopmentResearch Centre, in partnership with local andregional institutions and communities in theprovinces of Yunnan, Hunan, Guangxi andZhejiang in south China. A multidisciplinary,participatory approach was used and action-oriented research was carried out in the fieldover the period 1996-1998. The project includedseveral complementary small technologyresearch activities in different locations andthese are reported elsewhere.

This chapter synthesises the outcomes of thecase studies in Yunnan, Hunan, Guangxi andZhejiang. Research methods and biophysicaland socioeconomic settings are described. Thecauses of forest degradation are analysed and,based on the case studies, strategies necessaryfor rehabilitating degraded forest in south China

Table 1.1 Evolution of tree tenure and marketing systems in south China

Period System Ownership Timber marketing

Before 1950 Historical Bureaucrats, landlords, Free markettimber merchants andself-sustaining farmers

1950-1955 Land reform All farmers, including self- Free marketsustaining farmers andformer landlords

1956-1958 Socialist Collective ownership Quotas and pricestransformation and and production determined by the Stateagriculturalcollectivisation

1958-1980 People’s communes Collective ownership Quotas and pricesand production determined by the State

1979-1983 Household-based Family forest plots (ziliushan); Self-determined production:agriculture (Contract responsibility forest land compulsory delivery systemresponsibility) and (zerenshan); collective forestpeople’s communes

1984 to present Household-based Family forest plots, responsibility Price controls lifted, butagriculture forest land collective forest; taxes and fees increased;

shareholding system government monopoly onprocurement

1999 to present Logging ban in Same as above No timber harvesting isnatural forest allowed from collective and

private natural forests

Source: Sun (1992) and Liu (2001).



are suggested. Chapters 2 to 5 provide moredetails of the research and outcomes of each casestudy.

Research MethodsConventional scientific approaches in forestryand agroforestry have often seen technologiesdeveloped independently of the socialdimensions in which they must be adopted.Farmers on the other hand consider themanagement of woodlots, and trees grown inassociation with crops and livestock as part oftheir integrated farming system. To meet theireconomic needs they manage their land in aholistic manner aware of the interactions of theseveral biological components and thecomplexity of the social and economicenvironment that provides both constraints andopportunities. Integrated farming systemscombine agriculture, forestry, horticulture,animal husbandry, aquaculture, as well as otherbiological production, into an interconnectedsystem (Li 2001).

As the aim of this research was to seek toimprove the livelihoods of farmers byaddressing the problems of degraded forests andforest lands it was necessary to apply methodsthat could encompass the complex biological-social-economic system in which farmersmanaged their forests as part of an integratedfarming system. Rocheleau (1999) has reviewedthe many methods that have emerged in recentyears to address the complexities of conductingagroforestry research and especially the need toapply the methods of social science. Sheparticularly refers to the use of participatoryprocesses and practices and the need to bridgesocial, economic and ecological sciences.

We drew heavily on social sciencemethodologies especially in the diagnosis ofproblems and the design of options to addressthem. It would have been desirable to includesocial scientists directly in the research butinstitutional difficulties precluded this option. Thecompromise was that the forest researchers in theproject received training in social sciencemethods from the Ford Foundation in China at

the start of the project. A logical procedureconsisting of three stages: diagnosis, design anddelivery, called ‘Tri-D’ in this research, was used.

Diagnosis, Design and DeliveryDiagnosis and Design adapts farming systemapproaches and Rapid Rural Appraisal (RRA)or Participatory Rural Appraisal (PRA) toidentify problems and to design and test forestryand agroforestry options (Raintree 1987, Nair1993). This procedure has been used commonlyin community-based agroforestry researchoutside China but was a new approach in forestryin China. Participatory rural appraisal orparticipatory rapid rural appraisal has beendescribed by Chambers (1994 a, b, c) andChambers and Guijt (1995).

Diagnosis aimed at achieving a goodunderstanding of local biophysical andsocioeconomic settings, problems (extent andcauses of forest degradation and effects of forestdegradation on livelihoods), and constraints toforest rehabilitation. Design, the second stageof the research process, aimed to identifyeffective strategies or options and form policyrecommendations for rehabilitation andsustainable management of degraded forestsbased on the results of diagnosis. In the deliverystage new options that had been identified weretested for their appropriateness and effectivenessand policy recommendations were made.

The approaches and methods used within theTri-D process included:• participatory approach• multidisciplinary approach• triangulation• attention to indigenous knowledge• conventional forestry research methods, and• action research.

Participatory ApproachIn recent years there have been notable examplesof successful collaboration between scientificresearch organisations and rural communitieswhich have been assisted by a greater emphasison participation, partnership and negotiation inthe research process (Zhu 1997, Rocheleau



1999). Participation is now widely advocatedand accepted as a strategy for rural developmentand natural resource management. Participationcan be interpreted in various ways butmobilisation and empowerment of the localcommunities are two basic elements (Oakley1987).

This project emphasised the participatoryapproach with mobilisation, empowerment andinteraction as key elements of the researchprocess. Participation took place at all the stagesof diagnosis, design and delivery. We involvedfarmers and their families, researchers andextension staff, and government officials andfacilitated interaction between them to developand test options and strategies to address forestdegradation and improve livelihoods.

Farmer participation was emphasised asfarmers have good knowledge of theircircumstances, problems and constraints. Theyare also holders and beneficiaries of forests thatneed rehabilitating, and hence are able to makea very significant contribution to thedevelopment of effective strategies. During theresearch process nearly 400 households wereconsulted and 43 groups interviewed as part ofthe diagnosis. A subset of householdsparticipated in identification and testing ofrehabilitation options.

More than 50 researchers and extension staffhad some level of involvement in the research.Their participation was on the basis that theycould play an important role in developingappropriate strategies, providing informationand extension services, and facilitatingcommunication between participatinggovernment officials and farmers. This isextremely important in mountainous areaswhere farmer access to information andextension services is poor. About 130 localgovernment officials were involved to someextent. Their involvement was encouragedbecause they can facilitate implementation ofoptions identified by helping to raise investmentfunds and resolving administrative conflicts thatoften appear in resource management.

DiagnosisThe diagnosis phase involved detailed planningof the research, intensive data collection,analysis of data and specification of the potentialinterventions. PRA tools used at this stageincluded:• secondary data collection• participatory observation• household survey (semi-structured interview,

complemented by questionnaire survey)• key informant interview (semi-structured

interview)• group interview/meetings.

A large amount of secondary data wascollected from relevant government agencies atcounty level such as the Forestry Bureau,Agricultural Bureau, Statistical Bureau,Meteorological Bureau, County Archives;township governments; and village offices. Datawere collected on climate, soil, land uses, forestresources, the role of forest in local livelihoods,forest products, forest tenure and management,marketing and taxation of forest products, treeplanting policy, demography (population, labourforce, their education), wealth/income, villagesocial organisations and infrastructure.

Participatory field observation was carried outimmediately after secondary data collection. Itwas undertaken jointly by researchers, severalsenior villagers and village leaders who werefamiliar with village circumstances and history.The field visits observed and discussed land uses,types and cultivation patterns of crops, types offorest, historical changes in the forest, extent offorest degradation, and village infrastructure.

Household survey aimed to identifyproblems and constraints to forest rehabilitationat household level. Sample households wereselected through random sampling from allhouseholds in the research villages. Householdsurvey was conducted mainly in the form ofsemi-structured interviews plus a questionnairesurvey. Each household interview lasted 1.5-2hours. Topics discussed during householdinterviews usually included family population



and its age distribution, number of familymembers at labour age, major livelihood meansincluding sources of family income, familyexpenditure by activity, land uses, forest areaand volume, forest degradation, and constraintsto forest rehabilitation.

Key informant interviews were semi-structured, aiming to analyse problems at villagelevel. Topics covered population, labour forceand gender distribution of labour, education,land uses, crop calendar, village socialorganisations and governance institutions, forestresources and tenure, village regulations onforest management, timber prices, householdcollaboration mechanism for forest production(e.g., shareholding system for forestmanagement), and extension and credit services.

Group meetings examined questions ofrelevance to the whole village. In each researchvillage, representatives of five or seven groupswere interviewed. In the case of five groups, theywere village leaders, women, elders, middle-aged people and youths. In the case of sevengroups, they were village leaders, elders of 60years old or over, married middle-aged andyoung people, unmarried young people, women,the poor and the relatively wealthy. Interviewswere semi-structured, each lasting 2-3 hours.Each interview involved 4-12 people; most ofthem had not been involved in the householdsurveys.

Design (Planning)Participating researchers, farmers andgovernment officials worked together in aninteractive manner to identify options andstrategies for forest rehabilitation and evaluatethe role of existing technologies and indigenousknowledge. After a number of intensivediscussions, in which technical, economic andsocial factors were considered, several optionswere identified for each type of degraded forest.Farmers then chose the option or options bestsuited to their specific conditions andcircumstances.

Delivery (Testing)This phase involved planning the on-farmtesting, refining the technologies andapproaches, dissemination of results, andfeedback from users. Researchers andgovernment officials played a part in thisimplementation phase, but farmers had the mostimportant role. They tested options in theirforests, fields, home gardens and workshops.Participating farmer households representedthose who had different forest resource holdings,types of degraded forests and lands, and wealthlevels.

Species’ identification is a good example ofparticipatory implementation. The PRA surveysrevealed that villagers had not sufficient choiceof well-adapted species for profitable treeplanting. Based on this information, tree speciesnot previously grown in the area were selectedfrom both outside China and other parts of Chinaand tested their suitability for local ecologicalconditions and potential profitability. This wasundertaken at all the sites but was extremelyimportant at one site (Yuanmou, Yunnan) wherethe availability of a well-adapted and profitablespecies is key to degraded land reclamation. Theprocess was participatory and collaborative, withlocal communities providing land for fieldexperiments and being responsible for tendingseedlings.

Multidisciplinary ApproachForest degradation is related to bothsocioeconomic and biophysical factors andrehabilitation needs an integrated solution.Achieving a good understanding of causes offorest degradation and developing effectiveoptions or solutions often require amultidisciplinary approach. Each of the researchteams consisted of specialists from differentdisciplines. For example, the Hunan researchteam consisted of experts in forest economics,forest resource management, forest ecology,silviculture, and botany. Social scientists andagricultural experts were consulted as required.



TriangulationTriangulation was used as a research tool in dataanalysis to check the reliability of the datacollected. Supplementary survey or furtherobservations were made to validate data ifinconsistencies appeared in information fromsecondary sources, field observation, householdsurvey, key informant interview, group meetingor conventional forestry methods.

Indigenous KnowledgeIndigenous knowledge was used in developingsome options for rehabilitating degraded forestsand lands. Local people often had existingknowledge that was just as useful and effectiveas more recently developed techniques. Forexample, in Hunan, villagers developed asustainable system of establishment andmanagement of Chinese fir plantations severalhundred years ago (Menzies 1988, Yu 1994) butthis technology was abandoned for many yearsas government officials dictated plantingpractices.

Conventional Forestry ResearchMethodsWhile innovative approaches such as PRA wereemphasised, conventional forestry methods andtechniques such as forest vegetation survey, soilsurvey and expert consultancy were also used.

Forest Vegetation SurveyForest vegetation survey and forest inventorywere undertaken in sample plots of varying sizeat all research sites. Information was gatheredon forest types, species composition and forestproductivity (inventory), the extent of forestdegradation including historical changes inbiodiversity and forest biomass, and causes offorest degradation. Researchers andknowledgeable farmers together made surveysof historical changes to forests.

Soil SurveySoil survey was made in different forest typesand at different levels of degradation tounderstand relation between forest degradation

and changes in soil properties and fertility. Type,texture, depth, fertility of soil and its capacityfor holding water and fertility were observedand measured by routine methods.

Expert ConsultancyExpert consultancy was used to help with theidentification of causes of forest degradation andoptions for reclamation. Consultation took placewith those specialists with expertise in relatedfields or who had research activities in theresearch locations. The specialists were notmembers of the research teams of this projectbut they expanded the range of expertise andcontributed to developing options to addressparticular problems.

Action ResearchOptions identified to rehabilitate forests andforest lands and improve livelihoods were testedthrough an action research approach whichinvolved farmer participation as part of thecommunities’ own research and developmenteffort. Researchers and local governmentofficials facilitated this research in various ways.

Location of Research SitesThe main factors considered when selecting sitesfor the case studies were:• types and extent of degraded forest/land and

potential to address the causes of forestdegradation

• economic development level and the extentof dependence of local people on the forestand the forest products used and traded. Thiswas necessary to determine how people atdifferent levels of development market theirforest products and their attitude towardsforest rehabilitation, and

• environmental conditions, e.g., rainfall andhumidity vary east to west and differenttechnical interventions need to be developedfor areas with different climatic conditions.Tree species must be selected carefully foradaptation to site conditions.Case studies were carried out in one county

in each of four provinces (Map 1.1). They extend



Yunnan

Yuanmou

Guangxi

Hunan

Cangwu

Lin'an

Huitong

Zhejiang

from the eastern seaboard (Zhejiang) to theupper catchment area of the Yangtze River onthe high altitude plateau in the southwest(Yunnan) and together they represent areasonable sample of south China in the aspectsincluded in this study. Initially eight sites wereselected for study to provide a better sample butthis number had to be reduced due to costfactors. The sites selected were in:• Cangwu County, Guangxi• Huitong County, Hunan• Lin’an County, Zhejiang• Yuanmou County, Yunnan.

All the counties were in hilly or mountainousareas. Cangwu County is situated in eastGuangxi; Huitong County in the west of Hunan;Lin’an County in northwestern Zhejiang; andYuanmou County in central Yunnan in the uppercatchment area of the Yangtze River.

Research in Yuanmou was undertaken at bothcounty level and village level, while the studiesin the counties of Cangwu, Huitong and Lin’an

were largely at village level, though somecountywide issues were addressed to provide agood understanding of context. Village-levelstudies were carried out in:• Shanxin village in Cangwu County• Dongxi and Xiangjian villages in Huitong

County• Chenjiakan, Hongqiao and Sahngfeng

villages under the jurisdiction of GaohongTownship in Lin’an County

• Laofan, Xiaocun, Moke, Bingyue villages inYuanmou County.

Biophysical andSocioeconomic Settingsof South China

Topography and Extent of ForestLandsChina’s land area is 9.6 million km2 of which6.6 million km2 (69%) is mountainous. As theproportion of mountainous and hilly topography

Map 1.1 Location of the four provinces where study sites were located



in the south is higher than in the north, forestryis the primary land use and ‘forest land’ makesup most of the land area of south China. Morethan half of the 70% of people in rural areas insouth China live in mountainous areas and theyrely heavily on forests for timber, NWFPs andfuelwood for their livelihoods.

Mountainous and hilly areas comprisebetween 80% and 90% of the counties in theresearch project and the proportion of forest landis from 66% in Yuanmou County to 83% inShanxin village, Cangwu County.

ClimateMost of south China typically has a subtropicalclimate but rainfall and humidity vary markedlywithin the zone. Three research sites are humidor subhumid with a mean annual rainfall of 1516mm (Guangxi), 1415 mm (Zhejiang) and 1264mm (Hunan). In contrast, the climate ofYuanmou (Yunnan) is dry and hot and isclassified as semiarid. Its mean annual rainfallis 629 mm, and potential evaporation is six timesthe rainfall. Moreover, almost 90% of rainfallcomes during a short wet season (June toOctober). The temperature is as high as or higherthan in other research areas.

The climatic variation has implications forthe identification of technical solutions for forest

rehabilitation. The humid or subhumidenvironment is good for tree growth and fornatural regeneration of secondary forest but theharsh dry climate of Yuanmou makes successfultree growing a significant challenge.

Diversified Forests and PlantationsPrimary forest vegetation in most of southChina has been replaced by naturallyregenerated, secondary forests and man-madeplantations (Table 1.2). Secondary naturalforests include evergreen broadleaf forest,mixed coniferous and broadleaf forest, andvarious coniferous forests. There are majorplantations of Chinese fir (Cunninghamialanceolata), bamboo (e.g. Phyllostachyspubescens), and a range of tree species forNWFPs (e.g. Cinnamomum cassia andCamellia oleifera).

Evergreen broadleaf forest occurs at all thestudy sites, including Yuanmou. It is thedominant vegetation type in Shanxin village inCangwu county, but makes up a small proportionof forest area in the other three sites. All thebroadleaf forests at these sites are secondary inorigin.

Masson pine (Pinus massoniana) forestoccurs naturally in most provinces of southChina, and is present in Lin’an, Huitong and

Table 1.2 Woody vegetative cover on designated forest land (% of total land area)

Vegetative cover Yuanmou Shanxin, Dongxi, Lin’anYunnan Cangwu Huitong Zhejiang

Guangxi Hunan

Evergreen broadleaf forest 56 10 7Mixed forest 25Forest* and plantation 6Masson pine forest 33Chinese fir plantation 44 10Bamboo forest/plantation 20Tea plantation 4Other tree species for NWFPs 2 22Sparse bushes 46No woody vegetation 14 2

Forest land (%) 66 83 76 76

* Mixed forest, pine forest and evergreen broadleaf forest.



Cangwu counties but is considered absent inDongxi village, Huitong and Shanxin villages(Table 1.2) because they have individual treesbut not stands of this pine.

Chinese fir has been planted in many areas ofsouth China where Masson pine occurs althoughits best growth is observed on the more fertileand deeper soils of lower slopes and valleybottoms. It has been planted in Lin’an, Huitongand Cangwu counties but there are only scatteredtrees on the land of Shanxin village, Cangwu.

Plantations designated ‘NWFP’ areestablished for tree products other than wood.Although bamboo produces many non-woodforest products, it is treated separately becauseof its relative importance. These plantations area major source of cash income for villagers.Many tree and shrub species are used with thespecies and extent of NWFP plantations varyingbetween localities. For example, NWFPplantations accounted for only 2% of total landarea in Shanxin village, and major species wereCinnamomum cassia, Illicium micranthum andbamboo. Cinnamomum cassia is a medicinalplant, and fruits of Illicium micranthum (staranise) are a common spice in Chinese food.Bamboo is planted for both shoots and timber.The proportion of NWFP plantations was muchhigher in the two villages of Huitong, andcomprised about 25% of land area, mostly inhome gardens. Species used are Camelliaoleifera (tea oil), Vernicia fordii, Castanea sp.(chestnut), and fruits (orange, plum and pear).Bamboo is planted for both shoots and timber.In the Lin’an villages NWFP plantations madeup 24% of total land area, with tea and bambooas the main species. Bamboo is largely managedto produce shoots for export and localconsumption.

In Yuanmou, the majority of forest land iscovered with sparse bushes, 6% of forest landby poor quality mixed forest and tree plantations,and 14% has no tree cover at all. Forest cover isthe lowest of the four sites and degradation offorest land is more serious than at the other sites.

The variation of forest types between sitesis mainly a result of differences in the degree of

human intervention and it has implications forrehabilitation. Rehabilitation of existing,degraded forests in areas such as Cangwu,Huitong and Lin’an is possibly moreeconomically feasible and socially desirablethan reclaiming forest land without tree cover.However, where most of the tree cover hasalready been removed, as in Yuanmou, effortshave to focus on planting trees, shrubs andgrasses to restore the seriously deterioratingenvironment.

Role of Forest in Local LivelihoodsRural people in south China rely on forestproducts for income generation and on-farmconsumption and benefit from ecologicalservices of forests. Natural forests andplantations are managed for timber, NWFPs andfuelwood. The extent of local people’sdependence on forest and types of forestproducts they collect vary betweencommunities. This variation in dependence onforests inevitably results in differences incommunity attitudes to tree planting and forestmanagement. In Cangwu, Huitong, and Lin’an,there is heavy reliance on forests for incomegeneration as well as for timber and fuelwoodfor on-farm use. If villagers in these areas arenot interested in tree planting and sustainableforest management, it suggests that there is someproblem that needs to be resolved. In contrast,villagers in Laofan village, Yuanmou have othersources of income and domestic energy and mayhave little interest in tree planting.

Shanxin Village, Cangwu County,GuangxiForests are very important to the villagers inShanxin for generating cash income. Theevergreen broadleaf forest provides wood forfarm tools and is the source of many NWFPsand local fuelwood needs. Timber is harvestedin mixed forest for various on-farm uses. Beforethe late 1970s, they managed evergreenbroadleaf forest mainly for fuelwood and forwood for making charcoal. These products wereexported to cities in the Pearl River Delta and



Hong Kong. Subsequently, market demand forthese products diminished greatly due to theavailability of alternative energy sources, suchas electricity and natural gas. However, newmarkets appeared when many wood-processingenterprises requiring small-diameter wood wereestablished locally. Villagers also generateincome from NWFP species such asCinnamomum cassia, Illicium micranthum andbamboo. Now small-diameter wood, NWFPsand fuelwood are important sources of cashincome for local farmers. It was estimated thatthe contribution of forestry sub-sector toShanxin’s agricultural income was about 20%in the mid-1990s.

Huitong County, HunanVillagers manage Chinese fir and NWFPplantations to provide products for trade andthese are a major source of cash income forthem. Chinese fir timber is also harvested foron-farm consumption. Villagers also collectNWFPs and fuelwood from evergreen broadleafforest. It was estimated that forestry activitiescontributed about 30% of farmer income in thevillages of Dongxi and Xiangjian in 1993.

Lin’an County, ZhejiangForests in Lin’an play a greater role inemployment and income generation than in theother counties in the research project. Villagersmanage Chinese fir plantations, Masson pineforest, and evergreen broadleaf forest for timberfor trade and on-farm consumption. They managebamboo and tea plantations as cash crops. Therehas been a boom in developing bambooplantations and bamboo shoots are the majorsource of cash income for local farmers. Theyare so profitable that many farmers have extendedthe planting to unsuitable sites, resulting in poorproductivity and site degradation through soilerosion. Villagers also collect fuelwood fromevergreen broadleaf forest.

Laofan Village, Yuanmou County, YunnanMost people in Yuanmou generate income bycultivating vegetables in winter. It is less

profitable to plant trees and income cannot begenerated in short period of time. Little naturalforest remains in the area and there are fewplantations. Previously some trees were grownfor fuelwood but now there are other energysources such as electricity and gas available.Tree management is therefore relativelyunimportant and currently makes littlecontribution to local livelihoods.

Forest Tenure and ManagementTimber forests in China are controlled bygovernment agencies or enterprises, collectives,share-holding groups and individual households.Before the nationwide forestry reform, launchedin the early 1980s, most forests in China wereowned and managed by the collective(commune, production brigade and productionteam), and they hence were called ‘collectiveforests’ (Table 1.1). The reform allocated userights to denuded or non-forested collective landto individual households as their family plots(ziliushan in Chinese) in an attempt to encouragethem to plant trees and develop plantations. Thecollective also ‘allocated’ most of existingcollective forests to households (or a group ofhouseholds) to manage. As a result, often bothcollective management and household-basedmanagement of forest exist in one village, withthe latter being dominant form in terms ofmanaged forest area. In some cases, householdmanagement is only form of management as allforests and forest lands were allocated toindividual households (Yin 1994, Liu 2001).

Throughout China the state directly controlsabout 40% of forest land, and the collective 60%during the early 1980s when the forest reformbegan. In the ten provinces of south China,however, 90% of forests were collective owned,and the other two provinces (Sichuan andYunnan) had about 65%. After the forest reform,most of these collective forests were underhousehold management. For example, farmerhouseholds in Xiangjian village and Dongxivillage of Huitong manage 92% and 66% of forestland respectively, while households in the threevillages of Lin’an manage 85% of forest land.



Demography and Infrastructure

DemographyThree common features of demography in southChina were evident in the study villages.• Villages are agricultural communities and

rely heavily on agriculture (food cropping,forestry, and animal husbandry).

• A high proportion of the population hasreached labour age (women 16-54 years old,and men 16-59 years old), which means inmany families labour is available for forestrehabilitation from the underemployed.

• Most villagers have only primary and juniormiddle school education, which is aconstraint for securing off-farm employmentand external income generation.

Laofan village, Yuanmou has a smallpopulation. In 1995 there were 441 persons in108 households, with 73% of the population inlabour force of which 93% was engaged inagriculture.

Shanxin village, Cangwu had a population2570 in 521 households (1995). About 64% ofthe population had primary school leveleducation, 32% junior middle school educationand about 3% was illiterate and very fewvillagers had senior school education. In thisvillage 43% of the population was of labour age,the majority of whom were involved inagriculture. Only those with senior middleschool education and a proportion of those withjunior middle school education have sought parttime or full time off-farm employment.

Xiangjian village, Huitong had a populationof 1170 in 335 households (1995). About 40%of people had primary education, 40% attendedjunior school and about 4% went to seniormiddle school. The remainder was illiterate. Ofthe 75% of villagers at labour age, most wereinvolved in agriculture, and less than 10% hadoff-farm employment. Demography of Dongxivillage in Huitong was similar to that ofXiangjian.

The population of Gaohong township, Lin’anwas 10 513 in 1995. The education levels and

employment opportunities differ from the otherresearch sites. A large proportion of thepopulation had primary or junior middle schooleducation, but about 8% of villagers completedsenior high school. People of labour agecomprised 65% of the population. Only 45% ofthe labour force in Gaohong was working inagriculture, with 28% in secondary industry and27% in tertiary industry.

CommunicationsLaofan village is only 7 km from the YuanmouCounty town. A sealed road links it to the countytown that in turn has good road and railwayconnections with provincial capital cities(Kunming in Yunnan, and Chengdu in Sichuan).This provides villagers with ready access tomarkets for vegetables and other agriculturalproduce.

Shanxin village has relatively ready accessto transport. Six of its village household groupsreside along the highway passing through thevillage and 11 are connected by unsealed roadto the highway. However, the village is ratherisolated in terms of telecommunication, forexample, the village office had no telephoneconnection when this research was conducted.

The villages of Xiangjian and Dongxi areabout 25 km from Huitong County town. Bothvillages have ready access to both road and railtransport.

Gaohong is near Lin’an county town that hasa high standard highway to Hangzhou, capitalof Zhejiang province. Telecommunications aregood and many households have television. Thisenables farmers to obtain information andmarket signals more easily than farmers at theother sites.

Although generally mountainous and hillyareas have poor transport infrastructurecompared to those in the plains, in this studyalmost all the sites have relatively good accessto transport, which suggests this is not aconstraint to forest rehabilitation. Communitiesin mountainous areas do however vary in theirease of access to marketing and technologyinformation.



Economic StatusAmong the four sites, Lin’an was most advancedin economic development. Cangwu and Huitongwere less developed and Yuanmou farmers hadthe lowest income. For example, average percapita annual income of the three study villagesin Lin’an was about 1800 Yuan in 1995,compared to about 1540 Yuan in the villages ofXianhjian and Dongxi in Huitong. Average percapita annual income of villagers in Shanxinvillage, Cangwu was 980 Yuan in 1994, with80% of households unable to produce sufficientfood for themselves.

Extent of Forest DegradationForest degradation was interpreted as declinein productive capacity, reduced biodiversity, andin some cases deterioration of plantation health.The indicators used to assess degradationdiffered from one type of forest to another, andin a few cases, indicators for same type of forestare different across sites. Existing indicatorswere used as the aim was to concentrate onidentifying causes of forest degradation andintervention options for rehabilitation.

This research examined the level ofdegradation in the following types of forests andforest lands:• Evergreen broadleaf forest• Masson pine forest• Chinese fir plantation• Eucalypt plantation• NWFP plantations, and• Non-forested forest land.

The majority of forests was assessed asdegraded or seriously degraded. This isindicative of the situation in much of southChina. As a result, those rural people who relylargely on forest are now able to harvest lesstimber and non-wood forest products, whichmeans a decline in their income from forest andaccordingly a decline in their quality of life ifthey cannot find alternative sources ofemployment and income. The agro-ecologicalenvironment has become poor with potential

negative off-site effects such as local floodingand siltation. Forest degradation also means lossof biodiversity in natural forest. The situationin the different forest types is summarized belowand more details are provided in the case studiesin Chapters 2-5.

Evergreen Broadleaf ForestThree indicators were used to assess the state ofdegradation of evergreen broadleaf forest:• forest composition, or tree species diversity

indices (Simpson Index and Shannon-WienerIndex)

• biomass (standing timber volume/stock andfuelwood output)

• mean stem diameter at breast height.

The assessments showed most evergreenbroadleaf forests (70-85%) are degraded orseriously degraded. For example, 85% of thisforest type at Shanxin village was degraded,10% seriously (Chapter 3). Degraded andseriously degraded forest comprised 79% and69% of forest area of Xiangjian and Dongxivillages respectively in Huitong County(Chapter 4). Nearly 80% of this forest in thethree villages of Lin’an County was degraded(Chapter 5).

The degradation of evergreen broadleafforest is reflected in the decreasing number oftree species and families, simpler foreststructure, and reduced standing biomass (woodvolume), compared with its previous state andwith normal, undegraded forest. Degradedevergreen broadleaf forest often comprisesimpoverished and ecologically unstable stands.The cases of the villages of Cangwu, Huitongand Lin’an clearly show that the number ofspecies in the forests has been significantlyreduced. High value timber species for housebuilding and furniture are greatly diminished innumber and fast-growing, lower-value species,such as Castanopsis fissa and Quercus griffithii,have become dominant species in degradedforest. In seriously degraded forest few treespecies remain and many trees have been



reduced to a shrubby form. Seriously degradedforest of Shanxin village has become shrub landand bare land. In the three villages of Lin’an,evergreen broadleaf forests were in goodcondition in the 1960s and 1970s, with a densecanopy (80-90%), but now they are in poorcondition with a more open canopy (30-50%).

In Shanxin, standing biomass of the degradedevergreen broadleaf forest declined from about100 t ha-1 in the early 1950s to 65-100 t ha-1 inthe early 1980s to 30 t ha-1 in the late 1990s.The biomass is much lower than the 280 t ha-1

of forest with the same dominant species in anearby reserve.

Fuelwood output of Shanxin’s evergreenbroadleaf forest is now consistently less than20 t ha-1 (on about a 7 year cutting cycle),compared to the 30 t ha-1 or more from slightlydegraded forest. In the villages of Lin’an, meandiameter at breast height of evergreen broadleafforest declined from 10-12 cm in the 1960s and1970s to 6 cm currently.

Masson Pine ForestTwo indicators were used to assess thedegradation of Masson pine forests:• standing timber volume• mean diameter at breast height (dbh) and

mean tree height

Most Masson pine forest is degraded. InLin’an, pine forest was dense in the 1960s and1970s, with an average dbh of 26-28 cm, a meantree height of 11-12 m, and a timber volume of105-120 m3 ha-1. In contrast, by the mid-1990s,average dbh decreased to 8-10 cm; tree height,to 6-7 m; and timber volume to 20-30 m3 ha-1.

Chinese Fir PlantationIndicators used for Chinese fir plantations weresimilar to those for Masson pine forest. Theywere:• standing timber volume• mean diameter at breast height and mean tree

height• mean annual increment

Although probably in the best conditionamong all timber forests and plantations, manyChinese fir plantations are also degraded. TheMinistry of Forestry indicator for a fast-growingChinese fir plantation is 12-15 m3 ha-1 meanannual increment (MAI) depending on the areaof cultivation (Department for Reforestation,Ministry of Forestry 1982). Most Chinese firstands now are young and middle-aged and theirproductivity is lower than that of a normalChinese fir stand. For example, in Lin’an, 20-year-old plantations have a mean dbh of 10 cm,a mean tree height of 7 m and a mean annualincrement of 2.7 m3 ha-1, well below the nationalstandard.

Eucalypt PlantationThe only eucalypt plantation assessed was ofEucalyptus exserta in Laofan village, Yuanmou.The main indicator was growth rate.

Although this eucalypt is one of the mosttolerant species of poor soil conditions, the standis irregular and lacks vigour. Apart from the slowgrowth rate, there are almost no understoreyplants and there is continuing development ofnumerous soil erosion gullies. Both theplantation and the land are severely degraded.It is recognised that most species of Eucalyptusare not effective in controlling soil erosion indry areas, because their open crowns providevery little shading of the soil and their rootscompete very effectively with understorey plantsfor the small amount of water available. Thissuggests that a species with a dense spreadingcrown that is either deep rooting or has a lowwater requirement would have been a betterchoice on this eroded soil and in this dry climate.

NWFP PlantationsIndicators for NWFP plantations were decidedbased on the purpose of plantation establishmentand included:• productivity• uniformity of the plantation (bamboos)• plantation health.



The extent of degradation in NWFPplantations was variable. Tea plantations inLin’an and some bamboo plantations plantedoff-site were seriously degraded. In the studyvillages of Lin’an about 20% of Lei bambooplantations was degraded, with lowerproductivity, uneven mother bamboo individualsat different ages and higher incidence ofdiseases, compared to those of a normalplantation. Nearly half of bamboo forest usedfor producing dry shoots was also degraded.

Non-Forested Forest LandTwo indicators were used to assess the extentof degradation of this type of forest land byexamining how much forest left, how poor theyare, and how serious soil erosion is in the site.The indicators were:• per cent of land with vegetative cover, and• extent of soil erosion.

Yuanmou, Yunnan, was the site with thegreatest area of this type of land. While about66% of the land area of the county is designatedas forest land, only 6% was covered with forests,the majority had bushes, grass or no vegetativecover at all. Bush vegetation was characterisedby few species and with individuals sparselydistributed. Heavy rains in the short wet seasonresult in severe erosion of the unprotected soil.Degradation is a serious problem in Yuanmouand significant efforts are required for ecologicaland environmental restoration. Chapter 2describes this situation in greater detail. Othersites had relatively small areas of non-forestedforest land, e.g. 2% in Lin’an.

