international water asian development bank …oar.icrisat.org/3683/1/460-2003.pdfthe opinions...

Asian Development BankInternational Crops ResearchInstitute for the Semi-Arid Tropics

International WaterManagement Institute

Citation: Wani, S.P., Maglinao, A.R., Ramakrishna, A., and Rego, T.J. (eds.) 2003. Integratedwatershed management for land and water conservation and sustainable agricultural production in Asia:proceedings of the ADB-ICRISAT-IWMI Project Review and Planning Meeting, 10–14 December 2001,Hanoi, Vietnam. Patancheru 502 324, Andhra Pradesh, India: International Crops Research Institute forthe Semi-Arid Tropics. 268 pp. ISBN 92-9066-466-5. Order code CPE 150.

Abstract

Erratic rainfall, land degradation, soil erosion, poverty and burgeoning population characterize the dryregions in Asia. To develop sustainable natural resource management options for increasing the agriculturalproductivity and income of the rural poor in these regions, a new Integrated Farmer Participatory WatershedManagement Model was developed by ICRISAT in partnership with the national agricultural researchsystems (NARS). This model was applied at selected benchmark locations in Asia by ICRISAT throughthe project RETA 5812 “Improving management of natural resources for sustainable rainfed agriculture inAsia”, funded by the Asian Development Bank (ADB). The challenge for catchment research is togenerate technologies and management systems that is now being addressed by the Management ofSoil Erosion Consortium (MSEC) project RETA 5803 “Catchment approach to managing soil erosion inAsia”. This project is funded by ADB and the International Water Management Institute (IWMI) serves asthe facilitator. The workshop “Integrated Watershed Management for Land and Water Conservation andSustainable Agricultural Production in Asia” was held to review these projects. Forty-five scientists fromChina, India, Indonesia, Laos, Nepal, Philippines, Thailand, and Vietnam participated in the workshop.The objectives of the workshop were to:(1) Review the progress made at benchmark watersheds/catchments and synthesize the findings fromthe work of the projects RETA 5812 and RETA 5803; (2) Discuss work plans for 2002, identify emergingissues and future strategies for sustainable use of natural resources for improving rural livelihoodsthrough new initiatives; and (3) Discuss watershed development and management technologies. Duringthe period, a one-day workshop on Watershed Methodologies was also organized. The research papersbased on the work conducted for three years at different benchmark sites in Asia are covered in thispublication. The multi-country and multi-institutional research findings about watershed/catchmentmanagement reported here will serve as a valuable resource for the researchers, policymakers andstudents working in the area of sustainable management of natural resources.

The opinions expressed in this publication are those of the authors and not necessarily those of ICRISAT or AsianDevelopment Bank (ADB) or International Water Management Institute (IWMI). The designations employed and thepresentation of material in this publication do not imply the expression of any opinion whatsoever on the part of ICRISATor ADB or IWMI concerning the legal status of any country, territory, city, or area, or of its authorities, or concerning thedelimitation of its frontiers or boundaries. Where trade names are used this does not constitute endorsement of ordiscrimination against any product by ICRISAT or ADB or IWMI.

Integrated Watershed Management for Landand Water Conservation and Sustainable

Agricultural Production in Asia

Proceedings of the ADB-ICRISAT-IWMIProject Review and Planning Meeting

10–14 December 2001, Hanoi, Vietnam

Editors

S P Wani, A R Maglinao, A Ramakrishna, and T J Rego

International Crops Research Institute for the Semi-Arid TropicsPatancheru 502 324, Andhra Pradesh, India

Asian Development Bank0401 Metro Manila, 0980 Manila, The Philippines

Management of Soil Erosion Consortium (MSEC)International Water Management Institute

P O Box 2075, Colombo, Sri Lanka

2003

iii

Acknowledgments

This workshop was organized with the active support of several organizations/institutions. Weacknowledge the help from scientists and authorities from the following organizations/institutions forenabling successful conclusion of the review and planning meeting: Central Research Institute forDryland Agriculture (CRIDA), Hyderabad, India; Jawaharlal Nehru Krishi Vishwa Vidyalaya(JNKVV), Indore, India; Indian Institute of Soil Science (IISS), Bhopal, India; BAIF DevelopmentResearch Foundation (BAIF), Bhopal, India; District Water Management Agency (DWMA) (formerlyDrought Prone Area Programme) and Andhra Pradesh Rural Livelihoods Programme (APRLP),Andhra Pradesh, India; International Water Management Institute (IWMI); Khon Kaen University(KKU), Khon Kaen, Thailand; Department of Agriculture, Bangkok, Thailand; Department of LandDevelopment (DLD), Khon Kaen, Thailand; Vietnam Agricultural Science Institute (VASI), Hanoi,Vietnam; National Agriculture and Forestry Research Institute (NAFRI), Institut de recherche pour ledéveloppement (IRD), Laos; Nepal Agricultural Research Council (NARC) and International Centrefor Integrated Mountain Development (ICIMOD), Nepal.

Many individuals contributed to the success of the workshop. We sincerely thank Drs T D Long,Deputy Director General, VASI, Thai Phein and T D Toan of National Institute for Soils and Fertilizers(NISF), and M/s N V Viet and D D Phai for all the logistical support for the Workshop. We are thankfulto the Asian Development Bank (ADB) for providing financial support through RETA 5812“Improving Management of Natural Resources for Sustainable Rainfed Agriculture” and RETA 5803“Catchmeat Approach to Managing Soil Erosion in Asia” for conducting this workshop. TheOrganizing Committee sincerely thanks Drs C D Phat, Ministry of Agriculture and Rural Development(MARD), Vietnam, D J Bandaragoda, IWMI, and William D Dar, ICRISAT for delivering the openingand inaugural addresses and also for extending their full support. We are indebted to Drs R C Sachan,K L Sahrawat, and Piara Singh for reviewing the manuscripts. M/s K N V Satyanarayana andY Prabhakara Rao undertook the various secretarial tasks graciously and their help is acknowledged.We are thankful to Ms Sheila Vijayakumar for invaluable assistance for editing the manuscripts and toMr K N V Satyanarayana for incorporating corrections and page-setting the proceedings.

Workshop Committee

S P Wani (Chair) A R Maglinao (Co-chair)T D Long A RamakrishnaThai Phein

Local Organizing Committee

T D Long (Chair) T D ToanThai Phein N V VietA Ramakrishna D D Phai

v

Contents

Foreword ........................................................................................................................ .............. vii

Welcome Address ............................................................................................................ ............. 1T D Long

Catchment Research: A Valuable Support for Integrated Land and WaterResources Management ............................................................................................ ........... 3D J Bandaragoda

Integrated Natural Resource Management: The Key to Prosperity and Peacein the Drylands ... .................................................................................................................. 7W D Dar

Inaugural Address ............................................................................................................ ............ 10C D Phat

Improving Management of Natural Resources for Sustainable Rainfed Agriculturein Asia: An Overview .................................................................................................... ...... 13S P Wani, H P Singh, T D Long, and Narongsak Senanarong

The Management of Soil Erosion Consortium (MSEC): Linking Land andWater Management for Sustainable Upland Development in Asia ................................. .... 31A R Maglinao and F Penning de Vries

Feasible Solutions for Sustainable Land Use in Sloping Areas .............................................. ..... 42L Q Doanh and L V Tiem

Productivity and Resource Use Management of Soybean-based Systems ina Vertic Inceptisol Watershed .................................................................................... .......... 50Piara Singh, S P Wani, P Pathak, R Sudi, and M S Kumar

Nutrient and Water Management Studies for Increasing Productivity ofSoybean-based Systems in Operational Scale Watersheds ............................................ ...... 65A K Mishra, K P Raverkar, R S Chaudhari, A K Tripati, D D Reddy, K M Hathi,K G Mandal, S Ramana, and C L Acharya

Minimizing Land Degradation and Sustaining Productivity by Integrated WatershedManagement: Adarsha Watershed, Kothapally, India ........................................................... 79S P Wani, T K Sreedevi, P Pathak, T J Rego, G V Ranga Rao, L S Jangawad,G Pardhasaradhi, and Shailaja R Iyer

Conjuctive Use of Water Resource Technology and Extension in Improving Productivityof Rainfed Farming: An Experience at Lalatora, Madhya Pradesh, India ...... ..................... 97B R Patil, A B Pande, S Rao, S K Dixit, P Pathak, T J Rego, and S P Wani

Efficient Management of Natural Resources: A Way to Sustain Food Security in theRainfed Sloping Lands of Northern Vietnam ................................................................ ...... 109T D Long, A Ramakrishna, H M Tam, N V Viet, and N T Chinh

.

vi

Improving Management of Natural Resources for Sustainable Rainfed Agriculture inNortheastern Thailand ................................................................................................ ......... 123Narongsak Senanarong, Precha Cheuwchoom, Chanvid Lusanandana,Somchai Tangpoonpol, Suppachai Atichart, Arun Pongkanjana, Pranee Srihaban,Thommasak Singhapong, Aran Patanothai, Banyong Toomsan,Sawaeng Ruaysoongnern, and T J Rego

Improving Management of Natural Resources for Sustainable Rainfed Agriculturein Ringnodia Micro-watershed .............................................................................................. 134R A Sharma, O P Verma, Y M Kool, M C Chaurasia, G P Saraf, R S Nema,and Y S Chauhan

Use of Satellite Data for Watershed Management and Impact Assessment ............................ ..... 149R S Dwivedi, K V Ramana, S P Wani, and P Pathak

Statistical Analysis of Long Series Rainfall Data: A Regional Study in Southeast Asia .............. 158J P Bricquet, A Boonsaner, T Phommassack, and T D Toan

Factorial Analysis of Runoff and Sediment Yield from Catchments in Southeast Asia ............... 163T Phommassack, F Agus, A Boonsaner, J P Bricquet, A Chantavongsa,V Chaplot, R O Ilao, J L Janeau, P Marchand, T D Toan, and C Valentin

Land Use Characteristics and Hydrologic Behavior of Different Catchments in Asia ................ 171M G Villano and R O Ilao

Nutrient Loss and On-site Cost of Soil Erosion Under Different Land UseSystems in Southeast Asia ........................................................................................... ........ 186F Agus and Sukristiyonubowo

Soil Erosion and Farm Productivity Under Different Land Uses .............................................. .. 194T D Toan, F Agus, C M Duque, and A Chanthavongsa

Economic Incentives and the Adoption of Soil Conservation Practices ..................................... . 198E Biltonen

Consortium Approach in Soil Management Research: IWMI’s Experience in Southeast Asia .... 209A R Maglinao and F Penning de Vries

A Consortium Approach for Sustainable Management of Natural Resources in Watershed ....... 218S P Wani, T K Sreedevi, H P Singh, T J Rego, P Pathak, and Piara Singh

Improving Rural Livelihoods through Convergence in Watersheds of APRLP .......................... . 226T K Sreedevi

Integrated Water Resources Assessment at the Meso-scale: An Introduction to theWater and Erosion Studies of PARDYP ........................................................................ ...... 232R White and J Merz

A Dynamic Soil Erosion Model (MSEC 1): An Integration of Mathematical Model andPCRaster-GIS ............................................................................................................. ......... 241A Eiumnoh, S Pongsai, and A Sewana

Group Photograph ............................................................................................................... ......... 252Participants ......................................................................................................................... .......... 253

vii

Foreword Natural resources throughout the world, particularly in Asia where demographic pressures are veryhigh, are under severe threat. The need to improve the management of natural resources for meeting thefood, feed and fuel needs of the ever-increasing population, is urgent. Since 1999, the AsianDevelopment Bank (ADB) has supported the efforts of the International Crops Research Institute forthe Semi-Arid Tropics (ICRISAT) to address these concerns through RETA 5812, Improving theManagement of Natural Resources for Sustainable Rainfed Agriculture.

The project has covered substantial ground over the last three years, establishing benchmarkwatersheds in India, Thailand and Vietnam, and bringing about substantial increases in the productivityof the rainfed systems. The project has also achieved a significant milestone through closecollaboration with another ADB-supported project, Catchment Approach to Managing Soil Erosion inAsia, executed by the Bangkok office of the International Water Management Institute (IWMI). Thesetwo projects have brought new life to the management of watersheds by promoting participatoryintegrated management at the community level, and by assessing the impact of these efforts on soilerosion.

I am very pleased to note that these two projects have joined hands through a final workshop onIntegrated Watershed Management for Land and Water Conservation and Sustainable AgriculturalProduction in Asia, in addition to a Watershed Methodology Workshop. Scientists from seven Asiancountries and three international institutions working in Asia on the management of natural resourcescontributed to the proceedings of the workshop, which were published jointly by ICRISAT and IWMI.The detailed papers of both workshops, which include detailed case studies from seven Asiancountries, constitute a very valuable source of information on methodologies for managing naturalresources.

The ADB is to be commended for supporting this very important area of research. The ProjectManagers, Dr Suhas P Wani of ICRISAT and Dr Amado Maglinao of IWMI, have put a great deal ofeffort into coordinating these two multi-country projects. Having had the opportunity to monitor theprogress of the project in India, Thailand and Vietnam, I am happy to state that the results of three yearsof painstaking research of both projects have been coherently assembled in this publication. I amconfident that these proceedings will serve as a valuable resource for researchers, developmentworkers, policymakers and students of natural resource management. November 2003 William D Dar

Director GeneralICRISAT

1

Welcome Address

T D Long1

1. Vietnam Agricultural Science Institute (VASI), Hanoi, Vietnam.

The Integrated Participatory Watershed Develop-ment Program promoted under the AsianDevelopment Bank (ADB) assisted project (RETA5812) very well addressed the above constraints withemphasis on:• Simultaneous development of land, water, and

biomass resources in light of the symbioticrelationship among them.

• Integrated farming systems approach.• Meeting food, fodder, and fuel requirements of the

human and livestock population that depend onthese resources.

• Ensuring environmental sustainability along witheconomic viability by promoting low-costtechnologies.

• Improving land productivity by promotingimproved agronomic practices and input use.

• Releasing population pressure on land by creatingnon-farm employment.

• Development of local institutions for futuremanagement through participatory approach.The central thrust of the research was to enhance

productivity of land and water resources on the basisof a scientifically defined watershed that connotes ageographical unit rather than economic administrativeunits like household or village. It has also ensured thatthe whole range of stakeholders, from land users topolicy makers, are involved in the generation andpromotion of improved land use practices. Benchmarkwatersheds managed by Vietnam Agricultural ScienceInstitute (VASI) and National Institute for Soils andFertilizers (NISF) are serving as good demonstrationsites for the farmers and other stakeholders to developsustainable farming technologies on sloping landsthrough participatory approach.

We would like to congratulate ICRISAT andIWMI as well as VASI and NISF project teammembers for their innovativeness and pioneeringspirit that has provided the project with veryimportant information on the agronomic advisability,

On behalf of the local organizing committee, I wish toextend a very warm and hearty welcome to thescientists and experts from the International CropsResearch Institute for the Semi-Arid Tropics(ICRISAT), International Water ManagementInstitute (IWMI), China, India, Indonesia, Laos,Nepal, Philippines, Sri Lanka, Thailand and localdignitaries from Danish Agency for DevelopmentAssistance (DANIDA), Research and TechnologicalExchange Group (GRET), Centre de cooperationinternationale en recherché agronomique pour ledéveloppement (CIRAD), Centro Ricerche Fiat(CRF), and Bioseed who have come here toparticipate in the workshop on Integrated WatershedManagement for Land and Water Conservation andSustainable Agricultural Production in Asia (ADB-ICRISAT Annual Project Review and PlanningMeeting and 6th MSEC Assembly).

The uplands are a fragile environment characterizedby sloping lands that are prone to erosion, with lownatural soil fertility and declining forest cover. Theyare currently threatened with ecological degradation,which is already severe in many areas. As theburgeoning populations are expanding into steeperand more fragile areas in the uplands, morecatchments are threatened with severe soil erosion,declining soil productivity, and environmentaldegradation. Degradation of natural resources in thewatersheds now poses a great threat to the economiesof many developing countries and more so in thedeveloping world as the livelihoods of the ever-growing populations depend on these resources.

On-site soil loss reduces soil fertility in terms ofchemical, physical, and biological depletions. Thesesoil changes in turn are reducing crop yield, farmincome, and household food security. The off-siteeffects of soil erosion often have broader economic andenvironmental implications including sedimentation,flooding, and reduced water quality resulting inreduced living conditions of the people.

2

economic feasibility, and social acceptability of whatincreasingly appear to be very promising techno-logical possibilities for a large number of farmers innorthern Vietnam.

But much more important, it is our opinion that theproject is rapidly moving into a position of leadershipin northern Vietnam. If it continues its present course ofinnovation for another 3–5 years, it could well becomethe state-of-the-art integrated watershed project innorthern Vietnam, the project under which virtually allothers in northern Vietnam will be placed to learn aboutsoil and water conservation, recuperation, integrated

nutrient management, sustainable cropping systemsand how to apply science and technology-leddevelopment to significantly improve prospects forreducing poverty, food insecurity, and malnutrition inthe sloping lands of northern Vietnam.

I wish that the deliberations in the next few dayswould be very stimulating and thought provoking andhelp develop effective research plans. I once againwelcome all the participants to the beautiful countryof smiling faces and wish you all a pleasant andproductive stay in Hanoi, Vietnam.

Thank you.

3

1. International Water Management Institute (IWMI), Southeast Asia Regional Office, Bangkok, Thailand.

Catchment Research: A Valuable Support for Integrated Landand Water Resources Management

D J Bandaragoda1

better policies and administrative structures forsustainable development and management of thesetwo intrinsically interlinked resources.

An integrated effort in research, development, andmanagement of land and water coincides with thevalued idea of sustainability. As defined by FAO in1990, “Sustainable development is the managementand conservation of the natural resource base and theorientation of technological and institutional changein such a manner as to ensure the attainment andcontinued satisfaction of human needs for the presentand future generations.” Conservation and propermanagement of land, water, and plant and animalgenetic resources for agriculture, forestry, andfisheries are the key criteria of sustainability.

Both land and water resources are often used inunsustainable ways, resulting in the degradation ofmost catchments. Fulfilling the short-term economicdemand of the increasing rural population oftenoutweighs the concern for a more sustainable use ofthese resources. Recent estimates by the InternationalFood Policy Research Institute (IFPRI) indicate thatabout 40% of world’s agricultural land is seriouslydegraded. In South and Southeast Asia, about 46% ofthe total land area is affected by human-induced soildegradation, and of this, about 90% has experiencedsome decline in agricultural productivity. This poses achallenge to develop and apply catchmentmanagement strategies that will harmonize the use ofcatchment resources, particularly land and water, toproduce the desired goods and services without anyadverse effect on the environment. Soil degradation isa biophysical process, but its primary cause caninvariably be traced to socioeconomic and politicalforces. Therefore, the real challenge of catchmentresearch is to attract the attention of policy makers, anaspect which is yet to be fully realized.

Land-Water InteractionsThe linkage between land and water is well known,particularly in the context of agriculture. However, inmost policy, administrative, and scientificdeliberations, land and water have often been treatedin two distinct sectors. In the process, some importantland-water interactions have tended to be ignored, orconsidered in a fragmented manner. Many issuesfocusing on sustainable management of naturalresources such as those related to desertification,erosion, nutrient depletion, floods, drought, andprotection of watersheds and coastal areas tend to beanalyzed and understood either solely as land issues,or as water issues. “Water, a reflection of land use”,illustrates how human actions in changing thelandscape and their side effects influence water flowsand pathways, causing various chemicals to beintroduced into water, which in turn affect landfertility and thereby cause many social problems(Source: Falkenmark et al. 1977). Water in anycatchment has passed through upstream land and“carries the chemical memories from that journey”,while land needs access to safe water to beproductive. Thus, the catchment experiences aconstant interaction between the hydrological processand their ecological effects.

Catchment research, such as the efforts undertakenby the International Crops Research Institute for theSemi-Arid Tropics (ICRISAT), International WaterManagement Institute (IWMI), and their partners, isimportant in understanding these phenomena. Thisresearch also underscores the usefulness ofconsidering a catchment as the appropriate context inwhich land and water interactions can be effectivelyresearched in an integrated manner. An integratedapproach is more likely to help in the designing of

4

Catchment as a Unit of AnalysisShifting the focus of analysis from irrigation or otherland and water use systems to the river basin, or thecatchment, is helpful in many ways. For instance, weare able to capture most of the interventions by landand water users, as the catchment is the area wherethey interact and live. It also helps to include in theanalysis the interactions among various resource usesand user groups. In the process, a catchment approachgreatly helps in understanding better theenvironmental, social, and economic influences thatimpinge on the productivity of land and watermanagement. In a basin or a catchment context, inter-related issues on quantity and quality of land andwater resource use, surface and groundwater,upstream influences and downstream effects, use ofother inputs such as fertilizer, and disposal of wasteand surplus water, can all be more easily and morecomprehensively analyzed. Participation of a largernumber of stakeholders can be sought, and naturalresources planning can be more effectively carriedout. As can be seen in emerging results of catchmentresearch, the broader view is able to capturedimensions which are not normally included in asystem management approach, such as the causes(and not only the effect) of land degradation, waterscarcity, water quality, land and water relateddisputes, and inequitable benefits.

Non-physical Issues Related toCatchment ResearchIn conducting catchment research, in addition to thephysical aspects related to land and water use, thereare some non-physical (institutional and policy)concerns that need to be considered. I would like tohighlight some of these.

Appropriate coordination mechanisms

Land and water resources are managed and used bymultiple groups. Left to themselves, each group tendsto give priority to its own needs, as coordinatingmechanisms do not often exist on the basis ofcatchments. Consideration of all users and uses withina catchment facilitates the introduction of such inter-sectoral coordination.

Strong community participation

The need for greater coordination calls for a higherlevel of participation in resource management fromall the resource users. Issues related to local naturalresources can usually be identified clearly and moreeasily by the local community. A long-termassociation with the catchment area enables the localpeople to gain a very intimate knowledge about thenatural phenomena and the related constraints onresource use. Therefore, the pooling of theirknowledge is essential.

Collective action

The stakeholders may collectively derive somesynergic benefit from being able to integrate theirefforts. Effective participation and collective actionin resource management, however, depend on thedegree of awareness of important technicalconsiderations. Some efforts in social organization toestablish manageable groups within the communityhelps not only in bringing about this requiredawareness through various capacity buildingmeasures, but also in developing capability forcollective action through effective participation.Dissemination of catchment research is greatlyhelped by such organized user groups. Effectivecommunity participation leads to an empowerment ofthe people, enabling them to take their own decisionin agreed framework of rules. In a genuinelyparticipatory approach, outsiders can play only afacilitating role. Helping the local people to helpthemselves, through increased awareness about thephysical, ecological, and socioeconomic aspects ofcatchment management would be the best option weresearchers and extension workers have forgenerating some impact of our work.