Causes of Forest Degradationin South ChinaThe four case studies provided insights into thecauses of forest degradation in south China. Inmost instances several interacting factors areresponsible. Factors change over time anddominant factor(s) in a given period may bereplaced by other factors as context changes.These factors can be grouped into two broadcategories. One group includes aspects of forestpolicy, development policy and othersocioeconomic factors; the other mainlyinappropriate forest practices (Table 1.3).

Policy and Socioeconomic FactorsPolitical and socioeconomic factors affectforestry development and some of the effectshave been highlighted in reviews by Li et al.(1988), Richardson (1990), Sun (1992), Yin(1994) and Zhu (1997). Of the severalgovernment policies that have been linked todeforestation and forest degradation perhaps themost devastating was the ‘Great Leap Forward’development policy in the late 1950s. Howeverpolicies such as the forestry tenure reform ofthe early 1980s also had the unintended outcomeof deforestation.

Radical Development PolicyIn 1958, the government launched the ‘GreatLeap Forward’ campaign and one of thecampaign’s objectives was to realiseindustrialization in China with steel output asthe major indicator. It was planned that China’ssteel output would surpass that of Great Britainin 15 years. To achieve the goal, people were

Table 1.3 Causes of forest degradation

Policy and socioeconomic factors Forest practices

Development policyBreakdown of forest management systemTenure insecurityOver-harvestingMarket environment and external conditions

Poor harvesting methodsSuboptimal silvicultural practices (species-sitematching, multiple cropping, monoculture, burning).



mobilised to make iron and steel using backyardfurnaces throughout China. The furnaces failedto produce usable iron and steel and wastefullyconsumed a very large amount of fuelwood andcharcoal. Another aspect of the policy was thecall for self-sufficiency in grain and this led toforested land being cleared and cultivated(Richardson 1990). As this was a countrywidecampaign it impacted massively on forests andcaused a significant forest decline that can beconsidered as the first disastrous deforestationin modern China.

Examples from our research sites lendsupport to this conclusion. In Shanxin, CangwuCounty, a large number of people from theneighbouring town came to the village to cuttrees and make iron and steel. Together withlocal villagers they established about 190charcoal making sites. Consequently, patches ofhigh quality evergreen broadleaf forests werecleared and many other dense forests becamepoor quality secondary forest with low stocking.This also occurred at our sites in Yunnan, Hunanand Zhejiang.

Breakdown of Forest ManagementSystemWhen the forest management system breaksdown, deforestation is very often a consequence.This happened in many places in Chinaespecially during two periods: during the ‘GreatProletarian Cultural Revolution’ and after theforest reform in the early 1980s. The CulturalRevolution, initiated in 1966 and lasting adecade, paralysed almost the whole governancesystem from national to village level. It resultedin the breakdown of the management system ofcollective forests in many areas. The collapseof forest management was not a failure of thesystem itself but was the result of the abnormalpolitical and policy environment.

The effects were reported from the Shanxinvillage in Guangxi where the leadership of thevillage lost control over forest management andillegal logging increased considerably. From1966-1968, fuelwood and charcoal output inShanxin and neighbouring villages experienced

a peak reflecting uncontrolled harvesting offorest products.

Significant deforestation during the forestreform in the early 1980s was also related tocollapse of the forest management system,through tenure insecurity and risk aversion byvillagers (as discussed below). Managementbreakdown explains why numerous trees andstands that remained under collectivemanagement at that time were cleared, whiletenure insecurity accounts for the harvesting offorest designated for household management.At the beginning of the forest reform, Laofanvillage retained its eucalypt plantation ascollective property but nobody was givenauthority to manage it. Consequently, villagerscut down almost all the trees that were plantedin the mid-1960s.

Tenure InsecurityIn modern China, government interventionsthrough a series of policies and decrees soughtto implement farm forestry and encouragefarmers to participate in developing forestresources. There were frequent, radical changesin forest tenure during the period from the early1950s to the early 1980s. A majority of forestswas the private property of rural households inthe first half of the 1950s. These private forestswere collectivised between 1955-1957, andbecame the property of communes in 1958. Theywere then owned and managed exclusivelycollectively until the early 1980s.

Late in 1978, China began its agriculturalreform. The primary feature of this reform wasthe abolition of collective agriculture and theintroduction of household-based agriculture.Before the reform, officials of the commune andproduction brigade made most decisionsconcerning land use and management, andagricultural production on the collectivelymanaged lands. Farmers usually did notparticipate in the decision-making process.Through the reform, use rights to agriculturallands were distributed to individual farmhouseholds, while the village collectiveremained the owner of land. Along with the



transfer of land use rights, households gainedincreased authority to make decisions about landuse and management and agricultural productionand marketing. Village collectives reservedsome rights as the owners of land. As a result,farmers’ motivation to increase agriculturalproduction improved considerably.

Nationwide forestry reform in the early1980s followed the pattern of the agriculturalreform. Use rights to denuded or non-forestedcollective land were distributed to individualhouseholds as their family plots in an attemptto encourage them to plant trees and developplantations. The collective retained ownershipof the land, but farmer households have landuse rights and own any trees they plant on theland. At the same time, the collectives also‘allocated’ existing collective forests to farmerhouseholds to manage. In this case, the collectiveowned the forest but transferred responsibilityfor forest management to individual households.Income from the forest is shared between thecollective and households that manage the forestunder a wide range of terms agreed by bothparties.

The state of tree and land tenure and itsinfluence on farmers’ revenue directlyinfluences farmers’ attitudes to engaging inforestry as part of their farming operations (He1995). It is arguable that tenure insecurity orfrequent changes in policy for forest tenure inChina over the three decades to the early 1980swas responsible for the deforestation that tookplace immediately during the forest reform inthe 1980s and hence contributed considerablyto forest degradation. Between 1977 and 1987there was an overall reduction in standingvolume of forests in south China from 49.6 m3

ha-1 to 43.7 m3 ha-1 (Yin 1994).Significant deforestation took place

immediately after the adoption of the household-based forest management system. The root causeof the deforestation was a complete lack offarmer confidence in tenure security, a sensedeveloped from their experience in the periodfrom the 1950s to the 1980s, rather than thehousehold management system itself. With

perceived insecurity farmers immediatelyharvested trees distributed to them to realise theirvalue. They were not convinced that thegovernment would not reverse its policy so riskaversion strongly influenced their decision toharvest the trees (Yin 1994). All the study sitesin Yunnan, Hunan, Guangxi and Zhejiangexperienced deforestation at this time. Thepremature harvesting led to a huge loss of forestand many of the remaining forests were severelydegraded by over-logging.

The long period needed for timber togenerate income for investors requires tenuresecurity. If the security is not there then farmerswill take a lower risk option. For example,farmers in Lin’an were very enthusiastic aboutbamboo, because they can obtain income fromit in just two or three years. This behaviourconforms to that reported in studies on propertyrights and investment incentives in other partsof the world (e.g. Fortmann and Bruce 1988,Besley 1995). It follows that if there is doubtabout tenure security, longer-term tree crops fortimber production are likely to receive poormanagement and become degraded, and areunlikely to be replaced after harvesting.

Over-HarvestingAlthough over-harvesting took place in the early1980s due to tenure insecurity, generally itresults from the excessive dependence on forestresources. Local villagers rely heavily on foreststo generate cash income from timber andNWFPs, while timber and fuelwood are alsoharvested for on-farm and domestic use.Fuelwood is the main or only source of energyfor cooking and space heating, in mostcommunities, especially those living inmountainous areas where temperatures arelower.

Forestry contributed to about 30% offarmers’ income in the villages of Dongxi andXiangjian in Huitong County. In Cangwu,Guangxi, sales of fuelwood and charcoal to citieson the Pearl River and Hong Kong were a majorsource of cash income from the 1950s to thelate 1970s until market demand disappeared as



new other energy sources became available.Fuelwood still is the primary source of energyfor local villagers and the small-diameter woodis in demand for new wood processing plants.

If over-harvesting is to be halted and forestsallowed to recover, villagers need to havealternative sources of income. This can be achievedthrough either farm forestry development or non-farm and off-farm employment, or a combinationof both.

Market Environment and ExternalConditionsThe market environment includes governmentinterventions to the market such as harvestingquotas, pricing and taxation policies andmarketing services. External conditions relateto local constraints such as resource supply andthe demand for the forest products.

Government regulations on wood harvestand trade impact negatively on timber forestestablishment and management in China. Acutting permit is required to cut trees and quotasto control the quantity of timber harvested in aregion are specified by the government. Timberharvesting must be carried out at a time andlocation, and in an amount specified by thecutting permit. To some extent, this limitsfarmers’ authority and makes them unable torespond to market signals. For example, theymay not be able to harvest and sell their woodwhen prices are high due to the inability to obtaina permit. In most cases, the forestry department,or other institutions to which it delegatesauthority to purchase timber from farmers andvillage collectives, monopolizes the timbermarket. Consequently, timber prices are skewedand farmers receive prices lower than those onthe free market. Present wood marketingpractice may encourage illegal logging and salesand discourage farmers from managing timberforests sustainably. The study in HuitongCounty, Hunan, concluded that restrictiveregulations on timber harvesting and marketingand the high level of taxation and fees on timber

sales were disincentives to farmer to invest insustainable management of plantations ofChinese fir and other timber species (Chapter4). In comparison, the market environment forNWFPs and bamboo shoots is much better asfarmers are able to sell them freely.

Poorly designed mechanisms for benefitdistribution can discourage investment in forestmanagement and lead to forest decline. Datafrom our surveys show timber income isdistributed between outside stakeholders(government and forest department) and localcommunity/households. The outsidestakeholders receive income in the form of taxesand fees. For example in Huitong, Hunan, theprice for Chinese fir timber is about 600 Yuanm-3 from which 295 Yuan m-3 is deducted forgovernment taxes and fees including the forestdepartment’s fee for the reforestation fund(Chapter 4). Returns to the farmer are low as ittakes at least 20 years before Chinese fir is largeenough to harvest. In contrast, tax on bambooshoots is less than 1% of sale price. This isanother reason why farmers have littleenthusiasm for timber forest management.

Degradation or deforestation may also resultfrom poor local socioeconomic conditions.Where there is reduced demand for a productfarmers may respond by cutting down the trees(He 1995). In Yuanmou, Yunnan, forests werecut down and not replanted when the localdemand for fuelwood disappeared andagricultural crops provided an alternative sourceof income (Chapter 2). In the case of teaplantations in Lin’an, Zhejiang (Chapter 5),when market demand and prices for tea declined,farmers neglected their tea plantations andconsequently the productivity decreased. Ifindicators other than output are used, thenstrictly speaking it may not be appropriate toassess such a plantation of tea or other NWFPspecies as ‘degraded’.



Forest Practices

Harvesting MethodsInappropriate logging methods result from over-dependence of villagers on timber/wood for bothfarm-use and income generation. Over the pasttwo decades most timber stands in south Chinahave been young or middle-aged with a lowtimber volume. In the absence of mature forests,farmers logged the best quality, large-diametertrees from immature stands (badamao inChinese) to meet their timber needs. Thisselective logging also occurred because saleprices for large-diameter logs are high. Farmersin Cangwu, Huitong and Lin’an counties alladopted the logging practice that resulted in adysgenic selection in their forests. In Cangwu,increasing demand for wood for new localindustries resulted in cutting cycle being reducedfrom 8 years to 5-6 years.

Poor harvesting occurred in bamboo forestfor producing shoots. Farmers harvest earlysprouting and vigorous shoots and leave latesprouting, weak or unhealthy shoots as motherbamboos. This is detrimental to healthy andproductive bamboo forest. They also breakshoots off by hand instead unearthing them withhoes or other tools, a practice believed tocontribute to lower quality bamboo shoots.Demand and prices for low quality shoots arelow and accordingly local farmers were notenthusiastic about investing their time andmoney in properly managing bamboo forest.

Silvicultural PracticesTraditional cultivation practice for Chinese firwas a sustainable system that includedestablishing modest-sized plantings on suitablesites, allowing land to lie fallow after two orthree successive crops, and intercropping withfood crops or cash tree crops. However, thesetraditional technologies were abandoned forvarious reasons after about 1960. From this time,extensive Chinese fir plantations were oftendeveloped on inappropriate sites, and fallowsand mixtures were not used. This changedsilvicultural practice has resulted in lower soil

fertility and forest land degradation (Fang 1987,Sheng 1991, Yu et al. 1992), and plantationproductivity of later rotations is low. A furtherfactor in productivity decline in later rotationsmay be the destructive effects on soil fertilityof burning and intensive site preparationpractices prior to planting Chinese fir. Burninghas been a common practice in Chinese firplantations because it provides short-termbenefits to soil fertility, prevents weed growth,and makes planting easier. There are howevermany records of loss of site productivity intropical plantations due to the loss of soil organicmatter due to burning (e.g. Goncalves et al.1997). Observations at our research sites areconsistent with these literature reports, e.g. inthe Hunan villages average MAI of Chinese firplantations has decreased from 10.5 m3 ha-1 inthe first crop to 9.2 m3 ha-1 in the second cropand 6.0 m3 ha-1 in the third crop. These figuressuggest that the decline in tree productivitybecomes very significant in the third crop.

There has been considerable governmentpressure for farmers to plant Chinese fir withoutthe necessary attention to situating theplantations on more fertile, deeper soilsoccurring on lower slopes and valley bottomswhich this species is known to require forsatisfactory growth. In Huitong (and otherareas), Chinese fir plantations were oftendeveloped on shallow, infertile soils on upperslopes. The majority had low productivity andabout half were classified as degraded (MAI =6-10.5 m3 ha-1) or seriously degraded (MAI <6m3 ha-1) (Chapter 4).

Species-site matching is also a problem inbamboo plantations. In Lin’an, Zhejiang,farmers have established extensive Lei bambooplantations on steep slopes and ridges withinfertile, shallow soil and other lands not suitablefor the species. Bamboos planted on these sitesgrow slowly and plantation productivity is low.Planting Lei bamboo on steep slopes and ridgeshas caused serious soil erosion, and some extentbiodiversity loss where natural forests have beencleared for the plantations.



A mulch cultivation technology for bamboosdeveloped in the late 1980s to stimulate shootsto sprout earlier than normal has also causedproblems. Lei bamboo usually sprouts afterChinese New Year but mulching increased soiltemperature and shoots sprout before and duringChinese New Year when prices are much higher.This technology is profitable but unsustainableas it results in several bamboo diseases andunhealthy bamboo plantation including smallerstem diameters, a higher proportion of oldmother bamboos and increased flowering. Boththe application of the mulch technology and poorsiting of plantations are motivational factorsbecause high profitability is a great incentivefor farmers to plant bamboo.

The underlying causes of these methods andpractices leading to forest decline are over-dependence on forest products for livelihoods,farmer motivation, inappropriate guidelines offrom government agencies on plantationdevelopment and limited availability ofinformation on appropriate technologies andmethods.

Forest Rehabilitation in SouthChinaThe above analysis indicates that successfulrehabilitation of degraded forests and forest landneeds to address several issues. While someissues are site-specific, others will apply to muchof south China. They will require action to: 1)reduce dependence on forest resources; 2)overcome constraints to rehabilitation; 3)improve the motivation of farmers by removingdisincentives; 4) improve technical guidelinesfor plantation development; 5) build and raiseenvironmental awareness of farmers and othergroups; and 6) identify and develop effectiveinterventions for rehabilitating degraded forestsusing innovative approaches.

Reduce Dependence on ForestProductsAn important cause of forest degradation in Chinais over-harvesting of forest products driven byover-dependence on forests. Many farmers rely

heavily on timber/wood for both farm-use andincome generation. Over-exploitation explainsmuch of the degradation of timber forest,especially that observed in evergreen broadleafforest in Cangwu, Huitong and Lin’an. Both over-harvesting and selective logging immature standshave contributed to forest decline.

Ways must be found to reduce dependenceon forest products to support livelihoods if thepresent degraded forests are to be rehabilitated.These can include developing fast-growingplantations and intensifying or developing newfarm, non-farm and off-farm activities andemployment. Some possibilities weredemonstrated at the research sites e.g. neemplantations in Yuanmou, and pig raising andorchard development in Cangwu. In Lin’an,where education levels were higher, there weregreater opportunities for employment in non-farm and off-farm industries.

Overcome Constraints to ForestRehabilitationThere are many site-specific constraints tosuccessful rehabilitation of degraded forests, butsome constraints are more general. Experienceat the research sites highlighted the importanceof size of forest holdings, current access toforest, access to technical information andadvice, and capital for investment.

Fragmentation of Forest HoldingsFragmentation resulted from inappropriateimplementation of policy for the forest reformin the early 1980s. When forest land and forestwere distributed to individual households, toensure equal distribution they were divided intomany tiny plots based on tree species and age,land quality, distance of forest/land fromvillage etc., and matched to family size. As aresult, a household was allocated several verysmall plots in different locations and so userights to forest/land in one location are ownedby a number of households. For example, inShangfeng village of Lin’an County, severalhouseholds each have 10 or more plots offorest, while the average holding is only 2.3



ha per household (Chapter 5). This situation isvery common in south China.

Land fragmentation makes it difficult todevelop an integrated plan to rehabilitate adegraded forest. For example, mountain closurefor natural regeneration of secondary forest andplantation establishment in small catchment iscost-effective and technically easy to adopt, butit needs common action of many households.Mountain closure was adopted in three of ourresearch sites. In Lin’an, household forest farmsand a shareholding system were emerging asways to solve the fragmentation problem. In theformer case, a household expands its forest landarea through leasing land use rights from, or inpartnership with, other farmers. In the latter, allparties involved pool their land, funds and/ortechnology to manage forest and establishplantations. They subsequently share benefitsgenerated from the forest based on the shares theyhold through their resource input. A shareholdingsystem was also adopted by villagers in Huitongbut was not acceptable in Shanxin village. InShanxin farmers were more comfortable with anoption that did not change the form of foresttenure or involve redistribution of forest benefits.Each household with forest land in the closed arearetained control of their trees and will harvestproducts only from their own plot(s) of land whenthe mountain is reopened.

A point that must be made regarding commonaction in forest rehabilitation is the form ofcommon management. It is households whoshould decide what forms (e.g. shareholdingsystem, household forest farm or other models)to be adopted based on their socioeconomiccircumstances. At our sites decisions on thisseemed to be related to the communities’ levelof social and economic development and theextent of their exposure to market forces. InLin’an, farmers have achieved considerableeconomic advances and were more amenableto a shareholding system or household forestfarm that involves leasing of land use rights. Ashareholding system was not acceptable inShanxin village, Cangwu, which is much lessdeveloped, more isolated and the farmers more

conservative. In Huitong, which is moredeveloped economically than Shanxin, theshareholding system as an option for commonmanagement of forest often was also rejectedby local communities and farmers, though thegovernment made great efforts to promote it.

Observations from these sites and elsewheredemonstrated that common management is notonly is able to solve the problems of forest holdingfragmentation but also is more cost effective thanindividual household management where, forexample, each household must allocate a personto look after its scattered plots. It is necessary topool resources or to transfer land use rightsbetween households to realise commonmanagement for natural regeneration of forest butit must be achieved through negotiation and inresponse to market forces rather than externallyimposed by the administration.

Access to ForestsVillagers in mountainous areas rely heavily onforests for their welfare and so need some accessto forest resources when mountain closure isidentified as an option for natural regenerationof forest. If all their forest is closed, they willbe unable to collect fuelwood and NWFPs,harvest timber, or graze their animals. If thisoccurs they may have no cash income to buymanufactured goods or to pay school fees andmedical expenses. Full closure of a mountainarea for regeneration is therefore not acceptableand will not be successful.

Recognising the problems for full mountainclosure, partial closure was proposed for thestudy villages of Huitong and Shanxin villagewhen this strategy was identified for forestrehabilitation. Half of a mountain area wasclosed during a given time, while the other halfwas open for use. When the closed half reopensafter a specified period (several years), the otherhalf of mountain will be closed for rehabilitation.In Shanxin village, several households had allor the majority of their forests in the designatedclosed area so an exchange of forest land wasarranged with those households with majorityof their forest outside the closed area. The



arrangement made it possible for all householdsto have access to forest resources.

Implementation of partial mountain closurewas effective in the study villages in Cangwu,Huitong and Lin’an, which suggests it could beapplied in the context of small scale forestry insouth China.

Lack of Technical InformationIt was evident from our research that villagecommunities did not have ready access toinformation about alternative technologies thatcould help them diversify their sources ofincome. In Lin’an, farmers left tea plantationsuntended when market demand for general teadeclined and they were not able to identifyopportunities for alternative, profitable forestproducts. In Huitong, there was a need foragroforestry technologies and to identify newspecies or varieties with market potential thatwere adapted to local ecological conditions.These are beyond farmers. The situation wassimilar in Shanxin village when villagers triedto develop other sub-sectors of agriculture andnon-farm activities to reduce pressure onexisting forests. In Yuanmou, lack of a speciesthat was adapted to local conditions and thatcould be grown profitably was one reasonvillagers were not enthusiastic about plantingtrees.

These problems were identified and resolvedin the research sites by tapping into the broaderexperience and knowledge of project researchersand local research and development institutions.In Lin’an, there was a government extensionorganisation at township level and farmerdemonstration households with primaryresponsibility to demonstrate successfultechnologies to their fellow farmers but poorcommunication between them. The projecthelped to improve the agricultural extensionservices by establishing village technical groupsto assist the flow of information.

Villagers have good understanding of theirproblems but they need help in identifyingoptions to solve them. Public extension serviceshave the potential to make great contribution to

communities in remote and isolated rural areasbut need new initiatives to improvecommunication. Unless this occursrehabilitation of degraded forests will be difficultas villagers will find it hard to find alternativesources of income.

Shortage of Capital for InvestmentShortage of funds is a constraint most farmersface when they want to regenerate naturalforests, establish plantations or undertake otheragricultural and non-farm activities. This was ageneral problem at the research sites. In Shanxin,one of difficulties in implementing the‘mountain closure’ strategy was how to coveroperational costs. In the Huitong villages,farmers had to change their plan to replacedegraded Chinese fir plantations with chestnutsand oranges because they could not raise fundsfor these species from external sources and werenot able to cover all the costs themselves. Finallythey decided to plant Ginkgo biloba, a tree usedto provide medicinal products, because theforestry department at that time provided loansonly for that species.

Some of these problems in the research siteswere solved by financial support by the projectand special allocations by local governments butthis is not a solution that can be replicatedelsewhere. Improved rural credit services, suchas the microfinancing arrangements of GrameenBank and other rural financial institutions whichhave already shown promise in some areas inChina, are needed.

Farmer MotivationIt is essential to harness interest of allstakeholders, but especially rural communities,in rehabilitation and sustainable management ofdegraded forests in south China. Without theiractive involvement, there is little likelihood ofsuccess because so much of forest land in theregion is managed either by households or byvillage collectives. Our research, and that ofothers, shows that incentives are related to policyfactors, marketing of forest products, andcomparative economic benefits. At policy level,



tenure insecurity, over-taxation on timber andrestrictive regulations on wood harvest and tradeare factors that have been disincentives tofarmers to invest in timber tree planting andforest management. It also is important tomanage incentives as over-enthusiasm forplanting of a particular crop can have a negativeimpact on forests, as in the case of widespreadplanting of bamboo for shoot production inLin’an.

Tenure InsecurityTenure insecurity has been not only one offactors responsible for deforestation and forestdegradation in China but also a disincentive forfarmers to be involved in forest rehabilitation.The frequent change in policy for forestownership over the three decades to the mid-1980s led to a great lack of confidence amongvillagers in tenure security. With perceivedinsecure tenure, villagers did not manage forestssustainably and did not want to invest inestablishing timber plantations or managingtimber forests that need a long period to generateincome. In contrast, farmers were extremelyenthusiastic over bamboo and other NWFPplantations that can generate income in just afew years. The behaviour conforms to thatreported by others on property and investmentincentives in other parts of China and elsewhere(e.g. Fortmann and Bruce 1988, Besley 1995).

The tenure issue also involves the rights ofowners to dispose their property. A frustrationamong Chinese farmers is to have trees,recognised by law and policy as their property,which they are not permitted to harvest becauseof environmental concerns. The logging ban innatural forests in most areas of China (Yang2001) is an example. Such a ban may benecessary for the country’s environmentalbenefits but then farmers should be properlycompensated for their contribution. If this doesnot occur they will again lose confidence inpolicy for forest tenure and have no incentiveto manage their forests.

Over-Taxation on TimberDistribution of forest benefits, if not properlydesigned and implemented, can be a significantdisincentive for farmers and other investors.Timber income is distributed between outsidestakeholders (including government, forestdepartments and timber companies) andcommunity/individual households. Our datashow that local government and the forestdepartment captured about half of the turnoverof wood trade in the form of tax and fees; costsof logging and timber transport are estimated atabout 30-35% of sale price; and farmers receivedonly the remaining 15-20%. Profits to farmersfrom timber were nearly zero whenestablishment and management costs are takeninto account. Shares of timber income weresimilar in Lin’an, Zejiang, with outsidestakeholders seizing half of gross timber income(trade price). In the case of Chinese fir log inHuitong, government captured 25% of tradeprice; forestry department about 15%; timbercompany around 35% (including costs oftransport and profits); and farmers, 25% ofwhich 9% is logging cost. Sun (1992) refers toother studies that show the farmer captures onlya very small part of income from timber sales.As a result, farmers shift to develop NWFPplantations such as bamboo (e.g. Ruiz-Perez etal. 1996), fruits, nuts and others, because tax onthese products is very low, e.g. tax on bambooshoots is less than 1% of sale price.

The reduced level of benefits to farmers dueto over-taxation of timber in part explains whythey have little enthusiasm about timberplantation establishment and forestmanagement. It is also clear that where the statehas reduced its influence on production andmarketing of other crops farmers have beenmotivated to increase production.

Regulations on Wood Harvest and TradeWood harvesting is highly regulated in Chinathrough a system of cutting quotas and permits.The logging quota set by government, ratherthan market supply and demand, is the mostimportant factor determining China’s annual



timber production (Sun 1992). A cutting permitis required to cut trees irrespective of whetherthey belong to private households or collectives.Timber must be harvested at a time, location, andin the amount specified by the cutting permit.When the quota is filled for a county or town,then no more cutting permits will be issued thatyear. Quotas are limited and obtaining permits isa difficult, costly and time-consuming process.The regulation is discouraging because it inhibitsfarmers from responding quickly to marketsignals. They may not be able to harvest and selltheir wood when prices are high due to failure toobtain a permit. It also provides an incentive forillegal logging and selling.

Current wood marketing regulations are adisincentive. The timber market is a monopolywith only the government forest department, orinstitutions designated by the forest department,entitled to purchase timber from villages andhouseholds. Consequently, farmers may obtainprices much lower than market prices, while thestate timber company benefits. In comparison,the market environment for NWFPs is muchbetter as farmers can sell their products freely.This, in part helps explain why farmers are moreenthusiastic about NWFPs than timber. Tomotivate farmers to establish and manageforests, monopolised purchases by the forestdepartment and other regulations must bereviewed and modified to give farmers moreflexibility and authority in wood marketing.

Market Demand and PricesMarket demand and prices for forest productshave a strong influence on farmers’ motivation,especially in the case of NWFPs. When demandand prices are low, farmers have little incentiveto invest time and money managing their forestsproperly. For example, farmers in Lin’an lefttheir bamboo forest for dry shoot production andtea plantations untended because demand for thetwo products was low. Conversely, high pricesencourage farmers to develop plantations andmanage forests. Farmers in Lin’an were soenthusiastic about Lei bamboo that they plantedthe species on unsuitable sites, a practice that

resulted in plantations with low productivity andassociated soil erosion. While this is clearlyundesirable, it is not uncommon for properspecies-site matching to be neglected whenprices for a commodity are high.

Comparative Advantage in EconomicBenefitsFarmers will plant trees and manage their forestswhen their products are more profitable thanproducts from other agricultural activities. Farmersin Laofan village were not interested in treeplanting because cultivation of winter vegetablesis much more profitable and is their major cashincome. The income is generated within a yearand so they rely little on trees and plantations fortheir livelihoods. When tree planting is profitableor has the potential to be more profitable than otheragricultural activities, farmers will take actionimmediately as the rapid diffusion of neem treesin Yuanmou and other areas attests.

This is a factor that must be taken into accountwhen the government urges rural communitiesto restore environmental conditions throughplanting trees and grass in environmentallysensitive areas such as the upper reaches of theYangtze River.

Raising Environmental AwarenessHarnessing self-interest of stakeholders isessential but not sufficient. Althoughprofitability was the major driving force forfarmers to be involved in forest establishmentand management, the research found thatvillagers are also not active in planting treesbecause there is a lack of awareness ofenvironmental benefits of forest and treeplantations. Generally they did not recognisedirect benefits from improved agroecologicalconditions and reduced soil erosion. Farmers inYuanmou showed little interest in revegetatingbare hilly land even though there was evidencethat rising salt levels in the soil was beginningto affect the productivity of their more fertilecrop lands (Figure 2.1). This finding raises aneed for education among villagers to build andraise their environmental awareness. They may



participate more actively in forestry if they areconvinced environment restoration is not onlybeneficial to others but also to themselves.

Improve Plantation DevelopmentGuidelinesGuidelines by forest authorities on plantationdevelopment may need to be modified to preventforest plantations declining in productivity. Inthe 1960s-1980s, some forest authorities wereso ambitious to establish extensive plantationsthat they ignored the silvicultural principle ofspecies-site matching and practices to maintainsoil fertility. Plantations developed under theguidelines are poor and with low productivity.The inappropriate silvicultural technologies ofChinese fir in Huitong (site mismatch, multiplecropping [duodai lianzai in Chinese],monoculture plantations, burning and intensivesite preparation practices) are mainly a result ofpoor guidelines rather than a lack of appropriatetechnology. However, further research and theestablishment of demonstration plots with newtechnologies may be necessary to show theeffects of burning logging slash on the long-termproductivity of plantations. Research tounderstand the effects of site management onChinese fir and other tropical species inplantations has begun (e.g. Nambiar et al. 2000).

Developing Effective Interventionsfor Forest RehabilitationForest degradation is a complex process andresults from many inter-related factors. Asdiscussions and analysis of the research projectresults have shown, these factors involve policy,sociology and economic aspects as well astechnical issues. Diagnosis of problems (extentand causes of forest degradation) and constraintsto rehabilitation interventions and livelihoodenhancement require more innovativeapproaches than the conventional, singlediscipline approach usually applied in China.

Our research demonstrated that aparticipatory and interdisciplinary approach isvery effective. All stakeholders, i.e., villagers,researchers and government officials, played an

important role in developing strategies. Villagersknow their own circumstances and localconditions and can make significantcontributions to the identification of problemsand designing interventions to address them.Involvement of professionals can provideexpertise and new ideas, and improvecommunication between government agenciesand villagers. Government officials areimportant for forest rehabilitation because theirparticipation will help them make effectivepolicies and their networking with governmentagencies will help villagers to gain access topublic resources such as extension services, ruralcredit and market information.

Three lessons can be learnt from the researchexperience to encourage active participation ofthe three parties: 1) building partnerships toaddress problems requires all participants to trusteach other; 2) responsibilities of each of theparties involved and the potential benefits mustbe clearly specified at the start of discussionsand self-interest harnessed to encourageparticipation; and 3) farmers and theirindigenous knowledge must be respected andrecognition given to their ability to participatein and contribute to identification andimplementation of effective solutions. Illiteracyshould not be equated with ignorance. Withoutsuch preconditions, it is impossible to achievegenuine farmer participation when normallyfarmers are rarely involved in the design andplanning strategies for sustainable forestmanagement.

A multidisciplinary approach is effective indeveloping interventions and strategies for forestrehabilitation and associated improvements inthe livelihoods of rural communities. Ourresearch recognised the complementarity of theskills of social scientists, economists and rangeof forestry researchers, but the research teamsmay have performed better with greater inputsfrom social science. We found that adoption ofa multidisciplinary approach required: 1) teammembers to commit to cross-disciplinarycollaboration with members working withintheir disciplines, but contributing to overall



outputs; 2) members of different disciplines tolisten to and respect others’ opinions and to thinkmore broadly than their own discipline; and 3)a team leader with skills to guide amultidisciplinary team, manage disciplinaryconflicts and negotiations, and synthesise viewsand findings from various stakeholders.

ReferencesBesley, T. 1995. Property rights and investment

incentives: theory and evidence fromGhana. Journal of Political Economy 103:904-937.

Chambers, R. 1994a. The origins and practiceof participatory rural appraisal. WorldDevelopment, 22: 953-969.

Chambers, R. 1994b. Participatory ruralappraisal (PRA): analysis of experience.World Development 22: 1253-1268.

Chambers R. 1994c. Participatory ruralappraisal (PRA): challenges, potentials andparadigm. World Development 22: 1437-1454.

Chambers, R. and Guijt, I. 1995. PRA – Fiveyears later: where we are now? Forests,Trees and People Newsletter No. 26/27: 4-14.

Chen, B. 1992. Status, causes, and improvementstrategy of soil degradation in timberplantations in China. In: Sheng, W. (ed.).Research on site degradation of timberplantations. China Science and TechnologyPress, Beijing. (In Chinese)

China Statistical Bureau.1993. China statisticalabstract, Beijing.

Contreras-Hermosilla, A. 2000. The underlyingcauses of forest decline. CIFOR OccasionalPaper No. 30, Center for InternationalForestry Research, Bogor, Indonesia

Department for Reforestation, Ministry of Forestry(ed.) 1982. Technologies for high yieldChinese fir plantations. China ForestryPublishing House, Beijing. (In Chinese)

FAO (2001). Global forest resources assessment2000. Main report. FAO Forestry Paper

140. Food and Agriculture Organization ofthe United Nations, Rome.

Fang, Q. 1987. Impacts of many crops ofChinese fir in succession on soil fertility andtree growth. (Chinese). Forestry Science 23:389-397. (In Chinese)

Fortmann, L. and Bruce, J.W. (eds.) 1988.Whose trees? Proprietary dimension offorestry, Westview Press, Boulder,Colorado, USA.