Security of property rights

Similarly, another important institutional requirementfor sustainable land and water management is thesecurity of property rights. The reason is that asustained effort in catchment management requires alonger-term and secure property rights for individualsand groups as benefits would accrue over a longperiod of time. Stakeholders would be encouraged toinvest on various catchment management measures,

5

such as land contouring, tree planting, constructingcheck-dams, and bund protection, only if propertyownership and land tenure are adequately secured.

Reliable information base

For effective coordination and institutional integrity,resource management within the catchment must beessentially knowledge-driven. As the physical andsocial characteristics of a river basin system or acatchment are not easily discernible, an effectivecatchment management effort should necessarily relyon a sound database. Data on the physical, social,environmental, economic, and institutionalparameters of the catchment need to be collected andmaintained. The different actors and sectors usingland and water resources within the catchment shouldbe able to understand and assess the requirements ofone another. The most critical consideration inestablishing an information base is its location, i.e.,how well it can be placed so that all the involvedparties could have easy access to the information theyneed. Also the methods of collecting the information,which ensure appropriate levels of reliability areimportant.

Policy support for resources management

In a river basin or a catchment, a large number ofindividuals and groups are involved in using land,water, and environmental resources for differentsectoral purposes. An integrated approach in land andwater resources management can be achieved onlywithin a helpful policy environment. A coordinatedset of policies, rules, and regulations is essential tocover all natural resources within the catchment, andalso an effective coordination among the agenciesrelated to the management of each natural resource.For example, the land laws should be in keeping withthe laws related to the use of surface and groundwater,and the laws for environmental protection.

MSEC as an Important CatchmentResearch ProgramSoil erosion is considered a major cause of landdegradation and quite a few studies have beenconducted to address the problem. The challenge for

catchment research is to generate technologies andmanagement systems that are widely accepted andmaintained. This is the major task that is now beingaddressed by the Management of Soil ErosionConsortium (MSEC) project with funding supportfrom the Asian Development Bank (ADB).

MSEC uses a network approach to theorganization and implementation of soil erosionmanagement research. The approach provides amechanism that engages different scientists andresearch institutions to work together in a coordinatedand participatory mode. As mentioned above, the unitof analysis is at the level of a catchment to be able tocapture both the on-site and off-site effects of erosion.Research planning and implementation is undertakenthrough consultation among concerned nationalagricultural research and extension services(NARES), international agricultural research centers(IARCs), advanced research institutions (ARIs), non-governmental organizations (NGOs), and farmers.

The NARES play the central role in theconsortium, particularly in the participatory research,but with a broad responsibility for underpinningapplied and strategic research as well. Typically incatchment research, partnerships play a very vitalrole, and the efforts so far made in the MSEC projectare commendable.

The Role of IWMIFor those of you who may still be wondering whyIWMI is now involved in MSEC, and in this meeting,I would like to mention that as of 1 April 2001, theInternational Board for Soil Research andManagement (IBSRAM) ceased to exist and its workcontinues as part of IWMI’s science program. TheIBSRAM staff have become part of IWMI’s newSoutheast Asia Regional Office based at KasetsartUniversity in Bangkok, Thailand. This work includesthe supervision of the MSEC, which serves as theexecuting agency for the ADB project RETA 5803(Catchment Approach to Managing Soil Erosion inAsia). I must mention at this point that IWMI greatlyvalues the network tradition inherited from IBSRAM,and the opportunity of being able to collaborate with anumber of such experienced research partners. WhileIWMI serves as the consortium secretariat andfacilitator, it also has an important responsibility to

6

ensure a good quality research content in the MSECactivities. To strengthen this relationship, IWMI hasembarked on establishing appropriate institutionallinkages with partner countries, and looks forward tofulfilling its obligations, both towards the respectivecountries, and the donors, who essentially expect highlevel research quality through MSEC activities.

IWMI, being one of sixteen centers of theConsultative Group on International AgriculturalResearch (CGIAR), is basically a researchorganization. IWMI’s mission is to improve waterand land resources management for food, livelihoods,and nature, and in this respect, IWMI conducts aworldwide research and capacity-building program toimprove water and land resources through bettertechnologies, policies, institutions, and management.

IWMI has chosen five research themes as its keyinstruments to address the need for strategic prioritysetting in the institute and to assure thematicintegration of research agendas across physicallocations. The five themes are:1. Integrated Water Resource Management for

Agriculture;2. Sustainable Smallholder Land and Water

Management Systems;3. Sustainable Groundwater Management;4. Water Resource Institutions and Policies; and5. Water, Health, and Environment

The MSEC activities are mainly associated withIWMI’s Theme 2: Sustainable Smallholder Land andWater Management Systems. Research under thistheme concentrates on identifying the promisingsmallholder innovations and evaluating them togetherwith partners to understand how they work and whattheir impacts are. It seeks to understand theconditions under which the high potential smallholderpractices are viable, and then support their uptake indeveloping countries and regions.

Next StepThe first phase of the ADB-supported MSEC projecthas been going on for three years now and with ADB’sapproval has been extended for another year. Under

IWMI’s management, MSEC will stay as a majorprogram, particularly in Southeast Asia. As we planfor the project’s continuation, we anticipate a muchstrengthened program with the integration of land andwater management concerns in catchment research.This integration is the very essence of IWMI’sproposal for a second phase of MSEC submitted toADB for consideration of funding.

Notably, most CGIAR centers such as the CentroInternacional de Agricultura Tropical (CIAT), Center forInternational Forestry Research (CIFOR), InternationalCenter for Research in Agroforestry (ICRAF),ICRISAT, International Food Policy Research Institute(IFPRI), International Rice Research Institute (IRRI),and IWMI have recognized the value of carrying outresearch on and in catchments. For us at IWMI, weexpect that the outputs of the present catchment researchthat we are doing will provide valuable inputs to scale upthe application of research results and technologyoptions to much larger catchments and to the bigger riverbasins that we are very much involved in. This researchwill fully capture the interactions among the on-site andoff-site users of land and water resources and provide amore comprehensive basis to come up with theresolution of the competing demands of these users.These are all in support of IWMI’s vision of “ImprovingWater and Land Resources Management for Food,Livelihoods and Nature”. We hope to have a veryproductive collaborative program with all MSECpartners in the years to come.

This joint activity between ICRISAT and IWMI isa recognition of the value of collaboration amongdifferent CGIAR centers, working together towards acommon goal. The Global Challenge Programs thatthe CGIAR now advocates emphasize thecollaboration among centers and zero in on the valueof synergy and complementarity. As IWMI preparesto participate in these inter-center programs, we lookforward to see more research on the linkage betweenland and water management in the near future.

I hope that this week-long activity will be able togenerate valuable information and further interest inthis aspect.

Thank you.

7

Ladies and gentlemen!I extend my warmest welcome to the Honorable

Vice Minister of Planning, Cao Duc Fat, Dr Ngo TheDan, President, Scientific Committee, Nguyen VanBo, Chairman, Scientific Department, Le Hung Quoc,Chairman, Extension Department, Dang Kim Son,Director, Information Center, Nguyen Dinh Huong,Deputy Chairman, International CooperationDepartment, Ministry of Agriculture and RuralDevelopment, and Mr D J Bandaragoda, RegionalDirector, International Water Management Institute(IWMI), who are with us in this joint project planningmeeting. Let us recognize their presence by givingthem a loud applause. I also welcome Dr T D Long,Deputy Director General (DDG), VietnamAgricultural Science Institute (VASI), Dr Le QuocDoanh, DDG, VASI, Dr Thai Phien, and the delegateswho have come from different countries and themedia staff from Vietnam.

For some of you, this may be the first visit toVietnam, a beautiful and historic country with veryhospitable people. While you are here, you will alsohave an excellent opportunity to see an example ofintensive agriculture. I speak on behalf of theInternational Crops Research Institute for the Semi-Arid Tropics (ICRISAT), a non-profit, apolitical,international organization for science-basedagricultural development. Established in 1972, it isone of the 16 centers of the Consultative Group onInternational Agricultural Research (CGIAR).ICRISAT’s new vision is “improved well-being of thepoor of the semi-arid tropics (SAT) throughagricultural research for impact.”

Our new mission is “to help the poor of the SATthrough science with a human face and partnership-based research and to increase agriculturalproductivity and food security, reduce poverty, andprotect the environment in SAT production systems.”Basically, we aim to improve and sustain agricultural

Integrated Natural Resource Management: The Key toProsperity and Peace in the Drylands

W D Dar1

1. International Crops Research Institute for the Semi-Arid Tropics (ICRISAT), Patancheru 502 324, Andhra Pradesh, India.

productivity in some of the harshest environmentsthat account for about two-third of the world’scultivable land.

We serve more than 800 million people in 48countries, the poorest of the poor, who live in the drytropics of Asia and Sub-Saharan Africa. We aresupported by more than 50 governments, foundations,and development banks. Our strategy is to focus oncomparative advantage, develop a competitive edge,and enhance strategic partnerships. The situation ischallenging as this region is heavily populated andpressure on natural resources is very severe. Unlesswe protect our soil and water resources, they will bedegraded and will not be able to support the rurallivelihoods in the SAT. ICRISAT has a mandate toimprove the livelihoods of the millions of poor livingin the SAT through increased agriculturalproductivity through integrated genetic and naturalresource management.

Water is the life line of the people living in the SATregion and unless it is managed, its continuingdepletion will endanger the survival of the peopleliving in this region. The main source of water in SATis the monsoon rain, which generally occurs asdownpours resulting into excess water during therainy days. Downpours cause severe soil erosion andalso take away nutrient rich top soil along with therunoff, which causes severe damage to the naturalresources. Subsequently, dry spells follow during thecrop growing season. The appropriate way to managethese problems is the adoption of watershedmanagement. Watersheds are not only the units formanaging the rainwater but also are the convergingpoints of various rural activities of millions of thepoor living in them.

The Asian Development Bank (ADB) andICRISAT share a common vision of alleviatingpoverty in Asia and ADB’s valuable support toICRISAT is through our project “Improving

8

Management of Natural Resources for SustainableRainfed Agriculture” that has gone a long way. I havea special concern for this project since it showcasesour credo of science with a human face operating inthis project. Our approach of integrated participatorywatershed management has attracted the attention ofnot only the researchers and farmers but also toppolicy makers in India and Vietnam.

The integrated watershed management approachinvolves participation of all stakeholders in thewatershed program. It demands teamwork andeffective cooperation. Traditionally watersheds havebeen viewed as hydrological units to conserve soiland water. However, this view has not benefitedfarmers. Instead we must adopt the integratedwatershed management approach. Here, all naturalresources in the watershed are nurtured properly sothat livelihood sources are effectively sustained.

The world is changing and in a borderless globaleconomy, things done locally can have global impact.Thus, farmers in the drylands not only must increaseproductivity of their farm but also becomecompetitive. Working together, we must fight povertyand hunger especially in the drylands. They are theroot causes of the political and social instability wefrequently see today. Global disorder could behandled effectively by local efforts in the watershedsby increasing productivity and rural incomes.

A recent study conducted by ADB indicated thatinvestments in rainfed areas are as productive as thosein favorable irrigated areas. The same study alsohighlighted a need to develop rural infrastructurewhich will have direct impact on alleviating ruralpoverty and improving rural livelihoods. Theintegrated watershed management approach will helpto manage the natural resources efficiently andeffectively so that the rural livelihoods can beimproved substantially through convergence ofvarious activities in the watershed.

ICRISAT has a 29-year experience in watershedmanagement. It is time we now share experienceswith farmers. Hence, ADB’s timely support hashelped ICRISAT and the national agriculturalresearch systems (NARS) in Asia to develop asuitable model for sustainable natural resourcemanagement of watershed. The concerned efforts of ateam of scientists led by Dr S P Wani have shownexcellent results in this project. Dr Ian Johnson,

Chairman, CGIAR who visited Adarsha watershed inKothapally, Andhra Pradesh, India said: “This is anexcellent example of the value of partnerships,research and action in the field as well as thededication of ICRISAT staff to small and poorfarmers.” Recently, a Planning Commission Memberof the Government of India also visited Kothapally.He was highly impressed with the visible impact inthis watershed. In fact, he asked details for discussionin the commission for possible replication in otherwatershed programs.

Our important partners in Kothapally watershedare the Central Research Institute for DrylandAgriculture (CRIDA) and Drought Prone AreaProgramme (DPAP). Dr H P Singh (CRIDA) and MsT K Sreedevi (DPAP) are with us in this meeting. Ourother partners from other benchmark sites in India,Thailand, and Vietnam are also here. We are happythat together, we have brought ICRISAT and NARSexperiences in watershed management to benefit ourpoor farmers. Ten other villages in Andhra Pradeshare adopting this model in the Andhra Pradesh RuralLivelihoods Programme (APRLP). We are lookingforward to promote this model in five districts inAndhra Pradesh to alleviate poverty by improvingrural livelihoods.

Kothapally is not an isolated watershed model, butis just one of the benchmark sites operated in thisproject. I have personally visited benchmark sites inVietnam and Thailand. When you visit Thanh Hawatershed in Vietnam you will see and experience thehappiness of the farmers whose incomes haveincreased two-fold during the last three years. I metDeputy Prime Minister Mr Nguyen Cong Tab duringthis visit, and he was very satisfied with our work. Iam happy to announce that the Thanh Ha watershedhas helped in generating support from the Vietnamesegovernment for natural resource managementresearch. VASI and Danish Agency for DevelopmentAssistance (DANIDA) have also recognized theimportance of watershed research.

Likewise, I am happy to note that ADB’s supporthas made a big difference to the rural poor in thebenchmark watersheds in India, Thailand, andVietnam. Now the challenge is for us to translate oursuccess into a broad-based movement of naturalresource management. We can do this through theconsortium approach adopted in this project. For a

9

peaceful world and a better tomorrow, naturalresource endowments must be managed and usedsustainably. ICRISAT vows to convert “Grey SATareas into green”. We are committed to achieve thisthrough “Science with a human face”. The support of

development investors such as ADB is very critical inthis effort. I am sure that we, as a team, will win thisfight to help attain a food secure, prosperous andpeaceful world for present and future generations.

Thank you all.

10

Inaugural Address

C D Phat1

1. Ministry of Agriculture and Rural Development (MARD), Hanoi, Vietnam.

• The creation of a new administrative body, theCommittee for Ethnic Minorities and MountainousAreas, to improve the coordination of existingprograms and new initiatives.

• Increased government funding for the planningand implementation of activities under the Decree327, “Master guidelines and policies to utilizeunoccupied land, barren hilly areas, forests,denuded land, beaches, and waterfronts”,implementation of which started in 1993.

• Requests for international assistance to increasethe resources for land and improve the quality ofthese efforts.However, awareness and further improvements in

planning and implementation of these programs isrequired.

Watershed Management and ErosionControl Problems

The importance of water resources managementthrough adequate watershed rehabilitation andconservation is increasing. Water is required toirrigate the lowlands, for hydroelectric energy, and fordomestic and industrial use of the country’s largepopulation. Watershed management is required tomitigate the effects of floods and drought and toprovide a livelihood for the large number of ethnicminority groups living in the mountains.

A national watershed management program doesnot exist yet. The new land management and landtenure policy aims at distributing both the agriculturaland forestry land to local people. This has createda situation in which grass root associations andself-help groups can take a leading role in thedevelopment and management of productive land.There is, however, a lack of well-tested integrated andflexible models and methods to effectively involve

At the outset, on behalf of the VietnameseGovernment I would like to convey my warmestgreetings to the honorable participants of thisworkshop from the International Crops ResearchInstitute for the Semi-Arid Tropics (ICRISAT),International Water Management Institute (IWMI),China, India, Indonesia, Laos, Nepal, Philippines, SriLanka, Thailand, Danish Agency for DevelopmentAssistance (DANIDA), Research and TechnologicalExchange Group (GRET), Centre de cooperationinternationale en recherché agronomique pour ledéveloppement (CIRAD), Centro Ricerche Fiat(CRF), Bioseed, and Vietnamese deputies who arepresent in this workshop. I also wish this seminarmeets with splendid achievements.

Market oriented reforms in the economic systemintroduced in Vietnam in late 1980s, including a landlaw, placed greater emphasis on families as the basicunit of production. Recently a number of newinitiatives have been taken; these demonstrateawareness and needs for the commitment to improvemanagement of land and water resources, includingthose in the uplands of Vietnam, which comprise overthree quarters of its total area. Upland developmentinvolving ethnic minorities and the settlement oflowland farmers in the hills has been going on for along time. But new emphasis is being placed of late.New initiatives in this direction are:• New legislation to strengthen the role of land

potential benefits for households in landmanagement (the Law on Land of 1993) and tosome degree, also in forest land management (theForest Resources Protection and Development Actof 1991).

• Various national action plans, including NationalForestry Action Plan, the National EnvironmentalAction Plan (Vietnam National EnvironmentAction Plan, 1993), and the National ConservationStrategy.

11

local communities in the sustainable developmentand management of watershed resources.

Until a few decades ago shifting cultivationpractices in the mountainous areas were not athreatening factor for the watersheds. Few localinhabitants, mostly ethnic minorities, used only thosesites, which were not much affected by erosion. Dueto population pressure and the resulting requirementfor more land for agriculture, the picture is nowrapidly changing. Besides a rapid increase in thepopulation, people migrated from the denselypopulated delta areas into the mountainous regions,especially in the lower areas. They often practice landuse patterns inspired by delta area agriculturaltechniques, but generally not suitable for these sites.This has resulted in the worst cases of erosion beingencountered on the low hills adjacent to denselypopulated areas.

Effective People’s Participation:A Must for SuccessIn most areas in Vietnam, watershed management anderosion problems can be mitigated throughappropriate soil and water management practices.The farmers, especially the indigenous communities(ethnic minorities) in most mountainous areas, have alot of experience and knowledge to use the land foragriculture, forest, and animal husbandry. We need toconsider their knowledge in developing improvedtechnologies for greater adoption of new innovations.

As stated earlier, the new land management andland tenure policy has created a situation in whichgrass root organizations can take a leading role in thedevelopment and management of land. Ways have tobe found of integrating sustainable technical, social,economical, and operational measures aimed atimproving the effective participation of people andguaranteeing access by small land users to sustainableland husbandry and conservation techniques and tothe related services and supplies.

This could be achieved through promotion andstrengthening of participatory methods and modelsand the progressive design and adoption by localcommunities of an integrated watershed managementapproach. Key to this strategy is not imposing anartificial reorganization of local social structures, butpromoting and supporting a participatory process to

be more effective for sustainable resource utilization,so as to contribute to better social, economic, andenvironmental living conditions.

The adoption of participatory methods andapproaches in the development and conservation ofrain-fed upland areas will only provide the desiredresults if adequate provisions are made to improveliving standards through appropriate management ofnatural resources, integration of cropping systems,forest management, animal husbandry, and ruraldevelopment.

Finally, watershed management programs do notrepresent a panacea, and cannot be everything to allpeople. In order to be more meaningful andmanageable the activities included in a watershedmanagement program should contribute to theobjective of improved overall natural resourcemanagement with a clear orientation towardsupstream/downstream use and conservation of waterresources. However, given the general situation ofresource degradation and poverty in the uplands, it isessential that a watershed program, in order to besustainable, contributes significantly to the economicand social well-being of the upland people throughimproved overall natural resources management. Ihope the deliberations in the next few days wouldconsider these aspects and come out with appropriateresearch plans.

We are very happy that the Asian DevelopmentBank (ADB)-funded projects implemented byICRISAT and IWMI with collaborative researchsupport from Vietnam Agricultural Science Institute(VASI) and National Institute for Soils and Fertilizers(NISF) are addressing these concerns veryeffectively. We wish that the projects continue thegood work with the necessary support from the ADBfor undertaking this important research.

The Ministry of Agriculture and RuralDevelopment (MARD) is making concerted efforts tomitigate constraints in agriculture and ruraldevelopment sectors and identify appropriatetechnologies for sustainable agriculture through closecooperation with various national and internationalinstitutions. We hope ICRISAT and IWMI makeVietnam as a partner in many more future projects inthe coming years in this endeavor and work closelysince we too firmly believe and acknowledgeICRISAT’s motto “Science with a human face”.

12

I hope this workshop helps foster partnershipresearch between MARD and ICRISAT and IWMIfurther and identify new areas for future researchcollaboration. I wish all the participants a pleasant

stay and hope you enjoy our warm hospitality. I onceagain wish that the workshop meets with splendidachievements.

Thanks for your attention.

13

Improving Management of Natural Resources for Sustainable RainfedAgriculture in Asia: An Overview

S P Wani1, H P Singh2, T D Long3, and Narongsak Senanarong4

Abstract

Limiting natural resources, erratic rainfall, land degradation, soil erosion, poverty, and burgeoningpopulation characterize the dry regions in Asia. Over-exploitation of natural resources in these areas tomeet the ever-increasing demand for food and fuel of rapidly growing population has led to environmentaldegradation and calls for initiation of immediate steps for optimal utilization of natural resources based onthe potential and limitations. To develop sustainable natural resource management options for increasingthe agricultural productivity and income of rural poor in these dry regions, a new integrated farmerparticipatory watershed management model was developed by ICRISAT along with NARS partners. Thisholistic approach includes new science tools, linking on-station research to on-farm watersheds, technicalbackstopping through consortium of institutions with convergence of livelihood-based activities. This newmodel was applied at selected benchmark locations in Asia by ICRISAT in partnership with NARS throughexecution of the project “Improving Management of Natural Resources for Sustainable RainfedAgriculture in Asia”. The broad objectives of the project were to enhance and sustain the productivity ofmedium and high water-holding capacity soils in the intermediate rainfall ecoregions of the semi-aridtropics of Asia and to develop environment-friendly soil and water resource management practices. On-station benchmark locations served as sites for strategic research and on-farm benchmark watershedsserved as farmer-managed sites for farmer participatory refinement and evaluation of sustainable naturalresource management options under varying socioeconomic and bio-physical situations. The on-farmwatersheds were provided with technical backstopping from ICRISAT and other consortium institutions.The monitoring and impact assessment in these locations reflected a higher technology adoption ofimproved soil and water conservation practices, and nutrient and pest management with increasedproductivity and incomes.

Reduction in production capacity of land due to windand water erosion, loss of soil humus, depletion ofsoil nutrients, secondary salinization, diminution anddeterioration of vegetation cover as well as loss ofbiodiversity is referred as land degradation. Landdegradation in arid, semi-arid, and dry sub-humidareas resulting from various factors includingclimatic vicissitudes and human activities is a causeof desertification. Though land degradation has beena problem in the past also, the pace of degradation hasgreatly increased in recent times due to burgeoningpopulation and enhanced means of exploitation of

natural resources. Seventy per cent of 5.1 billion ha(39.5% of land area) dryland areas worldwide isafflicted with one or the other form of landdegradation. Permanently degraded lands aregrowing at the annual rate of 6 million ha globally,which are affecting livelihoods of millions of poorpeople in the developing and poor countries. Theprocess of land degradation is seriously underminingtheir livelihood security leading to poverty,starvation, and migration.

A global assessment of the extent and form of landdegradation showed that 57% of the total area of

1. International Crops Research Institute for the Semi-Arid Tropics (ICRISAT), Patancheru 502 324, Andhra Pradesh, India.2. Central Research Institute for Dryland Agriculture (CRIDA), Santoshnagar, Hyderabad 500 059, Andhra Pradesh, India.3. Vietnam Agricultural Science Institute (VASI), Hanoi, Vietnam.4. Royal Department of Agriculture (DOA), Bangkok, Thailand.

14

drylands occurring in two major Asian countriesnamely China (178.9 million ha) and India (108.6million ha) are degraded (UNEP 1997). Acceleratederosion resulting in loss of nutrient rich top fertilesoil, however, occurs nearly everywhere whereagriculture is practiced and is irreversible. Thetorrential character of the seasonal rainfall createshigh risk for the cultivated lands. Of the estimated173 million t of sediment discharged into the oceansannually, Asia alone contributes nearly half of theload, even though the actual land area is just one-third(UNEP 1997). This is an eloquent testimony to theintensity of the process and the consequential damageto the producing ability of land. In India, erosion ratesof 5 to 20 t ha-1 (up to 100 t ha-1) are reported. In Indiaalone some 150 million ha are affected by watererosion and 18 million ha by wind erosion (UNEP1997). Thus, erosion leaves behind impoverished soilon one hand, and siltation of reservoirs and tanks onthe other. This degradation induced source of carbon(C) emission contributes also to far reaching globalwarming consequences. If the current productionpractices are continued, the Asian countries will facea serious food shortage in the near future.

In India, 65% of arable land is rainfed and theincreasing demand for food and feed has to be metfrom the increased production from the rainfed areas,as there is no scope for expansion of cultivable area aswell as irrigation facilities. The policy shift towardsrainfed lands is necessitated on social grounds as alarge majority of the rural community has subsistence-level existence with a sizeable component of peoplebelow the poverty line. The incidence of poverty is28% in the Asian developing countries, with highincidence of 35% poverty in India. The poverty indexin India is high and is around 40%. Although povertyin India has shown a decline over time, the absolutenumbers have increased substantially from 180 millionin early 1950s to over 350 million by the end of 1990s(Ryan and Spencer 2001). The different facets ofpoverty are malnutrition, non- or underemployment,poor reproductive health care and associated infantmortality, high birth rates, illiteracy, and a feeling ofhelplessness.

To minimize land degradation in selectedcountries of Asia, the Asian Development Bank(ADB) has supported a project on “ImprovingManagement of Natural Resources for Sustainable

Rainfed Agriculture in Asia” through RETA 5812. TheInternational Crops Research Institute for the Semi-Arid Tropics (ICRISAT) executed this project duringJanuary 1999 to June 2002 in partnership with thenational agricultural research systems (NARSs) ofIndia, Thailand, and Vietnam. The broad objectivesare to enhance and sustain the productivity of themedium and high water-holding capacity soils in theintermediate rainfall ecoregion of the semi-aridtropics (SAT) of Asia (parts of India, Vietnam, andThailand) and to develop environment-friendlyresource management practices that will conserve soiland water resources. The specific objectives of theproject are to:1. Characterize natural resource base and identify

physical and socioeconomic constraints toincreased sustainable cropping in the targetecoregion.

2. Apply and refine integrated cost-effective soil,water, and nutrient management (SWNM)practices based on the natural resourceendowments of the farmers.

3. Rehabilitate degraded medium to high water-holding capacity soils and study effects ofintegrated SWNM strategies on profitability andsustainability of the system.

4. Integrate and evaluate techno-economic feasibilityof promising strategies for crop intensification andreduction in soil degradation in the target Asianecoregion, to identify indicators ofunsustainability, and to learn lessons for extension/transfer of promising practices to other parts of theSAT.

Target Countries and EcoregionThe project has targeted the assured rainfallecoregion production systems in Asia for managingwater at community-scale watersheds. It occursprincipally in the eastern Deccan plateau in India andportions of central Myanmar, northeastern Thailand,northern Vietnam, and dry climatic areas of Indonesiaand sloping lands of the Philippines. Thailand has itsown share of land resource stresses; many of these arenatural, but accelerated by human activities on theland. Some have been specifically created throughmismanagement. About 30% of the land area is steepland. About 75% (386,000 km2) of land is vulnerable

15

to desertification. In Thailand the annual average soilerosion rate is 34 t ha-1 yr1 with more than 30% of thecountry affected by moderate to severe erosion.Approximately 45% of the land in the Philippines ismoderately to severely eroded. About 41% of the areais cultivated and is on land with slopes greater than18%. In Indonesia, one-third of the 57 million ha ofupland soils is classified as in critical conditionbecause of land degradation. Conservationtechnology will certainly contain some of thesedegradation processes but more work is required for itto be truly sustainable.

To develop sustainable natural resourcemanagement options for increasing the agriculturalproductivity and incomes of rural poor in this agro-ecoregion three target countries, India, Thailand, andVietnam, were selected. Five on-farm benchmarkwatersheds were selected in the three target countries(Fig. 1).

The ecoregion constitutes the heartland of rainfedagriculture in Asia. The rainfall is dependable (800–1300 mm) and the soils are medium deep to deep(>1 m depth) with medium high (150–200 mm) water-holding capacity. This ecoregion is mostly suited fordouble cropping through intercropping or sequentialcropping systems. It has a potential to become thegreen revolution area of rainfed drylands. Althoughland degradation and receding groundwater tables arecommonly observed in semi-arid tropicalenvironments, these are of particular concern in theintermediate rainfall ecoregion of central India,northern Vietnam, and northeastern Thailand. Thesoils in this ecoregion are prone to severe soil erosionbecause of their positional toposequence setting.Currently the rainfall use efficiency for cropproduction ranges between 30 and 45% and annually300 to 800 mm of seasonal rain is lost as surfacerunoff or deep drainage (Wani et al. 2002a). If the

Figure 1. Benchmark on-farm watersheds in target ecoregion of the selected countries.

1. Milli watershed, Lalatora micro-watershed (Vidisha, India)2. Ringnodia (Indore, India)3. Adarsha watershed (Kothapally, Ranga Reddy, India)4. Tad Fa (Thailand)5. Thanh Ha (Vietnam)

16

annual rainfall is properly utilized for crop production, itwould be sufficient to sustain not only the doublecropping but 200 to 300 mm of surplus rainwaterwould be available annually to recharge groundwaterresource. The recession of groundwater levels in theDeccan Plateau in India has shown that in spite of highrainfall there is mismanagement of water.

New Integrated WatershedManagement ModelA new model for efficient management of naturalresources in the SAT has emerged from the lessonslearned from long-term watershed-based researchconducted by ICRISAT in partnership with NARSs(Wani et al. 2002b). The important components ofthis integrated watershed management model are:• Farmer participatory approach through cooperation

model and not through contractual model.• Use of new science tools for management and

monitoring of watersheds.• Link on-station and on-farm watersheds.• A holistic system’s approach to improve

livelihoods of people and not merely conservationof soil and water.

• A consortium of institutions for technicalbackstopping of the on-farm watersheds.

• A micro-watershed within the watershed wherefarmers conduct strategic research with technicalguidance from the scientists.

• Low-cost soil and water conservation measuresand structures.

• Amalgamation of traditional knowledge and newknowledge for efficient management of naturalresources.

• Emphasis on individual farmer-based conservationmeasures for increasing productivity of individualfarms along with community-based soil and waterconservation measures.

• Minimize free supply of inputs for undertakingevaluation of technologies and farmers areencouraged to evaluate new technologiesthemselves without financial subsidies.

• Continuous monitoring and evaluation by thestakeholders.

• Empowerment of community individuals andstrengthening of village institutions for managingnatural watersheds.

Consortium partners

India• Indian Council of Agricultural Research (ICAR):

– Central Research Institute for DrylandAgriculture (CRIDA), Hyderabad

– Indian Institute of Soil Science (IISS), Bhopal• State agricultural universities (SAUs):

– Jawaharlal Nehru Krishi Vishwa Vidyalaya(JNKVV), Indore

• State Government Department:– Drought Prone Area Program (DPAP), Ranga

Reddy, Andhra Pradesh• Non-governmental organizations (NGOs):

– M Venkatarangaiya Foundation (MVF),Hyderabad

– Bharatiya Agro Industries Research Foundation(BAIF), Bhopal

Thailand

• Royal Thai Department of Agriculture (DOA),Bangkok

• Royal Thai Department of Land Development(DLD), Bangkok

• Khon Kaen University (KKU), Khon Kaen

Vietnam

• Vietnam Agricultural Science Institute (VASI),Hanoi

CGIAR

• International Water Management Institute(IWMI), Bangkok

Advanced research institutions• Michigan State University (MSU), East Lansing,

Michigan, USA• University of Georgia, Griffin, Georgia, USA

Strategic ResearchIn this project three on-station benchmark micro-watersheds (catchments) (Table 1) served as fieldsites to undertake strategic research on integrated soil,

17

water, nutrient, and pest management and also forcollection of necessary data for validating simulationmodels. The five on-farm benchmark watersheds(Table 2) served as farmer-managed sites for farmerparticipatory refinement and evaluation of sustainablenatural resource management options under varyingsocioeconomic and bio-physical situations. These sitesalso provided insight into socioeconomic constraintsfor adoption and evaluation of techno-economicfeasibility of different management options. At thesesites researchers and farmers jointly monitoredhydrological processes, soil loss, nutrient cycling, andincreased productivity at watershed level. In this papera brief overview of the progress in strategic and on-farm research at benchmark sites is given and detailedreports for individual sites are covered separately insubsequent papers.

Use of high-science tools for watershedplanning, development, and impactassessment

Simulation modeling

Crop simulation models help in evaluating theperformance of technologies under differentagroclimatic situations through scenario analysis andin identifying the major constraints for sustainingproductivity and also appropriate technologyapplication domains. Integration of remotely senseddata in the geographic information system (GIS)along with simulation models would increase ourability to conceptualize, and develop strategies tomanage the natural resources in the watershedsefficiently on sustainable basis.

Table 1. Benchmark sites for on-farm and on-station work in Asia.

AnnualWatershed Soils rainfall (mm)

On-stationICRISAT, Patancheru, India Vertisols and Vertic Inceptisols 800Indian Institute of Soil Science, Bhopal, India Vertisols 1140College of Agriculture, JNKVV, Indore, India Vertisols 960

On-farmAdarsha watershed, Kothapally, Ranga Reddy, India Vertic Inceptisols 760Milli watershed, Lalatora, Vidisha, India Vertisols and Vertic Inceptisols 1200Ringnodia watershed, Solsinda, Indore, India Vertic Inceptisols 1050Tad Fa watershed, Khon Kaen, Thailand Sloping mixed heavy soils 1300Thanh Ha watershed, Hoa Binh, Vietnam Deep Alfisols and sloping lands 1300

Table 2. Geomorphological characteristics of the on-farm watersheds.

Linear aspect of Relief aspect ofWatershed shape drainage network drainage basin

Area- Bifurca-Form perimeter tion Drainage Relief Relative

Watershed factor ratio ratio density ratio relief

Adarsha Watershed, Kothapally (India) 0.38 0.47 2.33 0.56 0.03 1.05Milli watershed, Lalatora, Vidisha (India) 0.36 0.43 2.00 0.38 0.04 1.09Ringnodia watershed, Indore (India) 0.66 0.47 1.33 0.66 0.52 1.77Tad Fa watershed, Khon Kaen (Thailand) 0.22 0.32 1.50 1.37 0.05 3.33Thanh Ha watershed, Hoa Binh (Vietnam) 0.64 0.39 1.50 1.36 0.06 3.33

Source: Pathak et al. (2002).

18

Crop growth simulation models in an integratedwatershed management approach provide anopportunity to simulate the crop yields in a givenclimate and soil environment. ICRISAT researchershave adopted DSSAT v 3.0, a soybean crop growthmodel, to simulate the potential yields of soybean cropin Vertisols grown at different benchmark locations.Mean simulated yield obtained for a location wascompared with the mean observed yield of the last fiveyears to calculate the yield gap. The results (Table 3)showed that there is a considerable potential to bridgethe yield gap between the actual and potential yieldthrough adoption of improved resource managementtechnologies (Singh et al. 2001).

the model performs well, with few exceptions, tosimulate crop growth, yields, and soil waterdynamics. Water balance components and nitrogen(N) balance components (N uptake, N fixation,denitrification, leaching, and N mineralization) havebeen quantified for a few seasons.

To evaluate the effect of different SWNMpractices on soybean intercropped with pigeonpeausing the simulation model we have validated theAPSIM model using the data sets generated from on-station watershed at ICRISAT. Using APSIM model,productivity and resource use of soybean/pigeonpeaintercropping system were simulated. The modelparameters for soybean and pigeonpea crops weredetermined by calibration using the observed data ofphenology, crop growth, and soil water dynamics for1998/99 and 1999/2000 seasons. The modelsimulated growth and development of both the cropssatisfactorily; however, adjustments were needed toset competition parameters for light and water.

Simulated water balance and the observed datashowed that significant amount of rainfall (25%) waslost as runoff during 1998/99 season. Simulatedrunoff matched the observed runoff data for both theflat and broad-bed and furrows (BBF) landforms(Table 4). Soil water dynamics for both the treatmentsduring the entire growing season also matched theobserved data satisfactorily (data not shown).Rainfall lost as deep drainage from the shallow soilduring 1998/99 season was significant (292 to 303mm). However, during 1999/2000 season, runoff anddeep drainage were negligible (Table 4). Total wateruse by the crop (459 to 465 mm) was met by rainfalland soil water depletion.

Remote sensing

Over-exploitation of natural resources to meet ever-increasing demand for food and fuel of rapidlygrowing population has led to environmentaldegradation and calls for initiation of immediate stepsfor optimal utilization based on the potential andlimitations. Information on the nature, extent, andspatial distribution of natural resources is aprerequisite for achieving this goal. Multispectralmeasurements made at regular intervals usingsatellites hold immense potential of providing suchinformation in a timely and cost-effective manner,

Table 3. Simulated soybean yields and yield gapfor the selected locations in India.

Mean Meansimulated observed Yield

yield yield1 gapLocation (kg ha-1) (kg ha-1) (kg ha-1)

Primary zoneBetul 2371 858 1513Guna 1695 840 855Bhopal 2310 1000 1310Indore 2305 1122 1183Kota 1249 1014 235Wardha 2997 1042 1955

Secondary zoneJabalpur 2242 896 1346Amaravathi 1618 942 676Belgaum 1991 570 14211. Mean of reported yields of last five years (1997–2001).

All the previous data of BW7 watershed atICRISAT, India from 1995 to 2000 on crop growthand soil water have been converted to formatssuitable for testing of the Agricultural ProductionSystems Simulator (APSIM) models of soybean-chickpea sequential and soybean/pigeonpeaintercropping systems. These models have beenvalidated for soybean sole crop and soybean/pigeonpea intercropping systems, which requiredsignificant amount of time for understanding and finetuning model parameters, particularly related tocompetition of light and genetic coefficients. Finally

19

and facilitates studying the dynamic phenomenonthat helps in assessing the effectiveness of theinterventions made in the watersheds.

In partnership with the National Remote SensingAgency (NRSA), Hyderabad, India, we are usingIndian Remote Sensing Satellites (IRS-1B/-1C and-1D) data for developing and managing watershedsefficiently as well as for monitoring the impact ofvarious interventions made in the watersheds.

Soil conservation measures taken up in the areagenerally result in (i) arresting soil loss; and(ii) improving soil fertility and moisture status, whichsubsequently leads to establishment or improvementin vegetation/biomass. The changes in the terraincondition of the watershed monitored using satelliteimages are described.

The Milli watershed in Lateri Block is located inthe northwest corner of Vidisha district in MadhyaPradesh in central India (Fig. 2). This watershedconsists of 35 villages, which are grouped into 17micro-watersheds. The Lalatora micro-watershed(725 ha) was selected for detailed monitoring ofhydro-meteorological measurements.

The change in biomass is reflected in theagricultural land use. The agricultural land use ofMilli watershed is portrayed in Figure 3. Since thesoil and water conservation measures were initiatedduring 1997, IRS-1C LISS-III data for rabi(postrainy) season of 1997 was used to deriveinformation on agricultural land use before thewatershed activities were initiated in this area. Tostudy the impact of the program on the land use, IRS-1C LISS-III data of rabi 2001 was used. The FalseColor Composite (FCC) derived from LISS-III dataof 1997 and 2001 is shown in Figure 4. Red color inFCC indicates vegetation cover whereas bluish greenor greenish blue indicates black soil bare/fallow. Acomparison of the vegetation cover during the period1997 to 2001 points to a significant increase of 269 hain the vegetation cover (3,402 ha during 1997 versus3,672 ha during 2001). Contrastingly, there has beenshrinkage in the fallow/barren lands. This analysisusing remotely sensed data by satellite provideddirect evidence in increased cultivation duringpostrainy season. Such an increase in area duringpostrainy season is mainly due to increased wateravailability in soil or in wells, which helped thefarmers to increase their cultivated area.

In partnership with NRSA we also studied thevegetation cover by generating the NormalizedDifference Vegetation Index (NDVI), which isessentially the ratio of the differences of the responsein the near infrared and red regions of the spectrum.Their sum values thus obtained range normallybetween 0 and 1.0. However, values of NDVI lessthan zero are also encountered which indicate barrenor fallow land. The NDVI images generated fromLISS-III data of 1997 and 2001 are shown in Figure 4.

A close look at the figure indicates a marginalincrease in the vegetation cover supporting thereby theobservation made on the agricultural land use. Whereasan estimated 31.5% of the geographic area of watershedhas been found to be in the NDVI range of 0.10–0.55 in1997, it has risen to 40.3% during 2001 demonstratingan increase of 9% in the greenery in the watershed. It is,indeed, interesting to note that in Lalatora micro-watershed where the soil and water conservationtreatments have been imposed, the vegetation cover hasimproved tremendously. As against only 149 ha ofvegetation cover during 1997–98, it has risen to 229 haduring 2000–01 registering thereby an increase of 80 haduring 3-year period (Fig. 4).

Table 4. Simulated water balance of soybean/pigeonpea intercropping system on flat and BBFlandforms on a shallow soil during 1998–2000seasons, ICRISAT, India.

Water balance Flat BBFcomponents (mm) shallow shallow

1998/99 seasonRainfall 1035 1035Runoff1 261 (266) 252 (250)Deep drainage 292 303Evapotranspiration 567 586Soil water change –73.8 –95.7

1999/2000 seasonRainfall 436 436Runoff1 0.11 (Nil) 0.11 (Nil)Deep drainage 0.9 4.4Evapotranspiration 459 465Soil water change –10.1 –20.9

1. Figures in parentheses are observed data.

20

Figure 3. IRS-1C PAN image of Milli micro-watershed showing agricultural land use.

Figure 2. Map showing Milli watershed and Lalatora watershed in Madhya Pradesh, India.

21

Impact of waterlogging on growth and yieldof soybean

Waterlogging studies conducted at IISS, Bhopal onsoybean-based systems revealed that waterloggingsignificantly influenced nodule number and drybiomass.

On-farm Farmer ParticipatoryResearch

Site selection

The process of site selection for on-farm research wasdone by a consortium of institutions, based on thedryland area, feasibility of the technology adoption(similar agroclimatic conditions), landholding size,socioeconomic conditions, and willingness of thefarmer, etc. The site selection for Adarsha watershed,Kothapally, Andhra Pradesh, India was done byICRISAT, DPAP, and MVF along with theinvolvement of the stakeholders.

ICRISAT, DPAP, and MVF together surveyed threewatersheds in Andhra Pradesh and selected Adarshawatershed. The total irrigable area was very less.There was more dryland with a large area underrainfed farming in the village. There was not a singlewater harvesting structure for human or livestock useat the time of survey in 1998, i.e., before the initiationof the project. As no interventions were made toconserve soil and water in this watershed, it wasselected to encompass the convergence. Adarshawatershed was selected after a committee meetingwith villagers in a “Gram Sabha” and villagersparticipated in the proposed watershed activities.

Participating groups

Different committees and groups were formed in thevillage and the villagers themselves selected leaders.The committee members participated from theinitiation of the watershed activity such as selectionof the site, implementation of the activity, andexecution and assessment of all the developmentalactivities within the watershed.

Figure 4. NDVI image of land use in Lalatora micro-watershed duringpostrainy season 1997 and 2001.

22

Baseline survey for constraint analysis inthe watershed

After the selection process, necessary information onthe environment and conditions of the village wascollected. Baseline data collection was done by boththe researchers and the stakeholders. The followinginformation was collected:• Socioeconomic status of the farmers and landless

people, crop productivities, inputs, and livelihoodopportunities.

• Soil, water, and nutrient management practicesfollowed by the farmers.

• Soil, climate, cropping systems, and input use. Thedata was assembled and analyzed.

• Production constraints, yield gaps, andopportunities for crop intensification. GIS mapswere prepared for different crops, soils, andcropping systems of the village.The results of the survey indicated that in

Kothapally village, dryland area was more (62.79%)compared to irrigated land (37.21%); literacy rate(35.74%) was low; and labor was scarce. There wasinverse relationship between land size and fertilizer/pesticide use. Crop yields were very low (1070 kg ha-1

for sorghum, 1500 kg ha-1 for maize, 190 kg ha-1 forpigeonpea) and there was not a single waterharvesting structure in the village. The villagers didnot undertake any income generating activities.

Soil and water conservation activities

To control erosion and restore productivity ofdegraded soils in the benchmark watersheds selected,several soil and water conservation options wereevaluated to conserve and harvest rainwater andincrease the productivity of the crops. These activitiesare important in maintaining, improving, andenhancing productivity of the crops. Widespreadadoption of improved practices is essential fordesertification control and restoration of degradedsoils. Engineering techniques of erosion control andrunoff management can be made more effective whenused in conjunction with biological control measures.In all the watersheds several soil and waterconservation activities along with biological controlmeasures were taken up both at farm and communitylevels.

Ex-situ conservation

• Grassed waterways• Water storage structures• Gully control structures• Field bunding

In situ conservation

• Shaping of the land reduces runoff; hence, the landis made rough by BBF and other similar landformtreatments.

• In the BBF method the beds of 1.05 m width and45 cm furrows are prepared at 0.4 to 0.6%gradient. The BBF method reduces runoff,conserves more water in the soil profile, and drainsexcess water safely away from the crops.