Gonçalves, J.L.M., Barros, N.F., Nambiar,E.K.S. and Novais, R.F. 1997. Soil andstand management for short rotationplantations. In Nambiar, E.K.S. and Brown,A.G. (eds.) Management of soil, nutrientsand water in tropical plantation forests.ACIAR Monograph No. 43, 379-417.Australian Centre for InternationalAgricultural Research, Canberra.

He, Q. 1995. Economics and socioeconomicsof farm forestry in south China. In: Cai,M. and Hu, S. (eds.). Integrated researchin farm forestry, 237-247. China Scienceand Technology Press, Beijing.

Institute of Forestry and Pedology. 1980. Acollection of papers on ecology researchof Chinese fir plantations. ChineseAcademy of Sciences, Shenyang. (InChinese)

Jiang, Z. 1997. On the strategic position offorestry in sustainable development.Forestry Economics 2 (1): 17-26.

Kaimowitz, D. and Angelsen, A. 1998.Economic models of tropical deforestation:a review. Center for International ForestryResearch, Bogor, Indonesia.

Li, W. (ed.). 2001. Agro-ecological farmingsystems in China. Man and the BiosphereSeries Vol. 26. UNESCO, Paris and TheParthenon Publishing Group New York,USA and Carnforth, United Kingdom.

Li, J., Kong, F. and He, N. 1988. Price andpolicy: the keys to revamping China’sforestry resources. In: Repetto, R. andGillis, M. (eds.) Public policies and misuseof forest resources, 205-246. Cambridge



Ruiz-Perez, M., Fu, M., Xie, J., Belcher, B.,Zhong, M. and Xie. M. 1996. Policy changein China: the effects on the bamboo sectorin Anji County. Journal of Forest Economics2: 149-176.

Sheng, W. 1991. The management ofCunninghamia lanceolata plantations inChina. In: Shi, K. (ed.). Development offorestry science and technology in China,108-114. China Science and TechnologyPress, Beijing.

Sun, C. 1992. Community forestry in southernChina. Journal of Forestry 90 (6): 35-40.

Xu, H. 1992. Dynamic characteristic of soilfertility of forest land and problems of soildegradation in timber plantations. In:Sheng, W. (ed.). Research on sitedegradation of timber plantations. ChinaScience and Technology Press, Beijing. (InChinese)

Yang, X. 2001. Impacts and effectiveness oflogging bans in natural forests: People’sRepublic of China. In Durst, P.B.,Waggener, T.R., Enters T. and Tan L.C.(eds.), Forests out of bounds: impacts andeffectiveness of logging bans in naturalforests in Asia-Pacific, 81- 102. RAPPublication 2001/08, Food and AgricultureOrganization of the United Nations,Regional Office for Asia and the Pacific,Bangkok.

Yang, Z., Zhuang, Z., Qin, D., Ran, G., Fu, W.1999. Afforestation technologies for waterconservation in arid-hot valleys ofYuanmou County. Bulletin of Soil andWater Conservation 19: 38-42. (In Chinese)

Yin, R. 1994. China’s rural forestry since 1949.Journal of World Forest ResourcesManagement 7: 73-100.

Yu, L. 1997. A study for the selection ofafforestation species for the dry-hot rivervalley in Yuanmou. Journal of SouthwestForestry College. 17(2): 49-54. (InChinese)

University Press, Cambridge. UnitedKingdom.

Liu, D. 2001. Tenure and management of non-state forests in China since 1950: a historicalreview. Environmental History 6: 239-263.

Menzies, N. 1988. Three hundred years oftaungya: a sustainable system of forestry inSouth China. Human Ecology 16: 361-377.

Nair, P.K.R. 1993. The diagnosis and design(D&D) methodology. In: An introductionto agroforestry, 347-356. Kluwer AcademicPublishers, Dordrecht, Netherlands andInternational Centre for Research inAgroforestry, Nairobi.

Nambiar, E.K.S., Tiarks, A., Cossalter, C. andRanger, J. (eds.) 2000. Site managementand productivity in tropical plantationforests: a progress report. Center forInternational Forestry Research, Bogor,Indonesia.

Oakley, P. 1987. “State of process, means orend? The concept of participation in ruraldevelopment ideas”. IDRC Bulletin, March1987. International Research andDevelopment Centre, Ottawa.

Raintree, J.B. 1987. The state of the art ofagroforestry diagnosis and design.Agroforestry Systems 5 (3): 219-250.

Research Team of Intensive Cultivation ofChinese Fir. 1992. General report ofresearch group for ‘National ResearchProgram on Intensive SilvicultureTechniques on Chinese Fir Plantation’.Chinese Science and Technology Press,Beijing. (In Chinese)

Richardson, S.D. 1990. Forests and forestry inChina: changing patterns of resourcedevelopment. Inland Press, Covelo, USA.

Rocheleau, D. 1999 Confronting complexity,dealing with difference: social context,content and practice in agroforestry. In:Buck, L.E., Lassoie, J.P., and Fernades,E.C.M. (eds.) Agroforestry in sustainableagricultural systems, 191-235. LewisPublishers, Boca Raton, Florida, USA.



Yu, X. 1994. Chinese fir silviculture. FujianScience and Technology Press, Fuzhou. (InChinese)

Yu, X., Ye, G., Lin, S., and He, Z. 1992. Anapproach to the culture system for Chinesefir. Journal of Fujian College of Forestry12(3): 259-263. (In Chinese)

Zhao, X. 1997. Forestry Action Plan forChina’s Agenda 21 (abridged). ForestryEconomics 2 (1): 3-16.

Zhu, Z. 1991. Forest ecology. China ForestryPublishing House, Beijing. (In Chinese)

Zhu, Z. (ed.) 1997. Participatory forestry inChina. International Academic Publishers,Beijing.


29

IntroductionThe dry and hot valley climatic type occursmainly in the southwest provinces of Yunnan,Sichuan and Guangxi. In Yunnan, this climatictype covers 13 000 km2, between latitudes 22.5°-28.0° N, and makes up 3.4% of the area of theprovince (Yang 1992). As the name suggests, itis a hot, dry climate providing harsh growingconditions for plants and the aridity is a criticalconstraint to tree planting and plantationestablishment.

Little forest vegetation remains in areas withthese climatic conditions. Forest land is oftenseriously degraded, has had its vegetation coverreduced to sparse shrubs or even bare land, andsevere soil erosion is common. Existing, limitedforests are degraded or seriously degraded withlow productivity, resulting from excessiveexploitation and unsustainable practices. Giventhe inadequate vegetation cover and serious landdegradation, environmental restoration in theareas will have to be achieved through theplanting of trees, shrubs and grass. Improvementof the deteriorating ecological environment andachieving economically and ecologicallysustainable development of forest lands is a greatchallenge to foresters/biologists, villagers, andlocal governments at different levels.

Research to identify tree and shrub speciesadapted to the harsh ecological conditions and

to develop effective planting technologies arein progress (Yu 1997, Yu et al. 1997, Zhou andZhang 1998, Li and Zeng 1999, Yang et al.1999). The research has focused on technicalinterventions, however, experiences show thatimproved technology practices alone are notsufficient to bring about an expansion of treeplanting and plantation establishment in thisarea. It is essential to understand why this is thecase and what other interventions and optionsare needed for effective reclamation of degradedlands and forests.

Our research was carried out in YuanmouCounty, Yunnan. Yuanmou has a typical dry andhot valley environment and areas of severelyeroded land. It has a mean annual temperatureof about 22°C; mean annual rainfall is only 630mm, while potential evaporation is nearly sixtimes the rainfall. Only 6% of Yuanmou’s landarea was covered with forests in 1993. The ma-jority of its designated forest land is covered withpoor quality bushes or grass and sometimes thereis no vegetation cover at all. Heavy rains in theshort wet season cause severe erosion of theunprotected soil and restoration of cover byplanting trees, shrubs and grasses is necessaryto address this problem.

This chapter presents research methods;describes biophysical and socioeconomic

Chapter Two

Reclaiming Degraded Forest Lands in the Dry,Hot Climate of Yuanmou County, Yunnan

Lai Yongqi1, Liao Shengxi1, Liu Dachang2, Liu Jun1, Su Jianrong1 and Zhang Yanping1

1 Research Institute of Resource Insects, Chinese Academy of Forestry, Kunming.2 Southwest Forestry University, Kunming P.R. China and Center for International Forestry Research, Bogor, Indonesia.Currently at The Mekong Institute, Khon Kaen, Thailand.

Authors are listed in alphabetical order.


30 Reclaiming Degraded Forest Lands in the Dry, Hot Climate of Yuanmou County, Yunnan

circumstances of the research site, providing asetting for the research; and examines the stateof forest land degradation. It discusses strategiesand interventions for reclaiming local seriousdegraded forest lands: incentives andenvironmental awareness of farmers, andappropriate species.

Research MethodsAs an action-oriented research project, thisresearch adopted a logical procedure fromdiagnosis through design to delivery. It used aparticipatory approach complemented by themore conventional inputs from expertconsultants. The participatory approachinvolved great attention to interacting with keystakeholders to develop the options andstrategies. This was a departure from top-downapproach used in traditional forest research andgovernance practices in China. Therefore, theresearch itself was an experiment withinnovative research methods for natural resourcemanagement and reclamation of degraded forestland.

Participatory ApproachVillagers, government officials and agriculturalextension agents and researchers in relevantdisciplines were actively encouraged to interactto analyse problems, to identify constraints andopportunities, and to develop appropriateoptions and strategies to address degradationproblems.

A number of PRA tools were used at thediagnosis stage, including secondary datacollection, household interviews, groupinterviews, and direct field observations. Inaddition, a participatory program of speciestesting was implemented at the delivery stage.

Secondary data were collected fromrelevant county government agencies such asthe Forestry Bureau, Statistics Bureau andMeteorological Bureau, from townshipgovernments, and from previous researchprojects and programs. The data includedbiophysical conditions and socioeconomicsettings such as climate, land uses, forest

resources, changes in forest tenure andmanagement, tree planting policies, role offorest in the local economy, demography, andextent of infrastructure development.

Household interviews were used to deter-mine villagers’ views on land degradation andtheir attitudes to tree planting and utilisation ofthe uncultivated land. A random sample of 33households was made and each household in-terview lasted 1.5-2 hours. The interviews wereconducted in the form of a semi-structured in-terview complemented by questionnaire survey.After a pre-testing phase, 20 questions cover-ing nine topics were used in the household in-terviews. The topics were (1) family size, la-bour, education and production activities, (2)land resources, (3) land use, (4) livestock andfodder, (5) household income, (6) householdexpenses for production, (7) household expensesfor non-production activities, (8) value of fixedassets, and (9) taxes and fees paid. We soughtto understand the level of interest in tree plant-ing, the perceived constraints to plantation es-tablishment and the incentives needed to encour-age planting.

Group interviews examined questions ofrelevance to the whole village. Five groups(village leaders, women, elders, youths, andprogressive villagers) were interviewed. Eachgroup consisted of 10-12 people and most hadnot been involved in household interviews. Eachgroup interview took 2.5-4.0 hours to complete.After pre-testing, 21 questions covering 10topics were discussed. The topics included thestate of forest land degradation, the views ofvillagers on two forms of tree planting in thearea (government chief model plantation andplanting by villagers), attitudes and constraintsto tree planting at the village level, andespecially the issues that were controversial orwhich individual households failed to explainadequately.

Direct field observations were made by agroup of two forest researchers and three farmerswho had good knowledge of their village andits lands. Findings from these observations wererecorded and maps were made. The data and


31Lai Yongqi, Liao Shengxi, Liu Dachang, Liu Jun, Su Jianrong and Zhang Yanping

maps cover a wide range of issues such as vil-lage boundaries, topography, land uses, irriga-tion, forest types, state of degraded lands, cropcalendar, transport, gender distribution of work,and indigenous knowledge.

A small program of field testing exotic treespecies not previously grown in the area wasinitiated based on the villagers’ view that theydid not have sufficient choice of well-adaptedspecies for profitable tree planting. Theprocess was participatory and collaborative.Local communities (villages) provided landfor field experiment and were responsible fortending seedlings. Yuanmou County ForestryBureau provided part of funding and made thenecessary arrangements. The research teamtook responsibility for obtaining tree seedsand provided technical support for nurseryestablishment and part of the funds. It alsotook advantage of the availability ofspecialists with international experience intree species selection and access to computer-based selection methods. Potential specieswere selected on the basis of their suitabilityfor reclamation and their potentialprofitability was assessed. A total of 15 exotictree and shrub species were introduced andtested for their adaptability to the dry and hotclimate and impoverished soils. Commitmentto the project varied between villages and thismay have influenced the maintenance andperformance of some trials.

Expert ConsultancyExpert consultancy, the usual approach toforestry problems in China, was used to helpwith the identification of options forreclamation. During the research process,consultations took place with Kunming-basedspecialists from the Research Institute ofResource Insects, Southwest Forestry College,Yunnan Academy of Forestry, YunnanInstitute of Forestry Planning and Design, andYunnan University. Local professional stafffrom Yuanmou County Forestry Bureau werealso consulted. Topics discussed included thestate of land degradation, technical and

socioeconomic constraints to tree plantationestablishment and strategies for successfulplantation development in Yuanmou.

Background Information onYuanmou County and LaofanVillageResearch was conducted at both county andvillage levels. Studies were undertaken in fourvillages: Laofan and Xiaocun villages of Zuolintownship, Moke village of Nengyi township, andBingyue village of Laocheng township. Thesevillages are situated either in the Yuanmou basinor the transitional zone from the basin to themountainous area.

Overall biophysical conditions and the socialsetting of Yuanmou county are described andLaofan village is used to illustrate village-levelconditions as there were similarities in climate,vegetation, land uses, socioeconomiccircumstances and land degradation in the fourvillages studied.

Yuanmou County

ClimateYuanmou county is a typical dry and hot valleyarea in central Yunnan and is located at 25°23’-26°06’N latitude and 101°35’-102°06’Elongitude (see Map 1.1 in Chapter 1). The meanannual temperature is 22°C, May is the hottestmonth with a mean temperature of 27°C, andthe coldest month is December with a meantemperature of 15°C. Mean annual rainfall is 629mm and annual evaporation is 3729 mm.Rainfall is not distributed evenly throughout theyear. Most rain falls during June to October. The7-month dry season, from November to May,averages less than 100 mm or 14% of the meanannual rainfall. (Table 2.1)

The dry climate makes successful treeplanting a challenging operation. It also makesnatural regeneration extremely difficult afterforests are harvested or cleared. The survivalrate of trees planted using planting technologiesthat are effective in other areas was very low inthis area during the period 1952-1988 (Yuanmou



County Forestry Bureau 1997). This in partexplains why forest cover is so limited and whyfew plantations have been successfullyestablished in Yuanmou.

Land useThe land area of Yuanmou is 201 970 ha. Twothirds of the area is designated as forest land;19% is food crop land; and 15% is grazing land,water bodies and settlement areas. In 1993 forestcovered only about 6% of Yuanmou’s area while47% of the land had poor quality bushes(Yuanmou County Forestry Bureau 1997).

Socioeconomic settingYuanmou County has 13 townships and had apopulation of 198 000 in 1998 with most peopledepending on agriculture for their livelihood.Agriculture, broadly defined as including cropproduction, animal husbandry, forestry andagricultural products processing, plays a veryimportant role in employment and incomegeneration of the rural population of Yuanmou.In 1998, agriculture contributed nearly 70% ofthe gross output value of society with the balanceto the economy contributed by themanufacturing, construction, transport, andtertiary industry sectors. Gross output value ofsociety is a measure of Chinese economy. It isthe sum of gross output value of agriculture,industry (manufacturing), construction,transport, and commerce, the five material

production sectors. The sum of gross outputvalue of agriculture and industry is called grossoutput value of industry and agriculture. Thegross output value of agriculture comprised:crop production 76%; animal husbandry about20%; and forestry slightly more than 3%.Although the forest contribution to the economyis usually systematically underestimated inChina, its very low contribution to the Yuanmoueconomy is in sharp contrast to the 66% of landarea allocated to the forestry sector in the county.

Winter vegetables, livestock and sugar caneare major sources of cash income for villagers,especially farmer households in the YuanmouBasin. Relatively high temperatures in wintermake vegetable cultivation possible andprofitable. People in Yuanmou began plantingwinter vegetables in 1978, and now these aremarketed in more than 150 cities throughoutChina and are the most important source of cashincome for many farmer households.

Forests contribute little to the materialwelfare of communities and villagers of theYuanmou Basin area. Alternative sources ofenergy such as coal and electricity have beensubstituted for the fuelwood on which villagerswere dependant for cooking and space heatingprior to the 1980s. Other profitable products,such as winter vegetables, have increasedpeoples’ income and enabled them to afford thenew types of energy. The railway from Kunming(capital city of Yunnan) to Chengdu (capital city

Table 2.1 Climatic data for Yuanmou (1956-1990)

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Year

Meantemperature 15.2 17.8 21.8 25.2 26.9 26.4 26.4 25.4 24.3 21.3 17.6 14.9 21.9(°C)

Rainfall 3 4 5 13 45 115 142 125 88 65 18 6 629(mm)

Evaporation 251 318 461 514 506 340 294 233 222 203 191 196 3729(mm)

Relative 44 37 32 34 44 60 66 70 67 66 61 53 53humidity (%)

Source: Yuanmou Meteorological Station (Yuanmou County Forestry Bureau 1997).



of Sichuan) goes through Yuanmou, and a goodroad system has been developed in the county.This infrastructure provides ready access tomarkets for local products. These factors havehad a major influence on the attitude of peopleto tree planting in the area.

Laofan Village

Land useThe land area of Laofan village is 115 ha. About50% of total area of the village is used for foodcrop production, including paddy fields,irrigated land and rain-fed land. However,farmers often do not cultivate food crops on therain-fed land because its productivity is low.Instead, they graze their animals on it. Plantationforest and denuded land make up about 40% ofthe village’s total area. The plantation forest isEucalyptus exserta that was first introduced tothis area in 1966. Farmers also graze theiranimals in the plantation and on the denudedland. Water bodies and residential areas makeup nearly 10% of total area. (Table 2.2)

Socioeconomic settingSituated in Yuanmou Basin, Laofan village is6.5 km from the Yuanmou County town. It is asmall village with a population of 441 (1995),comprising 108 households, and with 51% ofthe population female and 49% male.

Although it is not far from the county town,Laofan is a typical agricultural community, withcrop production and animal husbandryproviding primary employment and sources of

income. In 1995, the labour force of the villagewas 322 of which 93% was involved in agricul-ture. Labour age for rural women is 16-54 yearsold and for rural men 16-59 years old. Agricul-ture, broadly defined, contributed 94% of theannual gross output value of society of Laofan,while non-farm activities was only 6%. Withinthe agricultural sector, food grain and vegeta-ble production and animal husbandry are pri-mary contributors of the economy. Food grainand vegetable production contributed 67% ofgross output value of society and animalhusbandry 27%. The contribution of forestry togross output value of society was less than 1%.

A sealed road links Laofan to the county townand provides villagers with ready access to marketsfor vegetables and other agricultural produce.

Forest tenure and managementForest tenure and management in Laofan hasundergone similar radical changes to other partsof China (Liu 2001). Prior to the nationwideLand Reform Campaign in 1952, landlords(wealthy village households), religious groupsand clans owned forests in Laofan. All thesecategories of forests were confiscated during theLand Reform and redistributed equally to farmerhouseholds as private property. However,following the agricultural collectivisation - theestablishment of the advanced (agricultural)cooperatives in 1956 and the people’s communein 1958, forests became collective property andwere managed exclusively by the villagecollective until the early 1980s.

Table 2.2 Land use in Laofan Village

Paddy fields Rain-fed Plantation Denuded Water bodies Totaland irrigated land forest forest land and residentialland (huangshan) area

Area (mu*) 632 231 300 400 160 1723

% total area 36.7 13.4 17.4 23.2 9.3 100

Area head-1 (mu) 1.4 0.7 3.9

Source: field survey data.* 1 mu = 0.066 ha



Laofan experienced two periods of destruc-tive deforestation from the late 1950s to the early1980s. The iron and steel making drive in 1958-59 caused the first deforestation. At that time aconsiderable amount of wood was cut and usedto fuel numerous backyard furnaces. Thecollectivisation of forest resources made the wastepossible. The second deforestation happenedimmediately after collective agriculture wasreplaced by household-based agriculture (knownas the agricultural production responsibilitysystem) in the early 1980s. The replacement ledto a breakdown of the management system ofcollective forest. Forest remained a collectiveasset at that time but nobody was given authorityto manage it, especially to protect it from illegalcutting. Consequently, villagers cut down almostall the eucalypt trees that had been introducedand planted in Laofan in the mid-1960s. Thevillage leadership had to sell the remainingcollective trees to individual farmer householdsin the village in 1983.

As an area in the upper catchment of theYangtze River, Yuanmou is entitled to funds ofa large government-sponsored environmentrestoration program - the Yangtze ShelterbeltProgram (Conservation Forest on Upper andMiddle Reaches of Yangtze River) that wasinitiated by Ministry of Forestry in 1989 to planttrees and to protect existing forests. YuanmouCounty Forestry Bureau implemented theprogram in Laofan village. To encouragefarmers to plant trees, the village leadershipallocated collective forest land use rights equallyto individual households, and the ForestryBureau again promised the policy ‘whoeverplants the tree owns it’. A eucalypt plantationof 20 ha (300 mu, see Table 2.2) was establishedand it was the only plantation forest in Laofanwhen this research was conducted. The ForestBureau appointed a villager to guard theplantation and made regulations to protect it.Only branches can be harvested because theplantation was developed for soil erosion controland environmental services; tree harvest is notallowed unless a permit is obtained from theBureau, and grazing is prohibited in the

plantation. In practice, the farmers have ignoredthe grazing regulation.

Land Degradation Assessment

Yuanmou CountyOnly 6% of the land area of Yuanmou was cov-ered with forests in 1993. The existing naturalforests and plantation forests are degraded. Themajority (90%) of designated forest land haseither poor-quality bushes or no vegetation coverat all. The bush vegetation is degraded and thedenuded forest land is seriously degraded. Thissituation is a cause for concern and has both localand regional implications. It requires significantefforts for ecological and environmental resto-ration through planting trees, shrubs and grass.

Laofan VillageOur field surveys showed that much of the landin Laofan is degraded or seriously degraded,including the denuded forest land, plantationforest, and the rain-fed land (see Table 2.2above). The extent of degradation varied on thedifferent types of land but exceeded 50% of totalarea of the village.

Figure 2.1 Highly salinized land in a low lying area



The denuded forest land of Laofan is seri-ously degraded. It can be further divided intotwo sub-types: (1) land along a small stream thatis highly salinized (Figure 2.1) and (2) thedenuded forest land on which either only a fewplants of a small shrub, Dodonaea angustifolia,and grass occur or there is very little or novegetation cover. Heavy rains in the rainy seasonresult in severe erosion of the unprotected soiland there are numerous erosion gullies ofvarying length, width, and depth on the land(Figure 2.2 and Figure 2.3). Little organic mat-ter remains and fertility is very poor.

The eucalypt plantation is seriouslydegraded. Indicators are slow growth of trees,lack of understorey plants and continuingdevelopment of numerous soil erosion gullies(Figure 2.4). Although the eucalypt species inthe plantation is one of the most tolerant of poorsoil conditions, the stand is irregular and lacksvigour. Most species of Eucalyptus are lesseffective in controlling soil erosion in dry areas,because they provide very little shading of thesoil due to their open crowns and because theirroots compete very effectively with understoreyplants for the small amount of water available.This suggests that from perspective of soilerosion control and ecological benefits,appropriate species are those trees withcharacteristics such as dense spreading crownsand either deep rooting or low water use (Boland1997). This is also a lesson farmers in Laofanand beyond have now learned from their ownexperience. Although eucalypts may inhibitunderstorey development in dry areas,uncontrolled animal grazing exacerbates thesituation here and the erosion damage appearsto be increasing.

The rain-fed crop land is also degraded,though not as seriously as the denuded forestland and the eucalypt plantation. The degrada-tion is indicated by low productivity of the land.Farmers now either grow only sweet potatoes

Figure 2.4 Erosion damage appears to be increasingeven in the plantation area

Figures 2.2 and 2.3 Erosion of the unprotected soilresulting in numerous erosion gullies of varying length,width, and depth on the land

Figure 2.2

Figure 2.3



on this type of land, or in most cases no longercrop it. This is a type of degraded land to whichfuture reclamation efforts designed for dry andhot valleys in Yuanmou and other areas mustgive attention.

Land degradation is a deterioration of localagroecological conditions and has direct impactson local welfare. Land degradation in this areaalso contributes to the problems in YangtzeRiver. For example, soil erosion leads to siltingof the river, and the lack of forests and othervegetation increases the likelihood of disastrousfloods, such as those in 1999.

Clearly, there is a need to reclaim degradedlands in dry and hot valleys in Yuanmou andadjacent areas, both for improving localagroecological conditions and welfare and forenvironmental benefits and services in middleand lower reaches of Yangzte River. Hencereclamation of degraded lands in the area is ofnational significance.

Strategies and Interventionsfor Land ReclamationThis section discusses socioeconomic factorsand technical aspects of reclaiming degradedforest lands in the dry and hot climate zone byanalysing the ways in which tree planting isorganised currently in Yuanmou and drawinglessons and implications from our speciestesting. The purpose is identifying strategies anddistilling policy recommendations forreclamation efforts in the harsh ecologicalconditions.

Organisation of Tree PlantingPlantation establishment is organised in twoways in Yuanmou: (1) by villagers with orwithout government support, and (2) bygovernment, with the local government chieftaking overall responsibility. The secondcategory of establishment is called ‘governmentchief model plantation’ or ‘yangbanlin’.

Local communities and villagers have so farbeen the primary force of plantationdevelopment in Yuanmou, though governmentsat various levels have played a greater role sincethe late 1980s. Over the six years from 1989-1994, about 8400 ha of plantations wasestablished in the county, two thirds bycommunities and villagers, and the rest bycounty and township governments and villageleaders (Table 2.3).

Government chief model plantationA government chief model plantation is a highquality plantation established throughadministrative means by a given level ofgovernment using reforestation funds. Thegovernment chief at each level takes overallresponsibility for plantation establishment andmanagement. One purpose of ‘government chiefmodel plantations’ is ‘demonstration’ toconvince villagers that it is feasible to establishplantations successfully and increase forestcover in the harsh environment.

Table 2.3 Plantation establishment in Yuanmou (1989-1994)

Year Total planted Tree planting by villagers Government chief model plantationarea (ha)

Area (ha) % total Area (ha) % total

1989 796 656 82 140 181990 2096 1645 79 451 211991 1815 1403 77 412 231992 1046 299 30 747 701993 1658 1217 73 441 271994 1004 412 41 592 59Total 8415 5632 67 2783 33

Source: Yuanmou County Forestry Bureau (1997).



In Yuanmou, government chief model plan-tation establishment started in 1989. The Countygovernment chief, township government chiefsand village leaders must develop their modelplantations respectively. Their success in estab-lishing plantations is an indicator used to assesstheir performance and is used as a reference forpromotion. The chief model plantation is a strat-egy designed to effectively establish plantationsthrough better use of available funds, and clearand specific responsibility.

Responsibilities of each party involved areclearly set out. The government chief takesoverall responsibility for plantationestablishment, including selecting planting sites,making this land available, ensuring no conflictover the lands; coordinating villagers whoselands are used with the forest department andrelated agencies; making funds available; andmonitoring results of the tree planting. Forestdepartments take specific responsibilities forplantation establishment, including managingfunds, selecting sites and tree species, makingoperational schemes, organising planting,protecting trees planted, and initial evaluationof plantation development. Likewise, theperformance of participating forest officials inplantation development is recorded and used asa reference in promotion.

Lands for the establishment of governmentchief model plantation belong to villagers.Hence, by policy, trees planted are the propertyof villagers who own lands. However, the rightsof villagers to trees/plantations are very limited:Only branch lopping is allowed because suchplantations are developed largely for ecologicalservices. Villagers who contribute lands aregiven priority in planting, or tending andguarding as paid labour.

Capital investment in the reforestation isfrom three sources: local government budget;reforestation funds from central, provincial andprefectural governments; and contributions bypublic sector staff as their obligation of“Voluntary Tree Planting”. A policy made byChuxiong Prefecture that governs YuanmouCounty, requires the county government to

spend 1-2% of its yearly budget on forestry.Yuanmou received reforestation funds totalling1.6 million Yuan annually from central, provin-cial and prefecture governments during the pe-riod from 1993-1996. Contributions by publicsector employees are also an important invest-ment source of government chief model planta-tion establishment. Initially, people must planttrees in person to fulfil their obligation of “Vol-untary Tree Planting”, a requirement by law forall adult citizens. Later in Yuanmou, the require-ment was adapted to allow cash contributionsrather than planting trees in person. Each pub-lic sector employee contributes 20 Yuan a yearfor tree planting.

The amount of capital spending varies acrossdifferent levels of government chief modelplantations. Planned investment is 1500-1800Yuan ha-1 for model plantation of countygovernment chief; 1200-1500 Yuan ha-1 formodel plantation of a township governmentchief; and 750-1200 Yuan ha-1 for modelplantation of a village leader. However, actualinput in county chief model plantations in 1998was 2700 Yuan ha-1. About 20% of capitalinvestment is used for seeds and seedlings, 50%for labour for land preparation and tree plantingand the remainder for fertilisers and plantationtending and guarding.

Tending after planting is essential forsuccessful tree survival and growth in the localhot, dry climate. Recognising this, the forestdepartment recruits forest guards (fromvillagers) to tend and protect plantations forthree years after planting or five years inextremely difficult areas. Subsequently, localcommunities are responsible for plantationmanagement.

Experience over the last decade shows thatthis form of plantation development is effective.The approach was helpful in obtaining necessaryfinancial inputs, using them effectively andgenerating a comprehensive tree planting effort.As a result, the government chief modelplantation is an important way of organisingplantation development in Yuanmou. Over theperiod of 1989-1994, the area of chief model



plantations accounted for one third of total areaof plantations established. In two of the years,the ratio of chief model plantations to total plan-tations established in the year rose to more than50% (Table 2.3). This form of plantation devel-opment significantly improved the quality of treeplanting and increased seedling survival rate.Survival rates increased from about 50% (1989-1992) to over 80% (Yuanmou County ForestryBureau 1997).

Nevertheless, the government chief modelplantation is not sufficient as it mainly involvesgovernment effort. Government funds arelimited and government alone is unlikely toreclaim all the denuded and degraded forestlands that make up about 60% of the total areaof Yuanmou. Without complementary efforts bycommunities and villagers, there is littlelikelihood of revegetating Yuanmou andproviding environment services locally and toother areas along the Yangtze River.

Incentives are essential to encourage activeinvolvement of local communities and villagers,but current policies do not provide enough incen-tives for villagers to plant trees, especially timberspecies. Although villagers who contribute landsto the development of government chief modelplantations are owners of trees planted but theyhave very limited rights to cut trees. Villagers havelittle incentive to become involved in the devel-opment of timber plantations, as timber harvest-ing is highly regulated and they can expect littlereturn from their labour and capital investment.While strict restrictions on logging (timber har-vesting) may be necessary for the environmentalservices for Yuanmou and other areas of the Yang-tze watershed, there is concern that compensationbe paid to local communities and villagers whoprovide such services through their trees and plan-tations. The ongoing countrywide governmentprogram of ‘food for trees and grass’ (tuigenhuanlin1 in Chinese) does encourage farmers toplant trees. The program is progressing well, withfarmers asking a greater quota of tree planting thanthe government is providing.

Species choice is another issue that needs tobe addressed in government chief model plan-

tations. More than 10 species have been used inthese plantations in Yuanmou, including speciesof Eucalyptus, Acacia, Albizia, Leucaena, Pinus,Ziziphus oenoplia and Cajanus cajan. Timberspecies are dominant and most plantations aremonocultures. Large-scale, highly intensive,plantation forests are usually grown inmonocultures to maximise production of a de-fined product such as pulpwood. However, therisk factor in smaller-scale plantings by farm-ers and communities, which often have less cer-tain markets, can be reduced by using more thanone species and producing multiple products.The implication here is that more attention mustbe given to multipurpose trees when selectingspecies for planting. This way, villagers canexpect material benefits from plantation forestsof multipurpose species that provide environ-mental services and economic products, a pointthat is well demonstrated widely else where. Inparts of southern China Acacia mearnsii (blackwattle) has been planted on degraded lands. Thisnitrogen-fixing tree provides ground cover andcommercial products such as tannins, poles andtimber (Ho and Fang 1997). Thoughmonoculture plantations of multipurpose spe-cies with dense spreading crowns and deep root-ing can also provide protective cover of soil,there are other benefits for using mixed speciesplantings for soil conservation and rehabilita-tion plantings. For example, mixed plantingsmay provide more habitats and encouragegreater biodiversity than monocultures. Mix-tures including a nitrogen-fixing species mayhave better nutrient cycling than pure stands(Khanna 1997), and increased growth and pro-ductivity (DeBell et al. 1997).