• Contour planting on flat landform.• Bullock-drawn tropicultor developed by ICRISAT

is used by the farmers at Kothapally for planting,sowing, fertilizer application, and intercultivationpractices.

• Planting of Gliricidia is done by farmers on fieldbunds for stabilizing the bunds to conserve therainwater and soil. In addition these plants willgenerate N-rich organic matter for fieldapplication, which will augment the N supply forcrop growth. This reduces the dependence onmineral fertilizer N.

Integrated nutrient management

Vegetative bunds

Gliricidia was planted on field bunds to conservemoisture. The loppings were incorporated into thesoil to provide biologically fixed N and reduce theusage of chemical fertilizers. Gliricidia also addsvaluable organic matter to soil.

Nutrient budgeting and balanced fertilization trials

To study nutrient budgets at watershed level, astratified random sampling was done by dividing thewatershed into three toposequences. As per the farmholding size, the farms were selected for nutrientbudgeting studies. This approach enabled us tocalculate the nutrient budgets at watershed level and

23

also assisted in developing the balanced nutrientmanagement strategies for sustaining the productivityin the watershed. This study involved detailed soiland crop analyses. In addition detailed accounts ofnutrient inputs and outputs by the farmers weremaintained. Pilot studies were conducted at Milliwatershed, Lalatora and Adarsha watershed,Kothapally.

We selected a sample of 25 farmers of differentfarm sizes in Milli watershed, Lalatora. Five farmershaving <1 ha (small), six farmers having 1–2 ha(medium), and 14 farmers having >5 ha (large)landholding were selected. These numbers were inproportion to the respective size category in thewatershed. A field was identified in each selectedfarmer’s holding and monitored for nutrient inputsand outputs. All information about various nutrientinputs to different crops in the same field wascollected. At harvest, the crops were sampled foryield as well as for nutrient uptake (Table 5).

The data indicated that for all major crops andcropping systems the phosphorus (P) balance waspositive, while the balance for potassium (K) wasnegative. This is because almost all farmers applydiammonium phosphate (DAP) while no farmerapplied any K fertilizer. Since these Vertisols are richin K, deficiency of K may not occur in the near future.We presume there is buildup of P in the soil and there

is scope to reduce P application. It is quite interestingthat both the legume crops, soybean and chickpea, aremining N from soil. The wheat crop is fertilized morethan it takes N and P from soil. Poor farmers growsorghum crop without any input and the yield as wellas nutrient outputs are also very small.

In Adarsha watershed, Kothapally, balances for N,P, and K were computed in 15 farmers’ fields whereinimproved SWNM options along with conventionalpractices were followed. Balanced nutrient doseswere used for sustaining productivity. In this study Ninputs through rainfall and biological nitrogenfixation (BNF) by legumes have been computed.Rhizobium inoculation of pigeonpea and soybeanseeds was done to increase BNF. Positive cropresponses to specific nutrient amendments based onsoil analysis, e.g., boron (B) and sulfur (S)applications were done and increased yields wereobserved. Higher grain yields were obtained withimproved practices indicating considerable scope forsavings on N fertilizer.

The nutrient uptake by maize/pigeonpeaintercropping system was more in the improvedsystems with BBF as compared to that of flatlandform treatment. The N-difference and 15N isotopedilution methods were used to quantify Ncontribution of legumes through BNF using non-fixing control plants. Similarly, for the sole maize

Table 5. Nutrient budgets for selected crops and systems at Milli watershed, Lalatora, India.

Input (kg ha-1) Output (kg ha-1) Balance (kg ha-1)

Crop N P K N P K N P K

Sole cropSoybean 8 20 0 911 6 42 –37 14 –42Wheat 59 15 0 51 6 34 8 9 –34Chickpea 15 9 0 631 4 41 –16 5 –41Sorghum 0 0 0 16 2 26 –16 –2 –26Soybean-wheat systemSoybean 9 24 0 1051 8 52 –43 16 –52Wheat 62 15 0 52 7 37 10 8 –37

Total 71 39 0 157 15 89 –33 24 –89Soybean-chickpea systemSoybean 10 23 0 1021 7 45 –41 15 –45Chickpea 17 8 0 611 4 43 –13 4 –43

Total 27 31 0 163 11 88 –54 19 –88

1. 50% of N uptake in soybean and chickpea is presumed to be from biological nitrogen fixation.

24

crop, uptake of nutrients was more in BBF systemthan in flat landform. The nutrient balances based onthe available data sets showed that all the systems aredepleting K from soils and more P is applied thanremoved by the crops. Crop yields as well as thenutrient removal was more in BBF than in the flatlandform treatment. High negative N balance inmaize/pigeonpea BBF system (–55 kg N ha-1)indicates that the crop extracted more N from the soilwhen grown on BBF system than when grown on flatsystem (–48 kg N ha-1). In sole maize on flat system Nbalance was only –24 kg N ha-1. Similar trend wasobserved in all the cropping systems studied.Potassium balance was also influenced by landform.In BBF system (–40 kg K ha-1) the crop could extractmore soil K than in flat system (–29 kg K ha-1) inmaize/pigeonpea cropping system.

Best-bet options

The scientists from JNKVV, CRIDA, and ICRISATput together a best-bet option for soybean-basedsystems. This consisted of use of improved variety ofsoybean JS 335, seed treatment with Thiram alongwith Rhizobium and phosphate solubilizing micro-organisms, application of DAP at 50 kg ha-1, andintegrated pest management (IPM). In the first year,27 farmers evaluated this option for soybean covering40 ha. Average increase in soybean yield was 34%above the baseline/control plot soybean yield of 950kg ha-1. Detailed analysis of benefit-cost ratio for thefarmers who evaluated this option was worked outand the net profit was estimated at Rs 5575 ha-1.

Micronutrient amendments

Balanced nutrient doses were used for sustainingproductivity in these watersheds. Rhizobiuminoculation of pigeonpea and soybean seeds was doneto increase BNF. Positive crop responses to specificamendments based on soil analysis, e.g., B and Samendments were done at Kothapally and Lalatorawatersheds, which proved to be a success as increasedyields were observed.

Green manuring

The importance of leguminous green manures such asGliricidia in maintaining soil and crop productivity

has been widely accepted. Decomposition ofGliricidia loppings and nutrient release occur at afaster rate due to low C:N ratio. Most of the nutrientsespecially N and K are released within 5–10 days ofdecomposition.

Comparative evaluation of decomposition ofGliricidia and pigeonpea plant residues showed thatleaves of Gliricidia decomposed faster thanpigeonpea plant parts (leaves, stem, and roots).Highest N mineralization (119 mg N kg-1) occurredwith surface soil application of Gliricidia leavescompared to Gliricidia stems (93 mg N kg-1) at 150days of incubation.

Micro-enterprise: vermicomposting

Earthworms are used in vermicomposting as they arevoracious feeders and can transform organic wastesinto compost in a short span. Compost, which isprocessed by earthworms, makes good organicfertilizer as it contains auxins, a growth promoter forplants and also some natural antibiotics.Vermicomposting is a cost-effective pollutionabatement technology. Women self-help groups(SHGs) in Adarsha watershed, Kothapally andindividual farmers in Lalatora watershed haveundertaken vermicomposting as an incomegenerating activity. Farmers have evaluated responseof vegetable crops to vermicomposting and haveobserved significant increases in yields of tomato inKothapally.

Integrated pest management

Integrated pest management is the coordinated use ofpest and environmental information to design andimplement pest control measures that areeconomically, environmentally, and socially sound.Pesticides are used only when needed and when othercontrol methods will not prevent economicallyimportant pest injuries. The outcome of a sound IPMprogram is usually increased profits due to savingsfrom reduced pesticide application and increasedprotection of the environment. Insect pests continueto be the major problem in pulse production in Asia.Intensive use of pesticides leads to total crop loss.Complete dependency on chemical control for thepast three decades has led to unsatisfactory pestmanagement along with environmental degradation.

25

ICRISAT, along with the national agricultural researchand extension systems (NARES), NGOs, and farmersin the watershed conducted research to identifyenvironmentally sound and economically viable plantprotection technologies which reduce yield lossesand improve the income of the farmers. Farm surveysand participatory rural appraisals identified the non-availability of IPM components such asbiopesticides, Helicoverpa nuclear polyhedrosisvirus (HNPV), pheromones, and high pest tolerantvarieties. The farmers harvested six-fold increasedyields through better management of pests bycontrolling them with neem seed extract along withpest tolerant crop varieties and there was 6–100%reduction in pesticide usage. After thorough evaluationof the existing pest management options, acomprehensive integrated pest management package forchickpea and pigeonpea has been developed andevaluated through farmer participatory approach mode.Revitalizing the effective indigenous methods likeshaking off pod borers from pigeonpea plants and use ofneem for pest management is done in the watersheds.These indigenous methods are effective, cheaper, andenvironment-friendly. Installation of pheromone trapsfor pest monitoring is done every year. Bird percheswere also installed in the fields for birds to rest and feedon Spodoptera and Helicoverpa larvae.

The availability of good quality HNPV wasconsidered a prime component for spread of IPM.Village-level production centers were initiated tocater to the needs of farmers. Many farmers andextension workers from villages were given trainingon HNPV production, storage, and usage on differentcrops. The project handled by ICRISAT has given

high priority for training village-level scouts inidentifying various pests and their natural enemies indifferent crops before the cropping season, and assistedthem in monitoring throughout the crop period.

Impact of Consortium Model forWatershed Management on RuralLivelihoods

Improved productivities

At Kothapally, farmers evaluated improved cropmanagement practices along with improved landmanagement practices such as sowing on a BBFlandform; flat sowing on contour; and using improvedbullock-drawn tropicultor for sowing and intercultureoperations. Farmers obtained two-fold increase in theyields in 1999 (3.3 t ha-1) and three-fold increase in2000 (4.2 t ha-1) as compared to the yields of solemaize (1.5 t ha-1) in 1998.

Increased incomes

Along with the highest system productivity thebenefit-cost ratio of the improved systems was more(1:2.47) compared to the farmers’ traditional cotton-based systems (Table 6) (Wani et al. 2002b).

Biological nitrogen fixation using legumesin Tad Fa watershed

Most of the farmers in northeast Thailand applychemical fertilizers to their cash crops for high yields.Chemical fertilizers are one of the costliest inputs and

Table 6. Economics of cultivation of different crops at Adarsha watershed, Kothapally during cropseason 1999/2000.

Total Cost of Total Benefit-productivity cultivation income Profit cost

Cropping system (kg ha-1) (Rs ha-1) (Rs ha-1) (Rs ha-1) ratio

ImprovedMaize/pigeonpea 3300 5900 20500 14600 2.47Sorghum/pigeonpea 1570 6000 15100 9100 1.51TraditionalCotton 900 13250 20000 6750 0.50Sorghum/pigeonpea 900 4900 10700 5800 1.18Mung bean 600 4700 9000 4300 0.91

26

there is a need to identify other alternatives orsupplement sources to overcome nutrient constraints.There is not much scope to use farmyard manure(FYM) as farm animals have been replaced by farmmachines for draft purposes. The use of legumes inthe cropping system would certainly help to reducethe amount of chemical N fertilizer. To recommendwhich legume should be grown in which croppingsystem we have to know the amount of N fixed by fewlegumes which are suitable to grow in this ecoregionand how much benefit these legumes give to thesucceeding non-legumes. This information is crucialfor recommending reduction of fertilizer N to farmers.

Rice bean, black gram, sword bean, and sunnhemphave been evaluated for quantifying N2 fixation andthe benefits of legumes using 15N abundance methodand 15N isotope dilution method on farmers’ fields atBan Koke Mon located near Ban Tad Fa in Thailandwhere ICRISAT benchmark watershed is situated.The cropping systems of Ban Koke Mon are similar tothose of Ban Tad Fa:• Legume-cereal: Rice bean-maize, sunnhemp-

maize, sword bean-maize, black gram-maize.• Cereal-ceral: Maize-maize.

Growing black gram, rice bean, and sunnhemp inthe system would help in reducing N requirement forthe succeeding maize crop. The results showed thatthe actual realized benefits from legumes in terms ofincreased N uptake by the succeeding maize cropvaried from 5.3 to 19.3 kg N ha-1 whereas theexpected benefits from legumes through BNF and soilN sparing effect over a maize crop varied from 15 to64 kg N ha-1. These results demonstrated that it is notonly the quantity of N2 fixed that determines thebenefit to the succeeding crop but also the quality oforganic matter and N release pattern from the legume

residue. However, in the long term for sustaining landproductivity sword bean could play an important role.

Micronutrient amendments – a success atLalatora watershed

Detailed characterization of soils in Lalatorawatershed revealed that these soils are deficient in Band S and both these nutrients are critical foroptimizing productivity of soybean-based systems.Farmers were made aware of the results and somefarmers came forward to evaluate the response for Band S application in their fields along with theimproved management options. Ten kg of borax (1 kgB) ha-1 and 200 kg of gypsum (30 kg S) ha-1 wereapplied by the farmers. In 2000, all the farmersreported significant differences in soybean plantgrowth with B, S, and B+S treatments over the best-bet control treatment. Also, soybean yields wereincreased by 19 to 26% over the best-bet controltreatment (Table 7). The results indicated thatamendments with B and S not only increased soybeanyields over best-bet treatment but also benefited thesubsequent wheat crop without further application ofB and S. Farmers were so much impressed with theirexperimentation that for 2001 season they indented Band S for their use well in advance through the NGO,the BAIF Research Foundation.

The economic analysis of the on-farm trials 2000/01 showed that combined application of B and S gavemaximum benefit of Rs 26,454 followed by only B(Rs 26,609) and S (Rs 25,955) application alone. Allthese three treatments proved to be beneficial to thefarmers with 1:1.8 benefit-cost ratio as compared tocontrol traditional practices (1:1.3) followed by thefarmers.

Table 7. On-farm evaluation of soybean and wheat to boron and sulfur amendments in Lalatora sub-watershed,2000/01.

Grain yield (t ha-1)

Treatment Soybean Wheat Soybean + wheat system

Boron 1.87 (23.2)1 3.74 (40.6) 5.61 (34.2)Sulfur 1.81 (19.1) 3.50 (31.9) 5.31 (27.0)Boron + Sulfur 1.91 (25.6) 3.57 (34.2) 5.48 (31.1)Control 1.52 2.66 4.18(Best-bet treatment)1. Figures in parentheses indicate increase (%) over control.

27

Monitoring and impact of watershedmanagement at Adarsha watershed

Monitoring

To know the impact of watershed managementcontinuous monitoring of several parameters asdescribed below was done:• Changes in crops and systems in farmers’ fields

were monitored.• An automatic weather station was installed to

record the rainfall, maximum and minimumtemperatures, and solar radiation.

• Sixty-four open wells in the watershed were geo-referenced and regular monitoring of groundwaterlevels was done.

• Water quality was monitored in all the wells andalso from the water storage structures in thevillage. Sediment samples (silt) were alsocollected from the tanks to understand theprocesses of environmental degradation in thewatershed.

• Nutrient budgeting studies were also undertaken.• Runoff and soil loss were monitored by using

automatic water level recorders and sedimentsamplers.

• Satellite monitoring was done.• Pest monitoring was also carried out.

Impact

The management of natural resources has becomeeffective and the livelihoods of the rural people haveimproved. The impact is assessed based on thefollowing:• Improved greenery: An increase in vegetation cover

was observed; in 1996 the vegetation cover was129 ha and in 2000 it was 200 ha at Kothapally.

• Improved groundwater levels: Groundwater levelin the village significantly increased in Adarshawatershed.

• Reduced runoff and soil loss: Runoff was 12% ofthe rainfall in the undeveloped watershed while itwas only 6% in the developed watershed wheresoil and water conservation measures wereundertaken.

• Increased productivities: The crop productivitiessignificantly increased with improved croppingsystems and improved management practices. The

yield of maize crop recorded two- to three-foldincrease (3.3 to 3.8 t ha-1) when compared withbaseline yields (1.5 t ha-1).

• Increased incomes: Farmers’ incomes as well ascropping system productivities increased. Maize/pigeonpea cropping system could give 3.5 timesbenefit (1:3.5) than the traditional cotton system(1:1.5).

Improved land management options

In 2000/01, at Adarsha watershed, several farmersevaluated BBF and flat landform treatments forshallow and medium-deep black soils using differentcrop combinations. Farmers harvested 250 kg morepigeonpea and 50 kg more maize per hectare usingBBF on medium-deep soils than flat landformtreatment. Furthermore, even on flat landformfarmers harvested 3.6 t ha-1 maize and pigeonpeausing improved management options as compared to1.7 t ha-1 maize and pigeonpea using normalcultivation practices.

Of all the cropping systems taken up in Adarshawatershed, maize-chickpea sequential cropping(benefit-cost ratio of 1:2.85) and maize/pigeonpeaintercrop (benefit-cost ratio of 1:2.81) proved to bemore beneficial to farmers in terms of benefitincurred to farmers. Farmers could gain about Rs19,590 and Rs 17,802 with these systemsrespectively. Sorghum, chickpea, and pigeonpea solecropping systems also proved beneficial, whereassorghum, maize, and chickpea traditional systemswere significantly less beneficial to the farmers.

Shift in cropping patterns

A close perusal of the prevalent cropping system, itsacreage and previous history before watershedintervention by ICRISAT gives a precise picture ofhow watershed approach benefits the finalstakeholders, i.e., farmers. Before dissemination ofwatershed technology at Kothapally, the village waspredominantly a cotton-growing area. The spread ofcotton crop was 200 ha in 1998 in the village. Theother crops grown were maize, chickpea, rice,pigeonpea, sorghum, and vegetable crops.

The watershed intervention by ICRISAT andconsortium partners followed an integrated new

28

technological approach which encompassedimproved soil cultivation and water managementtechniques, and land and crop management practices(suitable varieties, intercropping, legume-cerealcropping systems, BBF system, interculturaloperations, Gliricidia planting along bunds andincorporation of Gliricidia loppings, HNPV sprays,and vermicompost technology).

After three years of watershed activity inKothapally, the acreage under cotton cultivationdecreased from 200 ha to 100 ha (50% decline) with asimultaneous increase in maize and pigeonpea area(Table 8). The acreage under maize and pigeonpeaincreased three-fold from 60 ha to 180 ha within threeyears. The acreage of other crops remained almost thesame. This substantial shift in the cropped area wheremaize and pigeonpea replaced cotton crop wasmainly due to increased net profit per hectare. Thecotton-based cropping system had higher cultivationcosts (higher inputs) with lesser net profits comparedto maize/pigeonpea, sorghum/pigeonpea, or maize/chickpea system. Adoption of legume-cereal cropcombination or rotation cropping increased the netprofit with less cultivation costs in the watershed area.

Land use planning for increased householdincomes in Thanh Ha watershed

Unlike other Asian countries, the landholdings ofVietnamese farmers are very small. The averagefamily holding in drylands is around 0.5 to 1 ha. It is,therefore, important that the farm is utilized in themost prudent way for higher household incomes andfood security. Efforts have been made to identifyappropriate crops and crop combinations in various

seasons for enhanced household incomes in thebackdrop of systems sustainability, soil health, andpotential for large-scale adoption and adaptation. Forexample, maize, groundnut, and soybeancombination gave higher incomes in spring whilemaize and groundnut and maize and soybean cropcombination appear to be better in autumn-winterseason. The traditional maize cultivation was not atall economical.

Crop performance differed significantly across theseasons. Spring season was more favorable in termsof grain yields and associated income gains thanautumn-winter season (Fig. 5). Among the cropssoybean performed better in spring and summer thanin winter season. Soils in the sloping lands are highlyvulnerable to erosion when cleared of vegetativecover and are subjected to various forms of landdegradation. Loss of humus rich topsoil left behindthe subsoil devoid of vital plant nutrients leads torampant infertility and poor water-holding capacity. Itis, therefore, important to identify crops that not onlyperform well on these soils but also help improve soilhealth over the years.

To find out the influence of land degradation oncrop productivity and profitability the grain yields ofsoybean, groundnut, mung bean, and maize based onthe location on the toposequence in the landscapewatershed were delineated. In general, higher grainyields and farm incomes were obtained in the lowerand middle part of the toposequence compared to thatin the top due to less degradation and better soilfertility. Farmers are incurring higher expenditure dueto increased fertilizer usage in top of thetoposequence. Among the crops groundnut can begrown successfully in top, middle, and lower parts of

Table 8. Cropping practices at Adarsha watershed, Kothapally.

Area (ha) after watershed activityArea (ha) before watershed

Crop activity (1998) 1999 2000 2001

Maize 60 80 150 180Sorghum 30 40 55 65Pigeonpea 50 60 120 180Chickpea 45 50 60 75Vegetables 40 45 60 60Cotton 200 190 120 100Rice 40 45 60 60

29

the toposequence while mung bean and soybeanneed high level of management in top of thetoposequence for obtaining good yields. This kind ofinformation would assist in appropriate land useplanning and development of targeted nutrientmanagement technologies for systems resilience andincreased household incomes.

Human Resource DevelopmentHuman resource development is an importantcomponent of integrated watershed managementmodel to train farmers, national researchers, anddevelopment workers and to empower them byenhancing their knowledge. Farmers are exposed tonew methods and knowledge for managing naturalresources through training, video shows, and fieldvisits to on-station and on-farm watersheds. Educatedyouth are trained in skilled activities such as HNPVproduction and vermicomposting. Micro-watershedwithin the main watershed serves as “an island” forlearning for the farmers. Special emphasis is given toeducate the women farmers to new managementoptions. The technical backstopping team is alwayshandy to the farmers for clarifying their doubts andseeking more information at their location. Farmersand landless families are trained and encouraged toundertake income generating activities in thewatershed which can be of help to sustain theproductivity at catchment/watershed level.

Specialized hands-on training courses/workshopswere held for NARS partners. Training courses fordepartment officials and farmers were conducted.Twenty-four apprentices from developed/developingcountries were trained. Important dignitaries andpolicy makers were also made aware of the watershedprograms. Training materials were developed:(i) Web page of the project; (ii) On-line datamonitoring system for the project; and (iii) CD-ROMtraining module for watershed management.

Way ForwardSpectacular gains made and lessons learned in thisproject must be encashed for integrated ruraldevelopment. Institutional, policy, and technologicalneeds for sustaining watersheds need to be in place.Sharing and transfer of knowledge of naturalresources management to NARS throughempowerment as well as a study of on-site and off-siteimpacts on sustainability and environment quality areessential. Second generation problems in watershedsneed to be addressed.

Acknowledgments

This work was conducted by consortium partners atvarious benchmark watersheds (on-station and on-farm) under the project RETA 5812 “ImprovingManagement of Natural Resources for SustainableRainfed Agriculture” financially supported by ADB.