Tree planting by villagersVillagers and communities have beenresponsible for the majority of plantationsestablished in Yuanmou. Even thoughgovernment chief model plantations wereinitiated in 1989, plantations developed byvillagers accounted for two thirds of totalplantation area in the years from 1989 to 1994(Table 2.3). Villagers established small woodlots



on denuded forest lands and on rain-fed croplands and were responsible for choosing thespecies and planting methods and for decidingthe extent of planting.

However, the area planted by villagers hasbeen far less than it could be and what is needed.Our surveys found four factors affectingplantation development by villagers. First andmost important is the disincentive for them toplant trees compared to other crops. Besidesrestrictions on their rights to harvest and markettimber products as discussed earlier, villagersin Yuanmou have more profitable investmentopportunities. Since the late 1970s, vegetablegrowing in winter has been a profitableenterprise and vegetables have been a primarysource of farmers’ cash income. Moreover, veg-etable growing can generate income within a

few months while returns from tree plantingcannot be expected in at least a few years. Also,it is not easy to grow trees profitably due to thelow productivity of denuded lands and excessivegovernment taxes and charges. While it is noteasy to change the productivity of the land,changes in policies on rights to harvest andmarket timber and a more favourable approachto taxation and charges could provide a greaterincentive for tree planting.

The second factor is the lack of capital fortree planting. The availability of funds is animportant feature contributing to the high qualityof government chief model plantations. Capitalinput was 2700 Yuan ha-1 for county chiefplantations in 1998. In sharp contrast,reforestation funds provided by the forestdepartment to villagers was only 300 Yuan ha-1.It is unfortunate that since 1989 most villagershave received little or no funding support fortree planting as almost all public funds allocatedfor tree planting have been used to developgovernment chief model plantations. Thissuggests a necessity to find a mechanism toovercome the funding constraints villagers facefor plantation establishment.

Thirdly, it is essential to identify sufficienttree species well-adapted to local ecologicalconditions that can be grown profitably.

Finally, villagers do not value tree plantingfor environmental protection. Villagers inYuanmou Basin and the transition zone now relylittle on forests as they have alternatives sourcesof fuel and building materials. They are not keento plant trees when they can expect little materialbenefit from the forest. Our surveys found thatin four study villages, most villagers did notconsider it in their interest to develop plantationforests for environmental services. They did notrecognise direct benefits from improvedagroecological conditions and reduced soilerosion that might result from revegetating thedegraded land. This finding implies a need todemonstrate potential direct benefits ofenvironmental restoration to villagers and localcommunities.

Food for Trees and Grass Program(tuigen huanlin)

The program started in 1999 after thedisastrous Yangtze floods in 1998. Farmersof upper catchment of the Yangtze River andupper and middle catchments of the YellowRiver are required to stop cultivating foodcrops and to plant trees and grass instead ontheir crop lands on steep slopes. To encouragethem to do so, the central government providesthem with food grain, cash subsidy andfunding for tree seedlings. Every year, farmersget 2.25 t of grain ha-1 once they have done soin the upper catchment of the Yangtze, and1.5 t ha-1 in the upper and middle catchmentsof the Yellow River; and cash subsidy of 300Yuan ha-1. They get 750 Yuan ha-1 for seedlingsat beginning of planting. The governmentprovides grain and cash subsidy for eight yearsin the case of plantation forest developedlargely for ecological services and five yearsin the case of NWFP plantations. In the lattercase, the hypothesis is that farmers will be ableto generate income from their plantations infive years.



Species SelectionIdentification of a sufficient number of specieswell-adapted to the dry and hot climate and ableto be grown profitably is one of the keys torestoring vegetative cover and rehabilitatingdegraded forest lands in Yuanmou. So animportant component of this research was toidentify additional species to those that arecurrently planted in the area.

Introduction of plant species to Yuanmou canbe traced back to the 1920s. Considerable workon introduction of species started in 1958 andabout 110 species were tested during 1958-1992(Yuanmou County Forestry Bureau 1997). Mostof the species came from other parts of Chinabut some were from other countries. Theyincluded both timber species and species fornon-wood products. The efforts of more thanfour decades did identify some well-adaptedspecies. However, with change in economicconditions and markets, some of these speciesare no longer profitable to grow in plantations.

Our research in Yuanmou introduced 15 newspecies to expand the range of species availablefor selection by farmers and to understand whatspecies farmers want and why, because onlythose species accepted widely by farmers areappropriate species. Species tested includedAzadirachta indica, Acacia spp., Eucalyptusspp., and Casuarina sp.

Experiments, demonstration andextension of neemNeem (Azadirachta indica) is a small tomedium-sized tree, usually about 15 m tall, witha spreading evergreen crown. It is thought tohave originated in the Assam-Myanmar regionand is distributed throughout the Indiansubcontinent. It has been cultivated for manyyears in Indonesia, Malaysia and Thailand. Morerecently it has been grown as an exotic in manyparts of the world. It is a multipurpose treeadapted to poor soils and hot, dry tropicalconditions. It grows best where there is an annualrainfall of 1000 mm but will tolerate as low as400 mm. It is grown for shade and shelter, timberand fuel, and to control erosion. In India, oil

from the seed is used for soap making but iswell known for its medicinal and insecticidalproperties. The species is fully described byAhmed and Idris (1997).

It was introduced into Yuanmou in 1995 atbeginning of this research project. Seedlingswere raised and planted in around village andhomesteads, and on the denuded forest lands atLaofan during the rainy season of 1996. A totalof 350 seedlings were planted. In August 1997,after one year of planting, the height of youngtrees averaged 2.6m, with a maximum of 3.5m;and diameter at ground level averaged 3.8 cm,with a maximum of 6.2 cm. In August 1998,after two years of planting, the height of youngtrees averaged 3.8 m, with a maximum of 5.9m; diameter at breast height averaged 6.4 cm,with a maximum of 10.2 cm; and crown sizeaveraged 2.0 x 2.0 m, with a maximum of 2.9 mx 3.0 m. It was also in the year that the treesflowered and fruited.

Four additional Azadirachta provenances(seed sources) were introduced to Yuanmoufrom Myanmar and Africa in 1997. Experimentswith these were made in Xiaocun village and atotal of about 0.5 ha was planted. A survey afterthree months of planting found that survival ratewas over 98%. Experiments with neem havenow been carried out in other counties of Yunnanprovince such as Yuanjiang and Jinggu and showsimilar encouraging results.

These experiments demonstrated the highpotential of this species for planting in Yuanmouand adjacent areas with similar ecologicalconditions. Of particular interest currently is theoil that has many biologically active compounds.Recognising the great market potential, severalprocessing enterprises have entered intocontracts with farmers in Yuanmou to purchasefresh Azadirachta seeds at prices of 1Yuan kg-1.The economic potential of neem has alsoattracted considerable attention fromgovernments and farmers. Governments atdifferent levels are now encouraging farmers toplant these species and villagers themselves areenthusiastic.



Azadirachta indica could be the mostsuccessful introduced species in Yuanmou. Eightplantations, with a total area of 37 ha, weredeveloped in Yuanmou in 1998. Of these, 17 hawere established with government capitalinvestment, while 5 plantations totalling 20 hawere established exclusively with farmers’money. This is in sharp contrast to the attitudeof farmers to other introduced species. Thedriving force of the rapid expansion of neemplantation is the economic potential of the neemtree products, especially the oil. The lesson fromthis experience is that tree species must be bothecologically suitable and economicallyprofitable to generate enthusiasm among farmersto invest their time and money to plant them.Planting for environmental improvementreasons alone is not sufficient incentive to attractfarmers’ interest.

Testing other speciesAustralian tree species have been successful asexotics in many parts of the world. Many aremultipurpose species adapted to harsh tropicalenvironments (Doran and Turnbull 1997). Usingcomputer-assisted climatic analysis (Yang et al.1996) the climate of Yuanmou was comparedwith that of Australia to identify species with ahigh probability of being adapted to the hot anddry local conditions.

Ten species, mostly from central Queensland,were identified. Some of the species have alreadyshown potential in dry areas in Vietnam, parts ofWest Africa and elsewhere. Seeds were providedby the CSIRO Australian Tree Seed Centre,Canberra and the experiment was established inXiaocun in 1998. The species are: Acaciadifficilis, A. holosericea, A. neriifolia, A.plectocarpa, A. torulosa, A. tumida, Casuarinacunninghamiana, Eucalyptus camaldulensis var.obtusa, E. exserta and E. punctata.

Initial observations show that performancevaried across these species. Eight species showinitial positive results, with the survival ratesand average height of seedlings comparable withthose of endemic species such as Albizia kalkoraand Phyllanthus emblica. Two species, Acacia

torulosa and Acacia tumida, failed in the nursery,probably because local nursery techniques werenot suitable for them. Some of these species fixatmospheric nitrogen and could be planted inassociation with neem, especially on highlydegraded soils. Acacia holosericea is alreadybeing used in mixed plantings with eucalypts inGuangdong province, China. Villagers stated thatno vegetation survived under eucalypts and thisis probably related to competition for water inthe dry environment. There would be lesscompetition if eucalypts were planted at widerspacing and this might permit acacias to beinterplanted and provide better ground cover anda potential source of animal fodder.

ConclusionsImprovement in the ecological environment andpeople’s welfare through planting trees in theseriously degraded forest lands with a dry andhot valley climate in southwest China is a greatchallenge. It needs the combined efforts of localcommunities and villagers, governments atdifferent levels, and the academic community.

Our research highlighted what others havebeen finding in other regions and counties,namely that both socioeconomic factors andtechnical aspects must be addressed toencourage active involvement of localcommunities and villagers in this task. Theformer includes economic returns villagers canexpect from tree planting, removal ofgovernment disincentives, increasedenvironmental awareness among villagers, andthe availability of capital for investment inreforestation. Appropriate technologies includeidentification of appropriate species and thedevelopment of planting technologies effectivein the harsh climatic and soil conditions.

Economic IncentivesReasonable economic returns are essential toencourage active involvement of villagers. Suchreturns can be achieved through several pathways.One is the adoption of multipurpose tree speciesfor non-wood products in plantations becausesuch plantations can generate material benefits



without harvesting the whole tree thusmaintaining environmental services. The rapiddiffusion of neem in Yuanmou and othercounties in Yunnan demonstrates that farmersare willing to plant trees if they can expecteconomic returns from their investments.Another possible incentive would be to haveappropriate compensation for villagers for theenvironmental services their plantation forestsprovide. Government policies can also have asignificant influence on the profitability offorestry plantations.

Government PoliciesWhile timber harvesting has to be controlledstrictly in the upper reaches of the Yangtze toreduce erosion and flooding, it is a disincentivefor farmers to invest in tree plantations. Yin(1994) recognised the disincentives ofgovernmental market controls and pricedistortions and such disincentives were evidentin Yuanmou. Changes in government policieson rights to harvest and market timber and amore favourable approach to taxation andcharges could provide a greater incentive for treeplanting. The ongoing countrywide governmentprogram of “food for trees and grass” is anexample of a positive approach to the problem.An expansion of the program in size or adevelopment in similar programs is required.

Availability of CapitalOne constraint that needs to be addressed isinsufficiency of capital for investment inplantation establishment. Currently villagers useany surplus income for fixed assets, such ashouse construction, television, or to pay formarriages. Investment capital will be requiredto encourage villagers to develop qualityplantations. Government reforestation fundshave been limited, and most of them have beenused to develop government chief plantationswith little support to villagers. Investment bygovernment in plantation establishment is publicspending and should be used to develop forestswith environmental services as the primary ob-jective. Investment by villagers should be used

to establish profitable multipurpose plantations.The case of rapid adoption by villagers of neemtrees shows that provided the economic incen-tives are there the more wealthy farmers canovercome capital shortage.

Appropriate TechnologiesTechnical difficulties that need to be overcomeinclude identification of sufficient appropriatespecies to provide a choice of options and thegeneration of economical and effective plantingtechnologies for the dry and hot area.Appropriate species means species that are well-adapted to local climatic and soil conditions andhave characteristics that help control soilerosion, such as dense spreading crowns andeither deep rooting or low water use, and arealso profitable to grow. The neem tree appearsto be a good example of the type of tree that isrequired. Technologies that are effective butcostly inhibit widespread adoption, as shown byplanting techniques developed by several otherresearch projects.

A participatory and collaborative approachto testing species and silvicultural techniqueshas advantages including the rapid uptake ofappropriate technologies. This has implicationsfor the design of future projects of landreclamation and forest rehabilitation.

Environmental AwarenessThere is a need for environmental education forvillagers to build and raise their awareness ofenvironmental issues. This implies a need forefforts to demonstrate potential direct benefitsof environmental restoration to villagers andlocal communities. They may participateactively in developing plantations if they areconvinced that environment restoration throughtree planting is not only beneficial to others butalso to themselves. Locally this may meancontrolling severe soil erosion, preventingsalinization of soil and improving land fertilityand productivity.



ReferencesAhmed, S. and Idris, S. 1997. Azadirachta

indica A.H.L.Juss. In: Faridah Hanum, I.and van der Maesen, L.J.G. (eds.). Auxiliaryplants. Plant Resources of South-East AsiaNo. 11. 71-76. Backhuys Publishers,Leiden.

Boland, D.J. 1997. Selection of species andprovenances for planting. In: Doran J.C. andTurnbull, J.W. (eds.). Australian trees andshrubs: species for land rehabilitation andfarm planting in the tropics. ACIARMonograph No. 24: 39-58. AustralianCentre for International AgricultureResearch, Canberra.

DeBell, D.S., Cole, T.G. and Whitesell, C.D.1997. Growth, development, and yield inpure and mixed stands of Eucalyptus andAlbizia. Forest Science 43: 286-298.

Doran, J.C. and Turnbull, J.W. (eds). 1997.Australian trees and shrubs: species for landrehabilitation and farm planting in thetropics. ACIAR Monograph No. 24.Australian Centre for InternationalAgriculture Research, Canberra.

Ho, C.K. and Fang, Y.I. 1997. Development ofblack wattle (Acacia mearnsii De Wild.)plantations in China. In: Brown, A.G. andHo, C.K. (eds.). Black wattle and itsutilisation. 83-88. Rural Industries Researchand Development Corporation, Canberra.

Khanna, P.K. (1997). Comparison of growthand nutrition of young monocultures andmixed stands of Eucalyptus globulus andAcacia mearnsii. Forest Ecology andManagement 94: 105-113.

Li, K. and Zeng, J. 1999. [A] Comparison ofwater physiological characteristics of 9planting species in Yuanmou dry and hotvalley areas. Yunnan Forestry Science andTechnology, 1999 No. 1: 70-74. (in Chinese)

Liu, D. 2001. Tenure and management of non-state forests in China since 1950.Environmental History 6: 239-263.

Yang, H., Booth, T.H. and Zuo, H. 1996.GREEN - a climatic mapping program forChina and its use in forestry. In: Booth, T.H.(ed.) Matching trees and sites. ACIARProceedings No. 63: 24-29. AustralianCentre for International AgricultureResearch, Canberra.

Yang, M. 1992. Classification of dry and hotvalleys in Yunnan. In: Zhao, Junchen et al.,(eds.) Preliminary exploration of economicsof dry and hot valley. 123-131. ChineseEconomic and Culture Publication Co.,Hong Kong. (In Chinese)

Yang, Z., Zhuang, Z., Qin, D., Ran, G., Fu, W.1999. Afforestation techniques for waterconservation in arid-hot valleys of YuanmouCounty. Bulletin of Soil and WaterConservation 19: 38-42. (In Chinese)

Yin, R. 1994. China’s rural forestry since 1949.Journal of World Forest ResourceManagement 7: 73-100.

Yu, L. 1997. A study for the selection ofafforestation species for the dry-hot rivervalley in Yuanmou. Journal of SouthwestForestry College 17 (2): 49-54. (in Chinese)

Yu, L., Shu, Q., Gao, J. 1997. A study onintroduction experiment of afforestationspecies for vegetation recovery in dry-hotriver valley of Yuanmou. Journal ofSouthwest Forestry College 17 (2): 25-29.(In Chinese)

Yuanmou County Bureau of Forestry. 1997.Yuanmou County Annals of Forestry,Kunming. (In Chinese)

Zhou, J. and Zhang, M. 1998. Study onquantitative selection of drought-resistantand heat-resistant afforestation species inYuanmou. Yunnan Forestry Science andTechnology 3: 32-36. (In Chinese)


IntroductionEvergreen broadleaf forest is typical vegetationin China’s extensive subtropical zone in mosthilly and mountainous areas to the south of theYangtze River. The majority of these forests aresecondary forests and most have been degradedto a greater or lesser extent. Only a few primarynatural forests exist in isolated locations(Ministry of Forestry 1996).

Degraded evergreen broadleaf forest ischaracterised by a reduced number of treespecies and much lower biomass than normalforest. In the forest there is a loss of biodiversityand off-site effects of forest degradation includeincreased flooding, lower water quality andsiltation of dams. These forests have receivedconsiderable attention from foresters and localcommunities but there is limited understandingof the extent and causes of their degradation andits impact on people’s welfare. Few effectivestrategies or solutions, especially overallpackages, for rehabilitation of these forests havebeen identified. Nevertheless it is clear thatdegradation of such forests is continuing and ifleft unchecked will have serious consequencesfor people depending on forests for theirlivelihoods and for the environment in general.Part of the problem has been failure to recognisethat technical ‘fixes’ alone are not sufficient and

more innovative multidisciplinary approacheswith participation of stakeholders are required.

Our research was carried out Shanxin village,Cangwu County, in eastern Guangxi. Evergreenbroadleaf forests comprise two thirds of forestarea of the village and most of the remainder ismixed coniferous and broadleaf stands. For over100 years the broadleaf forests of this and othervillages in the area have been managed largelyfor fuelwood and charcoal. These products wereconsumed on farm and exported to the PearlRiver Delta and Hong Kong to provide animportant source of cash income. Now 75% ofthe broadleaf forest is degraded and 10% hasdeteriorated into bush-covered land andgrassland. Forest degradation appears to haveresulted in significant decline in forest outputand to threaten local livelihoods andenvironmental services.

A participatory approach was used toexamine critically the extent and immediate andunderlying causes of degradation of evergreenbroadleaf forest in one village to achieve animproved understanding of the issues. Then weaimed to identify effective strategies/solutionsto reduce degradation and to test some of thesolutions in the village’s forests through actionresearch.

Chapter Three

Degradation and Rehabilitation of EvergreenBroadleaf Forest in Cangwu County, Guangxi

Liu Dachang1, Weng Qijie2, Zeng Jie2, Zheng Haishui2 and Zhou Zaizhi2

1 Southwest Forestry University, Kunming P.R. China and Center for International Forestry Research, Bogor, Indonesia.Currently at The Mekong Institute, Khon Kaen, Thailand.2 Research Institute of Tropical Forestry, Chinese Academy of Forestry, Guangzhou.



46 Degradation and Rehabilitation of Evergreen Broadleaf Forest in Cangwu County, Guangxi

This chapter presents results of our re-search. It briefly presents research methods,describes biophysical and socio-economicconditions of the village, and examines theextent and causes of degradation of evergreenbroadleaf forest. It discusses disincentives tomanagement of broadleaf forest for timber andother wood products; and identifies strategiesfor forest rehabilitation and analyses theireffectiveness based on results from testingthem at the study site.

Research MethodsAs part of a multi-location research project, theresearch in Shanxin village used similar methodsto those in the other study sites (Chapter 1). Thisrequired a participatory approach combined withthe logical procedure of diagnosis of theproblem, design of possible solutions and finallytesting of proposed solution through actionresearch.

Participatory ApproachParticipation is now widely advocated andaccepted as a strategy for rural development andnatural resources management. It is referred toas ‘Participatory Rural Appraisal’ (PRA)(Chambers 1994a, b, c, Chambers and Guijt1995). The participatory approach recognisesthat it is critical to conduct research in the localsocial context where differences in land, labourand markets can radically change theeffectiveness of proposed technologicalsolutions to forestry problems. In this research,participation involved farmers, their families,researchers and local government officials toaddress forestry problems. It is an approach thathas rarely been used in China.

Farmer participationIn contrast to a ‘top down’ approach, farmerinvolvement in diagnosis, design and deliverywas particularly stressed in this research. Morethan 100 households representing nearly 20%of households in Shanxin and a number ofleaders of village and village household groups

who are also farmers were consulted during di-agnosis stage. The overall aim of the survey wasto identify factors responsible for the forest deg-radation, to canvass possible solutions and tofind acceptable ways of implementing actionresearch to test potential rehabilitation strategies.The survey was undertaken in two periods ofabout one month duration. Survey methodsincluded common PRA tools such as secondarydata collection, farmer household interviews andgroup interviews/meetings. In the first survey54 households were consulted and 52households during the second survey. Thesehouseholds were chosen through randomsampling from village household groups atdifferent levels of economic development.Specifically, the 54 sampling households wereselected from two wealthier village householdgroups, four middle-income groups, and twopoor groups. Households were consulted largelythrough semi-structured interviewscomplemented by a questionnaire survey.During the first survey period, three groupmeetings were organised, each with the villageleaders, the leaders of village household groups,and representatives of farmers.

Views and opinions were listened to andrespected throughout the research process. Forexample, initially the researchers proposed ashareholding system when discussing naturalregeneration through mountain closure(fengshan-yulin in Chinese) and reclamation ofa small area of sloping land. However, farmersrejected the proposal. After much debate, thefarmers proposed and adopted a form ofcommon management involving no changes intenure. Similarly, the researchers worked withfarmers to develop options but farmers decidedwhether to adopt them or not.

Involvement of researchersMore than 10 researchers and extensionpersonnel were involved. They were from thenational Research Institute of Tropical Forestryand South China University of Agriculture,based in Guangzhou, and from local government


47Liu Dachang, Weng Qijie, Zeng Jie, Zheng Haishui and Zhou Zaizhi

services such as agricultural extension, forestbureau, and scientific and technical commission.Researchers and extension staff played an im-portant role in identifying appropriate options/strategies, and providing necessary extensionservices and information. This was extremelyimportant in this isolated mountainous areawhere farmers have poor access to informationand extension services. Participation of research-ers helped improve communication betweengovernment officials and the farmers.

Participation of government officialsOver 20 government officials from sixgovernment agencies, including county bureausof agriculture, forestry, and science andtechnology, and from township governmentparticipated in this research. Involvement ofgovernment officials helped facilitate successfulimplementation of strategies identified becausetheir participation made a significantcontribution to the raising of investment fundsand to conflict resolution.

Action ResearchThe research in Shanxin adopted a logicalproject procedure of diagnosis, design anddelivery. The researchers worked together withvillagers and government officials to achieve agood understanding of local forest resources andto identify the state and causes of forestdegradation and opportunities and constraintsto forest rehabilitation. On the basis of this, arange of activities was identified to address theproblems of rehabilitation of degraded forests.These activities were either direct, such as themountain closure to promote naturalregeneration, or indirect, such as planting treesto provide non-wood forest products (NWFPs)and developing yard activities to provide sourcesof incomes that did not involve harvesting ofproducts from the broadleaved forests. Theseactivities were then tested in and around thevillage to determine their appropriateness andeffectiveness, which is action research or thedelivery part of the research process.

Conventional Forestry ResearchMethods

Forest vegetation surveyConventional methods for forest vegetationsurvey were used to understand plant species,forest resources, and the extent of forestdegradation. The survey team consisted of fiveresearchers and two villagers. It was carried outin sample plots measuring 20 m x 20-50 m. Aplant inventory was made to documentbiodiversity, and the extent of forest degrada-tion was assessed using indicators such asfuelwood output and standing biomass of forestover time. To achieve an understanding of for-est degradation a comparison between the cur-rent state of forest and past forest conditions wasmade. This was based on available research dataand recall of senior villagers. For the same rea-son, a comparison between forest in Shanxin andprimary forest in a nearby reserve was alsomade.

Soil surveysThe survey team undertook soil surveys indifferent types of forests and in forests degradedto a different extent of to understand relationbetween changes in soil properties and forestdegradation. Type, texture, depth and fertilityof soil and its capacity for holding water andfertility were observed and measured.

Background Informationon Shanxin Village,Cangwu County

Location, Terrain and ClimateShanxin village is an administrative villagegoverned by Cangwu County It is located at23°19’31”-23°26’50”N latitude, 111°40’41”-111o50’00”E longitude in eastern Guangxi (Map1.1 in Chapter 1). It is close to towns and cities,being only 30 km from Wuzhou city and 45 kmfrom Cangwu County town.

The terrain is predominantly hilly with analtitudinal range of 55-300 m above sea levelbut mainly 110-250 m. The mountains are steep,



with slopes of 25-35 degrees in general and upto 45 degrees. This rugged topography makesfood cropping difficult and less productive. Theclimate is warm, rainfall abundant, and summeris much longer than winter. For the period 1959-1992, mean annual temperature was 21°C, andmean annual rainfall 1516 mm. Most of therainfall (80%) comes in spring and summer, andautumn and winter are very dry. Although thelands and climate are quite favourable for treegrowth, the hilly topography increases the costof logging and wood transport.

Forest ResourcesShanxin has an area of 2940 ha. Data in Table3.1 shows that 83% of the area is forest whereascropland is only 4%. About 13% of the area hasother uses, such as grazing and habitation.Primary forests in Shanxin were naturalevergreen broadleaf forests. Now, due to long-term over-harvesting and other humaninterventions, secondary natural forest hasbecome the most common type of vegetation inthe village and adjacent areas. Currently, theforests can be grouped into three broadcategories: evergreen broadleaf forest, mixedconiferous and broadleaf forest, and NWFPplantation (Table 3.1).

Evergreen broadleaf forestEvergreen broadleaf forest is common inGuangxi and is the typical vegetation insubtropical areas of several other provinces in

south China. In Shanxin it comprises two thirdsof forest area (Table 3.1). This natural foresthas become less rich in species compositionbut still includes many species. Our surveysindicate that the most common tree species inthis forest type are: Castanopsis hickelii,Castanopsis fissa, Castanopsis fargesii,Lithocarpus fenzeliana, Quercus griffithii,Cinnamomum camphora, Sassafras tsumu,Michelia macclurei, Schima superba,Schefflera octophylla, Cratoxylum ligustrinum,Altingia chinensis, Madhuca subquincuncialis,Rhodomyrus tomentosa, Phyllanthus emblicaand Melastoma candidum. Miscanthusflorodulu and Pteridium aquilinum are the mostcommon herb species.

Evergreen broadleaf forest in Shanxin hasbeen managed mainly for fuelwood and woodfor making charcoal, although it also providesmaterials for farm tools and NWFPs. Until thelate 1970s, some fuelwood and charcoal wereconsumed on farm and the remainder exportedto cities in the Pearl River Delta and Hong Kong.Demand for these products has graduallydiminished due to the availability of alternativeenergy sources such as electricity and natural gas.However, a new demand for wood has appeared.Many wood processing enterprises, especially theWuzhou Wood Factory, have been establishedsince the 1980s, and these enterprises needsignificant amounts of small-diameter wood. Inaddition, fuelwood is still the primary source ofenergy for local rural communities.

Table 3.1. Land use in Shanxin village

Types of land Area (ha) % Total % Total forest area

Total land area 2940 100Total forest 2446 83 100 Evergreen broadleaf forest 1462 67 Mixed forest 728 30 NWFP plantations 76 3Crop land 103 4Other (grazing land, settlement area) 391 13

Source: Cangwu County Forestry Bureau and Guangxi Institute of Forestry Inventory and Planning.



Mixed forestMixed forest of conifers and broadleaf trees ac-counts for 30% of the forest area of Shanxin (Ta-ble 3.1). Dominant species in this type of forestare Masson pine (Pinus massoniana) and somebroadleaf tree species. Masson pine is natural andsecondary in origin. There are also plantations ofChinese fir (Cunninghamia lanceolata) but theyare very limited in extent comprising less than1% of forest area. The mixed forests are man-aged mainly for timber production, although resinis tapped from the Masson pine.

NWFP plantationsThese plantations are limited in area and makeup only 3% of the forest (Table 3.1). They aremainly plantations of Cinnamomum cassia,Illicium micranthum and bamboo.Cinnamomum cassia is a medicinal plant, itsbark is used directly in Chinese medicine andits leaves and branches are distilled to produceoil. Fruits of I. micranthum are a common spicein Chinese food. Bamboo is managed for timberto make baskets for farming use, as buildingmaterial and for paper pulp (Figure 3.1). Ediblebamboo shoots are also produced. Plantationsof Cinnamomum (Figure 3.2) and Illicum ac-count for 1.7% of forest area of the village; andbamboo, 1.4%. In addition, Camellia sp. isplanted for tea oil in several hamlets.

Shanxin is rich in forest resources on a percapita basis. On average, each villager has about1 ha of forest land and 8.5 m3 of timber.However, the endowment varies significantlyamong hamlets of the village. For example, percapita forest area is only 0.27 ha and 0.35 harespectively in two hamlets but is 1.9 ha inanother hamlet.

Contributions of forest to incomegeneration and villager livelihoodsShanxin village is a typical agriculturalcommunity, and if those who find off-farmemployment are excluded, farmers contributethe majority of the production of the village. Thecontribution of different sub-sectors ofagriculture has varied over time. Forestry hasbecome more important in income generationsince the early 1990s, with its share increasingfrom less than 10% in the early 1990s to nearly20% in the mid-1990s. In this period it becameas important as animal husbandry in the village.In contrast, the share of cropping decreased fromover 80% to about 60% over the same period,although it is still the largest sub-sector (Table3.2). Fuelwood is the sole source of energy forcooking and space heating for local people; and

Figure 3.1 Bundles of bamboo stems for a local paperfactory in Cangwu County

Figure 3.2 Leaves of Cinnamomum cassia harvestedfor oil production



sales of fuelwood, small-diameter wood andNWFPs are important sources of cash income,while the crops produced are largely for on-farmconsumption.

Demography, Infrastructure andWealthShanxin has a population of 2570 in 521households (1995) with 43% of the populationat labour age. The dependant ratio is not high,so, poverty is not much related to this factor.About 64% of the population has primary schoollevel education; 32% has junior middle schooleducation; and nearly 3% is illiterate. Fewvillagers have senior middle school education.Almost all those with senior middle schooleducation and a large proportion of those withjunior middle school education seek part-timeor full-time off-farm employment in urban areasof Guangdong and Guangxi.

The village has relatively good access totransport. The highway from Wuzhou to HexianCounty passes through the village and providesfor it access to vehicle transport. Six of 24 villagehousehold groups reside along the highway and11 of them are connected by dirt road to the high-way, though seven groups are far from the mainroad or have no access to motorised transport. Inaddition, the Xijiang River (a branch of the PearlRiver) provides transport between Wuzhou city

(30 km from Shanxin) and many cities in the PearlRiver Delta and Hong Kong.

The village is rather isolated in terms oftelecommunication. The village office had noaccess to telecommunication and officials had torely on oral or written messages when they neededto contact leaders of the village household groups.

Shanxin village is poor compared with theother 20 villages within the same township.Average per capita gross output value ofagriculture of the village was 77% of townshipaverage; and average per capita income about65% (Table 3.3). Average per capita output offood grain was also lower than the townshipaverage due to a smaller area suitable for cropproduction and poorer land productivity. As aconsequence, 80% of households were not ableto achieve food sufficiency and had to buy grainto make up the deficit.

Extent and Causes ofDegradation of EvergreenBroadleaf Forest

Extent of DegradationIn this chapter, two indicators - biomass ha-1 andforest composition – are used to assess the forestin Shanxin village. It is compared with itsprevious state and with an undegraded forest ina nearby reserve.

Table 3.2 Gross output value of agriculture in Shanxin by subsector (1991-94)Subsector 1991 1992 1993 1994

Yuan % Yuan % Yuan % Yuan %000s 000s 000s 000s

CroppingForestryAnimal husbandryFisheryVillage total

Yuan capita-1 Yuan capita-1 Yuan capita-1 Yuan capita-1

CroppingForestryAnimal husbandryFisheryTotal

Source: Shanxin Village Office, Annual Statistical Reports of Shanxin Village, Wangfu Township, 1991,1992, 1993, 1994.

6945986

1840

74364

2071

1016

751233242

21228

89428427421454

1802153223

32182

837

100.1

100

1911163533

3.62611

736

200.1

100

1930597621

6.03155

611920

0.2100

2293728703

5.63730

622019

0.2100



Our surveys in 1997 showed that 75% ofevergreen broadleaf forest in the village is de-graded, 10% severely degraded and about 15%slightly degraded. The degraded forest com-prises impoverished and ecologically unstablestands equivalent to those classified as‘overlogged forest’ by Banerjee (1994). Itsoutput of fuelwood is less than 20 t ha-1 on abouta 7-year cutting cycle. The seriously degradedforest has become shrub land or bare land. Itcan provide little fuelwood, compared to the 30t ha-1 or more from slightly degraded forest.

Biomass of Shanxin’s forests has declinedremarkably over the last five decades. Based onthe recall of senior villagers, standing biomassof the forest in the early 1950s averaged about100 t ha-1 and could reach 150 t ha-1. It declinedto about 65-100 t ha-1 in the early 1980s andfurther to only 30 t ha-1 in the late 1990s (Table3.4) so the past 20 years saw a substantialreduction in the biomass. Wood output hasshown a similar trend. Yield decreased from 50-60 t ha-1 in the early 1980s to about 20-30 t ha-1

in the late 1990s.Biomass of the evergreen broadleaf forests

in Shanxin is also sharply lower than that offorest with same dominant species in a nearbynatural reserve. Survey showed that forestbiomass in the reserve is 280 t ha-1, about eighttimes that of the forests in Shanxin. This furtherdemonstrates the severity of degradation of thevillage’s forests.

Although Shanxin’s forests are degraded orseverely degraded, soil fertility in the forests didnot show a similar trend. Our survey data

showed that soil fertility between the degradedforests in Shanxin and the primary forest in thereserve did not differ significantly. This suggeststhat land in the degraded forests still has greatpotential for forestry production.