Figure 5. Performance of crops in spring and autumn-winter seasons, Thanh Ha, 2000.(Note: M = maize, SB = soybean, GN = groundnut, and MB = mung bean.)

30

The financial assistance provided by the ADB isgratefully acknowledged. We gratefully acknowledgethe team of scientists, NARSs partners, developmentworkers, and farmers for conducting strategic and on-farm participatory research.

ReferencesPathak, P., Wani, S.P., Piara Singh, Sudi, R., andSrinivasa Rao, Ch. 2002. Hydrological characterizationof benchmark agricultural watersheds in India, Thailand,and Vietnam. Global Theme 3. Report no. 2. Patancheru502 324, Andhra Pradesh, India: International CropsResearch Institute for the Semi-Arid Tropics. 52 pp.

Ryan, J.G., and Spencer, D.C. 2001. Future challengesand opportunities for agricultural R&D in the semi-aridtropics. Patancheru 502 324, Andhra Pradesh, India:International Crops Research Institute for the Semi-AridTropics. 83 pp.

Singh, P., Vijaya, D., Chinh, N.T., Aroon Pongkanjana,Prasad, K.S., Srinivas, K., and Wani, S.P. 2001. Potential

productivity and yield gap of selected crops in the rainfedregions of India, Thailand, and Vietnam. Natural ResourceManagement Program Report no. 5. Patancheru 502 324,Andhra Pradesh, India: International Crops ResearchInstitute for the Semi-Arid Tropics. 52 pp.

Wani, S.P., Pathak, P., Tam, H.M., Ramakrishna, A.,Singh, P., and Sreedevi, T.K. 2002a. Integrated watershedmanagement for minimizing land degradation andsustaining productivity in Asia. Pages 207–230 inIntegrated land management in dry areas: proceedings of aJoint UNU-CAS International Workshop, 8–13 September2001, Beijing, China (Zafar Adeel, ed.). Tokyo, Japan:United Nations University.

Wani, S.P., Sreedevi, T.K., Singh, H.P., Pathak, P., andRego, T.J. 2002b. Innovative farmer participatoryintegrated watershed management model: AdarshaWatershed, Kothapally, India – A success story! Patancheru502 324, Andhra Pradesh, India: International CropsResearch Institute for the Semi-Arid Tropics. 24 pp.

UNEP (United Nations Environment Programme). 1997.World atlas of desertification. Second edition. London,UK: Edward Arnold Pub. Ltd. 79 pp.

31


Continued land degradation brought by soil erosion inthe sloping lands has been a major constraint insustaining upland agriculture and food security inmost of Asia. Farming and other economic activitieshave become environmentally unsustainable causingdeleterious on-site and off-site effects. A few studieson soil erosion and soil conservation have beenundertaken, but results have not yielded sustainableland management options that can provide reasonablereturns without further degrading the resource baseand the environment. Greenland et al. (1994) made areexamination of approaches to research onsustainable land management and recommended anew research paradigm providing an organizationalmodel that engages scientists and research institutionsto tackle a common goal through a participatory,interdisciplinary, and community- and catchment-based approach. This led to the establishment of theManagement of Soil Erosion Consortium (MSEC) asone of the four consortia under the soil, water, andnutrient management (SWNM) system-wide initiativeof the Consultative Group on InternationalAgricultural Research (CGIAR).

In 1998, MSEC initiated a research project on soilerosion management in six countries in Asia with

support from the Asian Development Bank (ADB).The project aimed to develop and promotesustainable and socially acceptable community-basedland management options for sloping uplands througha participatory and interdisciplinary approach at thelevel of a catchment. This paper presents an overviewof the progress of the project in Indonesia, Laos,Nepal, Philippines, Thailand, and Vietnam. Ithighlights the progress of the project in catchmentresearch and summarizes its accomplishments in theother components of information dissemination andcapacity building vis-à-vis the outputs expected. Italso discusses the project’s strategy in governanceand management.

Program Objectives and ExpectedOutputsThe objectives of the program are to:• Develop sustainable and acceptable community-

based land management systems that are suitablefor the entire catchment.

• Quantify and evaluate the biophysical,environmental, and socioeconomic effects of soilerosion, both on-site and off-site.

The Management of Soil Erosion Consortium (MSEC): Linking Land andWater Management for Sustainable Upland Development in Asia

A R Maglinao and F Penning de Vries1

Abstract

A case study of soil erosion research under the Management of Soil Erosion Consortium (MSEC), a system-wide initiative of the Consultative Group on International Agricultural Research is presented. The studyaims at developing and promoting sustainable and socially acceptable community-based landmanagement options through a participatory and interdisciplinary approach. The study has beenundertaken in 6 countries namely Indonesia, Laos, Nepal, Philippines, Thailand, and Vietnam underdiverse agro-ecological situations. In each country up to 6 catchments were selected, the area of eachranging from 63 ha in Laos to 139 ha in Indonesia. The paper describes the program objectives, expectedoutputs, approach to program implementation, and research progress and discusses the results of erosionlosses in relation to land uses, catchment size, and nutrient depletion. The study links the land and watermanagement systems which were identified, in consultation with farmers, for conservation of naturalresources and improving soil fertility and lists some of the accomplishments.

32

• Generate reliable information and preparescientifically-based guidelines for improvement ofcatchment management policies.

• Enhance capacity of the national agriculturalresearch and extension systems (NARES) inresearch on integrated catchment management andsoil erosion control.The program focuses on three major components

to address the stated objectives. These are:• Catchment research to evaluate the effects of

different land management practices on water andnutrient flows in selected representativecatchments.

• Capacity building of participating NARES inresearch on integrated catchment management andsoil erosion.

• Dissemination of research results for enhancedadoption of land management technologies and formore accessible information as concrete basis fordecision making.Outputs from the activities are expected to be

generated in the first three years, but for some, alonger timeframe is needed. In fact, the consortium isenvisioned for a period of at least 10 years. Theexpected outputs are given in the project logicalframework and summarized as follows:• Decision support tools and guidelines based on a

better understanding of the on- and off-site effectsof soil erosion.

• Alternative technologies and land managementsystems that are socially and institutionallyacceptable to the communities in the catchment.

• Methodology for assessment of impacts andobtaining participation of farmers and otherstakeholders in the management of catchmentswhich includes policies that will improve themanagement of catchments by the localgovernment and the communities.

• Information and communication strategies toeffectively disseminate the results of the researchto the farmers and other land users.

• Enhanced NARES capacity in integratedcatchment management research.

• Improved program management for catchmentmanagement research.

Program ImplementationMSEC uses a new approach to the organization andimplementation of soil erosion management research.The approach provides a mechanism that engagesdifferent scientists and research institutions in acoordinated and participatory mode at the catchmentscale. Research planning and implementation isundertaken through consultation among concernedNARES, international agricultural research centers(IARCs), advanced research institutes (ARIs), non-governmental organizations (NGOs), and farmers.The NARES play the central role in the consortium,particularly in the participatory research, but with abroad responsibility for underpinning applied andstrategic research as well (Fig. 1). The InternationalWater Management Institute (IWMI) serves as theconsortium secretariat and facilitator. Project and

Figure 1. The research continuum showing the role of different groups in the implementation of MSEC.(Source: Craswell and Maglinao 2001)

33

institutional linkages are also pursued to establishthis partnership at the country level.

The study catchments were selected usingcarefully defined criteria and methodologicalguidelines developed for the purpose (IBSRAM1997). Monitoring stations equipped with automaticwater level recorders, manual staff gauges, sedimenttraps, automatic weather stations, automatic sedimentsamplers, and manual rain gauges were installed inthe catchments to collect hydrological and erosiondata. In addition, monitoring of the socioeconomicparameters and agricultural practices of the farmerswas likewise undertaken. The detailed methodologyused in carrying out the activities in the catchments isdiscussed in the individual country reports presentedin the 5th MSEC assembly (Maglinao and Leslie2001).

The best-bet land management options wereidentified in consultation with the farmers. Theinformation gathered from the monitoring of thebiophysical and socioeconomic data were explainedto the farmers during the discussion. The identifiedoptions were implemented by the farmers withtechnical assistance from the researchers. Regularmonitoring of the effect of the introduced options isunderway.

Progress in Catchment Research

Catchment profiles

The experimental catchments range from 63 ha in Laosto 139 ha in Indonesia with up to four smaller micro-catchments representing different land usesdelineated within (Maglinao et al. 2001). Mostcatchments have slopes ranging from 12 to 80%, andan average annual rainfall ranging from 1,080 to2,500 mm (Table 1). In the Philippines and Thailand,water flows in the creeks only during the rainy season.The catchments are dominated by annual cash cropswith some patches of perennials and are cultivatedprimarily by ethnic minorities. In general, the modelcatchments represent a resource management domainwith common biophysical and socioeconomiccharacteristics common to the marginal slopinguplands (Craswell and Maglinao 2001).

In most catchments, the farmers who farm in theareas live in the village outside the catchment. It is

only in the Philippines where most of the farmerssettle within the catchment. Land use rights areprovided to the farmers in Vietnam and Laos whilethose in the other sites are either shareholders orowners. A number of research and developmentinstitutions have been collaborating with the projectin all areas.

The sub-catchment in the Philippines is small (0.9ha) (Table 2). The sub-catchments in Indonesia areprimarily cropped either with upland annual crops orperennials, primarily rambutan. In the Philippines,the sub-catchments represent a combination of thearea cultivated to maize, vegetables, or potato andgrasslands with a small settlement area in one of thesub-catchments. In Vietnam, the sub-catchments arecropped with cassava, either monocropped orintercropped, but with areas of natural grass stillpresent. In Laos, a large part of the area of the sub-catchments is under rotating cultivation or bushfallow. Annual upland crops also predominate in thecatchments in Thailand and Nepal.

Erosion and land use

The existing land management practices in Indonesiainfluenced the degree of soil erosion in the differentsub-catchments within each of the three catchments.Except in Nepal, soil loss is higher in areas moreintensively cultivated with upland crops than thosewith perennials or left under grass cover (Table 2).This confirms the initial observations seen from thesame catchments a year before (Maglinao et al.2001). In Indonesia, sediment yield was highest in thesub-catchment MC-1I dominated by upland annualcrops, with soil loss of 6.7 t ha-1. This is presumablybecause of minimal soil surface litter and little canopycover of the catchment. On the other hand, the othersub-catchments, MC-2I and MC-3I, planted toperennials (primarily rambutan), lost relatively lessamount of soil (only about 1 t ha-1) during the sameperiod and yielded considerable amount of sedimentonly during the middle part of the rainy season(January).

In the Philippines, observations recorded duringApril 2000 to March 2001 showed the effect of landuse on soil erosion. The smallest sub-catchment MC-4P, which has a higher percentage of cultivated area,gave the highest soil loss of 53.9 t ha-1. The lowest soil

34

Ta

ble

1.

Pro

file

desc

rip

tio

n o

f th

e M

SE

C c

atc

hm

en

ts i

n p

arti

cip

ati

ng

co

un

trie

s.

Gen

eral

des

crip

rion

B

abon

H

uay

Pano

M

asra

ng K

hola

M

apaw

a H

uay

Yai

D

ong

Cao

Ba

sic i

nfo

rm

ati

on

Cou

ntry

In

done

sia

Lao

s N

epal

Ph

ilipp

ines

T

haila

nd

Vie

tnam

Pr

ovin

ce

Sem

aran

g L

uang

Pra

bang

C

hitw

an

Buk

idno

n Ph

rae

Hoa

Bin

h L

atitu

de

7°22

’25”

S

19°5

1’10

’’ N

27

°49’

N

8°02

’50’

’ N

18

°13’

20’’

N

20°5

7’40

’’ N

L

ongi

tude

11

0°29

’02”

E

102°

10’4

5’’

E

85°3

2’30

” E

12

5°56

’35’

’ E

10

0°23

’40’

’ E

10

5°29

’10’

’ E

E

leva

tion

(m)

390–

510

400–

700

650–

1400

10

80–1

505

400–

480

125–

700

Cat

chm

ent s

ize

(ha)

13

9 63

12

4 91

71

96

N

o. o

f eq

uipp

ed c

atch

men

ts

and

sub-

catc

hmen

ts

5

95

55

5

Bio

ph

ysi

ca

l a

ttrib

ute

s

Slop

e (%

) 15

–75

30–8

0 40

–60

8–35

12

–50

40–6

0

Geo

logy

and

land

form

B

asal

tic la

va

Shal

e, s

chis

t G

neis

s, s

chis

t

Bas

alt,

pyro

clas

tics

Silts

tone

, san

dsto

ne

Schi

st

Rai

nfal

l (m

m)

2,50

0 1,

403

2,20

0 2,

537

1,07

7 1,

500

Soils

In

cept

isol

U

ltiso

l, E

ntis

ol

Ince

ptis

ol, A

lfis

ol

Ulti

sol,

Ince

ptis

ol

Alf

isol

, Ulti

sol

Ulti

sol

Veg

etat

ion

and

land

use

R

ice,

mai

ze,

ram

buta

n

Fore

st, b

ush

fallo

w, r

ice,

m

aize

, job

’s te

ars

Fore

st,

gras

slan

ds, r

ice,

m

aize

, mill

et,

pota

to

Fore

st, o

pen

gras

slan

d, m

aize

, po

tato

, veg

etab

les

Soyb

ean,

mun

g be

an, t

amar

ind

Cas

sava

, ric

e,

mai

ze, t

aro,

gr

ound

nut

Hyd

rolo

gyPe

rman

entf

low

(wat

er f

low

s ye

ar

roun

d)

Pe

rman

ent f

low

(w

ater

flo

ws

year

ro

und)

Perm

anen

t flo

w

(wat

er f

low

s ye

ar

roun

d)

Inte

rmitt

ent f

low

(w

ater

flo

ws

only

du

ring

rai

ny s

easo

n)

Inte

rmitt

ent f

low

(w

ater

flo

ws

only

du

ring

rai

ny s

easo

n)

Perm

anen

t flo

w

(wat

er f

low

s ye

ar r

ound

)

cont

inue

d

35

T

able

1. c

ontin

ued.

Gen

eral

des

crip

rion

B

abon

H

uay

Pano

M

asra

ng K

hola

M

apaw

a H

uay

Yai

D

ong

Cao

S

ocio

eco

no

mic

att

rib

ute

s

Popu

latio

n H

ouse

hold

(H

H)

40

5

80

54

7

0

5

0

3

8 P

erso

ns

1,81

2

427

354

155

3,

655

196

Eth

nic

grou

p -

Lao

The

ung

(92%

);

Lao

Lum

(2%

)

Gur

ung;

Gha

rti;

Bra

hmin

; C

hhet

ri/T

haku

ri

Tal

aand

ig

Tha

iK

inh

(40%

);M

uong

Lan

d te

nure

O

wne

rs,

shar

ehol

ders

St

ate

owne

d;

land

use

rig

ht

With

cer

tific

ate

of o

wne

rshi

p;

leas

ed

Priv

ate

owne

r L

and

use

title

L

and

use

righ

t

Ann

ual i

ncom

e (U

S$)

372

296

72

6 62

7 77

4

On

farm

64

%

70%

41

%

- -

96%

O

ff f

arm

36

%

30%

59

%

- -

4%

Dom

inan

t cro

ps

Ram

buta

n, lo

wla

nd

rice

, upl

and

crop

s U

plan

d ri

ce, j

ob’s

te

ars

Mai

ze, r

ice,

m

illet

, mus

tard

, le

gum

es

Veg

etab

les,

mai

ze

Soyb

ean,

mun

g be

an

Cas

sava

, ric

e,

mai

ze,

grou

ndnu

t

Agr

icul

tura

l pr

actic

es

Tw

o cr

ops

in o

ne

year

Sh

iftin

g cu

ltiva

tion

Tw

o or

thre

e cr

ops

a ye

ar

Tw

o cr

ops

in

one

year

T

wo

crop

s in

on

e ye

ar

Tw

o cr

ops

in

one

year

36

Table 2. Land use and soil loss in the micro-catchments in different countries.

Soil loss1

Micro-catchment Area (ha) Land use (t ha-1)

Indonesia (15–75% slope)MC-1I 3.2 50% annual upland crops, coffee and nutmeg 6.7

on the upper slopesMC-2I 2.0 Rambutan and some bare plots 0.8MC-3I 38.5 Rambutan 1.0

Laos (30–80% slope)MC-1L 1.2 69% rotating cultivated land, 31% teak 0.5MC-2L 19.5 76% rotating cultivated land, 6% upland rice 0.6MC-3L 13.3 80% rotating cultivated land, 12% forest 0.0MC-4L 18.6 61% rotating cultivated land, 11% job’s tears, 2.1

10% forest, 7% upland riceMC-5L 8.8 53% rotating cultivated land, 35% forest, 2.8

8% upland riceMC-6L 1.7 56% rotating cultivated land, 13% forest, 31% teak 2.0

Nepal (40–100% slope)MC-1N 72.6 Mixed (45% upland, 5% lowland, 20% shrub, 0.08

30% forest)MC-2N 39.6 Mixed (60% upland, 10% shrub, 30% forest) 0.14MC-3N 11.5 Mixed (23% upland, 2% lowland, 35% shrub, 0.09

40% forest)MC-4N 1.6 Upland cultivated (100%) Traces

Philippines (8–35% slope)MC-1P 24.9 20% cultivated, 80% falcata, grassland 0.1MC-2P 17.9 40% cultivated, 60% grassland/forest 0.7MC-3P 8.0 10% settlement, 15% cultivated, 75% natural grass 1.0MC-4P 0.9 40% cultivated, 60% grassland 53.9

Thailand (12–50% slope)MC-1T 11.6 47% soybean-mung bean, 47% tamarind 0.1MC-2T 9.8 78.2% soybean-mung bean, 13% shrub 1.6MC-3T 3.2 94% tamarind, shrub 1.0MC-4T 7.1 51% soybean-mung bean, 23% mango, tamarind 0.4

Vietnam (40–60% slope)MC-1V 4.8 67% monoculture cassava, 33% natural grass 5.2MC-2V 9.4 24% cassava intercrop, 59% cassava monoculture, 4.3

17% natural grassMC-3V 5.2 Cassava intercrop 3.9MC-4V 12.4 26% cassava intercrop, 74% natural grass 2.01. Period of observation: Indonesia – March 2000 to February 2001; Laos – May to September 2001; Nepal – March to September 2001;

Philippines – April 2000 to March 2001; Thailand – June to September 2001; Vietnam – January to August 2001.

37

loss was in the sub-catchment MC-1P, which has alower percentage of cultivated area and a larger areacovered with grasses. The sub-catchment MC-3P,which has the lowest percentage of cultivated area butwith some settlement within, yielded a higher soilloss. The relatively higher soil loss in this sub-catchment which has 10% built-up area may beattributed to erosion from the foot trails and roadnetwork (Duque et al. 2001). Using a simulationmodel, Ziegler et al. (1999) showed that roadsgenerate runoff sooner during an event, and havegreater discharge values than other surfaces.Sediment transport was also greater. Footpathsemerged as important areas of accelerated runoffgeneration on agricultural fields that otherwiserequire large amount of rainfall to produce runoff.

In Vietnam, the data collected from January toOctober 2001 showed that among the sub-catchments, MC-1V (predominantly cassavamonoculture with some natural grass) had the largestsoil loss of about 5.2 t ha-1. The least was from MC-4V (predominantly natural grass and cassavaintercropping) at 2.0 t ha-1. The larger soil loss fromMC-1V (primarily cassava monoculture) whencompared with MC-3V (cassava intercropping)shows the effect of cassava intercropping system asopposed to cassava monoculture. At its peak growth,cassava provides only about 47–56% soil cover andmixed cropping or intercropping can increase thisprotection. The effect of natural grass in the sub-catchments was also manifested. Natural grassenhances infiltration, reduces runoff and runoffvelocity, and consequently reduces soil loss.

In Laos, observations made from May to October2001 showed that the micro-catchment with thesmallest proportion of rotating cultivated land andwith about 8% of upland rice (MC-5L) gave thehighest soil loss of 2.8 t ha-1 (Phommassack et al.2001). No erosion was observed in the micro-catchment with the largest proportion of rotatingcultivated land and about 12% forest (MC-3L).Contrarily, in Nepal lowest soil loss was observed inthe sub-catchment which was extensively cultivated(Maskey et al. 2001).

Soil erosion and farming operation

Farming operations in the field is related to how the landis managed and obviously can also affect soil erosion.

Observations in Indonesia showed that at the time ofplanting the upland crops (in November), rainfall ofnearly 600 mm produced sediment yield as high as about2 t ha-1. At this time, the soil surface was bare and the soilaggregates were loose because of tillage. With rainfallexceeding 400 mm in January, March, and April,sediment yield was greater than 1 t ha-1.

In Vietnam, at the start of the rainy season whenthe soil was still dry, there was not much runoff evenduring heavy rains. The amount of rainfall could havejust been enough to saturate the soil as rainfall ofmore than 300 mm in May did not result in significantincrease in runoff. But runoff sharply increased inJune. Incidentally, the overall cover density of thearea was also at its highest from June to August.

Soil erosion and catchment size

In the Philippines, the smallest sub-catchment (MC-4P) which incidentally has a large proportion ofcultivated area had the highest soil loss. In Indonesia,most of the sediments measured from the trap in thesmaller sub-catchments (MC-1I and MC-2I) were ofthe larger sized aggregates or particles (bed load)while for the larger sub-catchment (MC-1I), the finersediment (suspended load) dominated (Fig. 2). Thisreflects that during the erosion process, relativelysmall portion of soil aggregates was dispersed,especially for the MC-2I sub-catchment with notillage and with ideal cover. This also reflects that thesource of most sediment reaching the sediment trapwas relatively close to the trap and the larger thecatchment, the less the bed load contribution tosediment yield. In Laos, the measured erosion fromthe microplots reached as high as 78 t ha-1 comparedwith approximately 1 t ha-1 at the micro-catchmentscale. These observations reconfirm earlier reportsthat direct extrapolation of soil loss data from plotscale to small catchments and from small catchmentsto bigger catchments will lead to overestimation.

Nutrient depletion and off-site effects

Analysis of the soil eroded from the catchment inVietnam clearly showed that a large quantity of plantnutrients is carried away with the sediments. Thecatchment has lost a total of 740 kg organic matter, 39kg nitrogen (N), 31 kg P2O5, and 80 kg K2O. In one ofthe micro-catchments in the Philippines, 3.4 t organic

38

matter, 0.1 kg extractable phosphorus (P), and 7.9 kgexchangeable potassium (K) were lost with 62 t oferoded soil measured from May 2000 to August 2001(Duque et al. 2001). The data clearly show thatfarming without soil conservation results in soil andnutrient losses that could further result in lower cropyields and productivity. It is anticipated that withproper soil management and land use, soil andnutrient losses could be minimized.

One visible off-site effect of erosion is thesedimentation downstream due to the transport of soilfrom the uplands. An initial valuation of this effect atthe site in the Philippines was done by valuing thecost of dredging the silted irrigation canals of theManupali River Irrigation System (Carpina et al.2001). Since 1995, a total of 81,724 m3 of sedimentswas estimated to have been transported to the systemcosting about US$ 49,000 for dredging. Assumingthat 0.5% comes from the Mapawa site, it wasestimated to have contributed 409 m3 of sediments tothe irrigation system or an equivalent of US$ 245 ascost for dredging (Duque et al. 2001).

Although only some of the model catchments havenearby reservoirs where the effect of erosion onsedimentation can easily be assessed, initial attemptshave identified economic activities andenvironmental effects that could be studied toevaluate the effect of soil erosion off-site. The effectof erosion on the quality of the water that flowsdownstream and on the production of crops in thelowlands could also be assessed and valued. Relatedto the amount of soil loss, the amount of nutrients thatare carried away is a reflection of the degrading effect

of soil erosion. As the top soil is removed by watererosion, considerable amounts of plant nutrients arealso lost. This reduces the soil fertility resulting inreduced crop yields unless external nutrient inputs aresupplied. Thus the land management systems thatmust be introduced should be able to restore the lostfertility and increase farmers’ income.

Best-bet land management options

In most instances, the land management optionsidentified for introduction in the catchments werevariants of the contour hedgerow farming incombination with soil fertility management andanimal production. In the Philippines, the use ofnatural vegetative strips was identified by the farmersin Mapawa catchment. Naturally-growing grasses andsome agroforestry crops are used as hedgerows.Several farmers have already made use of thistechnique as a result of the promotion activity by theInternational Center for Research in Agroforestry(ICRAF) in the area. Adoption seems to be affectedby the tenure system of the farmers. About 50% of thelandowners (but none of the tenants) have adoptedsome conservation measures (Duque et al. 2001). Forthose who are interested but have not yet adopted, themajor reason is the cost of establishment.

In Indonesia, the option identified for Baboncatchment is a combination of fodder grass planted onalternate terraces of land currently used for annualupland crops and cattle fattening. The fodder grass isexpected to reduce erosion and serve as feed forlivestock. The identification of the option was based

Figure 2. Sediment output from different sub-catchments in Babon catchment, Indonesiafrom March 2000 to February 2001.

39

on lessons learned from elsewhere in Indonesia thatfarmers’ adoption and improvement of a conservationmeasure is determined by the economic contributionof the measure to the household economy. Farmersare attracted to a practice only if it promises economicbenefit.

Hedgerows of vetiver grass and Tephrosia candidain the alley cropping system with improved variety ofcassava have been introduced in the Dong Caocatchment in Vietnam. The technology interventionhas just been started. The farmers believe that thesystem will reduce runoff and soil loss; incorporationof hedgerow trimmings into the soil will increaseorganic matter content and thus improve soil fertility.

Other options that the farmers in the Philippineslooked at are the planting of pasture legumes duringfallow after growing potato, maize, or cabbageinstead of grass fallow for 3–4 years and plantingtiger grass and bamboo along the creek banks to serveas buffer. Tiger grass and bamboo are expected toprovide additional income as tiger grass is used forsoft broom and bamboo as props for the bananaplantation.

Presentation and discussion of the results ofmonitoring in the catchments with the farmers helpedin the identification of the land management optionsthat are more appropriate in the particular area. Whilethe farmers are aware of soil erosion and its negativeeffect, actual observations and the alarming figurespresented increased their appreciation of looking at alonger time horizon. As they are aware of thedeclining productivity of their land, they were alsointerested in fertility management. Of course, theirimmediate concerns are the benefits that they willgain in the short term. These concerns should be givenmore emphasis in introducing any interventions intheir farms.

Modeling and extrapolation

The soil erosion and hydrology model (PCARES)developed by Dr E Paningbatan of University ofPhilippines, Los Banos (UPLB) was tested usingMSEC data from the Philippines and found to needsome modifications. ICRAF agreed that MSEC alsolooks at a similar model that they have developed.The model was found to work using the data inputsrequired by the other model. The model was

demonstrated during the MSEC assembly in Hanoi,Vietnam. Further analysis of secondary data,complemented with the primary data collected fromthe different catchments, was done to produce mapsthat would be needed in the application of thePCARES and ICRAF models. These were usedduring the follow-up training conducted in Vientiane,Laos from 22 to 26 October 2001.

Other strategic research

Rainfall simulation studies in Thailand showed thatinfiltration rate increased while runoff coefficientdecreased with decreasing slope gradient. Runoffvolume decreased with increasing slope. Theseresults suggest that for convex landforms, the steepmid-slope zone can play the role of infiltration trapfor runoff water from upper gentler zone. This mayhave substantial impacts on flow volume generatedfrom small watersheds and on water quality.

The work of the University of Bayreuth in northernThailand showed that vegetation (and land use)influences the quality and rate of organic matter inputinto the mineral soil and is therefore one of the mainfactors controlling the composition of soil organicmatter (Moller et al. 2001). The conversion of foreststo cabbage cultivation resulted in enhancedbreakdown of soil organic matter as indicated by thelower contents of organic carbon and N in the mineralsoil in the latter land use. Reforestation with Pinus didnot lead to a significant buildup of organic matter inthe mineral soil.

Accomplishments

Tools and guidelines for improved decisionmaking and research implementation

MSEC’s emphasis not only on research but also onresearch methodology is expected to produce toolsand guidelines to support decision making andimproved implementation of MSEC researchactivities. One such output is an earlier publication bythe International Board for Soil Research andManagement (IBSRAM) which provides theguidelines for model catchment selection for MSEC.The site selection was based on criteria agreed uponby the consortium partners.

40

The minimum data sets for biophysical andsocioeconomic site characterization was preparedand employed. Protocols on the biophysical datacollection, analysis, storage, and retrieval werediscussed during the country visits of IWMI staff. Theexisting methodologies for the economic assessmentof soil erosion and nutrient depletion and on-farmtrials were adapted and applied in the MSEC sites.Existing soil erosion and hydrology models werereviewed to identify appropriate models applicable tothe MSEC work.

Alternative technologies andland management systems

The best-bet land management options introduced inthe farming systems in the catchments were identifiedin consultation with the farmers. These options havebeen elaborated earlier.

Enhanced capacity of the NARES

MSEC has so far conducted 10 training programs ontopics such as program management, participatoryapproaches, hydrology, geographic informationsystem (GIS) and modeling, and rainfall simulationdata analysis. More than 60 partners from 16institutions participated in these programs. Tengraduate students have been involved with the projectby conducting their research in the sites. Consultantshave also been tapped to provide assistance inaddressing more specific research topics.

Improved program management,monitoring, and evaluation

A better management of the program is expected to yieldmore usable tools and guidelines, relevant technologies,effective information dissemination, and significantinstitution development. MSEC envisions to optimizethe use of scarce research resources by strengtheninglinkages and collaboration with related projects andinstitutions. Regular visits to the MSEC sites are still amajor activity to monitor progress and anticipateproblems in implementation. The regular monthlymeeting of the MSEC group at IBSRAM has providedbetter interaction and sharing of ideas within theorganization. In 2001, 11 monthly meetings were held.

Summary and ConclusionPast research and development efforts have not beenable to provide sustainable solution to landdegradation problems, and soil erosion has remaineda major constraint in improving the living conditionsof the people in the marginal and sloping uplands inAsia. MSEC with the funding support from the ADBhas implemented a research project to address suchproblems employing a new research paradigm.

Concrete outputs in terms of capacity buildinghave been achieved. Complete instrumentation of theexperimental catchments and training of NARESparticipants have been useful in initiating the researchwork in the field. The initial results from differentparticipating countries have shown some interestingobservations on the erosion and hydrologicalprocesses occurring in the experimental catchmentsand the factors that may affect them.

The consortium approach and the participatoryand interdisciplinary research methods could be apotential key to sustaining upland development inAsia. With stronger and continuing partnershipsamong stakeholders, it has added a new dimension tosoil erosion management, with the potential toenhance the adoption and sustainability of introducedinterventions. MSEC will continue to employ thisapproach and the promising outputs will further bevalidated at different scales of application andexpanded to a much wider area for greater impact.Under IWMI’s leadership, the results of theintegration of land and water issues are expected to behighlighted.

ReferencesCarpina, N.V., Duque, C.M., De Guzman, M.T.L., Ilao,R.O., Quita, R.Q., Santos, B.G., Tiongco, L.E., andYadao, R.S. 2001. Management of soil erosion consortium(MSEC): An innovative approach to sustainable landmanagement in the Philippines. Pages 187–214 in Soilerosion management research in Asian catchments:Methodological approaches and initial results –Proceedings of the 5th Management of Soil ErosionConsortium (MSEC) Assembly (Maglinao, A.R., andLeslie, R.N., eds.). Bangkok, Thailand: IWMI, SoutheastAsia Regional Office.

Craswell, E.T., and Maglinao, A.R. 2001. A catchmentapproach to research on soil erosion in the marginaluplands of Asia. Pages 151–162 in Sustainable agriculture:

41

Possibility and direction. Proceedings of the 2nd Asia-Pacific Conference on Sustainable Agriculture, 18–20October 1999, Phitsanulok, Thailand (Suthipradit, S.,Kuntha, C., Lorlowhakarn, S., and Rakngan, J., eds.).

Duque, C.M., Tiongco, L.E., Quita, R.S., Carpina, N.V.,Santos, B., Ilao, R.O., and De Guzman, M.T. 2001.Management of Soil Erosion Consortium (MSEC): Aninnovative approach to sustainable land management in thePhilippines. Annual report submitted to IWMI, October2001.

Greenland, D.J., Bowen, G., Eswaran, H., Rhoades, R.,and Valentin, C. 1994. Soil, water, and nutrientmanagement research – A new agenda. IBSRAM PositionPaper. Bangkok, Thailand: International Board for SoilResearch and Management.

IBSRAM (International Board for Soil Research andManagement). 1997. Model catchment selection for theManagement of Soil Erosion Consortium (MSEC) ofIBSRAM. Report on the Mission to Thailand, Indonesiaand the Philippines. Bangkok, Thailand: IBSRAM.

Maglinao, A.R., and Leslie, R.N. (eds.) 2001. Soil erosionmanagement research in Asian catchments: Methodologicalapproaches and initial results – Proceedings of the 5th

Management of Soil Erosion Consortium (MSEC)Assembly. Bangkok, Thailand: IWMI, Southeast AsiaRegional Office. 275 pp.

Maglinao, A.R., Wannitikul, G., and Penning De Vries,F. 2001. Soil erosion in catchments: Initial MSEC resultsin Asia. Pages 51–64 in Soil erosion management researchin Asian catchments: Methodological approaches andinitial results – Proceedings of the 5th Management of SoilErosion Consortium (MSEC) Assembly (Maglinao, A.R.,

and Leslie, R.N., eds.). Bangkok, Thailand: IWMI,Southeast Asia Regional Office.

Maskey, R.B., Thakur, N.S., Shreshta, A.B., and Rai,S.K. 2001. MSEC: An innovative approach to sustainableland management in Nepal. Pages 187–214 in Soil erosionmanagement research in Asian catchments:Methodological approaches and initial results –Proceedings of the 5th Management of Soil ErosionConsortium (MSEC) Assembly (Maglinao, A.R., andLeslie, R.N., eds.). Bangkok, Thailand: IWMI, SoutheastAsia Regional Office.

Moller, A., Kaiser, K., Wilcke, W., Maglinao, A.R.,Kanchanakool, N., Jirasuktaveekul, W., and Zech, W.2001. Organically-bound nutrients in soils of small watercatchments under different forest and agrosystems innorthern Thailand. Pages 73–84 in Soil erosion managementresearch in Asian catchments: Methodological approachesand initial results – Proceedings of the 5th Management ofSoil Erosion Consortium (MSEC) Assembly (Maglinao,A.R., and Leslie, R.N., eds.). Bangkok, Thailand: IWMI,Southeast Asia Regional Office.

Phommassack, T., Chanthavongsa, A., Sihavong, C., andThonglatsamy, S. 2001. An innovative approach tosustainable land management in Laos. Annual reportsubmitted to IWMI, October 2001.

Ziegler, A.D., Giambelluca, T.W., and Sutherland, R.A.1999. Field rainfall simulation experiments to investigaterunoff generation and sediment transport on mountainousroads in Northern Thailand. Presented at the MethodologyWorkshop on Environmental Services and Land UseChange: Bridging the Gap between Policy and Research inSoutheast Asia, 30 May to 2 June 1999, Chiang Mai,Thailand.

42

Feasible Solutions for Sustainable Land Use in Sloping Areas

L Q Doanh and L V Tiem1

Abstract

Of 33 million ha of Vietnam’s natural land, only 8.6million ha are relatively flat; the rest are sloping lands,which cover 74% of Vietnam’s total land. In addition,the average land for cultivation per person is lowest inthe world. At the same time expansion of cultivatedarea in the flat delta region has almost exhausted.Thus, expansion of cultivation in the sloping lands in asustainable way is essential.

Sloping lands in the tropical and humid areas likeVietnam is an unsustainable environment forcultivation. Also, when the forest vegetative cover isdecreased, the threat for erosion and soil lossincreases. In the past, when the population pressurewas low, it was possible to maintain the natural fertilityby practicing shifting cultivation with 8- to 10-yearcycle of no cultivation (coefficient of <12%) and highforest cover. There was a balance between the loss ofnatural fertility due to cultivation and gain ofregenerated fertility due to the regenerated forest inthe long fallow period. As the population pressureincreased, the fallow period for fertility regenerationwas decreased and finally became zero, which meansthe cultivation coefficient was 100% and thus thethreat of erosion and soil loss was the highest (Table1). Emphasis should be laid on the selection ofcropping system on the sloping lands anddevelopment of the appropriate method of cultivationthat maintains soil fertility, guarantees farm income,and meets the investment level of the local people.

Constraints of the Sloping Lands

Erosion and leaching

Erosion and leaching are regular threats to the slopinglands and cause loss of surface fertile soil andnutrients, followed by acidulation of the soil. Theseeffects are severe if the cultivated land gets exposedwithout any vegetative cover, or land is cultivated justbefore the rainy season. The degree of erosiondepends on many factors such as rainfall, landerosion, length of the slope, level of the slope, plantcover, and measures for soil protection.

The inappropriate methods of cultivation followedin many regions such as burning vegetation andplowing of land just before the rainy season causeserious erosion and leaching because the newlyplowed lands can easily be eroded. Erosion causessoil loss both in quality and quantity, land degradationand leaching, loss of nutrients, and acidulation of thesoil, and increases aluminum toxicity.

Soil degradation

Due to the destruction of the forest and annual regimeof cultivation, burning trees for cultivation of foodcrops leads to serious deterioration of the slopinglands in many regions. Fast deterioration of soil limitsproduction in the sloping lands. The increase ofaluminum toxicity in soil results from soil acidulation.

1. Vietnam Agricultural Science Institute (VASI), Hanoi, Vietnam.

In Vietnam nearly 74.1% of land is sloping land and subjected to soil erosion and degradation dueto human interventions of cultivation of these lands. Unless some technologies are developed tocheck the soil erosion, the sustainability of crop production in these lands is uncertain. Somesolutions like intensive cultivation in the valley to reduce pressures on uplands, use of hybrid riceand maize instead of local cultivars, introduction of commercial perennial crops instead of foodcrops, and appropriate sloping land management practices have been discussed in detail.

43

In addition, there is decrease in the availability ofminerals such as phosphorus, potassium, calcium,magnesium, and zinc.

Drought in the dry season

It is difficult for the sloping lands to conserve water.Cultivation is dependent on rainwater and in the dryseason, there is always serious drought. In manyregions, there is not enough water for animals andhuman beings. Drought and shortage of water are themain problems in the sloping lands; if it rains a monthlater than expected, crop failure is common. Droughtin the dry season is due to the loss of forest anduncontrolled cultivation on the sloping lands.

The isolated position

Due to poor transportation and separated relief, manysloping areas become isolated resulting in lowexchange of commodities. This slows down theprocess of the shift of cropping pattern from shiftingcultivation by clearing the forest to cultivation of foodcrops or perennial commercial crops.

Poor infrastructure

The sloping areas usually lie in mountainous areas, farfrom the center of development. So infrastructure inthose areas are very poor, which in turn badly affecteconomic development.

High rate of poverty and low level ofeducation

The inhabitants of the sloping areas are mainly ethnicminorities and have high rate of poverty and low levelof education. The construction works for erosionprevention, water conservation, and growing of cropswith high commercial value demand more investmentand higher level of farming technology.

Depletion of plant cover

The traditional methods of cultivation turned largeareas into barren land. When the forest is destroyed forfarming of food crops, majority of area becomesacidified and infested with alang grass (Imperacyclindrica). A few years later, the people have toleave the land and look for new farming land. Thereduced forest cover affects the general ecologicalenvironment causing drought, flood in the plain, andflash flood in the mountainous area.

The Potential of Sloping Land

The potential for expansion of cultivatedland

The sloping land constitutes an important componentof agricultural production, covering 973 million ha(66%) out of 1500 million ha of arable land of theworld. In Vietnam, out of 33.1 million ha of natural

Table 1. Change in cultivation coefficient.

CultivationYear Land use pattern coefficient1

Before 1954 Traditional shifting cultivation (long fallow period) <20

1965 Traditional shifting cultivation (long fallow period) 20–25

1980 Transfer from traditional to non-traditional shifting cultivation 30–40

1985 Non-traditional shifting cultivation (short fallow period) 45–50

Present Non-traditional shifting cultivation (very short fallow period) 50–100Number of cultivation year

1. Cultivation coefficient (R%) = 100Total cultivation cycle

including fallow period (years)

×

44

land, 24.4 million ha is still unexplored (74%). Due tothe difficulty in exploration, the proportion ofcultivation on sloping land, and lower populationdensity in comparison with the plain areas, the slopingland still holds good potential for expansion ofcultivated land.

The forest potential

The forest is not only a precious natural resource ineconomic terms but also has high environmental valuein terms of water regulation and oxygen-carbonregulation. Most of the forest in Vietnam isconcentrated in the sloping area.

Potential of commercial crops

Most of the flat land is reserved for the cultivation offood crops to meet food demand. Most of the cropswith high commercial and/or export value such ascoffee, rubber, sugarcane, coconut, palm oil, andcocoa can be planted in the sloping land.

Energy potential

The high altitude and high volume of rain providesgood potential for hydropower from the sloping areas.

Livestock development potential

The sloping area is potentially good for thedevelopment of pasture to provide nutritional feed forlivestock. However, the development of pasture fieldsdepends on the customs and level of development ineach region.

Change in Land Use Systems inUpland Regions

Traditional shifting cultivation transformedto non-traditional shifting cultivation

Traditional shifting cultivation has been practiced forlong in history, but increase in population leads to theincreased demand for land for cultivation. Adoption offarming methods from the plain area by immigrantsmakes the traditional shifting cultivation graduallychange to non-traditional form of shifting cultivation

(Fig. 1). The differences between these two forms aregiven below:Traditional shifting cultivation:1. Minimum tillage2. Hole sowing (minimum tillage)3. Long fallow period4. Slow soil degradationNon-traditional shifting cultivation:1. Thorough land preparation2. Sowing on flat land or ridges by dibbling3. Short fallow period4. Rapid soil degradation

Expansion of continuous cultivation areas

Continuous cultivation does not have a fallow period.The shift from shifting cultivation to continuouscultivation increases cultivation coefficient and helpsto increase productivity and output of crops on thesloping lands but causes severe land degradation(Table 2). This process of change occurs very quicklytogether with the population increase because of theincreasing demand for food. There are two types ofcontinuous cultivation: monocultural continuouscultivation and rotated continuous cultivation. Apartfrom high economic value, continuous cultivationresults in erosion and leaching and decrease in soilfertility (Table 2).

Diversification of land use forms

Together with the demand from both the localpopulation and the market, the forms of sloping landuse have become more varied. In the past, the mainform of land use was forest and shifting cultivation.Now there are many new forms of land use such ascontinuous cultivation, house garden, and hill garden.In different regions, based on their climatic andsocioeconomic conditions, specific forms are created.The diversified forms of sloping land exploration aredescribed in Figures 2 and 3.

Some Feasible Solutions forSustainable Use of Sloping Lands

Intensive cultivation of rice in the valleyIn most sloping lands native communes practicemixed farming of “sloping land and flat fields”. In

45

Figure 1. Change in land use systems by increase of cultivation coefficient.

Table 2. Relationship between soil fertility and farming practice in uplands.

R1 Available P2O5 Total K2O Available K2OFarming practice (%) (mg 100g-1 soil) (%) (mg 100g-1 soil) pH

3 years cultivation + 20 8.8 1.9 8 6.412 years fallow

5 years cultivation + 33 4.8 1.7 17 6.010 years fallow

15 years cultivation 100 4.8 1.7 9 5.5(maize-cassava-beanrotation)

5 years cultivation of 100 1.6 1.5 27 4.8annual crop +10 yearsbanana

15 years cassava monoculture 100 1.0 1.4 8 5.01. R = Corelation coefficient.

46

Figu

re 2

. Far

mer

land

use

syst

em.

Figu

re 3

. Tra

nsfo

rmed

land

use

syst

em o

f the

pre

sent

.

47

fact, there are a very small number of communes withonly upland cultivation. The survey of somecommunes of the mixed farming group showscorrelation between the area of rice and the area ofupland fields. In the communes with larger area ofirrigated rice and high yield, the upland area undercultivation for food crop is reduced; this also dependson the availability of land in each commune. Forexample, three communes, Chieng Phu, Phat, and Thaiare in Chieng Pan village but the average upland areain each commune varies. The Thai commune has thelargest upland area (almost three times that of ChiengPhu) because the rice area of this commune is thesmallest. The yield of upland rice is the lowest. On thecontrary, in Chieng Phu the area of irrigated rice is notlarge, but the yield is high; therefore, the upland area isthe smallest. Intensive cultivation of rice in the valleyhelps reduce forest destruction and burning of forestsfor expansion of land for cultivation of food crops(Table 3).

The yield of rice in the mountainous communes isstill low, but the opportunity for intensive cultivationexists. Many rice varieties with high yield potentialcan be grown in the mountainous area. Manymountainous communes successfully planted hybridrice. In spite of the high fertility in the valleys, plantingwithout balanced fertilization (i.e., application of ureaonly) resulted in low yields. With the newtechnological advancement, such as new varieties withhigh yield potential, balanced fertilization,construction of small channels for irrigation, and useof intensive cultivation of rice in the valley area can becarried out successfully.