Forest composition has become less diverse.Senior villagers in Shanxin recall that treespecies preferred for house building andfurniture making such as Castanopsis hystrix,Erythrophleum fordii, Phoebe nanmu, andMadhuca subquincuncinalis were very commonin evergreen broadleaf forest in the early 1950s.Due to prolonged selective cutting, there are nownot many trees of these species left. Fast-growing species such as Castanopsis fissa,Quercus griffithii, Alangium chinense andSchefflera octophylla, are now very common sothe value of the forest is lower. In addition, as aresult of shorter interval between wood harvests,the proportion of large-diameter trees in theforests is very small.

Causes of DegradationForest degradation has received increasingattention throughout the world but convincingresearch on causes of forest degradation is stilllimited in China. Our research in Shanxinfocused on causes of degradation of evergreenbroadleaf forest to better understand the issueand to identify effective strategies forrehabilitation. The research indicated a numberof factors responsible for forest degradation.They are:• over-harvesting• inappropriate logging methods

Table 3.3 Agricultural economic performance in Shanxin and Wangfu township (1994)

Mean annual Average Gross output value of agriculturegross output annual income ’000s Yuanvalue of agriculture (Yuan capita-1)(Yuan capita-1)

Shanxin 1454 980 3730 2293 728 703 5.6Wangfu 1888 1509 5030 3161 511 937 421Shanxin as 77 65 74 73 143 75 1.3% of Wangfu

Source: Shanxin Village Office (1994), Annual statistical reports of Shanxin village and Wangfu township, 1994.

Total Cropping Forestry Animal Fishery Husbandry



Table 3.4 Changes in forest biomass

Forest type Biomass (t ha-1 )

Early 1980s* 1997**

Castanopsis fissa forest 98.5 33.5Schima superba forest 65.1Mixed forest of C. fissa and S. superba 77.2Castanopsis hystrix forest 33.7Mixed forest (I) 29.3Mixed forest (II) 6.8Shrubs 4.4Primary forest (natural reserve) 281

* Data from a survey on fuelwood by Guangxi Academy of Forestry.

** Data from our field survey.

• radical development policy and breakdownof collective forest management system, and

• tenure insecurity

Over-harvestingCangwu County has been a fuelwood andcharcoal production area for more than a centuryexporting these products to Guangzhou, othercities in the Pearl River Delta and Hong Kong.Overlogging, driven by local and outsidedemands for fuelwood and charcoal, has causedforest degradation in the county, includingShanxin. This took place largely during theperiod from 1954 to the late 1970s.

Although farmers sold these products before1954, the quantity marketed to outside areas waslimited, mainly due to the lack of transport in-frastructure. Construction of the highway fromWuzhou to Hexian in 1954 provided villagersin Shanxin with easy access to transport andfacilitated increased production and marketingof fuelwood and charcoal. These were primarysources of cash income for farmers in Shanxinand other villages in Cangwu and beyond atleast until the late 1970s. Since then demandfor them in Hong Kong and the Pearl RiverDelta declined significantly as other energysources became available. In 1977 Shanxinstopped making charcoal because of very lim-ited demand. However, there is still marketdemand for fuelwood, because this remains thesole source of energy for cooking and spaceheating in the area.

High consumption of fuelwood on-farm andlocal market demand places a huge pressure onthe forest. Our survey of 74 households inShanxin in 1995 showed these householdsannually harvested 940 t of fuelwood, nearly13 t per household. On-farm consumption was68% and 32% was marketed. About 70% ofhouseholds in the village sold fuelwood, andmore than half earned most of their cash incomefrom these sales. Each of 40% of householdsearned 1000 Yuan or more annually fromfuelwood selling. Shanxin village harvested6711 t of fuelwood in 1995. As noted earlier,average fuelwood output of 7-year-old forest isestimated to be around 30 t ha-1. This suggeststhat Shanxin needs to harvest more than 220 haof forests annually for fuelwood only. This isequivalent to about 13% of evergreen broadleafforest in the village.

A new demand for small-diameter woodappeared in the mid 1980s with theestablishment of a number of wood-usingenterprises. The largest is the Wuzhou WoodFactory, located in Cangwu County, whichproduces fibreboard and particleboard and needs90 000-100 000 t year-1. It is estimated thatcurrent demand in Cangwu County for small-sized wood is 180 000-190 000 t year-1. On thesupply side, the county is not able to meet thisdemand. The area of secondary broadleafforests in Cangwu is 67 600 ha, assumingharvesting is at 7-year intervals, so about 9660ha can be harvested a year. Experience



indicates that average output of small-diameter wood of 7-year-old broadleaf forestis only about 10 t ha-1. This suggests that CangwuCounty is able to supply about 96 000 t of smallwood at present and in the near future, or onlyable to meet the requirement of the WuzhouWood Factory. Clearly, there exists a substantialgap between demand and supply of wood in thiscounty. The situation has placed great pressureon the already degraded local forests.

Over-harvesting also occurs because of over-dependence of local communities on forestresources. Surveys showed that more than 80%of households depend on small wood and otherforest products as main sources of cash incomeand they have limited cash income from othersub-sectors of agriculture and non-farmactivities.

A possible solution is to increase small woodoutput by developing fast-growing plantationsand tapping other sources of income to reduceover-dependence on the degraded forests.Empirical evidence shows that there is a closeinterrelationship between dependence on forestand development of other sub-sectors ofagriculture and non-farm activities. Forests inFuzhu hamlet of Shanxin have been conservedand are in relatively better condition than thosein other hamlets, as village households in Fuzhuhave developed animal husbandry, fruit orchardsand food processing enterprises and do not relyheavily on forests for income generation.

Inappropriate logging methodsPoor logging methods, including selectivelogging, and shorter intervals betweenharvesting have contributed to the degradationof forests in Shanxin and other villages in thearea. Excessive application of selective cuttinghas resulted in less individuals of slowergrowing species preferred for house buildingand furniture making, while less valuable, lowerquality species, such as Castanopsis fissa andQuercus griffithii, have become dominant.Hence tree diversity and wood quality havediminished. Increasing demand for wood

resulted in cutting cycles being reduced fromabout 8 years to 5-6 years, which also meansthat slow-growing species that produce higherquality wood have disappeared from the forest.

Radical development policies andcollective forest management breakdownForest degradation in Shanxin also hassomething to do with the radical developmentpolicies of the late 1950s and a unique politicalcircumstance that caused the breakdown ofcollective forest management system in the late1960s.

In 1958, the Chinese Government launchedthe radical ‘Great Leap Forward’ campaign toestablish the people’s commune and attempt torealise country’s industrialisation with a visionthat China’s steel output would surpass that ofGreat Britain in 15 years. Throughout Chinapeople were mobilised and forced to make ironand steel using backward backyard furnaces.These furnaces used a very large amount offuelwood and charcoal but failed to produceusable iron and steel. Shanxin suffered from thisevent as a large number of people from theneighbouring town came to the village to cuttrees and make iron and steel. At that time 190charcoal making places were established in thevillage. Patches of higher quality evergreenbroadleaf forests were cleared and many otherdense forests became poor quality secondaryforests with low stocking. As fuelwood andcharcoal output peaked it was accompanied bythe first serious deforestation in Shanxin since1949.

In 1966, Chairman Mao initiated the ‘GreatCultural Revolution’. The campaign paralysedalmost the whole governance system fromcentral to village level and resulted in thebreakdown of management system of collectiveforests in many rural areas. The leadership ofShanxin village also lost control over forestmanagement at that time and illegal cuttingincreased significantly. In the period 1966-1968,fuelwood and charcoal output in Shanxin andadjacent areas experienced another peak, which



implies serious deforestation, though not assevere as occurred in the late 1950s in terms offorest stock logged.

Insecurity of forest tenureFurther significant deforestation anddegradation of forests took place in Shanxin andother villages following the distribution ofcollective-owned forests to individualhouseholds in 1982. This is shown by thesignificant decline in forest biomass over thelast two decades as previously noted.

Frequent changes in policy for forest tenureover the three decades to the early 1980s werethe underlying cause of this deforestation due toinsecurity of forest tenure (Liu 2001). As in otherparts of China, Shanxin village experienced radi-cal and frequent changes in forest ownership.Forest was private property of rural householdsin the first half of the 1950s, collectivised in 1956-1957 and became commune’s property in 1958,and owned and managed exclusively collectivelyuntil the early 1980s. The ownership of scatteredtrees around homesteads also experienced simi-lar change. They were collectivised in 1958, re-turned to individual households in the early1960s, and taken away again by commune fromfarmers in mid-1960s.

The frequent changes led to a great lack ofconfidence among farmers in tenure security.With perceived insecure tenure it made littlesense to them to manage forests sustainably.Instead, they immediately harvested treesdistributed to them to realise their value, areasonable response based on their pastexperience.

Disincentives to ManagingEvergreen Broadleaf Forestfor Commercial Timber andOther Wood ProductsThe case study in Shanxin showed thatgovernment policies for timber harvesting andtaxation on timber discourage farmers fromproperly managing timber forests. Local farm-ers preferred to sell all their wood as fuelwoodon the free market rather than directing their

small-diameter wood to the wood processingfactories even though demand is great and pricesfor small-diameter wood are higher than forfuelwood. This apparent aberration can be ex-plained largely by farmers’ response to the twogovernment policies.

Timber harvesting is highly regulated inChina. Usually, one must obtain a cutting permitto harvest timber no matter whether this is incollective-owned forests and plantations or inhousehold plantations. Furthermore, cuttingcertificates are tied to logging quotas. When anadministrative area has used up its quota in agiven year, it is not allowed to issue more cuttingcertificates that year. Experience shows thatfarmers are usually in a disadvantageousposition to get certificates for the quantity theywish to harvest and at the time they want toharvest it. Timing is important to ensure the bestprice is obtained. In Shanxin, however,regulations were not strictly followed forharvesting fuelwood.

Data indicate that local governments and theforest department collected and charged 14 kindsof tax and fees on timber in Cangwu County.Consequently, half of the turnover of woodtrading went to the outside stakeholders. As costsof logging and timber transport are estimated atabout 30-35% of market price for timber, itfollows that the farmers received only theremaining 15-20% of sale price. Profits offarmers from timber production were nearly zerowhen their costs of planting, tending andguarding of trees are taken into account. Takingpine timber an example. The price was about500 Yuan m-3 in 1997-1998, from which thefarmers received only 60-80 Yuan (12-16% ofmarket price). Although fuelwood prices wereonly 160 Yuan t-1, much lower than for timber,farmers received more income from it than fromtimber, because they were subject to littletaxation when selling fuelwood. This explainswhy farmers preferred to sell their wood asfuelwood.

Less restrictive regulations on timber har-vesting and less taxation on timber sales wouldprovide incentives to encourage farmers in



Shanxin and other villages to plant timber treesand manage timber forests more effectively byenabling a better return on the products.

Strategies for RehabilitatingDegraded Evergreen BroadleafForestDegraded and seriously degraded forests mustbe targets of rehabilitation efforts, because if leftuntreated their yields will continue to declineand restoration of their productivity will be timeconsuming and expensive. In their degradedstate they will not be able to provide the normalenvironmental functions of a forest. In contrast,slightly degraded forest has high potential to re-cover its productivity in a reasonable period oftime.

The most important causes of forestdegradation in Shanxin are over-harvesting andshort intervals between harvests. This problemcan only be addressed by reducing the farmers’dependence on the forest for income generation.The strategies identified to rehabilitate theseriously degraded and degraded evergreenbroadleaf forest included direct and indirectinterventions. Three broad strategies emerged:• natural regeneration through closing moun-

tain (fengshan yulin in Chinese),• plantation establishment, and• development of other sub-sectors of

agriculture and non-farm activities.The first strategy was designed to restore the

degraded forest by restricting access and leavingit to regenerate naturally for a sufficient period.The second was designed to reclaim theseriously degraded lands and expandopportunities for income generation to reducedependence on the existing forests throughdeveloping plantations, especially NWFPplantations, which could provide income in arelatively short time. The third strategy was todevelop agroforestry systems, including raisinganimals with trees, and non-farm economicactivities in household yards and on landsaround homestead. The purpose of this thirdoption was to expand opportunities for incomegeneration for households.

Natural Regeneration throughMountain ClosureNatural regeneration of forest through closingthe mountain or hillside has been a commonsilvicultural practice in China over the last fourdecades. It aims to promote regeneration ofsecondary forest through prohibiting orrestricting specified activities such as grazingon a tract of degraded mountainous land. It istechnically feasible since broadleaf trees readilysprout by coppice shoots and coniferous treesregenerate easily by seeds. The favourablesubtropical climate with warm temperatures andadequate rainfall provide ecological conditionsfor successful natural regeneration of secondaryforests. Many secondary forests have beensuccessfully regenerated in this way and thepractice is cost-effective and much cheaper thanestablishment of artificial plantations (Yang andOu 1987, Xu and Zheng 1994). Experience inother parts of Cangwu County also shows thatsecondary forest recovery begins two years afterclosing the area and this enables small diameterwood to be harvested in 7 to 8 years.

The experiment at Shanxin began in early1998 and the area closed was 670 ha. Informalvillage regulations on mountain closure specifythat participating households collect forestproducts only from their own patch of forest,and no tree removal, fire or grazing in the closedarea is permitted. However, several challengesof governance and management needed to beaddressed and some constraints overcome if themountain closure was to succeed in Shanxin.These challenges and constraints, and strategiesto address them are summarised in Table 3.5 anddiscussed below.

Collective action and tenure securityForest holdings became very fragmented as aresult of the inappropriate implementation of thepolicy for the distribution of collective forestland and forests to individual farmer householdsin China in the 1980s. In almost all cases, forestson a mountain slope now belong to manyhouseholds. Therefore, closing an area of themountain requires some form of collective



action. At first, we recommended a shareholdingsystem to farmers. Under the system, villagerswould pool their forest resources for commonmanagement and share benefits from the forestin accordance with resources they contribute.This involves a change in the form of foresttenure and benefit distribution. Farmers inShanxin rejected this proposal. Throughconsultation an option was identified that didnot change forest tenure and did not involveredistribution of benefits. Each household withforest land in the closed area remains the holdersof trees, has use rights to the land and willharvest products from their own land when themountain is reopened. Nobody is allowed tocollect forest products from the forest areas ofother households within the closed area.

Farmers adopted and implemented theoption. Two points are stressed. First is thattenure security is crucial to ensure success inclosing mountain reforestation. Farmers are verysensitive to tenure as a consequence of thefrequent changes in forest tenure over the pastseveral decades. They are still not confident ofsecurity of their use rights to land, which is oneof main reasons they are not keen to invest intree planting and sustainable use of forest re-

sources in China in general. This in part explainswhy farmers in Shanxin refused the initial ideaof a shareholding system for closing mountainregeneration. The other point is that there mustbe an agreement on an acceptable mechanismfor benefit distribution. Farmers refused the ini-tial proposal because they were concerned thatvillage leaders who hold power for benefit dis-tribution under a shareholding system wouldappropriate proceeds derived from their forestarea. Such a problem has occurred elsewhere inChina.

Current access to the forestIf mountain closure is to be an effective strategyfor forest regeneration, villagers’ current accessto forest resources needs to be addressed whilethe regeneration area is closed. Villagers in moun-tainous areas in China are forest-dependent, asin Shanxin. They collect fuelwood for self-useand wood and/or NWFPs to sell for a cash in-come. Options completely denying access to for-ests are impracticable and definitely not accept-able to farmers. In Shanxin, we worked with thefarmers to develop a plan that required half ofthe degraded evergreen broadleaf forest to beopen for use when the other half is closed. When

Table 3.5 Challenges, constraints and strategies for ‘closing mountain’ for natural regeneration

Challenges and constraints Strategies

Fragmentation of household holdings Common management of household forest resourceswithout any changes in forest tenure and benefitdistribution.

Maintaining access to forest when closing mountains - “Shifting closure of a mountain” at village level (onlyhalf of forest was closed at one time)

- “partial closure” at household level (just part offorest of any a household was closed at agiven time)

Encouragement of active farmer involvement Harness their self-interest and build their awarenessof the importance of mountain closure.

Set up an effective management system Establish a governance group, a management group,a supervision group,and appoint full-timeforest guard (s).

Overcome lack of operational funds County Forestry Bureau, Township Government, andthe research team jointly agreed to cover the costs.



the closed half re-opens after several years, theother half of mountain will be closed for reha-bilitation. At household level, some householdshad all or the majority of their forests in the des-ignated closed area. We addressed this issue byhelping coordinate an exchange of forest landsuch that these households exchanged part of theirforests with those with the majority of their for-ests outside the designated closed area. This wayno household had all their forests closed in onetime and so maintained current access to forestresources. Moreover, agreement was reached towaive children’s education tuition fees for sev-eral very poor households whose income woulddecrease as a result of reducing their access toforests.

Farmer participationThe Shanxin experience shows that farmerparticipation is essential to the successfulrehabilitation of degraded forests using themountain closure approach. Regulations,statutory or informal, and measures for forestmanagement will not work without their col-laboration. Only when their self-interest is har-nessed and their awareness raised of the impor-tance to their future income of rehabilitatingdegraded forests by natural regeneration usingmountain closure, will they be willing to be ac-tively involved.

Governance systemAn effective governance system is essential inensuring success in the mountain closurestrategy. The system we facilitated in Shanxinconsisted of a governance group at townshiplevel, a management group at village level, anda supervision group. Responsibilities of each ofthe groups are clear and specific. Thegovernance group is responsible for dealing withany illegal deforestation cases and for providingextension services. The management grouptakes its responsibilities through two keypersons: a forest officer and a forest guard.Duties of the forest officer are liaison andcommunication between the supervision group,the forest guard, and other members of the

management group. The forest guard isresponsible for patrol, forest fire control andprevention of damage by domestic animals. Thesupervision group consists of representatives offarmers with forest in the closed area. It isresponsible for monitoring performance of theforest officer and the forest guard and will reportany problems to the management group. In theevent that this group does not take timely actionon the problem the supervision group can reportproblem to the governance group. Informalregulations on forest management wereformulated to ensure forest protection andmaintenance and all farmers involved are requiredto help with this task.

Operational costsThere are operational costs that must be met.Natural regeneration through mountain closurerequires at least a full time guard to patrol forestsand prevent illegal cutting and grazing damage.This involves payment of a salary and othercosts. However, a majority of rural communitiesin mountainous areas are too poor to contributefunds to cover such costs. In the experiment inShanxin, the Cangwu County Forestry Bureau,the Wangfu Township Government and theresearch team agreed to pay for essentialoperational costs for the first period of mountainclosure. Farmer households will reimburse thesebodies when they harvest products from theregenerated closed forest. This money then willgo into a fund to cover the costs of next roundof mountain closing. It is the aim that economicsustainability of the system can be achieved byusing this rolling fund.

ProgressSince the mountain closure in Shanxin began inearly 1998, initial experience in the experimentsuggests that the arrangements are workingsatisfactorily. The forest guard fully assumed hisresponsibilities. The informal regulationsfarmers made have strictly been followed, withno illegal cutting or animal grazing in the closedarea. The experiment in Shanxin hasdemonstrated the potential for developing



arrangements suitable for poor villages with adevelopment level similar to that of Shanxin torehabilitate their degraded subtropical broadleafevergreen forests.

Plantation Establishment andManagementThere are about 160 ha of seriously degradedforest representing about 10% of the area ofevergreen broadleaf forest in the village. Furtherforest degradation will occur if measures are nottaken. However, villagers have few other sourcesof cash income apart from the forest. Over-dependence of local communities on forests fortheir welfare is a primary, underlying cause ofover exploitation. Plantation development is oneoption to improve livelihoods as it will provideadditional opportunities for income generation.It can also ameliorate ecological conditions wherethe land is degraded.

Development of plantations involves manysocio-economic and technical issues.Technically, different species may be needed fordifferent parts of a mountain slope. For example,adaptable timber species, such as Pinus spp.,may be appropriate on shallower, less fertilesoils on the upper slopes and ridges. More site-demanding NWFP plantations and agroforestrysystems may be ecologically suitable and moreprofitable on the middle and lower slopes. Frommanagement and policy viewpoints, successfulplantation development requires a mechanismto ensure implementation of collective actionand needs to provide incentives to farmers tobecome involved. Farmers’ incentives to plantNWFP are related, to large extent, to theavailability of appropriate species that areecologically suitable, with good market prospects(high demand and prices) and high economicreturns that can be expected within a few years.Three initiatives were designed to address theseissues in Shanxin:• reclamation of a small area of sloping land,• development and management by individual

farmers of NWFP plantations, and• exotic NWFP species experiment.

The first initiative gave attention to a

mechanism for collective action involving manystakeholders, the second examined factors thatencourage farmers to establish and manageNWFP plantations, and the third tried to identifynew species for NWFP with market prospectsand quick returns.

Reclamation of a hillsideAn experiment to improve productivity andvegetation cover on a seriously degraded hillsidewas initiated. Thirty-eight households ofShanxin village owned a piece of sloping landof about 10 ha in extent. All had to agree toparticipate in the collective planning effort.Technical actions had also to be integrated. Theagreed plan was to plant the native pine (Pinusmassoniana) for timber and soil and waterconservation on the upper part of the slope andNWFP trees (mainly Illicium verum andCanarium album) in an agroforestry system withfood crops on the middle and lower slopes.

As with the mountain closure for nativebroadleaf forests, the researchers initiallyproposed a shareholding system. The landowners rejected the idea as they perceived it asre-collectivisation and were concerned that theirinterests might be appropriated by villageleaders. An arrangement similar to that for themountain closure was finally adopted. Under thearrangement, each of the household would planttrees based on the integrated plan and collectproducts from their own patch of land. An oralagreement was made among all the householdsinvolved that grazing would not be allowed onthe area. The research team provided thehouseholds with tree seedlings and fertiliserswithout charge.

The planning took place in October 1996.Soon after agreement was reached the land wasprepared in late 1996 and trees, cassava, beans,green beans and corn were planted in March1997. The survival rate of Illicium verum was asatisfactory 90%, while that of Canarium albumwas only 50%, largely due to poor quality ofthe seedlings and late planting. However, 80%of the young trees were damaged by the brows-ing of domestic animals after the intercropped



crops were harvested in July and August 1997.This was a serious breach of the agreementreached by villagers.

What went wrong? It had been decided thatthe village leader would have overallresponsibility for ‘management’ of the area butno one, even on a part time basis, was assignedto tend, care and guard the tree seedlings.Moreover, no action was taken when the firstgrazing damage took place. Other villagersbegan to graze their cattle on the tree plantingarea when they found no discipline for thedestruction. Some of households therefore didnot observe the agreement on grazing and thevillage leader did not fulfill his responsibilityin punishing those whose animals damagedseedlings. The village leader had no incentiveto take action as he did not own land on the slopeand there was no reward for his participation.

The failure of the experiment shows fac-tors that work to affect common action/man-agement as has been the experience elsewhere(Arnold 1998). An agreement is important buta mechanism to ensure its strict observation isequally, if not more, important. Disciplinarymeasures are needed when someone breaks theagreement to encourage all people involved toobserve the rules. This is a platform for col-lective action of natural resource managementinvolving many stakeholders. Incentives areessential for the person responsible for disci-plinary measures. The person’s interest mustbe harnessed either as a stakeholder or throughother material returns commensurate with hisresponsibility.

The lesson drawn from this experience is thatan effective management mechanism is essentialin rehabilitating land owned by manyhouseholds.

Development and management byindividual farmers of NWFP plantationsWhile the slope rehabilitation involvedcollective action of natural resourcemanagement, another group of activities wasinitiated to assist individual farmers to developand manage NWFP plantations. The purpose

was to determine the critical factors involved inthe development and management of NWFPplantations by households or individuals. Twocase studies were conducted.

One study examined the tree managementby individual farmers of a collective forest farm.The forest farm had an area of 14 ha of hillyland. It was established and is owned by thetownship government. Pine (P. massoniana) wasplanted on the upper slopes, Chinese fir(Cunninghamia lanceolata) on middle slopes,and NWFP trees such as Cinnamomum cassia,Illicium verum, Canarium album and plums(Prunus saliana) on lower areas. The farmplanted the trees but contracted managementsuch as tending, caring and guarding to threefarmers. The farm covered salary of the threefarmers and other operational costs. In theagreement the farm paid the farmers a basicamount of money monthly and the balance atthe end of year based on their performance inmanaging the trees. The plantation has grownwell as a result of adequate tending and caring.Cinnamomum cassia planted in 1993 and 1994was ready for harvesting in 1998. Linking thefarmer performance to financial rewards was aneffective strategy.

The other case involved establishment ofsmall NWFP plantations by individualhouseholds. Six households plantedCinnamomum cassia, a species chosen by thefarmers. They stated that they planted thisspecies because they believed it has potentialfor high economic returns after 4 to 5 years inthe case of bark harvesting and in only 3 yearsfor leaf products (Figure 3.2). Market demandand prices for Cinnamomum products are high.In 1997-1998 the mean value of bark was 6000-7000 Yuan ha-1 and leaves from three-year-oldtrees yielded 2500 Yuan ha-1. Some smallfactories were established locally to process oilsfrom Cinnamomum leaves and branches.Implications from this activity are that marketdemand and reasonable returns in a short periodof time were the driving forces of plantationestablishment by farmers. It is very importantfor farmers to have information on markets and



to assess future trends in demand for theirproducts. Reliance on a single product can be aproblem if the market collapses.

Experiment with exotic NWFP speciesThe availability of species adapted to localecological conditions is critical for successfuldevelopment in NWFP plantations. Identifyingsuch species that also have good potential formarkets is an important ‘technical’ activity. Ourresearch in Shanxin also gave attention to theidentification of appropriate species by testingthe adaptability of selected rattan and bamboospecies.

Three rattan species (Calamus simplicifolius,C. tetradactylus and Daemonorops margaritae)were introduced to Shanxin village and a nearbyvillage in spring 1997. One household in eachof the villages was involved in the experiments.One household planted rattans in Micheliamacclurei plantation, while the other plantedthem in a secondary evergreen broadleaf forest.Each planted an area of about 0.8 ha. Preliminaryobservation suggests that the three species growwell in the local conditions and may be apotential source of income generation for localfarmers.

The experiment with bamboo was an attemptto identify a multipurpose bamboo species thatcould produce both shoots and timber. CangwuCounty has a long history of bamboo plantingbut almost exclusively for timber. Marketdemand for bamboo timber is shrinking so thereis a need for alternative sources of income. Amultipurpose bamboo species has potential toprovide a new opportunity for incomegeneration. Tissue cultured plantlets of fivebamboo species (Bambusa pervariabilis x[Dendrocalamus latiflorus x B. textilis], No. 1;Bambusa pervariabilis x [Dendrocalamuslatiflorus x B. textilis], No. 7; Dendrocalamuslatiflorus; Dendrocalamus brandisii; andBambusa oldhami) were planted in March 1997by two households. Performance varied greatlyamong the species, but the growth of two of themwas rather rapid for this area and they produceda large number of shoots. While more testing of

adaptability and management is necessary, thesenew introductions show considerable potentialfor expanded planting in the area.

Development of Other AgriculturalSub-Sectors and Non-Farm ActivitiesResults of our surveys in Shanxin indicated thatover-dependence of local villagers on forest wasbecause their sources of cash income other thanfrom the forest are quite limited. Food grain wasproduced largely for on-farm consumption; andeach household raised only 1-2 pigs and a fewchickens and ducks. Growing bamboo for localtraditional commodities such as buildingmaterials and baskets was practised but plasticand other materials are replacing these productsin the market.

Development of other agricultural sub-sectors and non-farm activities is an option todiversify sources of income for farmerhouseholds and to reduce over-dependence onexisting degraded forest. Other agricultural sub-sectors include cropping, animal husbandry(poultry, pigs etc.) and processing of agricul-tural and forest products. This group of strate-gies attempts to address forest degradation bytaking full advantage of house yards and landaround homesteads for non-forestry agriculturalactivities and non-farm industrial activities.Small processing industries suit farmerhouseholds because of low capital investmentand simpler techniques. Farmer households havethe flexibility to develop activities according totheir own circumstances. We worked with threehouseholds to determine whether it is feasibleto find other sources of income and to identifyconstraints to their development and strategiesto overcome them.

The first household developed anagroforestry system comprising fruit trees andpig raising on 4 ha of land. Part of the land wasan abandoned school property and the rest anabandoned collective orchard. The householdleased the land for 5 years for 7000 Yuan year-1

with a possibility for contract renewal. The fam-ily then planted persimmon (Diospyros kaki) andplum (Prunus saliana) and raised 80 pigs on the



land. Pig dung was used to generate methanegas for lighting and cooking, and as fertiliser.This was a typical household business as thehousehold was responsible land leasing, treeplanting and tending, and raising the pigs.

The second household developed anagroforestry system that included moreelements. It included fruit trees, food crops, andpoultry on a plot of 0.3 ha. With our technicalassistance, the family raised chickens and ducksin their yard, planted fruit trees such as citrusand persimmon on land around homestead, andintercropped the fruit trees with food crops suchas peanuts, cassava and beans. Preliminaryobservation shows that the family was able toraise and sell some chickens and achieve areasonably good output of food crops.

A third household was involved in non-farmactivities plus growing fruit trees. The familyhad already established a small factory makingchess pieces from local wood and was achievinga good economic return. Lack of technicalknowledge prevented the family fromeffectively using the land around the homesteadand it was left almost uncultivated. We encour-aged the household to take advantage of theirland and tap another source of income by plant-ing fruit trees on it. With our technical assist-ance, the family planted persimmons andintercropped them with cassava, beans andpeanuts.

In all three cases, we provided informationon market demand for goods, helped designvarious agroforestry systems and non-farmactivities, and provided advice on agroforestryand in some cases free seedlings. It is clearlypossible to expand sources of income for localfarmers by these means but there are severalconstraints to be overcome. There is a need foraccess to information on market demand forproducts on which basis agroforestry systemscan be developed. In addition, there is a needfor improved extension services, especially foradvice on development and management ofagroforestry systems and non-farm enterprises.There is also a need for improved rural financeservices since a lack of capital investment is also

a constraint to development of non-forestry ag-ricultural activities and non-farm economic un-dertakings.

Conclusions

Causes of Forest DegradationSeveral factors were responsible for thedegradation of evergreen broadleaf forest inShanxin village. The participatory approachadopted in the research greatly facilitated theidentification of these factors. The radicaldevelopment policy (Great Leap Forwardcampaign) in the late 1950s and the abnormalpolitical environment (the Cultural Revolution)in the 1960s to 1970s contributed to significantdeforestation in the Shanxin area. Lack ofconfidence among farmers in tenure security,resulting from frequent changes in forestownership over the three decades to the early1980s, caused another major deforestation earlyin the 1980s. Excessive logging and selectivecutting, driven by poor financial returns forforest products and over-dependence of thepopulation on the forest, have been a continu-ing cause of forest degradation. This depend-ence has been exacerbated by a lack of oppor-tunities for other income generation activitiesdue to poor access to market information andtechnical advances, and unavailability of ruralcredit.

Strategies and PolicyRecommendations on ForestRehabilitationThe evergreen broadleaf forests are very resil-ient and can recover provided they are givensufficient time. Reduction of the level ofdependence on the forests by identifying othersources of income is the key to reducing the levelof exploitation and giving them time to recover.Several strategies to generate income from ani-mal husbandry, agroforestry, and non-farm ac-tivities were identified. Natural regeneration,through mountain closure, was used to improvedegraded forests, while agroforestry was usedto revegetate seriously degraded forest land.



The action research in Shanxin demonstratedthat these strategies can be effective. Theymainly used existing technologies, althoughthere is potential for introducing new crops toproduce new products. Resolvingsocioeconomic and management problemsthrough active participation of the stakeholderswas a major factor in achieving positiveoutcomes. The research provided someinstructive lessons and implications.

Motivating farmersFarmer participation is key to the success innatural regeneration through mountain closureand the establishment and management ofplantations whether through common action orby individual households. To encourage thisparticipation, it is most important to harnessfarmers’ interest, especially through financialincentives.

In Shanxin, farmers could also appreciate thesomewhat longer-term benefit of improving theyield from the degraded evergreen broadleafforests. The prospect of financial benefits withina reasonable time from the production and salesof NWFP was also attractive to Shanxin farmersand the community. Such activities can bebeneficial even to those who only providelabour, as was shown by the three farmers whomanaged the plantation of NWFP on thetownship-owned forest. It is also clear thatresearchers attempting to identify new speciesfor farmers to grow should seek those that areboth ecologically suitable and have the potentialfor good returns in a short period of time.

Organization of common management ofhousehold resourcesBoth mountain closure and hillside reclamationinvolve common management of households’forest resources. The form of management (ashareholding system or common managementwithout changes in individual forest tenure andbenefit distribution), should be decided byrelevant stakeholders. Tenure issues are verysensitive and management skills limited inisolated and poor areas, as was the case in

Shanxin. In this situation a shareholding systemwas neither appropriate nor acceptable to farmers.

When common action for resourcemanagement, such as hillside reclamation andmountain closure, is needed, it is essential toensure that all people concerned observe agreedrules. Disciplinary measures should be specifiedand implemented to establish a platform ofcommon action.

Current access to forest resourcesVillagers in mountainous areas depend heavilyon the forest for income and other livelihoodneeds. It is not appropriate to completely denythem access to forest resources during therehabilitation process. The compromises agreedin Shanxin i.e. ‘shifting closure of a mountain’at village level (only half of the forest was closedat one time) and ‘partial closure’ at householdlevel (only part of the forest of a household wasclosed at a given time) show that the accessproblems are not insurmountable.