Hybrid rice and maize varieties

The cultivated land in the mountainous area has lowsoil fertility and the ethnic farmers either do not use oruse little fertilizer. Hence varieties that need intensivecultivation should not be introduced. However, it hasbeen proved that the hybrid rice and maize varietieshave been successfully cultivated in mountainousarea. Many hybrid varieties gave high yield and haveshown better resistance to pests than normal varieties.

Development of hybrid maize varietyin the mountainous area

In recent years, the area under upland rice in northernmountainous areas has reduced greatly. At the sametime, the area under maize has increased, especiallyafter introduction and expansion of hybrid maize. Inthree North Western provinces the area under maizeincreased by 36% (from 67,100 ha to 91,900 ha)during 1995–99 and hybrid maize occupies 76% oftotal area under maize. However, in nine provinces ofthe Red River delta, the area increase was only 1%(from 75,100 ha to 76,100 ha).

The replacement of upland rice by hybrid maize notonly brought about higher economic gains but alsohelped to reduce soil erosion and leaching because thecover of maize is better than rice. Maize rootspenetrate deeper and thus the crop has better droughtresistance. Normally, the land for upland rice can becultivated for 3 years; then it should be leftuncultivated or used to grow maniocs. Over expansionof maize area by destruction of the forest or growingmaize on sloping lands (>15% slope) should not be

Table 3. Average area of irrigated and upland rice per household in Chieng Pan, Son La province,Vietnam.

Ethnic Irrigated rice Paddy yield Upland rice Upland rice/group/ village (ha) (t ha-1 yr-1) (ha) irrigated rice Land use system

Kinh 0.09 10 0.40 5:1 Monoculture(Chieng Phu) (2 crops yr-1) continuous cultivation

Thai 0.25 8 0.80 3:1 Rotated cultural(Phat village) (2 crops yr-1) continuous cultivation

Kh Mu 0.07 3 1.10 15:1 Shifting cultivation(1 crop yr-1)

48

Intensive cultivated hill sugarcane

In comparison with upland rice and maize, sugarcanehas better adaptability. Sugarcane can be planted in theupland field left uncultivated for many years after rice,maize, or on soils depleted of fertility. This is becausethe roots of sugarcane can penetrate deeper than riceor maize. In a survey of the hill sugarcane growingcommunes of Ngoc Lac District, Thanh Hoa province,it was observed that sugarcane has higher economicvalue than other annual crops like cassava, uplandrice, and maize on the sloping lands (Table 4).

recommended. Intensive cultivation of hybrid maizealso demands higher amounts of fertilizers. Maizebeing an annual crop, cultivation of annual crops onsloping lands is not sustainable.

Development of hybrid rice in themountainous area

Due to the price subsidies by the state, the area ofhybrid rice in the valley fields and in the terrace fieldsin the mountainous areas is increasing rapidly. Thepercentage of valley field rice of the mountainousvillages is very low and due to the slope relief, it isvery difficult to expand the area of rice. Therefore, theonly way to increase food security is to increase riceproductivity in these areas.

Although it is not easy to transfer improvedtechnology of rice to the farmers in the mountainousarea, the results obtained so far in the mountainousarea confirmed the feasibility. In 1998 when hybridrice was not planted in Quan Than San (Sa Ma Caidistrict, Lai Chau Province), the people had enoughrice for food for 8 months; in the remaining monthsthey had to depend on maize. Now hybrid rice covers80% of their terrace fields. People have enough ricefor food and do not have to eat maize. The quantity ofmanure used for rice is much higher than before.Expansion of hybrid rice and increased fertilizer usagein the mountainous areas increase food security andmaintain soil fertility.

Securing food supply with commercialperennial plants on the sloping land

The market economy and expansion of agriculturalproduct trading allows shift in cultural practices.Earlier farmers were forced to grow rice on the slopingland when market was not available. Growing ofcommercial crops was possible with markets and thishas improved income generation and livelihoods. Onthe sloping lands, forest trees and perennial crops arethe best cropping patterns that can help reduce soilerosion. There are many mountainous communes thathave been successful with the cultivation ofcommercial perennial crops (i.e., fruit trees, industrialplants, medicinal herbs, etc.). Hill sugarcane andbamboo garden are best examples.

Table 4. Economic efficiency of different land usepatterns.

Income BenefitLand use pattern (US$ ha-1 yr-1) (US$ ha-1 yr-1)

Extensive cassava 200 27Intensive cassava 373 73Upland maize 573 373Intensive sugarcane 1067 533Bamboo 1000 853

Intensive cultivated hill sugarcane is different fromthe local sugarcane. In addition to the ditches dug asparallel lines, balanced application of nutrients mustbe ensured. Because sugarcane requires morenutrients than other crops, it is necessary to applyrequired quantities of nutrients. After sugarcane isharvested, the ratoon crop is covered to preserve soilmoisture and return part of the nutrients back to thesoil. Sugarcane is a semi-perennial crop, and needs tobe planted every three years. Therefore, the amount ofsoil lost through erosion is much less in comparison toother annual crops like upland rice, maniocs, beans,maize, etc. (Table 5). Even if there is no major changein consumption of sugarcane, intensive cultivated hillsugarcane is relatively sustainable and ensures bothincreased incomes and maintenance/restoration of soilfertility.

Bamboo garden

Bamboo is a forest crop having high adaptability. Itcan be planted on degraded land where other cropslike maize, upland rice, or sugarcane cannot grow. It

49

can also be planted on sloping lands with >25% slope.The survey carried out in some bamboo planting areasin the western districts of Thanh Hoa province showedthat economic value of bamboo is approximatelyequal to maize and sugarcane. The bamboo garden cancontrol erosion better and the soil loss is much less incomparison to annual crops (e.g., maize) and semi-perennial crops (e.g., sugarcane). Bamboo can retainwater better on the bamboo hill, when the plants arehigh enough to provide full cover; the groundwaterlevel also increased significantly. The humus contentalso increased after many years of planting bamboothough soil fertility improvement was not as high as inbroad-leaved crops.

Appropriate methods for sloping landcultivation

As mentioned earlier, due to the shortage of cultivatedarea, it is impossible to stop cultivation of food cropson the sloping lands when the population iscontinuously increasing. In the last few years, in theframework of the cooperation between the VietnamAgricultural Science Institute (VASI) and the Centre

de cooperation internationale en recherchéagronomique pour le développement (CIRAD),France, a study on agricultural system of themountainous area was initiated to identify promisingtechniques that meet the above mentionedrequirements. Many promising technologies wereidentified; e.g., mini-terraces, planting parallelhedges, and hedgerow cropping.

Strengthening the capacity of local peopleand officials

The technological innovations need to be widelyadopted in mountainous areas. Besides investment ininfrastructure, it is necessary to increase trainingactivities, information dissemination, and networkingactivities to create conditions for both local peopleand officials to improve their capabilities for properunderstanding and application of technologiesenvisaged for the sloping lands.

ConclusionsThe most difficult question that needs to be addressedimmediately is how to undertake production activitieson the sloping lands, which are easily subjected toerosion and leaching and have low soil fertility. Thefollowing solutions help address some of the concernseffectively:• Intensive cultivation in the valley to reduce the

pressure of upland exploration.• Use of hybrid rice and maize cultivars for high yields.• Securing food supply with commercial perennial

crops.• Developing appropriate methods for sloping land

cultivation.All technologies should be feasible solutions that

secure food supply and improve household incomes inthe mountainous areas and help to preserve soilfertility.

Table 5. Soil erosion from different land usepatterns.

Land use Slope Soil loss OM loss1

pattern (°) (t ha-1 yr-1) (t ha-1 yr-1)

Bare hill 18 40.28 -

Annual cassava 17 75.16 1.650

Broods 16 17.40 0.504

Sugarcane 17 14.46 0.376(2 years old)

Bamboo 18 9.12 0.200(9 years old)1. OM = Organic matter.

50

Productivity and Resource Use Management of Soybean-basedSystems in a Vertic Inceptisol Watershed

Piara Singh, S P Wani, P Pathak, R Sudi, and M S Kumar1

Abstract

Erratic rainfall and land degradation are the major constraints affecting productivity of soybean-basedsystems in central India. Operational scale watershed experiments were conducted for six seasons on aVertic Inceptisol at ICRISAT, Patancheru, India to study the effects of improved management on landdegradation, rainfall use efficiency, and the productivity of the soybean-chickpea sequential and soybean/pigeonpea intercropping systems. Improved management comprised of integrated nutrient management(additions of crop residues and Gliricidia loppings) and sowing on broadbed-and-furrow (BBF) system.The traditional management consisted of sowing on flat landform and no addition of external sources ofnutrients, except P application. These treatments were imposed on medium-deep and shallow phases of thesoil type. The BBF system decreased surface runoff (16% of rainfall) compared to the flat system (21% ofrainfall) with concomitant increase in deep drainage. Mean rainfall use efficiency was 70 to 73% acrosscropping systems and soil depths. Integrated nutrient management resulted in balanced soil N budget,whereas the traditional system showed a deficit of about 50 kg ha-1. Denitrification and leaching losseswere negligible. Landform treatments did not increase the crop yields significantly. Total productivity ofthe soybean-chickpea system ranged from 2.3 to 2.7 t ha-1 of seed yield and that of soybean/pigeonpeaintercropping system ranged from 2.1 to 2.3 t ha-1 over the years, still showing a yield gap of about 1 t ha-1 forthe Patancheru site. Simulated yield gap analysis for other nine sites in central India showed a mean yieldgap of 1.2 to 1.9 t ha-1 under rainfed conditions, which could be even more in good rainfall years. Thisstudy has shown the potential of the improved technology for higher productivity and efficient use ofnatural resources on a Vertic Inceptisol, which has potential applications in the target region of centralIndia.

1. International Crops Research Institute for the Semi-Arid Tropics (ICRISAT), Patancheru 502 324, Andhra Pradesh, India.

Soybean is grown on about 6 million ha in Indiamainly in the states of Madhya Pradesh, Maharashtra,and Rajasthan (FAO 2002). It is also grown in thestates of Uttar Pradesh, Andhra Pradesh, Karnataka,and Tamil Nadu (Fig. 1). Despite the increase inproduction and area under soybean in the country, itsproductivity has stagnated at less than 1.0 t ha-1 (Fig.2). Major increase in the area under soybean hasoccurred in Madhya Pradesh, where the annual rainfallspatially ranges from 800 to 1200 mm. The soils areVertisols and associated Vertic Inceptisols. Majorconstraints to the production of soybean-basedsystems in Madhya Pradesh are physical, chemical,and biological forms of land degradation. Soil erosionis particularly high in central India because Vertisols

and associated soils which predominate the landscapeare prone to sheet and gully erosion under tropicalmonsoon climate. Therefore, to sustain crop yields ofsoybean-based systems, it is essential that landdegradation is minimized, natural resources areefficiently used, and efficient cropping systems areintroduced that will optimally use the naturalresources. Considering these issues, a small watershedwas developed on Vertic Inceptisol at the InternationalCrops Research Institute for the Semi-Arid Tropics(ICRISAT), Patancheru, Andhra Pradesh, India withthe following objectives:• Evaluate the productivity of the selected soybean-

based cropping systems with improved andtraditional management on a Vertic Inceptisol.

51

Figure 1. Soybean distribution and its agroecology in India.

Figure 2. Area, production, and productivity of soybean in India.

52

• Evaluate the extent to which land degradation (soilerosion and nutrient depletion) can be minimized,productivity can be increased, and the othernatural resources can be efficiently used withimproved management.

Agroecology of the Patancheru Siteand the Target RegionThe annual rainfall (average of 30 years) ofPatancheru is 800 mm. In the past 20 years, the rainfallwas below normal in 8 out of 20 years. On an average3 years out of 5 are average rainfall years (<25%variation of normal rainfall), 1 out of 5 years is aboveaverage (>25% of normal rainfall), and 1 out of 5years are below average (<75% of normal rainfall).Since the average potential evapotranspirationrequired by any rainy season crop in Patancheru areais 600 mm, the below average years can be broadlytermed as drought years. The coefficient of variabilityof the annual rainfall is 27% based on rainfallobserved during the past 30 years.

The daily rainfall during the past seven years(1995–2001) clearly shows that at Patancheru therainfall is unevenly spread during the rainy season(Fig. 3). It is not uncommon to receive 50% of the totalseasonal rainfall in a few high volume, high intensityrainstorms. This amount of rainfall exceeds the waterintake rate of most soils at Patancheru when thesurface is fully wet. Rainfall then flows as surfacerunoff and causes extensive soil erosion and loss ofnutrients. Effective rainfall is thus a fraction of thetotal rainfall received. This leads to reduced rainfalluse efficiency caused by rainfall variability during therainy season.

Madhya Pradesh receives annual rainfall varyingfrom 800 to 1600 mm. Eighty percent of this isreceived from mid-June to mid-October. The rainfallincreases from 800 mm in the western parts of MadhyaPradesh to 1500 mm in the eastern parts (Fig. 1).Madhya Pradesh has six soil types ranging fromalluvial soils to deep black soils distributed over 12agroclimatic zones. Medium to deep black soilsreceive 800 mm to 1200 mm annual rainfall. Becauseof poor manageability of black soils the croppingintensity is low (117%). Soybean-based croppingsystems are mainly practiced. Various constraints of

these soils are low infiltration rate, low organic mattercontent, poor structural stability, and vulnerability ofsoil to erosion. Harvesting and recycling of water on awatershed basis is required for sustaining productionon these soils. Water balance of the two sites (Indoreand Bhopal) in Madhya Pradesh is compared with thatof Patancheru site (Fig. 4). The data shows thatalthough the length of water availability period islonger at Patancheru compared to Indore and Bhopal,there are 3 months of rainfall exceeding potentialevapotranspiration. Total rainfall in July and Augustreceived at Indore and Bhopal is greater than thatreceived at Patancheru. This indicates that theopportunities for water harvesting are greater at thesetwo sites in Madhya Pradesh compared to thePatancheru site. The problem of land degradationbecause of soil erosion is also greater at Indore andBhopal. Therefore, it is expected that technologicalconcepts developed at Patancheru site would havepotential application at these two sites in MadhyaPradesh.

Current and Potential Land UseSystems in the Target RegionVarious cropping systems are being followed in thetarget region of Madhya Pradesh where soybean crophas even greater potential in the region. Croppingsystems practiced are cotton-wheat, maize-chickpea,pearl millet-wheat/mustard, pigeonpea-chickpea, rice-chickpea/mustard/wheat, sorghum-wheat, and soybean-wheat/chickpea. Madhya Pradesh accounts for morethan 75% of total area and production of soybean inIndia. Soybean-wheat is a popular rotation in partiallyirrigated areas, while soybean-chickpea/safflower/lentil/linseed /mustard cropping systems are practicedin the rainfed areas. Soybean is also grown as anintercrop with medium-duration pigeonpea.

With improved management of black soils(Vertisols and Vertic Inceptisols), soybean is furtherexpected to replace high input requiring crops such ascotton, maize, pearl millet, and sorghum grown duringthe rainy season. Soybean being a legume crop, thereplacement of other crops by soybean is expected toresult in saving of chemicals such as mineral fertilizersand biocides, thus contributing to the alleviation ofenvironmental pollution.

53

Figure 3. Daily rainfall at ICRISAT, Patancheru, India during 1995 to 2001.

54

Figure 4. Water balance of Indore, Bhopal, and Patancheru sites.

55

Watershed Research at ICRISATThe Patancheru site (BW7) serves as a reference siteon a Vertic Inceptisol for strategic research on thesoybean-based farming system (Fig. 5). Because theVertic Inceptisols are relatively shallow in depth, thefocus of the technology for these soils was to rechargethe groundwater through land management andharvesting of excess rainfall in percolation tanks.Integrated nutrient management practices (legumes inthe system, biofertilizers, and chemical fertilizers) wasfollowed to meet the nutrient needs of the crops and tominimize the pollution of groundwater. Based on thetoposequential soil depth the 15-ha watershed wasdivided into two hydrological units: medium-deep(50–90 cm) and shallow (<50 cm). These twohydrological units were further divided into two unitson which two landform treatments were imposed:(1) broad-bed and furrow (BBF) with Gliricidiasepium on graded bands; and (2) flat landform withsowing on grade. Thus the watershed was divided intofour hydrological units and on each unit two croppingsystems were evaluated: (1) soybean/pigeonpeaintercropping system; and (2) soybean-chickpeasequential cropping system except in the first yearwhen only soybean-chickpea system was evaluated.This experiment has been conducted for six years from1995/96 to 2000/01 and data have been collected oncrop yields, runoff and soil erosion, water and nitrogen(N) balance, and groundwater recharging.

Surface runoff and soil erosion

Significant amount of runoff occurred in four out ofsix years of study (Table 1). Surface runoff, thoughvariable over the years, averaged about 25% of therainfall. Surface runoff was relatively more from themedium-deep soil than from the shallow soil. Becauseof greater time of concentration, total seasonal runoffand peak runoff rates were lower on the BBF landformcompared to those on the flat landform on both the soildepths. On the medium-deep soil, BBF landform on anaverage reduced surface runoff by 22% compared tothe flat landform, whereas on the shallow soil suchreduction in runoff by the BBF system was about 18%.Whenever surface runoff occurred, the peak runoffrates were lower on the BBF landform than on the flatlandform and did not differ significantly between the

two soil depths (Table 1). Soil erosion observedbetween the landform treatments over the seasons wasproportional to the amount of runoff observed and thepeak runoff rates. Total soil loss averaged over theyears was 2.4 to 2.5 t ha-1 on the BBF and 4.0 to 4.5 tha-1 on the flat landform. Maximum soil loss wasrecorded during 2000/01 season, which was 6.5 to 6.7t ha-1 on BBF and 11.1 to 12.0 t ha-1 on flat landform ontwo soil depths.

Simulated water balance of the soybean-chickpea sequential system

Water balance of the soybean-chickpea sequentialsystem was simulated using Decision Support Systemfor Agrotechnology Transfer (DSSAT) model (Tsuji etal. 1994). Model parameters for soil water balancewere calibrated using the observed surface runoff andsoil water dynamics data. Though the simulated runoffvaried across seasons and soil depths, it averaged 160mm for BBF medium-deep, 157 mm for BBF shallow,213 mm for flat medium-deep, and 196 mm for BBFshallow (Table 2). On an average surface runoffconstituted 16% of seasonal rainfall for the BBFlandform and 20% of seasonal rainfall for the flatlandform. This resulted in concomitant increase indeep drainage in both the soil types. For the medium-deep soil, average deep drainage for the croppingperiod was 135 mm for BBF (14% of rainfall) and 93mm for the flat landform (11% of rainfall) (Table 2).For the shallow soil, average deep drainage was 183mm for the BBF landform (21% of rainfall) and 139mm for the flat landform (16% of rainfall). Total wateruse by the soybean-chickpea sequential systemaveraged over the seasons ranged from 481 to 515 mmacross soil types and landforms (70 to 73% of rainfall)(Table 2).

Simulated water balance of the soybean/pigeonpea intercropping system

Water balance of the soybean/pigeonpea inter-cropping system was simulated using the AgriculturalProduction Systems Simulator (APSIM) (McCown etal. 1996) following the same approach as for thesoybean-chickpea sequential system. As theintercropping system is of longer duration than thesequential system the values of various water balance

56

Figure 5. BW7 watershed at ICRISAT, Patancheru, India.

Table 1. Seasonal runoff, peak runoff rates, and soil loss in BW7 watershed at ICRISAT, Patancheru, India.

Surface runoff Peak runoff rate Soil loss(mm) (m3 s-1 ha-1) (t ha-1)Rainfall

Year (mm) BBF1 Flat BBF Flat BBF Flat

Medium-deep1995/96 657 168 196 0.068 0.098 NR2 NR1996/97 961 232 263 0.120 0.137 3.2 4.91997/98 546 1 3 0.003 0.003 0 01998/99 1043 200 290 0.135 0.145 2.7 5.51999/2000 401 0 0 0 0 0 02000/01 1062 477 641 0.270 0.385 6.5 12.0

Mean 778 180 232 0.099 0.128 2.5 4.5

Shallow1996/97 961 130 194 0.109 0.235 1.8 3.71997/98 546 2 2 0.003 0.004 0 01998/99 1043 251 283 0.130 0.168 3.4 5.31999/2000 401 0 0 0 0 0 02000/01 1062 489 588 0.270 0.320 6.7 11.1

Mean 803 174 213 0.102 0.145 2.4 4.01. BBF = Broad-bed and furrow.2. NR = Not recorded.

57

components (runoff, deep drainage, and water use)were generally larger than those for the sequentialsystem (Table 3). However, when these water balancecomponents were expressed as percentage of seasonalrainfall, their values were similar to those simulatedfor the soybean-chickpea sequential system.

Groundwater recharge

Overall improvement in land management in BW7watershed resulted in increased groundwater recharge.Surface runoff and deep drainage water was capturedin surface tanks and dug wells in the watershed.During 1996, 1998, and 2000 there was significantrise in water level (5 to 6 m) in the wells situated at thelower part of the watershed (Fig. 6). Because of lowrainfall during 1999 the rise in water level in the wellswas small. This additional water availability in thewells helped provide supplemental irrigation to thehorticultural crops in the lower part of the watershed.Thus the overall rainfall use efficiency on watershedbasis was greater than 50% in most years.

Measured nitrogen balanceIntegrated nutrient management followed in theimproved system (sowing on BBF + Gliricidia onbunds + addition of compost) resulted in balanced Nbudget for the soybean-chickpea sequential andsoybean/pigeonpea intercropping systems (Table 4).In spite of N contributions through rainfall, biologicalnitrogen fixation (BNF), leaf fall, and roots, there wasa net loss of about 50 kg N ha-1 for both the croppingsystems in the conventional system (flat landformtreatment) during the first four years. In the improvedsystem, the application of Gliricidia loppings andcompost provided about 45 kg N ha-1 without affectingthe yield of crops in the nearby rows, thus balancingthe N budget. Pigeonpea derived up to 89%, soybeanup to 75%, and chickpea up to 42% of their Nrequirement through BNF.

Simulated nitrogen balance

Simulated N uptake and N fixation by the soybean/pigeonpea intercropping system was variable across

Table 2. Simulated water balance components of soybean-chickpea sequential system in varioustreatments on a Vertic Inceptisol, ICRISAT, Patancheru, India.