Improved rural servicesThere are opportunities to develop other sub-sectors of agriculture, agroforestry, and non-farm activities for income generation to reducepressure on existing forests. To do thiseffectively, it is necessary to overcome a rangeof constraints faced by farmers. Important firststeps suggested by the research in Shanxin arebetter access to market information and technicaladvances through improved agriculturalextension, and availability of rural credit/financefrom both public and private sources.

Government policiesThe current restrictive regulations on timberharvesting and high taxation on timber sales aredisincentives to farmers in Shanxin and othervillages to plant timber trees and manage timberforests more effectively. A review andamendment of such government policies couldenable farmers to achieve a higher return on theirforest products and would encourage them tomake greater efforts to manage their forests andreduce forest and land degradation.



ReferencesArnold, J E M. 1998. Devolution of control of

common pool resources to local communities:experiences in forestry”, Paper to the Meetingof the UNU/WIDER Project on Land ReformRevisited: Access to Land, Rural Poverty, andPublic Action, April 1998, Santiago, Chile.

Banerjee, A. K. 1994. Rehabilitation ofdegraded forests in Asia. World Bank,Washington DC.

Cangwu County Forestry Bureau and GuangxiInstitute of Forestry Inventory and Planning.1995. Cangwu County Forest ManagementPlan, Cangwu, Guangxi. (In Chinese)



Chambers, R. 1994c. Participatory ruralappraisal (PRA): challenges, potentials andparadigm. World Development 22: 1437-1454.

Chambers, R. and Guijt, I. 1995. PRA – fiveyears later: where are we now? Forests, Treesand People Newsletter No. 26/27: 4-14.


Ministry of Forestry. 1996. A summary of forestresources of contemporary China. Ministryof Forestry, Beijing. pp. 1-22. (In Chinese)

Xu, H. and Zheng, J. (eds.) 1994. A study onforest regeneration through closing mountain.China Forestry Press, Beijing. (In Chinese)

Yang, Z. and Ou, Z. 1987. Forest regenerationthrough closing mountain. China ForestryPress, Beijing. (In Chinese)


IntroductionHuitong County, Hunan, is in the subtropicalclimatic zone and its primary vegetation isevergreen broadleaf forest. Extensive Chinesefir (Cunninghamia lanceolata) plantations havebeen developed since the first cultivation of thisspecies in the county in 1368, and especiallyduring the last 50 years they have replacedevergreen broadleaf forest. As a result, nowChinese fir plantations cover 22% of the landarea and evergreen broadleaf forests 18%(Huitong County Forestry Bureau 1993a).Moreover, forests in Huitong, and in Hunanprovince and beyond are degraded with lessstanding timber volume, reduced speciesdiversity and declining soil fertility. Forestdegradation and a reduction in forest cover haveresulted in a significant decline in ecologicalservices by forests and have impacted negativelyon the livelihoods of forest-dependent people.They are now able to collect less timber and non-wood forest products (NWFPs) and their incomefrom forests is reduced.

There has been research on causes andprevention of degradation of nature forest andChinese fir plantations since the 1960s (Institute

of Forestry and Pedology 1980, Research Teamof Intensive Cultivation of Chinese Fir 1992,Chen 1992 and Xu 1992). Zhu (1991) examinedhuman impacts on changes in biodiversity andsoil fertility in secondary broadleaf forests.These studies centre on biological or technicalcauses of forest degradation and identificationof technology-oriented solutions.

Our research, carried in Dongxi village andXiangjian village in Huitong, targeted two majorforest types: evergreen broadleaf forest andChinese fir plantation. In collaboration withlocal officials and farmers the researchersexamined the state and causes of degradationand identified activities for forest rehabilitationrespectively for each forest type. We then testedsome of the activities to assess theireffectiveness. We also gave considerableattention to disincentives among farmers tosustainable management of their forests in anattempt to address forest rehabilitation at apolicy level.

As in other chapters, research methods andbiophysical and socio-economic settings of thetwo villages are described. The state of

Chapter Four

Rehabilitation of Degraded Chinese Fir Plantationsand Evergreen Broadleaf Forest in Huitong County,Hunan

Li Tiehua1, Liu Dachang2, Xiang Wenhua1, Xu Guozheng1 and Zeng Guangzhen1

1 Central-South Forestry University, Hunan.2 Southwest Forestry University, Kunming P.R. China and Center for International Forestry Research, Bogor, Indonesia.Currently at The Mekong Institute, Khon Kaen, Thailand.



66 Rehabilitation of Degraded Chinese Fir Plantations and Evergreen Broadleaf Forest in Huitong County, Hunan

degradation, factors responsible for degradation,and strategies for rehabilitation of evergreenbroadleaf forest and Chinese fir plantation arediscussed. Finally, the disincentives torehabilitation and sustainable management ofdegraded forests are highlighted.

Research MethodsThe research in Huitong used methods similarto those in the other sites of this research projectbut different from those in traditional forestryresearch in China. They included a logicalprocedure of action research; use of aparticipatory approach involving villagers,government officials and researchers and amultidisciplinary team of socio-economic andbiophysical scientists, complemented by someconventional research methods; and attention toindigenous knowledge.

Farmer Participation in at All Stagesof Project ProcessIn this research, the logical procedure startedfrom diagnosis through design to delivery.Farmer involvement, at all stages, wasemphasised. More than 500 villagers contributedto the diagnosis of problems and the subsequentdesign and testing of strategies and options.

DiagnosisThis phase aimed at obtaining reliable baselineinformation on the two villages and beyond, andachieving a good understanding of the extentand causes of forest degradation and constraintsto rehabilitation. Participatory rural appraisaltools used included:• secondary data collection• participatory observation• household interview (semi-structured

interview, complemented by questionnairesurvey)

• key informant survey (semi-structuredinterview), and

• group interview/meetings.

Secondary data were collected largely fromHuitong County Forest Bureau, Huitong County

Archives and the township government respon-sible for the villages where the research tookplace. Data were collected on climate, soil, landuses, forest resources, tenure and managementsystem of forest, taxation of forest products,population, education, labour force, income,village social organisations and infrastructure.

Participatory field observation was carriedout immediately after secondary data collection.We worked with several senior villagers andvillage leaders familiar with local circumstanceson field visits to observe and discuss land use,types of crops and their patterns of cultivation,types of forest, extent of forest degradation andvillage infrastructure.

Household surveys were mainly in the formof semi-structured interviews plus aquestionnaire survey. Sixty households wereinterviewed. Criteria for sample strata includedlevel of family wealth, family composition ofsex and age, and distance between householdand market and between household and villageoffice. Household surveys aimed to identifyproblems, constraints and opportunities athousehold level, and topics discussed includedfamily population and number of members atlabour age, land uses, forest area and volume,forest degradation, constraints to rehabilitation,major livelihood means, sources of familyincome, and family expenditure by activity.

Key informant interviews used semi-structured interviews and aimed at analysingproblems at village level. Topics coveredpopulation, education, labour force and genderdivision of labour, land uses, crop calendar,village social organizations and governanceinstitutions, forest resources, forest tenure,village regulations on forest management,timber prices, household collaboration regardingforest production (shareholding system, forinstance), and extension and credit services.

Group interviews/meetings also used semi-structured interviews. The first interviews werewith villagers at the two villages to validate thesecondary data and baseline informationcollected through the key informants. Thesecond interviews were with groups of villagers


67Li Tiehua, Liu Dachang, Xiang Wenhua, Xu Guozheng and Zeng Guangzhen

of different ages, women, and village leaders toidentify causes of forest degradation. Finally, ameeting was held with villagers, townshipgovernment officials, and forest officials fromHuitong County Forestry Bureau and HuaihuaPrefecture Forestry Bureau to further analysecauses of forest degradation.

Design and testingVillagers were involved in the identification andtesting of options for forest rehabilitation. Manydiscussions were held in 1996 and 1997 todevelop strategies and the villagers madedecisions on whether or not to adopt them. Forexample, when identifying species for homegarden development, researchers provided adviceand options for promising fruit varieties butfarmers selected those they considered to havegood market potential and were easy to store.

Participation of Government OfficialsParticipation of local government officials fromprefecture, county, and township was essentialbecause they had knowledge of problems andconstraints from an institutional and policy pointof view, had access to financial and extensionresources, and were able to coordinate projectimplementation activities. Over 50 governmentofficials were consulted and some of themparticipated directly in the research. Severalmeetings were organised with them at thediagnosis stage to discuss the circumstances ofthe two villages. At the design stage, there weremore meetings to identify options and strategiesfor forest rehabilitation.

Involvement of Researchers andMultidisciplinary ApproachThe research team comprised of seven special-ists in forest resource management, forest ecol-ogy, forest economics and silviculture. At thebeginning of the study, four team members at-tended training courses on the application ofsocial science research methods in naturalresource management.

Indigenous KnowledgeLocal people have rich local knowledge that canbe as useful and effective as modern technol-ogy. For example, villagers developed a sustain-able system of establishment and managementof Chinese fir plantations over the last 600 years(Wu 1984, Yu 1994). This, and other indigenousknowledge, was used in our research to developstrategies for rehabilitating degraded Chinese firplantations.

Conventional Forest Survey MethodsConventional forest survey techniques wereused for part of this research. Four investigationswere undertaken. Thirty sample plots were usedto gather information on types and speciescomposition of evergreen broadleaf forest, theextent and causes of forest degradation, and treegrowth and soil fertility under differentcultivation practices in Chinese fir. Otherinvestigations measured the growth of crops andtrees and changes in soil physical features andfertility in agroforestry systems.

Farmer TrainingLack of necessary skills among farmers is oneof the constraints to rehabilitation. Hence,improvement in farmer skills is of criticalimportance in forest rehabilitation in short termand sustainable resources management in thelonger term. Five training courses wereorganised, involving more than 250 villagers.

Background Information onDongxi and Xiangjian Villages,Huitong County

Location and Biophysical ConditionsHuitong County (26°40’00”-27°09’13” in Nlatitude, 109°26’48”-110°07’56”in E longitude)is in the west of Hunan province (Map 1.1 inChapter 1). Research was undertaken mainly intwo villages of the county, Xiangjian andDongxi, both of which administratively belongto Ma’an Township. They are near to each otherand about 25 km from the county town that is



the political and cultural centre of the county.Both villages have good access to transport as ahighway and a railway cross them.

Huitong has a subtropical monsoon climate.The frost season is short, both rainfall andrelative humidity are high, and there are fourdistinct seasons in a year. Climatic data (1957-1992) show a frost season of about 60 days,mean annual temperature is 16.6oC, mean annualrainfall is 1265 mm, 54% of which comesbetween April and July, with the greatest in May(212 mm) and the lowest in December (37 mm).Average annual evaporation is 1148 mm andrelative humidity 82%.

The villages of Xiangjian and Dongxi aresituated in a landscape of hills and lowmountains. The altitudinal range of the area is50-650 m above sea level. Forests cover themountains while in the valleys there are paddyfields, rain-fed crop lands and streams. Thestreams are the primary source of water forirrigation. There are two broad categories of soil:‘red soil’ makes up 40% and occurs in areasbelow 400 m, ‘yellow soil’ makes up 60% andoccurs in areas above 400 m. Most soil is acidicand deep, with organic matter content higherthan 1%. The subtropical climate and relativelyfertile soil are suitable for the growth of forest,food crop production and animal husbandry.

Land Uses and Forest ResourcesMost land is classified as forest and forest land.Xiangjian village with an area of 1234 ha, has73% forest land and 4% crop land with theremainder for other uses. Dongxi village has

1021 ha of land of which 75% is forest land,5% crop land (Table 4.1).

Crop landCrop land is used intensively with villagersusually growing one summer crop and onewinter crop annually. Rice is the main summercrop and this is the staple food in the villages ofthe area. Villagers also grow other food cropssuch as corn, sorghum and sweet potato, andcash crops such as cotton, fibre plants,watermelon, chilli and ginger. Winter cropsinclude rape seed, wheat and beans.

Chinese fir plantationsForests in the two villages consist of Chinesefir plantations, evergreen broadleaf forests, andcash tree crops. Chinese fir plantations are verycommon in the provinces of Hunan, Fujian,Guizhou, Guangxi, Jiangxi and Zhejiang and avery important source of industrial timber. Theyrepresent the largest forest type in Xiangjian andDongxi, making up about 35-45% of area (Table4.1) and 72% and 81% of total standing timbervolume of the two villages respectively.

Evergreen broadleaf forestEvergreen broadleaf forest of the two villages,Xiangjian and Dongxi, is secondary forestformed by natural regeneration. It is smallest inextent, comprises just about 10% of area (Table4.1) and 28% and 19% of standing timber vol-ume respectively in Xiangjian and Dongxi.However, these secondary natural forests aresignificant in local livelihoods and welfare. They

Table 4.1 Land use in Xiangjian and Dongxi

Land use Xiangjian Dongxi

Ha % Ha %

Total land area 1234 100 1021 100Forest land 910 73 775 75

Chinese fir plantations 451 36 451 44Evergreen broadleaf forests 168 13 105 10NWFPs 290 23 219 21

Food crop land 60 4 49 4Other (roads, water, buildings) 264 21 197 19

Source: Project survey 1995.



provide villagers with fuelwood and NWFPs,and timber for marketing and making farm tools.Our household surveys showed each householduses about 0.5-0.7 m3 year-1 of fuelwood. In1995, the volume of fuelwood used was largerthan the volume of commercial timber used inthe two villages, which is common in south andsouthwest China. Moreover, evergreen broadleafforest provides more environmental servicesthan Chinese fir and NWFP plantations.

Broadleaf trees are the dominant species inthese secondary forests, but conifers also occur.The most common broadleaf species include:

Castanea henryi, C. mollissima,Castanopsis fargesii, C. sclerophylla, C.tibetana, Cyclobalanopsis glauca, Quercusfabri, Cinnamomum camphora, C.porrectum, Litsea cubeba, L. coreana,Machilus lichuanensis, M. pauhoi,Neolitsea cambodiana, Liquidambarformosana, Dendrobenthamia japonica,Adinandra bockiana, Elaeocarpusjaponicus, Sloandea sinensis, Ilexchinensis, Diospyros kaki, Engelharatiaroxburghiana, Styrax suberifolius,Huodendron tibeticum, Sorbus folgneri,Photinia biauverdiana, Eriobotriacavalerieri, Daphyniphyllum oldhamii,Ulmus parvifolia, Tarenna mollissima,Clethra varbinervis, Meliosma flexuosa,Mallotus japonicus, Albizia macrophylla,Kalopanax septemlobus, Celtis sinensisand Ormosia henryi.

NWFP plantationsThese plantations are the second largest,making up about 20% of area in the twovillages respectively (Table 4.1). They are amajor source of cash income of local villagers.The main species are bamboo (Phyllostachyspubescens), tea oil (Camellia oleifera),Vernicia fordii, chestnut, orange, and plum.Bamboo has become an important species inHuitong because of new factories using it asraw material.

A large proportion of NWFP (cash tree)plantations are in villagers’ yards and/or

homesteads. Most plantations are part of anagroforestry system because villagers usuallyraise cows, buffaloes, pigs, chickens andducks in them. In recent years villagers haveraised goats in their plantations as well. InDongxi village households have an averageholding of 30 goats and derive more incomefrom animal husbandry than villagers inXiangjian (Table 4.2).

Forest TenureThere have been frequent, radical changes inforest tenure over the last 50 years. Before 1952,over 90% of forests in Xiangjian and Dongxivillages were the private property of landlordsand rich peasants. Private forests were confiscatedand then equally redistributed to all ruralhouseholds (including former landlords and richpeasants) in the early 1950s. These private forestswere collectivised in 1956 when advancedagricultural cooperatives were formed. They weretransferred in 1958 from advanced cooperativesto communes, larger collective institutions. In1961, the ownership of timber forest wasdevolved from commune to production brigade(former advanced cooperatives); and ownershipof cash tree plantations from commune toproduction team, a level lower than productionbrigade. These property regimes remained till theearly 1980s.

In the two villages, the Chinese government’sforest reform in 1982 allocated 40% of collectiveforestland/forest to farmer households as familyplots. Most of the remaining 60% wasdistributed to households as responsibility hillsin 1984. In total, 92% of forest land of Xiangjianvillage and 66% of forest land of Dongxi villagewas distributed to individual households. Thistenure pattern has so far been maintained andhousehold management is the dominant formof forest management.

Forest as a Major Income SourceXiangjian and Dongxi villages are typicalagricultural communities. Agriculture, broadlydefined, is the primary pursuit from which mostof total output value or farmer income is generated.



In 1993 agriculture generated nearly 90% offarmer income in Xiangjian, and even more inDongxi village (Table 4.2). Forestry, as a sub-sector of agriculture, contributed around 30%of farmers’ income in 1993. The contributionof forest to the economy in the two villages ishigher than in Yuanmou, Yunnan (Chapter 2)and Shanxin village, Cangwu, Guangxi (Table3.2, Chapter 3).

Social and Economic SettingDemography of the two villages is similar. Therewas a population of 1170 in 335 households inXiangjian village (1995). Ethnic minoritiesmake up more than slightly half of thepopulation: Dong nationality accounts for about37% and Miao 18%. There were 880 people oflabour age, with equal numbers of male andfemale. About 40% of people have primaryeducation, 40% attended junior middle schooland about 4% went to senior middle school. Theremainder is illiterate.

Dongxi village had a population of 782 in228 households (1995). Dong people make up42% and Miao 15%. There were 368 people oflabour age with 230 male farmers (62%) and138 female farmers (38%). About 43% ofvillagers have primary schooling, 43% attendedjunior middle school and 4% have a seniormiddle school. About 10% is illiterate.

Most villagers of labour age in the twovillages undertake food cropping, forestry andanimal husbandry. Less than 10% found off-farm employment (65 people in Xiangjianvillage and 15 people in Dongxi village). Data

on labour distribution by season showed thatvillagers harvest timber mainly in the periodwhen they are not involved in food production,but tree planting and camellia tending must takeplace in the same season as food production.

The two villages were at average level ofrural economy development for China as awhole in the mid-1990s. In 1993, average annualincome per person was 1302 Yuan in Xiangjianvillage and 1329 Yuan in Dongxi village. Innominal terms, in 1995 it rose to 1526 Yuan inXiangjian and 1545 Yuan in Dongxi, comparedthe national average of 1600 Yuan in that year.There were some households of Dongxi villageunable to produce sufficient food grain forthemselves. However, the wealth level washigher than that of Shanxin village, Guangxi(Chapter 3). Agriculture generated 80-90% ofincome (Table 4.2). While non- and off-farmemployment is encouraged, the forestry sub-sector should be able to make greatercontribution to local employment andlivelihoods.

Chinese Fir Plantations

Degree of DegradationThe Ministry of Forestry (MoF, now StateForestry Administration) set technical indicatorsfor fast-growing Chinese fir plantation by culti-vation region: mean annual increment (MAI) is12-15 m3 ha-1 in the central area of cultivationand 12 m3 ha-1 in other areas (Department forReforestation, MoF 1982). Our investigations

Table 4.2 Average income of villagers in Xiangjian and Dongxi villages

Xiangjian Dongxi

1993 1995 1993 1995

Source ’000 Yuan % ’000 Yuan % ’000 Yuan % ’000 Yuan %

Total income 1544 100 1650 100 1105 100 1208 100

Cropping 463 30 545 33 386 35 463 38 Forestry 559 36 428 26 343 31 205 17 Animal husbandry 337 22 337 20 312 28 408 34 Other 185 12 340 21 64 6 132 11

Sources: Data for 1993 from village statistics and for 1995 from the project’s farmer household surveys based on a sample of 21households in Xiangjian and 29 households in Dongxi. Comparisons of 1993 and 1995 data are not accurate because of differentdata collection methods.



in Xiangjian and Dongxi villages showed MAIof Chinese fir is 16.0 m3 ha-1 for the best planta-tions and 10.5 m3 ha-1 for average plantations.

Chinese fir plantations in the two villageswere classified as slightly degraded (MAI >10.5m3 ha-1) (Figure 4.1), degraded (MAI = 6-10.5m3 ha-1), and seriously degraded (MAI <6 m3

ha-1) (Figure 4.2). Based on these indicators, inXiangjian village 32% of Chinese fir plantationsis degraded and 32% is seriously degraded andDongxi village has 12% degraded and 36%seriously degraded (Table 4.3).

Causes of DegradationDegradation of Chinese fir plantations is due toseveral factors. The traditional cultivationpractice of Chinese fir is a sustainable systemwith components such as establishing modest-sized plantings on suitable sites, allowing landto lie fallow after two or three successive crops,and intercropping with food crops or cash treecrops. However, these traditional technologieswere abandoned for various reasons after about1960. From this time, extensive Chinese fir plan-tations were often developed on inappropriatesites, and fallows and mixtures were not used.In addition to changes in silvicultural practices,selective harvesting and forest tenure insecurityhave contributed to degradation of Chinese firplantations.

Site selectionChinese fir is adapted to relatively fertile, well-drained, moist soils on the lower slopes of hills.Forestry authorities have been so ambitious to

Table 4.3 Degradation of Chinese fir plantations

Degree of degradation Xiangjian Dongxi

ha % ha %

Slightly degraded 161 36 236 52Degraded 145 32 53 12Seriously degraded 145 32 162 36Total 451 100 451 100

Source: Project’s forest survey 1995.

Figure 4.1 Normal Chinese fir plantation

Figure 4.2 Degraded Chinese fir plantation

establish extensive Chinese fir plantations that theyignored the basic silvicultural principle of species-site matching. The consequence of planting Chi-nese fir on shallow, infertile soils on upper slopeshas been the slow tree growth and low productiv-ity of plantations. Our surveys in the two villagesshow Chinese fir plantations on poor sites with anMAI as low as 1.5 m3 ha-1 which is less than 20%



of the MAI of an average plantation. Inappropri-ate guidelines of the government forest departmentfor Chinese fir plantation development are a causeof this problem.

Multiple cropping (duodai lianzai inChinese)Chinese fir develops after evergreen broadleafforest is harvested or destroyed (Institute ofForestry and Pedology 1980). In its long historyof planting Chinese fir, Huitong Countydeveloped a traditional cultivation technologyincorporating fallow. The first crop of Chinesefir, planted after evergreen broadleaf forest washarvested, was followed by a second Chinesefir crop. After two crops, land was usually leftfallow for 10-20 years to recover soil fertility.The fallow practice has been discontinued inrecent decades. This changed silviculturalpractice has resulted in lower soil fertility andforest land degradation (Fang 1987, Yu et al.1992), and plantation productivity of laterrotations is low. Our surveys in the villagesindicated that average MAI of Chinese firplantations has decreased from 10.5 m3 ha-1 inthe first crop to 9.2 m3 ha-1 in the second cropand 6.0 m3 ha-1 in the third crop. These figuressuggest that the decline in tree productivitybecomes significant in the third crop.

Monoculture plantationsDegradation of Chinese fir plantation may alsobe related to the development of monoculturesinstead of the former practice of intercroppingwith either food crops or tree crops. There isexperimental evidence in the Huitong area thatsoil physical and chemical properties are betterand productivity higher in mixed stands ofChinese fir and evergreen broadleaf speciesMichelia macclurei than in pure Chinese firplantations (Chen and Zhang 1991). Thissuggests that intercropping with food crops and/or tree crops and mixed stands are options forimproving productivity and sustainability inChinese fir plantations.

Selective harvestingPoor harvesting practice is another factor in thedegradation of Chinese fir plantations. Since theearly 1980s, farmers in Huitong County (andelsewhere) have selectively harvested timberfrom large, vigorous trees and leaving the smalland weak trees. This practice resulted insparsely-stocked stands of low productivity. TheMAI of such stands is about 4.5 m3 ha-1, lessthan half that of normal stands. Adoption of thispractice is largely attributed to the very strongmarket demand and higher price for large-diameter timber.

Tenure insecurityForest tenure insecurity is a critical factor thathas contributed to forest degradation in Huitong,as well as in other areas of China. The frequent,radical changes in forest tenure, alreadydiscussed, have had at least two negativeimpacts. Significant deforestation took place inXiangjian and Dongxi villages immediately afterforests were distributed to farmer householdsin the early 1980s. Farmers, concerned thatgovernment would repossess forests and trees,responded by immediate harvesting. Accordingto a villager in Xiangjian village, over 70 ha ofChinese fir plantations established by one hamletof the village in the early 1970s were cut in theearly 1980s. Tenure insecurity has alsodiscouraged farmers from investing in treeplanting. Frequent changes of tenure are notconsistent with the long period of time requiredfor timber production. A cycle of traditionalcultivation of Chinese fir consists of twosuccessive crops plus a fallow period, whichtakes about 50 years. The negative impacts offorest tenure insecurity apply to Chinese firplantations and to other types of forest.

Strategies for RehabilitationThe above discussion suggests that strategies forimproving degraded forest land and theproductivity of degraded Chinese fir plantationsmust address silvicultural techniques, harvesting



methods and tenure security. Options identifiedfor improving silvicultural techniques includedsome agroforestry systems, development ofmixed species forest stands, and replacement ofChinese fir with multipurpose species throughplanting and/or natural regeneration.

AgroforestryAgroforestry is part of traditional cultivationsystem of Chinese fir in Huitong and other areasin south China, and still can be employed to helpcreate productive Chinese fir plantations today.Many types of agroforestry have been practised,including a taungya system in which Chinesefir was intercropped with food crops for the firstthree years after the trees were planted (Menzies1988, Fu et al. 1995) or with cash tree cropssuch as Vernicia fordii (a species with oil-bearing seeds). These systems were abandonedduring the period of collective agriculture fromthe 1950s to the early 1980s. Since thenintercropping with a range of crops in Chinesefir plantations has resumed in different provinces(Wu 1984, Yang et al. 2001).

Our research identified five agroforestrytechnologies to promote the establishment ofproductive Chinese fir plantations in Huitong.The first four were designed to use in youngChinese fir plantations and the fifth for olderplantations.• Chinese fir and Amaranthus sp. (grain

amaranth)• Chinese fir and Citrullus vulgarisschard• Chinese fir and Paspalum notatum• Chinese fir and Vernicia fordii• Chinese fir and Gastrodia elata

The first three options were tested in Chinesefir plantations of Lianshan Forest Farm ofHuitong County Forestry Bureau. These specieswere new to Huitong and it was necessary totest their suitability. The research team providedthe forest farm with seeds, fertilisers and atraining course, while the Forest Farmcontributed the land and labour for theexperiment.

The tests showed that these intercroppingoptions are beneficial to the establishment of

productive Chinese fir plantations through in-creased growth of the young trees. After oneyear, the average height of Chinese fir seedlingswithout intercropping was 1.3-1.4 m, incomparison, seedlings intercropped with P.notatum it was 1.6-1.7 m, with C. vulgarisschardit was 1.6 m, and with Amaranthus sp. it was1.4-1.5m. The increase can largely be attributedto an improvement in soil properties throughcultivation and added organic matter from leavesand stems. Grain amaranth takes up potassiumeffectively and so promotes cycling and use ofthis element. Soil organic matter in the Chinesefir plantation intercropped with C.vulgarisschard was 17% higher than Chinesefir plantation without the intercrop.

Growth of these species was variable. Grainamaranth, an annual plant with high proteincontent and a quality fodder for domesticanimals, grew poorly. Citrullus vulgarisschard,a species that can produce about 280 kg ha-1ofedible seeds in normal conditions, grew well.Paspalum notatum grew best, with a biomassof 5.5 kg m-2 of fresh grass. The C.vulgarisschard and P. notatum grew well enoughto show potential for intercropping with Chinesefir and to increase economic benefits.

Intercropping Chinese fir with V. fordii wasa traditional and well-developed agroforestrysystem (Wu 1984, Yu 1994), and remains apromising option. The mixture has twoimportant ecological benefits for Chinese fir.First, V. fordii is a tree that provides a canopycover able to improve moisture relations in theplantation by reducing radiation and wind speed.Second, it can improve physical and chemicalproperties of the soil. Data from the researchindicate that growing stock volume of theChinese fir-V. fordii mixture is 25-30% greaterthan that of Chinese fir monoculture.

Cultivation of Gastrodia elata is a complextechnology, and it involves the establishment ofArmillariella tabescens as the former lives onthe latter. Armillariella tabescens lives on a tree.Cultivation of Gastrodia elata builds on thisrelationship as Chinese fir plantation (and otherforests) provides a suitable site for G. elata. This



research did not test the option because it wasnot new in Huitong at that time. As an importantChinese medicine, Gastrodia elata cultivationcan be profitable. Data show that profits of thespecies stood at 20 000-90 000 Yuan ha-1 in thearea in the first half of the 1990s. As with manysuch products, there can be a marketing problemif promotion of the product results in overproduction.

Mixed standsAfter intensive discussions between farmers andthe research team and among farmersthemselves, they decided to plant bamboo indegraded Chinese fir plantations. This strategywas implemented in Xiangjian village, involvinga total of 20 households. Funds needed to plantbamboo were provided by the County ForestryBureau in the form of seedlings and the loanmust be repaid (including interest) when thebamboo is harvested.

It is too early to assess effects of the mixedplantation on soil fertility or plantationproductivity improvement. The choice ofbamboo for the mixture was based less on itsecological suitability than the fact that it cangenerate economic benefits quickly.

Replace Chinese fir with other treespeciesThis strategy aimed at rational use of land onwhich there are seriously degraded Chinese firplantations. These plantations are often the resultof planting Chinese fir on inappropriate sites orpoor harvesting practice. There are many speciesthat can be grown on sites unsuitable for Chinesefir. With the present pattern of land tenure inrural China, developing even a small plantationinvolves many households due to small size oflandholdings. Hence, this strategy not onlyinvolves choosing another species butaddressing the issue of common managementof household resources.

The establishment of a Ginkgo biloba plan-tation was designed and tested. The plantationwas developed on land in Xiangjian village witha seriously degraded Chinese fir plantation. The

plantation area was 17 ha and was managed by40 households.

A major factor that influenced selection ofthis species was the availability of funds forplantation development. In 1996 villagerswanted to replace the Chinese fir with chestnutson higher part of the mountain and oranges onthe lower slope. However, they could not raisefunds for these species from external sourcesand were only able to cover part of the costs ofestablishment from their resources. Although itwas suggested that they implement the optionstep by step with their funds, villagers wantedto reclaim all the land immediately. HuaihuaPrefecture Forestry Bureau was able to provideloans for G. biloba planting only and thevillagers agreed to this.

As in the case of mixed plantation development,potential market demand for products was mostimportant factor that influenced species selection.Villagers agreed on G. biloba because it is amultipurpose species with leaves, fruit and timberproviding potential for income generation. Inrecent years, the value of G. biloba as a medicinehas been widely recognised and many ginkgo-based food and medicinal products are beingdeveloped (Xing et al. 1998, CommonwealthAgricultural Bureau International 2000). Domesticand international demand for ginkgo products isgreat, and prices are high at present. For example,the price for ginkgo seeds is 40 Yuan kg-1 or higher,leaves sell for 2.4 Yuan kg-1 and the price for timberis about 6000 Yuan m-3. Moreover, income fromleaves can be obtained after only two or three years.

Villagers in the area have planted Ginkgobiloba for a long time but only aroundhomesteads. To develop a plantation on landbelonging to many households they had to poolland and practise common management. Ashareholding system was adopted with villagersgetting shares based on land contributed andgovernment agencies their shares through capitalinvestment and/or their work in managing theplantation. Land shares made up 20% of the totaland were distributed among the 40 households.Management shares represented 10%, half ofwhich are owned by the village committee and



forest guards and another half by representativesof Huitong County Forestry Bureau at townshiplevel and the Township Government. Theremaining 70% is owned by the investors(Huitong County Forestry Bureau and itsrepresentative at the township) who providedthe capital. Profits will be distributed among theshareholders based on shares they hold.

Each shareholder has clear responsibilities.The County Forestry Bureau’s representative attownship level is responsible for the tending(weeding and fertilising) and guarding theplantation. The village committee andrepresentatives of land shareholders assist withguarding of the plantation and implementationof the scheme for benefit distribution. A contractthat was made between the County ForestryBureau’s representative and the villagecommittee agreed that the ginkgo plantationmust be established within 3 years and generatebenefits in 10 years or the village, as landowner,will be entitled to use the land for other purposes.The farmers were allowed to cultivate food cropsin the plantation in the first 3 years when theland was not fully covered by tree canopy.

Evergreen Broadleaf Forests

Extent of DegradationWhen an evergreen broadleaf forest communitybecomes degraded, the number of tree speciesand families decreases and forest structurebecomes simpler. Tree species diversity indicese.g. Simpson Index and Shannon-Wiener Index(Department for Science and Technology,Ministry of Forestry 1994), were used to assessthe extent of degradation of secondary broadleaf

forests. Using the two indices, secondary ever-green broadleaf forests in the villages were clas-sified into:• normal or slightly degraded (Simpson Index

>0.8, Shannon-Wiener Index >3.0)• degraded (Simpson Index = 0.3-0.8,

Shannon-Wiener Index = 1.0-3.0), and• seriously degraded (Simpson Index <0.3,

Shannon-Wiener Index <1.0).The number of species is significantly

reduced in degraded forests. In seriouslydegraded forests few tree species remain andmany trees have been reduced to a shrubby form.About 70%-80% of evergreen broadleaf forestin the two villages is degraded and seriouslydergraded. (Table 4.4)

A consequence of degradation of secondarybroadleaf forests is a significant change in somesoil properties. Analyses showed soil porositywas less and there was a reduction in soil organicmatter and nitrogen in seriously degradedforests.

Causes of DegradationProlonged over-exploitation of forest resourcesis the major cause of degradation of evergreenbroadleaf forests of the two villages. Localresidents rely heavily on this type of forest fortimber, for sale and on-farm use, for fuelwoodthat is the primary energy source of the villages,and for non-wood forest products. They dependon forest to generate cash income from timberand NWFPs. Clearly, there is a need to reducedependence on broadleaf forest to haltdegradation and allow rehabilitation to takeplace. A more recent pressure on these forestsis the browsing effects of goats as goat raisingis now practised widely in the area.

Table 4.4 Extent of degradation of evergreen broadleaf forests in Xiangjian and Dongxi

Extent of degradation Xiangjian Dongxi

Area (ha) % Area (ha) %

Slightly degraded 35 21 32 31Degraded 126 75 40 38Seriously degraded 7 4 33 31Total 168 100 105 100

Source: Forest survey by this research, 1995.



Rehabilitation Strategies

Mountain closureEvergreen broadleaf forests usually occur onsteep slopes and in ravines. These areas areecologically fragile so it is easy to cause soilerosion if they are subjected to artificial planting.Natural regeneration through ‘mountain closing’is an ecologically sound and cost-effectivealternative. Our surveys indicated the cost of thistechnique was about 40 Yuan ha–1, much lowerthan for plantation establishment.

Mountain closure was a traditional practicein Huitong, and in other areas of south China,before 1950. In Huitong, every year or two alladults in each village gathered to discuss whicharea to close and what penalties would applyfor infringements. Forest owners contributedfunds and food for the meeting. All signed adocument recording the agreement. It usuallyincluded: no cutting of Chinese fir trees withouta permit; no fuelwood collection in broadleafforest at a given time of year; no tree theft; nograzing of domestic animals in newly developedareas; no clearing of forest for food cropproduction; and no collecting tea oil and Verniciafordii seeds before the mountain or forest is openfor that purpose. Tools or animals of offenderswere confiscated, and they were fined and hadto compensate for loss incurred. The offenderalso had to host a feast for all the villagers. Thegovernance mechanism worked well and therewere few offenders.

Today the situation is different and twoconstraints must be overcome in mountainclosure for natural regeneration. One is currentaccess of villagers to collect fuelwood and grazetheir domestic animals in evergreen broadleafforests. The other is the fragmentation of forestland use rights into very small patches resultingfrom inappropriate implementation of the forestreform policy in the first half of 1980s. Nowone household has several tiny patches of forestland in different locations and some kind of col-laboration among households or joint manage-ment is needed to implement the ‘closing moun-tain’ option.

The option was initiated in a forest of 133ha in the boundary area of Xiangjian and Dongxivillages in 1998. Partial closure and ashareholding system were used. Partial closuremeans closing part of the mountain during agiven period and leaving other part open. Theshareholding system was designed to addressforest holding fragmentation. Households withforest land in the closed area obtained sharesbased on their land and trees. In calculatingshares, both the area and quality of land and thestock of all trees with diameter at breast heightof 6 cm and over were taken into account. Aninventory of forest was made to estimate thestanding volume of each household’s trees.

A governance group, a supervision group anda forest guard group were set up to implementthe strategy. The governance group is responsiblefor making management regulations, distributingbenefits, coordinating relationship amonghouseholds involved and between households andvillage committee, and seeking support from thegovernment forest department. The supervisiongroup, consisting of villager representatives,oversees activities of the governance group andthe forest guard group. Duties of the forest guardgroup include control of forest fire and preventionof tree theft and illegal cutting. Benefits fromclosed forests, mainly income from timber, arethen distributed among all the stakeholders. Thegovernance group is entitled 10% of benefits,25% of income is used to pay forest guards, andthe remaining 65% is distributed to householdson the basis of the shares they hold.

Home gardensThis strategy was not designed directly for aparticular type of degraded forest. It wasintended to reduce excessive pressure on allforest types by generating income from outsidethe forest. This strategy is attractive because in-come can be produced quickly.

Home garden is a household-based land usesystem and an important component of theagricultural economy in Huitong and elsewherein south China. It is a traditional land use systemthat integrates agriculture, forestry, animal



husbandry, and processing of agricultural andforest products at one site. It takes full advantageof lands around homesteads to plant trees andraise animals to generate cash income. Thetenure of such land is usually clear and moresecure than for other types of land, so there isless conflict and a good opportunity forinvestment of capital and labour.

A survey of households and village leadersin the two villages in 1995 revealed that 129(43%) of 304 households in Xiangjian villagehave a home garden with an area of 0.03-0.05ha. In Dongxi village 139 households (73%)have a home garden mainly of 0.03-0.2 ha (butup to 0.8 ha). There are three types of homegardens in the two villages based on their maincomponents: (1) animal (goat) raising dominant,(2) a combination of animal raising and fruittrees, and (3) mixed tiny plantation of fruit andother cash trees.

Although widely adopted, home gardens inthe two villages had low productivity. Oneproblem was the inferior varieties of fruit andother cash trees that were planted so marketdemand was not great and prices were low. Forexample, the price for locally produced orangeswas 0.8 Yuan kg-1 or less in 1996-1997. Anotherreason was poor home garden management(including tending) with investment lower thanrequired. Adequate investment was estimated at6000-7500 Yuan ha-1 in young orchards and 12000-13 500 Yuan ha-1 when orchards are fruit-ing. Actual investment was only 1500-4800 Yuanha-1. This analysis suggests that identificationand propagation of species/varieties with goodmarket demand, greater capital investment, andimprovement of management skills among localvillage households must be addressed to developquality home gardens.

In response, new, better-quality varieties oforange and plum were introduced. The effortwith oranges achieved a positive result withsurvival rates of grafted varieties being 60-70%.In 1998, over 600 plum seedlings were providedfor eight households, four in each of the twovillage, and they were well established when thisreport was prepared. From several proposals,villagers chose these species/varieties because

of their market potential and ease of storage. Afruit technician from Huitong County FruitAssociation passed on grafting skills andprovided advice on fruit varieties. The researchteam provided free seedlings and graftingmaterials, which helped overcome the capitalconstraint. Skill development for home gardenoperations and management was promotedlargely through training courses. Four courseswere organised, covering topics such as pest anddisease control, pruning and fruit storage. Apartfrom skill development, these training coursesenabled some households in the two villages toestablish and maintain contacts with fruittechnicians in Huitong County town.

Disincentives to ForestRehabilitationThere are government policies and practices thatare disincentives to farmers to plant trees andmanage existing forests and plantationssustainably.

Taxes and FeesHigh taxes and fee charges on timber productsare very common in China. In Huitong County,in the case of Chinese fir log grade, localgovernment tax and fees amount to about aquarter of wood trade prices. The forestdepartment collects about 14% as reforestationfund but little is returned to farmers. Farmersobtain 16% of the log price after deductinglogging costs (Table 4.5).

In the 1990s, costs of planting, tending andguarding of Chinese fir plantations were 40-45Yuan m-3, which meant farmers obtained a netincome of about 80 Yuan m-3 (129 Yuan minus40-45 Yuan) for Chinese fir logs that took 20years to grow. In contrast, there are few taxesand fees on fruit and fuelwood. According to afarmer in Xiangjian village, he can gain anincome of 200 Yuan m-3 from fuelwood withless capital and labour investment. A farmer inthe village leased 2000 orange trees from hisvillage collective and earned an income of 20000 Yuan year-1. This in part explains whyfarmers are not enthusiastic about timberplantation establishment and are more interested



in fuelwood and NWFPs. It also explains whysome farmer households converted their timberforests into orchards, bamboo plantations orother cash tree plantations.

Harvesting and MarketingRegulationsRestriction of rights to harvest and marketforest products, especially timber, is anotherfactor that discourages farmers from plantingtrees and managing forests. Wood harvestingis highly regulated through a cutting quota andcutting permit system that extends throughoutChina. In Huitong, the County ForestryBureau and County Planning Commissionallocate the yearly cutting quota to townships,each of which is responsible for reallocatingquotas to villages under its governance.Village offices then allocate their quotas tohouseholds. Quotas are limited and obtaininga cutting permit is a difficult, costly and time-consuming process.

The timber market is a monopoly with onlythe government forest department, or institutionsdesignated by the forest department, entitled topurchase timber from villages and households.As a consequence, farmers receive prices thatare much lower than market prices. Forexample, the market price for Chinese fir logwas 800 Yuan m-3 in 1996, while state timbercompany obtained a profit of 185 Yuan, andfarmers about 130 Yuan (including costs oftending and guarding) (Table 4.5). In contrast,regulations on marketing NWFPs are much less

restrictive as farmers can sell their products inlocal free markets.

Conclusions

Causes of Forest DegradationDegradation of Chinese fir plantations andevergreen broadleaf forests in Huitong Countyis associated with many factors, includinginappropriate silviculture practices, poorharvesting methods, tenure insecurity and over-dependence on forest resources. Silviculturalmethods, such as planting Chinese fir onunsuitable sites, planting successive crops ofChinese fir without a fallow period andestablishing monocultures are technical issues butare promoted by government forestrydepartment’s planning and guidelines onplantation development. These guidelines oftenhave negative consequences and neglecttraditional management practices developed over600 years.

On Strategies to RehabilitateDegraded ForestsTo be effective, strategies for rehabilitation mustaddress causes of forest degradation and con-straints to rehabilitation. They must be overallpackages rather than technology ‘fixes’.

Silvicultural practicesIt is crucial to revise guidelines on plantationdevelopment on the basis of both new and

Timbergrade

Grossincome

Tax Refores-tation fund

Other Costs Logging costs

Profits Income

Localgovernment

Forestrydepartment

Transportand marketing

Farmers

Fees fortownship

and village

Table 4.5 Distribution of income from two grades of Chinese fir timber

Yuan m-3

Yuantiao* 580 104 119 76 15 80 60 50 77% 20 21 13 3 14 10 9 13

Log** 800 76 121 96 14 110 185 70 129% 10 15 12 2 14 23 9 16

Source: Huitong County Forestry Bureau, its representative office at township, and Ma’an Township Government* Less than 5 m in length and 8 cm in diameter.** Over 5 m in length and 8-12 cm diameter



indigenous knowledge to prevent the establishmentof Chinese fir plantations with low productivitypotential. The unsuitable sites planted with Chinesefir should be allocated to species better adapted tothem. Agroforestry technologies that ensuredsustainable management of Chinese fir over thelast six centuries, including intercropping trees withfood crops or cash tree crops, still have animportant role to play.

It is essential to reduce dependence of therural population on forest resources. Homegardens can provide income-generatingopportunities and reduce pressure on forests tosome extent. The reduced pressure should makeforest rehabilitation easier.

Improved rural servicesImproved rural credit services and expandedsources of funding are needed to overcomefarmers’ lack of capital to invest in rehabilitatingdegraded forests and production activities.Farmers’ access to commercial credit is limited,as they often cannot meet bank requirements forcollateral. One possibility is to introduce andadopt an innovative rural financial system, e.g.using the Grameen Bank approach. Another isto involve outside institutions to provide fundsfor development of plantations and/or homegardens. Capital input by Huaihua PrefectureForestry Bureau into ginkgo plantationestablishment in Xiangjian village and byHuitong County Forestry Bureau into bambooplanting are examples of such an approach.

Many farmers lack the skills and technicalknowledge to effectively manage forestry andagroforestry activities. This requires betteraccess to technical information and improvedextension services, including training courses.

Organisation of common management ofhousehold resourcesFragmented forest land holdings make it difficultfor farmers to adopt options such as naturalregeneration of broadleaf forests through‘closing mountain’ and tree plantationestablishment. Tenure issues are very sensitivebut joint forest management, such as a

shareholding system, can rearrange forest tenureto the advantage of the landowners. This mustbe achieved on a voluntary basis and with propergovernance or there will again be a crisis ofvillagers’ trust in tenure policy.

Policy Changes are EssentialRestrictive regulations on timber harvesting andmarketing and the high level of taxation and feeson timber sales are disincentives to farmer toestablish and manage plantations of Chinese firand other timber species. Farmers are notinterested in establishing timber plantationsbecause of the unsatisfactory returns from timbersales as a result of high taxes and fees. Therestrictive regulations discourage them frominvesting in tree planting and sustainablemanagement of forest. Changes to these policiesto give farmers greater freedom in marketingand to obtain a greater return from timberproducts are essential to encourage theestablishment and management of timber-producing forests and plantations.

Problem and Strategy IdentificationForest degradation is a complex process and therehabilitation of degraded forest requires adynamic partnership of villagers, technicalprofessionals and government officials.Villagers know their circumstances and localconditions and are able to make significantcontributions to the identification of problemsand constraints and the development ofstrategies and options to address them. They arethe means of achieving rehabilitation andsustainable management of degraded forest, andultimately the beneficiaries of it. Involvementof technical personnel can provide expertise andinnovation in developing effective strategies andimprove communication between governmentagencies and villagers. Government officials areimportant for forest rehabilitation, because theirparticipation will help make effective policiesand their networking with government agencieswill help villagers to gain access to public sec-tor resources such as technical extension, ruralcredit, and market information services.



ReferencesChen, B. 1992. Status, causes and improvement

strategy of soil degradation in timber plan-tations in China. In: Sheng, W. (ed.). Re-search on site degradation of timber planta-tions. China Science and Technology Press,Beijing. (In Chinese)

Chen, C. and Zhang, J. 1991. Study on the effectof mixed forest of Cunninghamia lanceolataand Michelia macclurei upon forestproductivity and ecological balance. In: Shi,K (ed.). Development of forestry science andtechnology in China. 170-177. China Scienceand Technology Press, Beijing.

Commonwealth Agricultural BureauInternational. 2000. Forestry Compendium,Global Module. CAB International,Wallingford, UK. (CD version)

Department for Reforestation, Ministry ofForestry (ed.). 1982. Technologies for highyield Chinese fir plantations. China ForestryPublishing House, Beijing. (in Chinese)

Department for Science and Technology,Ministry of Forestry. 1994. Methods forresearch on forest ecosystem in fixed site.China Forestry Publishing House, Beijing.(In Chinese)

Fang, Q. 1987. Impacts of many crops ofChinese fir in succession on soil fertility andtree growth. (Chinese). Forestry Science 23:389-397. (In Chinese)

Fu, J., Fu, M., Fang, M. and Wang, A. 1995.The main agroforestry models in the easternsubtropics of China. In: Cai, M. and Hu, S.(eds.). Integrated research in farm forestry.41-49. China Science and Technology Press,Beijing.

Huitong County Forestry Bureau, 1993a.Huitong County annals of forestry. Huitong,Hunan. (In Chinese)

Huitong County Forestry Bureau. 1993b. Dataof forest inventory of Huitong County 1991-1995. Huitong, Hunan. (In Chinese).

Institute of Forestry and Pedology. 1980. Acollection of papers on ecology research ofChinese fir plantations. Chinese Academy ofSciences, Shenyang. (In Chinese)

Menzies, N. 1988. Three hundred years oftaungya: a sustainable system of forestry inSouth China. Human Ecology. 16: 361-377.

Research Team of Intensive Cultivation ofChinese Fir. 1992. General report of researchgroup for ‘National Research Program onIntensive Silviculture Techniques on ChineseFir Plantation’. In: Sheng, W. (ed.). Researchon site degradation of timber plantations.China Science and Technology Press,Beijing. (In Chinese)

Wu, Z. (ed.). 1984. Chinese fir. China ForestryPublishing House, Beijing. (in Chinese)

Xing, S.Y., Tian, C.J., Wang, C.B., Guo, J. andWang, Y. 1998. Review of the silvicultureand kernel utilization of Ginkgo biloba L.World Forestry Research 11(2): 32-37.

Xu, H. 1992. Dynamic characteristic of soilfertility of forest land and problems of soildegradation in timber plantations. In: Sheng,W. (ed.). Research on site degradation oftimber plantations. China Science andTechnology Press, Beijing. (In Chinese)

Yang, X., Luo, J. and Li W. 2001 Agro-silviculture systems. In: Agro-ecologicalfarming systems in China. (ed. W. Li) pp.105-131. Man and the Biosphere Series Vol.26. UNESCO, Paris and The ParthenonPublishing Group New York, USA andCarnforth, United Kingdom.

Yu, X., Ye, G., Lin, S. and He, Z. 1992. Anapproach to the culture system for Chinesefir. Journal of Fujian College of Forestry12(3): 259-263. (In Chinese)

Yu, X. 1994. Chinese fir silviculture. FujianScience and Technology Press, Fuzhou. (InChinese)

Zhu, Z. 1991. Forest ecology. China ForestryPublishing House, Beijing. (In Chinese)


IntroductionZhejiang province is in the subtropical evergreenbroadleaf forest zone. Due to excessive humanintervention, however, most of evergreenbroadleaf forest has now disappeared and hasbeen replaced by secondary natural Masson pine(Pinus massoniana) forest and man-madeplantations of Chinese fir (Cunninghamialanceolata), bamboo, tea and other cash treecrops (NWFPs). Remaining evergreen broadleafforest is secondary in origin.

A significant proportion of forests, both timberand NWFPs, in the region is degraded or seriouslydegraded. Standing timber volume and qualityhas decreased and biodiversity is diminished.Local rural communities and other forest-dependent people are able to obtain lessenvironmental and ecological services and tocollect fewer products and generate less incomethan formerly. Farmer income in the mountainousarea is 75% of the provincial average andtherefore it is a great challenge to rehabilitate andmanage degraded forests and plantations togenerate income and improve livelihoods.

Few comprehensive studies have been madeon why forests in the region are degraded sounderstanding of these issues is still limited.Identification of effective strategies forsustainable management of degraded forests haslargely focused on technical solutions.

In response, our research, carried out in thethree villages of Gaohong Township of Lin’anCounty, Zhejiang, used a participatory approachand aimed at understanding root causes ofdegradation of timber forest, bamboo plantations(a major source of cash income) and teaplantation. We also examined disincentives andconstraints to forest rehabilitation, anddeveloped strategies for forest rehabilitation andimprovement in local livelihoods.

This chapter presents research methods andbackground information on the three villages,deals with three types of forests/plantations: tim-ber forest, bamboo and tea plantations by ana-lysing the causes of degradation and disincen-tives and constraints to their rehabilitation., Re-habilitation options are described and the devel-opment and extension of effective forestry tech-nology discussed. Lessons and implications aredrawn that may have wider application to sus-tainable management of degraded forests.

Research MethodsAs an action-oriented research project ourresearch in Gaohong followed a logical procedureof diagnosis, design and delivery, and stressedparticipatory and interdisciplinary approaches.Participatory Rural Appraisal (PRA) was used.PRA has emerged from a synthesis of Rapid RuralAppraisal, Agroecosystem Analysis, Diagnosis

Chapter Five

Rehabilitation and Sustainable Management ofDegraded Forests in Gaohong, Lin’an County, Zhejiang

Li Minghua1, Liu Dachang2, Shen Yueqin1, Wang Anguo3, Wei Xinliang1, Yu Shuquan1

and Zhou Guomo1

1 Zhejiang Forestry College, Lin’an.2 Southwest Forestry University, Kunming P.R. China and Center for International Forestry Research, Bogor, Indonesia.Currently at The Mekong Institute, Khon Kaen, Thailand.3 Lin’an County Forestry Bureau, Zhejiang.



82 Rehabilitation and Sustainable Management of Degraded Forests in Gaohong, Lin’an County, Zhejiang

and Design and other appraisal techniques withaction research and community organizationtechniques (Chambers 1994a,b,c; Rocheleau1999). This type of social science research wascomplemented by conventional forestrymethods such as soil survey, forest inventoryand landscape analysis.

Using a Participatory Approach

Participatory diagnosisA participatory rural appraisal (PRA) team wasassembled comprising 15 farmers, 15 technicalspecialists and researchers, and 5 governmentofficials. PRA tools such as secondary datacollection, participatory direct observation,household survey and group interviews wereused.

A large amount of secondary data wascollected from relevant government agenciesand technical institutions at county and townshiplevels. For example, information was derivedfrom Lin’an County Annals, Lin’an CountyAnnals of Forestry, Report on Forest Inventoryin Lin’an County and yearly statistical data.

Direct field observations were made byresearchers together with farmers. The farmershad good knowledge of local forest managementand historical changes in environment. Theywere encouraged to draw sketch maps to supporttheir observations.

As occurred at research sites in otherprovinces in this project, the primary methodfor farmer household survey was semi-structured interviews, complemented by aquestionnaire survey. Sample households wereselected through random sampling from 659households in three villages. About 350 villagersfrom 175 households were consulted during thesurvey. The villagers consulted representedabout 20% of the population in the three villages.

Representatives of seven groups of villagersin each of the three villages were interviewed.The seven groups were: senior villagers (60years old or over), village leaders, marriedmiddle-aged and young people, unmarriedyoung people, women, the poor, and therelatively wealthy. Twenty-one group interviews

were conducted, each involved 4-8 persons andlasted up to 2 hours.

Participatory planningDesign aimed at developing effective strategiesfor rehabilitation and sustainable managementof degraded forests based on identification andanalysis of problems and constraints at thediagnosis stage. We worked with farmers andgovernment officials in an iterative manner toidentify options and evaluate the role of existingtechnologies and indigenous knowledge. Aftera number of discussions, several options wereagreed for each type of degraded forest to allowfarmers to choose the option(s) that best suitedtheir specific conditions and circumstances.

Testing options/strategiesIn the delivery phase, more than 10 households,with support of researchers and forest extensionworkers, were involved in testing effectivenessof options identified. Participating householdsrepresented different economic developmentlevels and had different forest resources and typesof degraded forest. The options that provedeffective were then disseminated to villagersthrough training courses, printed materials, farmerhousehold visits, and the existing extensionnetwork, and mass media (cable TV and video tape).

Conventional Forestry Methods

Soil surveyField surveys were made to determine soil typesand their distribution and collect soil samples.Analysis on chemical properties of soil samplesfrom 29 transects in different forest types of thethree villages was made to assess soil fertility.

Forest vegetation surveysAn inventory was made in 14 sample plotsrepresenting different types of forestcommunities. Both tree layer and shrub layerwere surveyed in each of the sample plots. In thetree layer, species, diameter at breast height, treeheight, trunk height to first branch, canopy classand health were recorded. Names, numbers anddimension of shrubs were also recorded.


83Li Minghua, Liu Dachang, Shen Yueqin, Wang Anguo, Wei Xinliang, Yu Shuquan and Zhou Guomo

A second field survey focused on types andamount of forest products. The surveyconcentrated on historical changes in forestvolume, forest land and environment, and on theextent of forest degradation during the previous20-30 years. We (the research team) worked withseveral experienced villagers to assess thehistorical changes based on field survey, villagers’recall, and secondary data obtained from forestinventories respectively in 1983, 1989 and 1994.This survey is by its nature participatory.

Forest landscape analysisA landscape study was used to analysedegradation and features of forest tocomplement findings from other research.Criteria for the classification of forest polygonswere decided first. Plots were located on mapsduring the field survey, followed by a stratifiedstatistical analysis by forest type. The resultsprovided an estimate of area, dominant species,and features of each type of forest. Thisinformation assisted in assessing the extent offorest degradation.

Background Informationon Lin’an CountyLin’an County (29o56’-30o23’N latitude and119o11’-119o52’E longitude) is located innorthwest Zhejiang province (see Map 1.1,Chapter 1). Our research was mainly undertakenin the villages of Chenjiakan, Hongqiao andShangfeng, all of which are under jurisdictionof Gaohong Township, though some optionsidentified were also tested in the township-owned Gaohong Forest Farm and several othervillages in Lin’an.

Climate and SoilSituated to the south of the Yangtze River, Lin’anCounty has a subtropical climate, with fourdistinct seasons a year. Winter and summer are124 days and 114 days respectively, and springand autumn are each 64 days. Mean annual rain-fall is 1415mm.

Soil is typically red earth, with two subtypes:yellow-red soil and red soil. Yellow-red soil is

found on 55% of land area of the villages. Redearth is generally assessed as a fertile soil butdue to soil erosion on some land in recent years,profiles are becoming shallower and fertility hasdeclined. Some of this land degradation is a re-sult of planting Lei bamboo on steep slopes byfarmers. Soil fertility is also broadly related totypes of forest and age of trees on it with greaterfertility in natural forest than in man-made plan-tations, and under older stands.

Land Use and Forest ResourcesThere are 1128 ha of land in the three villages.The largest land use is forestry with the balancecomprising food crops, settled area and waterbodies. Forest land comprises 76% of total areaand crop land 10% (Table 5.1).

At present, 98% of forest land is covered byforests (Table 5.1). Masson pine forest andevergreen broadleaf forest are usuallyregenerated naturally whereas Chinese fir andtea are planted. Bamboo is either natural orartificial depending on the species. Plantationsof Chinese fir and Lei bamboo make up 32% offorest land or more than 80% of plantation areain the villages. There are only 43 ha of teaplantation.

All timber forests are young or middle-agedstands at present so the standing timber volumeis low. It averages 13.5 m3 ha-1 in Shangfengvillage, 20.7 m3 ha-1 in Chenjiakan and 24.3 m3

ha-1 in Hongqiao. Without mature forests, localfarmers practise selective logging of larger treesfrom these immature stands to meet their timberneeds. The practice has contributed todegradation of timber forest.

Local people use forest resources to meettheir varied needs. Masson pine forest andChinese fir plantation are primarily for timberproduction. Evergreen broadleaf forest is fortimber, fuelwood, NWFPs and ecologicalservices. Bamboo and tea are cash tree crops.Bamboo is planted primarily for bamboo shootsfor cash income and on-farm consumption. Teawas also a major source of cash income but fromthe late 1980s was replaced by bamboo shoots.Evergreen broadleaf forest contains most



biodiversity, while Masson pine forest, Chinesefir plantation, bamboo forest, and tea plantationare much less diverse.

Agricultural and Forest ProductsImportant agricultural products include rice,wheat, barley, watermelons, vegetables and rapeseed. Primary forest products are fresh bambooshoots, dried bamboo shoots, timber and tea.Food grains, especially rice, are produced bothfor on-farm consumption and sale to fulfilfarmers’ obligation to the state, while otheragricultural crops are mainly consumed on-farm.

Bamboo shoots now have become the mainsource of farmers’ cash income. About 80% ofshoots is marketed and 20% is consumed on-farm. Part of the fresh shoot crop harvested in anormal season is marketed, part is used on-farm,and the remainder is processed into dry shoots.About half of the dried shoots are marketed.

Timber was major source of cash income forfarmers in the villages. It has been replaced bybamboo shoots since the late 1980s. It is nowlargely for on-farm consumption, harvestedfrom household’s own plots and/or purchasedfrom forests under collective management. Thetimber market is monopolised by the state-owned timber company, which is one reasonfarmers have been less enthusiastic about tim-ber plantation establishment.

Although the market demand for quality teais still high, the quality of tea produced in thevillages is poor and the amount farmers are ableto sell is declining. Fresh leaves are sold tobuyers who come to the villages and then sellthem to tea processing factories. Most teaproduced in the villages is used locally.

Forest Tenure and ManagementAs in other parts of China, Lin’an County im-plemented the policy for forestry reform in theearly 1980s. The term ‘sanding’ is often used inChinese literature to refer to the forest reform.Village collectives distributed part of their for-ests and forest lands to village households formanagement, while retaining some under col-lective management. Forests/forest lands weredistributed to farmer households in two catego-ries: family plot (ziliushan in Chinese) and re-sponsibility hill (zerenshan in Chinese) (Liu2001) that are often called liangshan in Chineseliterature on forestry.

In Lin’an County, 122 200 ha, or 50%, offorest were distributed as responsibility hills; 42700 ha, or 17%, as family plots; and 81 900 haor 33% were under collective management.However, in the three villages nearly 80% offorests were distributed to rural households, andonly about 20% remained under collective man-agement (Table 5.2).

Table 5.1 Land use

Land uses Area (ha) % Total Area % Forest land

Total area 1128 100

Forest land 853 76Timber forests 561 50 65

Masson pine 372 33 43Chinese fir 117 10 13Evergreen broadleaf forest 72 7 8

Bamboo forests 229 20 27Lei bamboo* plantation 158 14 19Bamboo forest producing dry shoots** 71 6 8

Tea plantation 43 4 5Non-forested land 20 2 2

Crop land 115 10Others 160 14

Sources: Based on the field survey 1995.Notes: * Phyllostachys praecox f. preveynalis ** Including Phyllostachys nuda, Phyllostachys dulcis, and Phyllostachys iridenscens.



In the case of family plots, there is a divisionof use rights and land ownership. Use rights toland were distributed to farmer households onlong-term basis, while the collective remainedthe owner of land. The policy dictated that a landcertificate had to be issued to each householdwhen forest or forest land was allocated to them.Trees planted on family plots are privateproperty and both the land use rights and treeownership can be inherited. A cutting permit isrequired to cut trees on family plots, althoughin the villages this regulation has been strictlyapplied only to the harvesting of commercialtimber and not to harvesting for on-farm use.

Responsibility hill areas were distributed tofarmer households in the attempt to improve themanagement of existing forests. Benefits fromresponsibility hills are shared between thecollective as owner and the householdsassuming responsibility for forest management.A cutting permit is essential to cut trees onresponsibility hills.

The usual practice is for the village collectiveto appoint part-time or full-time forest guardsto look after forests remaining under collectivemanagement. Income from the forests goes tothe collective and is used to cover publicexpenses including part of village leaders’ salary.Village leaders must obtain cutting permits andlogging operation planning from the forestauthority to harvest timber from collectiveforests. Villagers are given priority to purchasetimber for their own use from collective-managed forests.

DemographyAlmost all the population of 10 513 in Gaohongtownship in 1995 were Han Chinese. Based onour random sample survey, gender distributionwas 51% male and 49% female. Those of la-bour age comprised 65% of the population. 45%of the labour force was working in primary in-dustry (agriculture, forestry, animal husbandry,and fishery and aquaculture), 28% in second-ary industry (manufacturing, construction andtransport) and 27% in tertiary industry (mainlyservices). A high proportion of the populationhas primary education and 14% is illiterate. Onlyabout 8% of villagers completed senior highschool education or higher level. Most ofmiddle-aged and senior villagers, who have hada low level of education, remain working inagriculture, while the better-educated youngergeneration prefers non- and/or off- farmactivities as their employment and sources ofincome.

In the past, the three villages were typicalsubsistence communities. Villagers producedfood grain and vegetables and raised animalsmainly for self-consumption, and managedtimber forest and tea plantation to generate cashincome. This pattern, however, has changedsince the late 1980s due to the availability ofother opportunities for employment and incomegeneration. Sources of farmer income havediversified to include bamboo, transportation,manufacturing, food crop production, trade, andanimal raising in descending order. In brief,villagers rely on less agriculture and moremarket related economic activities.

Table 5.2 Forest land by tenure

Village Forest land (%)

Household managed Collectively-managed

Family plots Responsibility hills

Chenjakan 66 19 15Hongqiao 56 9 35Shangfeng - 97 3Total 43 36 21

Source: Field survey 1995



Farmer income at the site is relatively higherthan that in other mountainous areas. However,there was a considerable disparity of incomeamong households. Data of income for the threevillages show that in the mid-1990s, income ofthe lowest income group was 18% of that of thehighest group in Shangfeng village, 32% inHongqiao village, and 46% in Chenjiakan.Interestingly, the disparity was not because of adifference in households’ landholding size butwas due to the different industries pursued byvillagers. In Shangfeng, in the highest incomegroup were those with part-time off-farmemployment; and in the lowest income groupwere those who pursued agriculture only. InChenjiakan, most households planted bambooand so the disparity of income among householdswas not as great as that in Shangfeng.

Overview of ForestDegradation in the VillagesOur surveys showed that a large proportion offorests in the villages was degraded. Thesedegraded forests covered about 390 ha,representing 46% of forest land. DegradedMasson pine forest made up 57% of degradedforest area. Most of the evergreen broadleafforest and tea plantations were degraded andnearly half of the bamboo forest producing dryshoots was also degraded. (Table 5.3)

From a landscape perspective, there are fourfeatures of forest degradation in the villages.• Evergreen broadleaf forest, formerly the typi-

cal vegetation in the region, is only 7% of

total area (Table 5.1). It is now secondaryforest with less species and shorter trees.

• The forest has become much more frag-mented with the majority of patches smallerthan 6 ha.

• Forest communities with a single species inthe tree layer have become dominant forestin the landscape, including Lei bamboo andChinese fir plantation, and Masson pineforest.

• Distribution patterns of some forestlandscape elements are not sound. Leibamboo plantations and Chinese firplantations established on steep slopes areirrational land use and lead to soil erosion.

Factors causing forest degradation andoptions for rehabilitation and management varyaccording to the different types of forest.

Degraded Timber Forests

Extent of DegradationAccording to general criteria of the forestdepartment, the following four types of forestsare degraded and need to be rehabilitated. Theyare:a) land with a canopy density of 10-20% and

with little likelihood to naturally regenerateinto stands;

b) young and middle-aged forests on anyunsuitable site;

Table 5.3 Degraded forests by type

Degraded forest type Area (ha) % total area of % area of samedegraded forests type of forest

Broadleaf forest 56 14 78Tea plantation 30 8 70Masson pine forest 224 57 60Bamboo forest for dry shoots 34 9 46Chinese fir plantation 15 4 13Lei bamboo plantation 33 8 20

Total 392 100

Source: Field survey 1996.



c) young and middle-aged forests with anannual growth of less than 3 m3 ha-1; and

d) forests suffering from serious diseases, pests,forest fires and human damage.About 60% of the Masson pine forests is

degraded. The 224 ha of degraded pine forestaccounts for 57% of degraded forest in theresearch area (Table 5.3). Generally, Massonpine forests are growing on sites suitable for thespecies. In the 1960s and 1970s there were densesecondary Masson pine forests with an averagediameter breast height (dbh) of 26-28 cm, amean tree height of 11-12 m and timber volumeof 105-120 m3 ha-1. When we surveyed theforests in 1995, the stands were less productivewith an average dbh of 8-10 cm, a tree height of6-7 m, and a timber volume of 20-30 m3 ha-1.

Nearly 80% of evergreen broadleaf forestsare degraded (Table 5.3). These forests aremainly on cool, north facing slopes that are steepand rocky, and with shallow soil. Dominantspecies include Castanopsis sclerophylla andCyclobalanopsis (Quercus) glauca, often mixedwith a few Masson pine trees. Evergreenbroadleaf forests were in good condition in the1960s and 1970s, with mean dbh 10-12 cm(maximum 20 cm) and a dense canopy of (80–90%). They are now in poor condition with meandbh 6 cm, mean height 6-7 m and a more opencanopy (30-50%).

Chinese fir plantations are in the bestcondition among the timber forests with only13% degraded (Table 5.3). The degradedChinese fir stands are young or middle-agedwith low productivity. For example, 20-year-oldplantations have an average dbh of 10 cm and aheight of 7 m, standing timber volume of about54 m3 ha-1, and mean annual increment of 2.7m3 ha-1. All these indicators are lower than thoseof a normal Chinese fir stand.

Factors Responsible for DegradationOur survey shows that immediate causes of tim-ber forest degradation included tenure insecu-rity and over-harvesting of forest, including poorharvesting practices.

Deforestation resulting from tenureinsecurityIn 1981, the government decided to allocate non-forested forest land and forests to farmerhouseholds to encourage tree planting andimprove management of existing forests.Contrary to the policy objectives, this forestreform was followed by increased harvesting inmany areas. Significant deforestation occurredin the mid-1980s, immediately after the adoptionof household-based forest management. Somefarmers cut trees immediately after they weredistributed to them. This occurred in the threevillages too. One underlying cause of thedeforestation was tenure insecurity in anuncertain policy environment due to frequentchanges in polices for forest and land tenure.China experienced radical changes in foresttenure and management from the early 1950sto the 1980s. The major change included fromprivate ownership and household-basedmanagement in the early 1950s to collectiveownership and management during thefollowing 25 years. With this experience it isnot surprising that farmers chose to harvest theirtrees rather than to manage them. This attitudeis consistent with observations made elsewherein China (Yin and Newman 1997, Liu 2001).

We found in our interviews with farmers thattheir misgivings about tenure security have notyet been dispelled completely. Despite policiesand law assuring farmers of forest tenuresecurity, uncertain factors remain at theimplementation level. For example, certificateswere not issued nor contracts made for forestsallocated to farmer households in Shangfengvillage. In other cases, use rights to land wereoften reallocated by local officials. This helpsexplain why many farmers are reluctant toestablish and/or sustainably manage forests.

Over-harvesting of forestForest degradation in the villages is also closelyrelated to poor harvesting practices. Timber hasbeen the major source of cash income in the areafor a long period of time and has resulted in over-harvesting. Over the past 20 years, farmers in



the research area have practised a dysgenic se-lection in the forests by removing the best qual-ity, large-diameter trees (badamao in Chinese)and leaving poor, often unhealthy, trees. Eventoday, local villagers practise selective loggingof larger trees from immature stands becausethere are no mature forests left. Similarly,fuelwood was the only source of energy forcooking and space heating until the 1980s. Thisdependence on fuelwood caused depletion of theevergreen broadleaf forest.

Disincentives and Constraints toRehabilitation

DisincentivesDisincentives to farmers to invest in timber treeplanting and forest management included:• tenure insecurity• over-taxation on timber and• restrictive regulations on timber harvesting

and trade.Perceptions of tenure insecurity can

explain why farmers are interested in bambooproduction but not timber growing. The longperiod needed for timber to generate income forinvestors requires tenure security. In contrast,farmers in Lin’an were incredibly enthusiasticabout bamboo, because they can obtain incomefrom it in just two or three years. There havebeen similar observations of farmers’ behaviourin relation to property rights and investmentincentives elsewhere (e.g. Fortmann and Bruce1988, Besley 1995).

Poorly-designed mechanisms for benefitdistribution can discourage investment. Datafrom our surveys show that timber income isdistributed between outside stakeholders(government and forest department) and localcommunity/households. The outsidestakeholders received about half of gross timberincome (trading prices) in the form of taxes andfees. For instance, the price for Chinese firtimber is about 600 Yuan m-3. From this thegovernment collected taxes and fees of about200 Yuan m-3 and the forest department 95 Yuanm-3 for the reforestation fund. Farmers received

the balance that had to cover costs of planting,tending, guarding and harvesting. As it takes atleast 20 years before Chinese fir is large enoughto harvest, returns to farmer investment are verylow. In sharp contrast, tax on bamboo shoots isless than 1% of sale prices. From a perspectiveof benefit sharing among stakeholders, thisexplains why farmers have little enthusiasmabout timber forest management and why over-taxation and fee charges are a disincentive totimber forest rehabilitation.

Regulations on wood harvest and trade alsoimpact negatively on timber forest establishmentand management. In China, a cutting permit isrequired to cut trees. Timber harvesting mustbe carried out at a time and location, and in anamount specified by a cutting permit. To someextent, this limits farmers’ authority and makesthem unable to respond to market signals. Forexample, they may not be able to harvest andsell their wood when prices are high due to theinability to obtain a permit. Present woodmarketing practice also discourages farmersfrom developing timber forests. In most cases,the forestry department, or other institutions towhich it delegates authority to purchase timberfrom farmers and village collectives,monopolises the timber market. Consequently,timber prices are skewed and farmers receiveprices lower than those on the free market. Incomparison, the market environment forbamboo shoots is much better as farmers are ableto sell them freely.

From this analysis, it is clear that at policylevel, three disincentives must be removed toencourage farmers to rehabilitate and managedegraded forests. (1) The government needs tofurther reassure farmers about tenure security,and tenure policy and law must be strictlyenforced. While transfer of use rights to forestland is necessary, it should be through marketmechanisms, such as leasing, rather thanreallocation by political forces andadministrative orders. (2) Tax and fee chargesmust be reduced so local communities/households can receive more benefits from salesof timber. (3) Monopolised wood purchases by



the forest department should be abandoned ormodified to give farmers more authority in thewood trade and to enhance their prospects forhigher gains from timber sales.

ConstraintsTwo major constraints to the rehabilitation oftimber forests need to be overcome. The firstconstraint is the fragmentation of forest landholdings resulting from the way collective non-forested forest land and forest was allocated toindividual households in the early 1980s. Whenthis land was allocated in the villages, villagersand local government officials emphasisedabsolute equity. Lands and forests were firstdivided into many tiny plots based on tree age,species, soil quality and distance from villageetc., and then matched to family size. As a result,land holdings are very fragmented so that onehousehold has many tiny plots in differentlocations, and a land or forest on one slopebelongs to many households. Severalhouseholds each have 10 or more plots of forestin Shangfeng village, while the average forestarea is only 2.3 ha household-1. This makes forestoperations more costly and difficult, and makesmountain closure for natural regeneration offorest almost impossible to organise. One forestguard can easily patrol an area of forest if aneffective management institution exists, butunder household management each householdmust allocate one person at least part time tolook after its many tiny, scattered forest plots.

Clearly, there is a need for an improvementin forest management arrangements. Householdforest farm and shareholding systems emergingin Lin’an (and other areas) offer somepossibilities to address the problem.

In household-based management, ahousehold expands the forest land areadistributed to it in the early 1980s throughleasing use rights to land from, or in partner-ship with, other farmers. It is more efficient tomanage the larger forest land area. In Lin’anCounty up to late 1998 more than 500 householdforest farms, managing an area of over 20 000ha, were established. Initial observations sug-

gest that this form of forest management is ef-fective in overcoming the constraint of land frag-mentation. Tenants have capital and labour tomanage a larger area, while those who lease theirland use rights out usually have insufficient la-bour and investment capital.

In the shareholding system, all partiesinvolved pool their land, funds or technologiesto establish plantations and manage forest. Theysubsequently share benefits generated based onthe shares they hold through their resource input.The system can take many forms based on therange of stakeholders, including collaborationamong households; between households and thevillage collective; among households, the villagecollective and stakeholders outside of villagecommunity; and between households andinstitutions outside of community. Experiencefrom the area shows that the shareholdingsystem can largely overcome constraints of landfragmentation, while addressing farmers’ lackof capital investment and expertise throughfunds and extension services from outsidepartners. The shareholding system has been usedto develop timber forest and more often cashtree plantations in the area and beyond.

The second constraint to rehabilitation is lackof expertise and knowledge of appropriatetechnologies for sustainable management oftimber forests among farmers. This requiresdevelopment of technologies suitable for localconditions and markets and the support of aneffective extension service.

Mountain Closure as a Strategy forRehabilitationNatural regeneration of secondary forest through‘closing mountain’ is a common, traditionalsolution to forest rehabilitation, largely becauseit is cost-effective. It was identified as a strategyfor rehabilitating degraded evergreen broadleafforest and Masson pine forest distributed inlocation far from the village settlements. Theoption was tested on about 260 ha of forests inHongqiao village. About 60 ha of the forestswere household-managed family plots. Villagersformerly collected fuelwood from the broadleaf



forests but now they depend less on fuelwoodfor cooking and space heating as gas andelectricity are available. This made it feasibleto implement the option.

Degraded Bamboo Plantationsand Forests

Lei Bamboo PlantationsLei bamboo (Phyllostachys praecox f.preveynalis) in the region could be planted ornatural but it is mainly planted. In this chapter,the term ‘plantation’ is used for this type ofbamboo. There has been a rapid developmentin Lei bamboo plantation over the last decadein Lin’an and beyond. As a result, shoots of thisspecies have replaced tea as an important sourceof income for farmers. However, degradationalso occurs in Lei bamboo plantation. About20% of Lei bamboo plantations in the threevillages are degraded (Table 5.3). They arecharacterised by low productivity, unevenmother bamboo individuals at different ages, anddiseases.

Degradation is strongly associated with twosilvicultural practices. One is the planting ofbamboos on inappropriate sites. Local farmerswere so enthusiastic that they establishedextensive Lei bamboo plantations on landsunsuitable for the species, such as steep slopesand ridges with infertile, shallow soil. These arethe sites where Masson pine forests occurnaturally. Bamboo plantations on these sitesgrow slowly and poorly, and have low output.Where Lei bamboo has been planted there hasbeen serious soil erosion, further landdegradation and biodiversity loss. A majorreason for development of bamboo plantationson unsuitable sites has been the great marketdemand for the shoots and high economicreturns from this crop.

The other silvicultural practice with anegative impact is the application of mulch, abamboo cultivation technology developed in thelate 1980s (Figure 5.1). Mulching stimulatedshoots to sprout earlier than normal. Lei bamboo

usually sprouts after Chinese New Year butmulching increased soil temperature and shootswere produced before and during Chinese NewYear when prices are very high. This profitablepractice unfortunately has adverse site impacts.Covering land with mulch for 3-4 months everyyear results in a significant increase in soilmicroorganisms that change the soil’s physicaland chemical properties. The changed growthenvironment makes the bamboo susceptible todiseases and makes it difficult to maintain a goodage class distribution of healthy motherbamboos. Zhou et al. (1998) found thetechnology changes bamboo stands in manyways, resulting in smaller stem diameters, ahigher proportion of old mother bamboos andincreased flowering. The longer mulch is usedthe more mother bamboos flower. When landwas covered with mulch for 3 years insuccession over 10% of mother bamboosflowered. Jin et al. (1998a, b) examined impactsof long-term mulching on the growth, structureand distribution of underground rhizomes. Theyfound that the number of rhizomes increases buttotal length of rhizomes and percentage ofyoung, vigorous rhizomes (2-4 years old)decline, compared with unmulched plantations.There were 181 buds m-2 in the whole rhizomesystem of a mulched bamboo plantationcompared to about half that number in an

Figure 5.1 Intensive culture of Lei bamboo usingmulching in Lin’an County



unmulched plantation. Most of rhizomes occurin the upper layer of soil in mulched bambooplantations but are found below 30 cm depth inunmulched plantations. Diseases often infectrhizomes of mulched bamboo plantations.

Two rehabilitation options were identified.The first was to replace degraded plantationswith less site-demanding species. In 1996, anexperiment was carried out to incorporate timberspecies such as Masson pine and cash trees suchas Myrica rubra and chestnuts into a bambooplantation area.

The second solution identified for thisproblem was to modify the practice to minimisethe adverse effects on the mother bamboos bynot mulching every year and by controlling thesoil temperature. This option was first tested ina farmer’s field in Shangfeng village and at theGaohong Forest Farm. The plantation wasmulched for 2 years during which not manyshoots were retained as mother bamboos,followed by 1 year during which it was notmulched and more shoots were retained asmother bamboos. Ideally, a bamboo plantationshould have 18 000 bamboo individuals ha-1, anaverage dbh of 3.5 cm and over 70% of motherbamboos of 3 years old or younger. Soil tem-perature was reduced to help develop healthymother bamboos and to prevent diseases by (1)using mulch materials that does not raise thetemperature so much and reducing the thicknessof mulch to about 20 cm, (2) reducing mulchduring the second half of the sprouting period,and (3) shortening length of the mulching pe-riod.

Preliminary results of the testing werepositive with an increase in both the number ofbuds on rhizomes and the weight of individualshoots. This could lead to an increase in healthymother bamboo individuals and an improvementin quality of the bamboo shoots. The results ofthe tests resulted in adoption of the modifiedmulch technology by many households and thiswill potentially alleviate degradation of Lei bam-boo plantations.

The problems of mulching Lei bamboohighlight the importance of developing forestrytechnologies that are not only profitable but also

have with minimum adverse impacts on agro-ecological conditions. This is necessary becauseforest operations are often undertaken in eco-logically fragile areas.

Bamboo Forest for Producing DryShootsBamboo forest for producing dry shoots occurson upper slopes. Formerly, forests on these landswere mixed forest of Masson pine, broadleafspecies and small bamboo groves. After thehousehold-based forest management wasadopted in the early 1980s, farmers cleared allthe pines and broadleaf species and developedmonocultures of bamboo. Nearly half of thesebamboo forests were found degraded with lowproductivity (Table 5.3).

The research found that degradation of theseforests was closely related to inappropriate shootharvesting methods. Surveys in Shangfengvillage show that local farmers harvested earliersprouting and vigorous shoots and leaving latersprouting, weak or unhealthy shoots as motherbamboos. This selective harvesting isdetrimental in the longer term to healthy andproductive bamboo forest. Farmers also degradethe bamboo forest by breaking shoots off byhand instead of unearthing them with hoes orother tools.

Poor shoot processing technologycontributed to the poor quality of bamboo shootproducts. The processing was a manualoperation in household workshops and therewere no uniform standards. This produced thedry shoots of highly variable quality. Theprocessing method also resulted in wastage offresh shoots. Some households produced 1 kgof dry shoots from 4 kg of fresh shoots, andothers used 10 kg of fresh shoots.

Prices for poor quality bamboo shoots werevery low and bamboo forest management wasnot profitable. Consequently, more than 80% offarmer households did not tend their bambooforests. If the quality of bamboo shoots can beimproved by adopting appropriate technologiesfor harvesting and processing shoots, farmerswill have incentives to properly manage de-graded bamboo forest.



Based on this premise, we worked withfarmers to identify and test several options.• To adopt harvesting methods such as

retaining vigorous shoots sprouting in theintermediate sprouting period as motherbamboos; renewing one third to a quarter oftotal mother bamboo individuals (4500-6000bamboos ha-1 year-1); clearing all bamboosthat are older than 5 years; and maintaining18 000-24 000 bamboos ha-1 evenly spacedto generate healthy bamboo forests.

• To adopt tending technologies such asclearing weeds, other tree species, and oldand diseased bamboo individuals; usingherbicide in July and August; scarifying thesoils in the bamboo forests in November andDecember; and applying fertilisers twice ayear.

• To improve technology and standards forshoot processing.These options were tested in bamboo forests

and workshops of several households in the threevillages at first and demonstrated to otherhouseholds after they were proved appropriate.The test results were encouraging. The adoptionof appropriate methods for shoot harvestingtogether with improved tending of bambooforest significantly improved the health ofbamboo forests, as shown by increased unitoutput of shoots. Output of dry shoots increasedfrom 60 kg ha-1 to 600 kg ha-1 and theimprovement in shoot processing resulted inadded value. As a result, on average eachhousehold now generated 6000 Yuan year-1 fromdry shoots. About 20% of households inShangfeng village adopted some of thesetechnologies in 1998-1999. This suggests thereare good prospects for an improvement inmanagement of bamboo forests for dry shootproduction.

Degraded Tea PlantationsTea was formerly one of three primary productsin Lin’an County and an important source ofincome for farmers. At that time, tea plantationswere in good condition and their output of freshleaves was as high as 7500 kg ha-1. As a conse-

quence, Lin’an County ranked third in Zhejiangprovince in tea output in the early 1980s. Sincethen 70% of tea plantations in the county havebecome seriously degraded (Table 5.3) and theoutput of fresh leaves has dropped dramaticallyto less than 1500 kg ha-1.

The “degradation”, or low productivity, wasclosely related to the disinclination of farmersto manage tea plantations because: 1) marketshare of Lin’an tea declined considerably asmore tea plantations were establishedthroughout China; 2) developing tea plantationsthat produce general tea is much less profitablethan planting Lei bamboo. Data collected fromthe villages indicated that returns to investmentin Lei bamboo could be 10 times that for teacultivation. Farmers responded by producingbamboo shoots and neglecting tea.

Similar to the case of timber forest,landholding fragmentation is a constraint toeffective and profitable management of teaplantation. In the early 1980s Hongqiao villagedistributed its tea plantations to householdsacross the contours of the hills so that everyhousehold received an equal share of both goodand poor lands. This pattern makes tending,pruning and fertiliser application difficult andless effective. It is also not good for tea makingbecause processing superior tea requires acertain scale of plantation area that is hard toachieve under the fragmented landholdingpattern.

We worked with farmers to develop twooptions. One was conversion of tea plantationinto bamboo plantation where site is suitable forbamboo and production of high quality tea. InHongqiao village, where land is more suitablefor Lei bamboo than in the other two villages,part of the tea plantations have been replacedby bamboo, with results being positive. In aninnovative approach an exotic variety of qual-ity tea, ‘white tea’, was introduced to theGaohong Forest Farm in 1997. This tea is largelyfor export, with prices as high as 2000 Yuan kg-1

(US$240 kg-1). About 2 kg of white tea was pro-duced from an area of 0.06 ha in 1997, whichsuggests the venture could be profitable.



The other was an expansion of tea area man-aged by a household as a production unit. Thiswas to be achieved by leasing use rights to landfrom other households. In Hongqiao village, forexample, a household rented tea plantation landfrom several other households, obtaining userights to a single tea plantation area of 0.33 ha.It also was to be achieved by combining severaltiny pieces of land owned by several householdsinto a larger piece managed by one householdthrough voluntary exchange among themselves.

Development and Extensionof Forest TechnologyForest technology has an important part to playin the rehabilitation and sustainable managementof degraded forests. It is essential to generateappropriate technologies with farmers and thensupport their dissemination. Three questionsthen arise: (1) What are appropriate foresttechnologies? (2) How they can be generated?(3) What are effective ways to disseminatetechnical information to stakeholders?

Appropriate Technologies for Reha-bilitationRogers (1982) argued that agriculturalinnovations have five characteristics that affecttheir rate of adoption:• relative advantage (productivity gains)• compatibility (ecological compatibility to

local conditions)• complexity (consistent with local capacities

to adopt; including low risk in adopting)• trialability; and• observability

This list is helpful in reviewing the experi-ences in Lin’an. The mulch technology for Leibamboo shoot production was rapidly adoptedby 90% of households in the villages because ithas high productivity gains. The returns on in-vestment in applying the method are predict-able and farmers quickly benefit from increasedincome. The more predictable the benefits of anew technology the more rapidly farmers willadopt it. The technology is simple and was de-veloped on the basis of existing practice, and

hence involves just a minor modification of thetraditional method. All villagers can understandand use it whether they are literate or not. Mulch-ing involves little risk and so takes into accountthe vulnerability of poor farmers. Most poorfarmers are not innovators because they cannotafford to take risks. The mulch technology fitsmost of Roger’s criteria but it must also be sus-tainable. Unfortunately, the mulch technologyfor Lei bamboo is profitable in short-term butwas not sustainable in the long run.

Development of AppropriateTechnologyParticipatory and interdisciplinary approachesare effective ways to generate appropriatetechnologies. Farmer involvement is particularlyimportant because they fully understand localecological conditions, socio-economiccircumstances, and their technology needs.Technical options identified must be tested withfarmers in their fields. In this study farmers wereinvolved in at all stages of research process, fromidentification and analysis of problems,constraints and options through planning totesting of options and solutions. They helpedthe researchers improve their understanding oflocal situations and contributed to the generationof effective strategy packages for therehabilitation and sustainable management ofdegraded forests of the villages. The researchersused their knowledge of biological processesand experience from other places to contributeto problem identification and solutiongeneration. The research approaches offer amodel for the generation of forest technologyin general. It supports the observations by otherson participatory technology development (e.g.Rao 1985, Waugh et al. 1989, Scoones et al.1993, Chambers 1994a,b,c).

Extension of Forest TechnologyEffectiveness of pathways for information dis-semination may vary across user groups andareas. A pathway that works well in an area witheasy access to transport and mass media maynot be effective in more isolated localities.



Our surveys in the villages found that farm-ers obtain technical information through neigh-bours, relatives and friends, visits to outside ar-eas and television programs in descending or-der of importance. They indicated they expectto continue to rely heavily on neighbours, rela-tives and friends, and visits to outside areas inthe future for technical information, probablybecause they can see the technologies beingadopted. They are less confident in those tech-nologies shown on television because their suit-ability for local conditions may not have beentested. This emphasis on local observation sug-gests the importance of extension methods suchas demonstration plots and demonstration house-holds in the transfer of forest technology.

The surveys also showed three areas of forestextension that need improvement:• mismatch between the role of the existing

technical organisations and farmer demandfor extension services;

• insufficient number and limited role ofdemonstration households; and

• lack of extension organisations at villagelevel.There are several technical organisations

providing extension services at the townshiplevel, including the Society of Science andTechnology, forestry station, BambooAssociation and Association of Forest Farms.The township Society of Science andTechnology is responsible for research andtechnology related affairs in the township. Thetownship forest station is the lowest level of theForest Department and has forest extension asone of its responsibilities. The Bamboo Asso-ciation is made up of bamboo growers and theAssociation of Forest Farms comprises, on avoluntary basis, those involved with townshipforest farm, village forest farms and householdforest farms plus technical professionals. Bothare non-government organisations and were setup to provide members with technical adviceand opportunities for information sharing andexchange. Although they have much potential,these four organisations are not effective in pro-viding extension services for villagers as only

8% of households interviewed knew these or-ganisations existed in the township.

Demonstration households play an importantpart in forest extension in general, and ininformation dissemination in particular, giventhe farmers’ preference for this type ofinformation source. Several demonstrationhouseholds have been appointed in Gaohongtownship but they are insufficient for the rolethey are supposed to play and lacked institutionalsupport.

Based on this analysis, we tried to improveforest extension in the area by:• appointing additional demonstration

households and empowering existingdemonstration households;

• establishing experimental and demonstrationplots;

• providing short training courses for farmers;• enhancing the bamboo and forest farm

associations;• setting up village technical groups, and

assisting the formation of a network ofextension services comprising townshiptechnical organisations, village technicalgroups and farmer households.

Three additional demonstration householdswere appointed and one experimental anddemonstration plot established in each villageto test options identified for sustainablemanagement of degraded forests and as a wayof technology transfer. Existing demonstrationhouseholds were empowered through (1)granting them membership of technical societiesat township level or some societies at countylevel, and (2) providing them with trainingcourses and opportunities for involvement inadaptive research and with some experimentaland demonstration projects.

Short training courses were provided forabout 3300 villagers and 150 township officialsin Lin’an County, covering a wide range of top-ics such as sustainable rural development, sus-tainable forest management, non-wood forestproducts, and other specific forest technologies.The courses assisted dissemination of technical



information and development of appropriatetechnologies.

The Bamboo Association and Association ofForest Farms were enhanced through promotingmembership and improving their rules,including requirements for regular meetings.The aim was to support them as functionalorganisations and to improve their role inextension, materials supply and productmarketing.

Lack of technical organisations at villagelevel was addressed by establishing a technicalgroup in each of the three villages as a linkbetween farmer households and the technicalorganisations at township level. Core membersof the village technical group are demonstrationhouseholds, and other members arerepresentatives of farmers who are innovatorsor risk-adverse in adopting technologies. Thevillage technical groups were effective incommunicating with farmers because it isindigenous and consists of only farmers. Theyalso provided an opportunity for interactionbetween innovative and less-innovative farmersand an opportunity to transfer technologies to awide range of farmers. The technical networkhelped farmers obtain more information andfacilitated technology transfer to them fromgovernment organizations on the one hand. Italso provided feedback to the governmentorganizations on the other hand.

Conclusions

Causes of Forest DegradationThere are more types of forests and a higherproportion of plantations in Zhejiang as a resultof prolonged intensive economic activities thanat the other three sites of this research. Theseforests provide villagers with manyopportunities for employment and incomegeneration but all forest types are now degraded.

Factors causing forest degradation in Lin’anCounty varied across forest types. Major causesof timber forest degradation includedeforestation, over-dependence on forest, and

inappropriate methods of harvesting.Underlying causes of deforestation weredevelopment policies such as the ‘Great LeapForward Campaign’, and tenure insecurityresulting from land reform policies from the1950s to the early 1980s. Degradation of Leibamboo plantations was directly related to poorsilvicultural techniques including plantingbamboo on unsuitable sites and using theprofitable but unsustainable mulchingtechnology. The underlying cause is huge marketdemand and high prices for Lei bamboo shoots.Conversely, the shrinking demand and fallingprices for dried bamboo shoots and teadiscourage farmers from properly managing thetea plantations and bamboo forest resulting inlow productivity. Degradation of bamboo forestproducing dry shoots is exacerbated by poorshoot harvesting methods.

Rehabilitation StrategiesThe research provided insights into the problemsof rehabilitation of degraded forests and someways to address them. Timber forest rehabilitationmust address disincentives to farmers (tenureinsecurity, over-taxation, and restrictiveregulations on product harvest and trade) andconstraints (landholding fragmentation, and lackof expertise and technology among farmers).Landholding fragmentation is also a constraintto tea plantation management. The action researchin Gaohong demonstrated that some of theseconstraints can be overcome.

Government policiesFarmers’ incentives to plant trees and manageforest involve a number of factors, includingtenure security; and favourable policies forharvesting, marketing, and taxation of forestproducts.

Tenure insecurity is an underlying factor inforest degradation and a disincentive torehabilitation. It discourages farmers fromplanting trees and managing existing forestssustainably. Deforestation after the adoption ofhousehold-based forest management in the mid-1980s is closely related to lack of farmer



confidence in the security of tenure of forestsallocated to them, a lack of confidence resultingfrom the radical, frequent changes in polices forforest and land tenure during the period from the1950s to the early 1980s. Perceptions of tenureinsecurity resulting from the many policy changespartially explain why farmers are more interestedin bamboo production than timber growing. Thelong period needed for timber to generate incomefor investors requires tenure security. Althoughcurrent policies and law assure farmers of foresttenure security, uncertain factors remain at theimplementation level and these need to beaddressed by government.

The current restrictive regulations on timberharvesting and high taxation on timber sales aredisincentives to farmers in Gaohong to planttimber trees and manage timber forestssustainably. A review and amendment of suchgovernment policies could enable farmers toobtain a better return on their forest productsand would encourage them to make greaterefforts to manage their forests and reduce forestand land degradation. Farmers in the area werehighly enthusiastic about bamboo planting ratherthan timber forest because they do not haveproblems with harvesting, selling or hightaxation of the product.

Common management of householdresourcesThe development and acceptance of householdforest farms and shareholding systems in Lin’anCounty shows that farmers can find acceptableways to overcome the problems caused by landholding fragmentation. Our research thereprovided examples of successful commonmanagement of household land resources tomake a tea plantation of larger scale feasible andmore cost effective. It was also possible to getagreement for mountain closure to enable thearea to be naturally regenerated with trees.

Appropriate technologiesTechnology has an important role to play inrehabilitating degraded forests and plantations.For instance, development of profitable and

sustainable technology for Lei bamboo planting,production of quality tea through introducing newvarieties and processing methods, improvementof quality of dry shoots by adopting appropriateharvesting and processing techniques.Technologies promoted for forest rehabilitationmust be suitable for local ecological conditionsand socioeconomic circumstances, and bothprofitable and sustainable. A technology is notappropriate if it is profitable but not sustainable,as illustrated by the Lei bamboo mulch technology.Improving product quality by technical means mayhelp increase market share of products and soencourage farmers to provide better managementof their forests and plantations.

Improved rural servicesThere are opportunities to improve themanagement and profitability of forests that arecurrently degraded. To do this effectively, it isnecessary to overcome a range of constraintsfaced by farmers. Important first steps suggestedby the research in Lin’an are better access offarmers to technical advances and marketinformation through improved services. This canbe achieved by establishing village technicalgroups to facilitate communication betweenfarmer households and the technical organisationsat township level. Such means are also effectivein improving communication among farmers,which provides an opportunity for technologytransfer. This is so because the village technicalgroups formed part of a technical network thathelps farmers obtain more technology and marketinformation from government organisations anddisseminate among themselves and is able toprovide feedback from farmers to the governmentorganisations.

Further ways of improving extension anddemonstrating new technologies can include:appointing additional demonstration householdsand empowering existing demonstrationhouseholds; establishing experimental anddemonstration plots; providing short trainingcourses for farmers; and enhancing local non-governmental organisations such as bamboo andforest farm associations.



ReferencesBesley, T. 1995. Property rights and investment

incentives: theory and evidence from Ghana.Journal of Political Economy 103: 904-937.



Chambers, R. 1994c. Participatory ruralappraisal (PRA): challenges, potentials andparadigm. World Development 22: 1437-1454.

Fortmann, L. and Bruce, J.W. (eds.) 1988.Whose trees? Proprietary dimension offorestry, Westview Press, Boulder, Colorado,USA.

Jin, A., Zhou G., Zheng B., Chen Z. and FengW. 1998a. An effect of cultivation in mulchedand protected Phyllostachys praecoxplantations on its rhizome, Journal of BambooResearch 17 (4): 36-39. (In Chinese)

Jin, A., Zhou G., Zheng B. and Zhao X. 1998b.A preliminary study on degenerativemechanism of Phyllostachys praecox standplanted in protected site. Journal of FujianCollege of Forestry 19 (1): 94-96. (In Chinese)


Rao, Y. S. 1985. Building success throughpeople’s participation. Unasylva 37 (147):29-35.

Rocheleau, D. 1999. Confronting complexity,dealing with difference: social context, contentand practice in agroforestry. In: Buck, L.E.,Lassoie, J.P. and Fernandes, E.C.M. (eds.)Agroforestry in sustainable agriculturalsystems, pp. 191-235. Lewis Publishers, BocaRaton, Florida, USA.

Rogers, E. M. 1982. Diffusion of innovations(3rd ed.), The Free Press, New York, USA.

Scoones, I., Clark, J., Matose, F., Phiri, C.,Hofstad, O., Makoni, I. and Mvududu, S.1993. Future directions for forestry extension.In: Bradley and McNamara (eds.) Living withtrees: policies for forestry management inZimbabwe. Technical Paper No. 210, WorldBank, Washington DC.

Waugh, R. K., Hildebrand P. E. and Andrew, C.O. 1989. Farming system research andextension. In: Compton, J.L. (ed.), Thetransformation of international agriculturalresearch and development. Lynne RiennerPublishers, Boulder, Colorado, USA.

Yin, R. and Newman, D. 1997. Impacts of ruralreform: the case of the Chinese forest sector.Environment and Development EconomicsNo 2, 291-305.

Zhou, G., Jin A., Zheng B., Fang W. and Yu W.1998. [A] Preliminary study on compositionof lei bamboo in protected plot. Journal ofZhejiang Forestry College 15 (2) :111-115. (InChinese)


Rehabilitation ofDegraded Forests to Improve Livelihoods of Poor Farmers in South China

EditorLiu Dachang

Editorial BoardZhu Zhaohua, Cai Mantang, Liu Dachang and John Turnbull

CAF

Degradation of forests and forest lands is a problem in many parts of theworld and is particularly serious in south China. Chinese forest policyreforms in recent years have enabled rural households to generate incomefrom forests, to own the trees they have planted, and have offered newopportunities to manage forests sustainably. Rehabilitation of degradedforests and forest lands is one of the possible pathways to improvelivelihoods of poor farmers and others in the rural communities.

This report documents the results of four case studies in south Chinain which farmers, local officials and researchers analysed the problems ofdegraded forests and forest lands, and formulated options for theirsolution. Opportunities to improve forest management and people'slivelihoods are dependent on overcoming a range of biophysical,socioeconomic and political constraints. Action research was used toimplement and test some of the options identified. The experience andanalysis should be of value for researchers, resource managers andgovernment officials in China and elsewhere to address poverty andenvironmental concerns through a multidisciplinary, participatory andholistic approach.

ISBN 979-8764-98-6

Rehabilitation of Degraded Forests to Im

prove Livelihoods of Poor Farmers in South China

cover2.qxd 6/30/03 11:29 AM Page 1

dachang, l. 2003. rehabilitation of degraded forests to improve

Documents