Surface runoff (mm) Deep drainage (mm) Crop water use (mm)Rainfall1


Medium-deep soil1995/96 653 81 92 133 137 517 512(12)3 (14) (20) (21) (79) (78)1996/97 973 231 272 165 172 559 563(24) (28) (17) (18) (57) (58)1997/98 532 1 3 0 1 563 565(0) (1) (0) (0) (100) (100)1998/99 876 171 259 215 126 518 510(20) (30) (25) (14) (59) (58)1999/2000 401 8 18 0 0 395 406(2) (5) (0) (0) (98) (100)2000/01 1251 466 630 295 121 541 499(37) (50) (24) (10) (43) (40) Mean 781 160 213 135 93 515 509(16) (21) (14) (11) (73) (73)

Shallow soil1995/96 653 77 85 234 215 487 481(12) (13) (36) (33) (75) (74)1996/97 973 142 207 327 271 496 503(15) (21) (34) (28) (51) (52)1997/98 532 2 4 38 19 565 558(0) (1) (7) (4) (100) (100)1998/99 876 228 269 206 152 487 487(26) (31) (24) (17) (56) (56)1999/2000 401 16 40 0 0 400 388(4) (10) (0) (0) (100) (97)2000/01 1251 479 574 296 177 459 472(38) (46) (24) (14) (37) (38) Mean 781 157 196 183 139 482 481(16) (20) (21) (16) (70) (70)1. During the cropping season.2. BBF = Broad-bed and furrow.3. Values expressed as percentages of rainfall are given in parentheses.

58

Table 3. Simulated water balance components of soybean/pigeonpea intercropping system in varioustreatments on a Vertic Inceptisol, ICRISAT, Patancheru, India.

Surface runoff (mm) Deep drainage (mm) Crop water use (mm)Rainfall1


Medium-deep soil1996/97 966 234 268 184 198 545 534(24)3 (28) (19) (20) (56) (55)1997/98 531 1 0 0 0 544 538(0) (0) (0) (0) (100) (100)1998/99 1035 229 304 262 223 577 568(22) (29) (25) (21) (56) (55)1999/2000 436 1 1 0 0 460 465(0) (0) (0) (0) (100) (100)2000/01 1248 473 622 251 97 579 567(38) (50) (20) (8) (46) (45)

Mean 843 187 239 139 104 541 534(17) (21) (13) (10) (72) (71)

Shallow soil1996/97 966 137 207 360 299 521 514(14) (21) (37) (31) (54) (53)1997/98 531 0 0 23 16 543 536(0) (0) (4) (3) (100) (100)1998/99 1035 291 291 267 257 574 565(28) (28) (26) (25) (55) (55)1999/2000 436 0 0 0 0 468 460(0) (0) (0) (0) (100) (100)2000/01 1248 478 583 269 162 545 537(38) (47) (21) (13) (44) (43)

Mean 843 181 216 184 147 530 522(16) (19) (18) (14) (71) (70)1. During the cropping season.2. BBF = Broad-bed and furrow.3. Values expressed as percentages of rainfall are given in parentheses.

seasons depending upon weather and amount of totaldry matter produced by the system (Table 5). The Nuptake by the intercropping system approximatelyranged from 240 to 270 kg ha-1, whereas N fixationapproximately ranged from 170 to 250 kg ha-1. Therewas no significant difference between the twolandforms and the two soil depths for plant N uptake

and N fixation by the soybean/pigeonpeaintercropping system. Although denitrification andleaching of N were greater in the BBF treatment inboth the soil depths because of higher rainfallinfiltration and deep drainage, these constitutedinsignificant amounts of N losses from the soil to be ofany environmental concern (Table 5).

Table 4. Nitrogen (N) contributions and balance (kg ha-1) of soybean-based cropping systems in a VerticInceptisol watershed, ICRISAT, Patancheru, India, 1995–981.

Soybean-chickpea Soybean/pigeonpea

Description BBF Flat BBF Flat

Total N uptake 197 198 220 214

Total N loss (runoff + deep drainage) 13 17 14 17

N additions (rainfall, fallen leaves,roots, and BNF) 165 168 200 183

N additions (compost, Gliricidia loppings) 45 0 44 0

N balance 0 –47 +10 –491. BBF = Broad-bed and furrow; BNF = Biological nitrogen fixation.

59

Figu

re 6

. Rai

nfal

l dis

trib

utio

n an

d gr

ound

wat

er le

vel i

n th

e w

ell i

n B

W7

wat

ersh

ed.

60

Table 5. Simulated nitrogen (N) balance (kg ha-1) of soybean/pigeonpea intercropping system on a VerticInceptisol, ICRISAT, Patancheru, India.

Plant N uptake N fixation Denitrification Leaching

Year BBF1 Flat BBF Flat BBF Flat BBF Flat

Medium deep soil1996/97 259 263 189 200 0.95 0.56 3.03 0.0141997/98 242 239 169 168 0.60 0.72 0 01998/99 270 269 218 228 0.33 0.20 0.001 01999/2000 263 259 201 209 0.62 0.47 0 02000/01 271 269 253 251 0.024 0.023 0 0

Mean 261 260 206 211 0.50 0.39 0.61 0.002

Shallow soil1996/97 259 262 200 212 0.3 0.08 2.1 0.021997/98 242 239 174 175 0.14 0.10 0.001 01998/99 270 268 218 230 0.047 0.015 0.006 0.0021999/2000 271 269 222 220 0.072 0.077 0 02000/01 265 262 231 244 0.03 0.002 0.005 0 Mean 261 260 209 216 0.12 0.05 0.42 0.0041. BBF = Broad-bed and furrow.

Crop yields

The landform treatments did not significantly affectthe yields of the component crops of the two croppingsystems (Tables 6 and 7). Soil depth did not influencethe yield of rainy season crops in most years, except inlow rainfall years the yields were higher on medium-deep soil. Major effect of soil depth was on theproductivity of chickpea crop, which established andgrew on the receding soil moisture. Chickpea yieldswere generally higher on medium-deep soil than onshallow soil. Pigeonpea yields were not significantlyaffected by soil depth. Across years and landforms,sole soybean yields ranged from 0.9 to 2.4 t ha-1 onmedium-deep soil and 1.0 to 2.3 t ha-1 on shallow soil(Table 8). Chickpea yields ranged from 0.5 to 1.5 t ha-1

on medium-deep soil and 0.4 to 1.0 t ha-1 on shallowsoil. Total productivity of the soybean-chickpeasequential system ranged from 1.9 to 3.7 t ha-1 onmedium-deep soil and 1.5 to 3.3 t ha-1 on shallow soil.Average productivity of the soybean-chickpea systemon medium-deep soil was 2.7 t ha-1 and on shallow soilit was 2.3 t ha-1.

As expected, soybean yield in the intercroppingsystem was less than that observed in the sole system.Across seasons, soil types, and landforms the

intercropped soybean yield ranged from 0.7 to 2.0t ha-1 and was slightly higher on medium-deep soilthan on shallow soil (Table 7). Pigeonpea yield acrossseasons, soil types, and landforms ranged from 0.5 to1.4 t ha-1, giving an average yield of 0.9 t ha-1

irrespective of the soil type and landform. Totalproductivity of the soybean/pigeonpea intercroppingsystem ranged from 1.2 to 2.9 t ha-1 across soil typesand seasons, giving an average productivity of about2.1 to 2.3 t ha-1 on shallow and medium-deep soils,respectively (Table 7). In the drought year of 1999/2000 the total productivity of the soybean/pigeonpeasystem was greater than that of the soybean-chickpeasystem when only about 400 mm of seasonal rainfallwas received.

Long-term simulation of water balanceand crop yields

Long-term analysis using simulation models and cropweather data of 26 years (1974–2000) have shown thatin 70% of years total seasonal runoff ranged from 35to 269 mm for shallow soil and 70 to 320 mm formedium-deep soil (Table 8). Deep drainage beyondthe rooting zone ranged from 60 to 390 mm forshallow soil and 10 to 280 mm for medium-deep soil.

61

Table 6. Seed yield (t ha-1) of soybean-chickpea sequential system in various treatments on a VerticInceptisol, ICRISAT, Patancheru, India.

Soybean Chickpea Soybean and chickpea

Year BBF1 Flat SE2 BBF Flat SE BBF Flat SE

Medium-deep soil1995/96 1.7 1.9 0.06 0.5 0.6 0.02 2.2 2.5 0.051996/97 2.1 2.4 0.07 1.5 1.4 0.13 3.6 3.7 0.171997/98 1.0 0.9 0.06 1.5 1.1 0.12 2.5 2.0 0.111998/99 1.6 1.6 0.10 1.5 1.3 0.12 3.1 2.9 0.201999/2000 1.8 1.7 0.06 0.1 0.2 0.04 2.0 1.9 0.112000/01 2.4 2.1 0.12 0.9 0.7 0.08 3.3 2.8 0.18 Mean 1.8 1.8 - 1.0 0.9 - 2.8 2.6 -

Shallow soil1995/96 1.6 1.5 0.06 0.4 0.4 0.05 2.0 1.9 0.121996/97 2.3 2.3 0.07 1.0 1.0 0.13 3.3 3.3 0.171997/98 1.1 1.0 0.06 1.0 1.0 0.12 2.1 2.0 0.111998/99 1.5 1.7 0.10 1.0 0.8 0.12 2.5 2.6 0.201999/2000 1.7 1.5 0.06 0.1 0.1 0.04 1.8 1.5 0.112000/01 1.7 1.8 0.12 0.5 0.4 0.08 2.1 2.2 0.18 Mean 1.7 1.6 - 0.7 0.6 - 2.3 2.3 -1. BBF = Broad-bed and furrow.2. SE = Standard error (±).

Table 7. Seed yield (t ha-1) of soybean/pigeonpea intercropping system in various treatments on a VerticInceptisol, ICRISAT, Patancheru, India.

Soybean Chickpea Soybean and chickpea

Year BBF1 Flat SE2 BBF Flat SE BBF Flat SE

Medium-deep soil1996/97 1.5 1.8 0.07 0.9 1.1 0.16 2.4 2.9 0.171997/98 0.7 0.7 0.06 0.5 0.6 0.06 1.2 1.3 0.111998/99 1.1 1.1 0.10 1.4 1.2 0.12 2.4 2.4 0.201999/2000 1.4 1.5 0.06 0.8 0.8 0.08 2.2 2.3 0.112000/01 2.0 1.8 0.12 0.9 0.8 0.05 2.9 2.6 0.18

Mean 1.34 1.38 - 0.90 0.9 - 2.22 2.30 -

Shallow soil1996/97 1.5 1.7 0.07 0.9 1.0 0.16 2.4 2.7 0.171997/98 0.8 0.7 0.06 0.7 0.6 0.06 1.5 1.3 0.111998/99 1.0 0.9 0.10 1.4 1.1 0.12 2.4 2.1 0.201999/2000 1.5 1.3 0.06 0.7 0.6 0.08 2.2 1.9 0.112000/01 1.6 1.6 0.12 0.9 0.7 0.05 2.5 2.4 0.18

Mean 1.28 1.24 - 0.92 0.8 - 2.20 2.08 -1. BBF = Broad-bed and furrow.2. SE = Standard error (±).

62

Total productivity of the soybean-chickpeasequential system was 3 to 4.1 t ha-1 on shallow soiland 3.5 to 4.7 t ha-1 on medium-deep soil in 70% ofyears (Table 9). Total productivity of the soybean/pigeonpea intercropping system was 2.9 to 4.2 t ha-1

on shallow soil and 3.1 to 4.3 t ha-1 on medium-deepsoil. These results show the potential of theenvironment and technology for achieving higheryields provided the natural resources are managedproperly. Comparing the potential yields withobserved yields of the two cropping systems obtainedin the BW7 watershed, it becomes evident that in spiteof high yields obtained in the watershed a yield gap ofat least 1.0 t ha-1 still exists.

Potential productivity and yield gap ofsoybean growing locations

To assess the scope for increasing productivity ofsoybean in the major soybean-growing region ofIndia, potential productivity and yield gaps were

Table 8. Simulated surface runoff and deep drainage, using weather data of 26 years (1974 to 2000) forshallow and medium-deep Vertic Inceptisols at ICRISAT, Patancheru, India.

Landform Shallow soil Medium-deep soil

Runoff in 70% of years (mm)Flat 60–269 80–320Broad-bed and furrow (BBF) 35–190 70–280

Deep drainage in 70% of years (mm)Flat 60–330 10–245Broad-bed and furrow (BBF) 80–390 25–280

Table 9. Simulated yield potential (t ha-1) of soybean-chickpea sequential and soybean/pigeonpeaintercropping systems for shallow and medium-deep Vertic Inceptisols at ICRISAT, Patancheru, India1.

Crop Shallow soil Medium-deep soil

Soybean-chickpea systemSoybean 2.2–3.0 2.2–3.0Chickpea 0.5–1.5 0.8–1.9Soybean and chickpea 3.0–4.1 3.5–4.7

Soybean/pigeonpea systemSoybean 1.8–2.1 1.9–2.1Pigeonpea 1.0–2.3 1.2–2.3Soybean and pigeonpea 2.9–4.2 3.1–4.31. In 70% of years using weather data of 26 years (1974 to 2000).

assessed using the CROPGRO-soybean simulationmodel. Based on the yield gaps the locations or theregions could be targeted to bridge the yield gap. Thisanalysis was performed for 10 locations in India forwhich the soils and historical weather records wereavailable. These locations are Raisen, Betul, Guna,Bhopal, Indore, Kota, Wardha, Jabalpur, Amaravati,and Belgaum (Table 10). The potential yields (waterlimited) varied from year to year because of weathervariability. There were large differences in maximumand minimum obtainable yields for a location. Meanyield obtained for a location was compared with themean observed yield of the last five years to calculatethe yield gap. Simulated mean yield for Raisen andWardha was greater than 2500 kg ha-1, while for Betul,Jabalpur, Bhopal, and Indore it ranged from 2000 to2500 kg ha-1. For other locations the simulated meanyield ranged from 1200 to 2000 kg ha-1. The yield gapfor various locations ranged from 235 to 1955 kg ha-1.The yield gap was minimal for Kota where sufficientarea is under irrigation. For Raisen, Betul, Bhopal,

63

Indore, and Wardha, the mean yield gap ranged from1183 to 1955 kg ha-1. However, in some years greateryield gap is expected as indicated by the maximumobtainable yields. Based on the maximum obtainableyield, the yield gap ranged from 1812 to 2930 kg ha-1

for various locations. These results show that there is aconsiderable potential to bridge the yield gap betweenthe actual and potential yield through adoption ofimproved resource management technologies.

Extension of BW7 Watershed WorkIn BW7 watershed we have successfully shown thatintegrated watershed management technology hasresulted in crop intensification of the soybean-basedsystem as well as increased use of excess rainfallstored in surface ponds and dug wells. The BBFsystem decreases surface runoff and soil erosion thusreducing soil degradation. The target region ofMadhya Pradesh has similar agroecology as thePatancheru watershed area, except that rainfallintensity and amounts in July and August are greaterthan those observed at Patancheru. This is expected tocause severe soil erosion and waterlogging at Bhopaland Indore watershed sites. Therefore, the BBFsystem of land surface management is expected toperform better in controlling runoff and soil erosion,

Table 10. Simulated soybean yields and yield gap for the selected locations in India.

Mean MeanSimulated yields (kg ha-1)

Mean Yieldsowing harvest observed yield2 gap

Location date date Minimum Maximum Mean SD1 (kg ha-1) (kg ha-1)

Raisen 22 Jun 11 Oct 393 4670 2882 1269 - -Betul 19 Jun 8 Oct 924 3296 2141 603 858 1283Guna 30 Jun 14 Oct 342 2916 1633 907 840 793Bhopal 16 Jun 8 Oct 805 3064 2310 615 1000 1310Indore 22 Jun 10 Oct 760 4588 2273 939 1122 1151Kota 3 Jul 16 Oct 0 3188 1165 936 1014 151Wardha 17 Jun 6 Oct 1824 3955 3040 640 1042 1998Jabalpur 23 Jun 11 Oct 1132 2477 2079 382 896 1183Amaravati 18 Jun 8 Oct 440 2624 1552 713 942 610Belgaum 17 Jun 30 Sep 858 2943 1844 629 570 12741. SD = Standard deviation.2. Mean of reported yields of five years (1996/97 to 2000/01).

and to alleviate waterlogging of heavy clay soils.Groundwater recharging at these two sites could beimproved by constructing percolation tanks or gullyplugging.

Future Work PlanThere is a need to continuously monitor the long-termsoil, water, and nutrient management on sustainabilityand soil quality changes in the watershed, particularlysequestration of carbon. There is also a need to quantifythe economic losses due to land degradation andexamine how these losses can be minimized. As wateravailability in the watershed has been increased due tosurface ponds and wells, there is a need to quantify theoverall use efficiency including water use by thehorticultural system. The digital terrain modeldeveloped at the Michigan State University, USA willbe evaluated for this watershed. The work done at BW7watershed will be scaled-up for extending the benefitsof watershed management to rural communities. Thenational agricultural research system (NARS) scientistsworking in other watersheds will be trained in theinstrumentation for data recording. This watershed shallcontinue to serve as a training ground for governmentofficials working in the area of watershed managementand the farmers in the region.

64

ReferencesFAO (Food and Agriculture Organization of the UnitedNations). 2002. Website: http://www.fao.org

McCown, R.L., Hammer, G.L., Hargreaves, J.N.G.,Holzworth, D.P., and Freebairn, D.M. 1996. APSIM: a

novel software system for model development, modeltesting and simulation in agricultural systems research.Agricultural Systems 50:255–271.

Tsuji, G.Y., Jones, J.W., and Balas, S. (eds.) 1994. DSSATV3. Honolulu, Hawaii, USA: University of Hawaii.

65

Nutrient and Water Management Studies for Increasing Productivity ofSoybean-based Systems in Operational Scale Watersheds

A K Mishra, K P Raverkar, R S Chaudhari, A K Tripati, D D Reddy,K M Hathi, K G Mandal, S Ramana, and C L Acharya1

Abstract

Madhya Pradesh is well endowed with high moisture holding Vertisols and assured rainfall (800–1600mm yr-1) and is the heartland of dryland agriculture in India. The current productivity of soybean in Indiais 1 t ha-1 whereas the yield potential of soybean is 3 to 3.5 t ha-1. At the Indian Institute of Soil Science(IISS), Bhopal, Madhya Pradesh, an operational scale watershed was developed to study nutrient andwater management options for increasing productivity of soybean-based systems. Waterlogging duringrainy season significantly reduced soybean yields. The excess runoff water 300–400 mm could beharnessed and used as life saving irrigation for increasing productivity of soybean-wheat system. Duringnormal rainfall years broad-bed and furrow (BBF) landform treatment alleviated waterlogging andincreased productivity of soybean. The BBF landform stored more moisture in soil than the flat on grade(FOG) treatment; BBF also had low cone penetration resistance of soil than the FOG treatment. The BBFtreatment also recorded reduced runoff (10.6%) as against the runoff from FOG (18.6%). Application ofgreen manure through alley cropping of Gliricidia increased productivity of soybean-based systemsthrough increased water use and improved soil fertility and reduced mineral N needs.

1. Indian Institute of Soil Science (IISS), Nabi Bagh, Berasia Road, Bhopal 462 038, Madhya Pradesh, India.

Madhya Pradesh is the heartland of dryland agriculturein India which produces 4.4 million t of soybean on 4.4million ha. The soils are Vertisols and associated soilswith a good water-holding capacity, with annual rainfallvarying from 800 to 1600 mm and a potential for doublecropping on stored soil moisture or with supplementalirrigation. Most of the rainfall (85 to 90%) is receivedduring the monsoon (rainy) season with a rainfall peakin June to August resulting in waterlogging and sheeterosion in the region. In spite of spectacular growth insoybean area since 1970s soybean productivity is onlyabout 1 t ha-1.

Large areas are kept fallow during the rainy season.The main constraints for low soybean yields in theregion are inappropriate soil, water, and nutrientmanagement (SWNM) practices followed by thefarmers, availability of quality seeds of soybeanvarieties with suitable maturity duration, lack of creditfacilities, and large landholdings. To address theissues related to increasing productivity of soybean-based systems, strategic research on SWNM for

sustaining productivity of soybean-based systems wasundertaken at the on-station watershed of the IndianInstitute of Soil Science (IISS) in partnership with theInternational Crops Research Institute for the Semi-Arid Tropics (ICRISAT) through the AsianDevelopment Bank (ADB) supported project (RETA5812) during 1999–2002. IISS, Nabibagh, Bhopal islocated in the heartland of soybean production area(23°18’–23°20’ N, 77°24’–77°25’ E, and altitude 490m) of Madhya Pradesh. A 12-ha watershed wasdeveloped on Vertisol in the campus in 1999.

Weather and Soil CharacteristicsAnalysis of historical weather data (1980–99) revealedthat the average annual rainfall is 1130 mm and rangedfrom 694 mm in 1992 to 1521 mm in 1982. The rainfallis mainly received (89%) during June through October(Fig. 1). The annual evapotranspiration (ET) of Bhopalarea is about 1500 mm, of which ET for the periodfrom June through October is 650 mm. The actual ET

66

for rainy season crop is about 550 mm. Thus, on anaverage 300–400 mm rainwater is lost as surfacerunoff and deep drainage indicating good potential toharvest rainwater for securing and extending thegrowing season by providing supplemental irrigationand recharging the groundwater.

A reservoir of 2.5 ha-m capacity with 2 m depthwas developed in the watershed to capture the runoff

The pH increases slightly with soil depth and varies ina narrow range from 7.8 to 8.2 (slightly alkaline).Electrical conductivity (EC) is higher at 0–30 cm thandeeper soil depths (40–90 cm) beyond which thevalues increase to 0.15 dS m-1.

Weather data are collected using an automaticweather station. The wind speed is relatively high insummer and rainy months and gradually declines to a

from the experimental watershed. A part of watershedarea (1.5 t ha-1) was developed as broad-bed andfurrow (BBF) and flat on grade (FOG) landtreatments. These plots are equipped with H flumesand automatic runoff recorders and sediment samplersfor monitoring runoff and sediment. Gliricidia leafmanure farm was developed for providing Gliricidialeaves to experimental treatments. Gliricidia plantsare planted on the field bunds. General soil andclimatic characteristics are given in Table 1. The soilprofile is uniform with respect to soil moistureconstants (Fig. 2) except that at 60 cm depth,saturation and 0.1 bar moisture percentages are high.

minimum (1.0 m s-1) in November. In winter fromDecember to January, the northerly wind blows at lowaverage speed ranging from 1.0 to 2.0 m s-1. Whennortherly wind is replaced by southwesterly wind inJune, the speed increases steadily to a maximum ofabout 4 m s-1. The region being characterized as semi-arid tropics receives heavy showers in rainy season(Fig. 1). The winter months are generally calm and dryexcept for rare light showers. The spring season startswith a sharp rise in temperature from Februaryonwards to a maximum air temperature (about 40.5°C)in May. The soil surface temperature reaches a highrange of 33–35°C and shrinkage cracks develop on the

Figure 1. Mean monthly rainfall at Nabibagh, Bhopal in Madhya Pradesh, India.

67

soil surface in summer months. The air temperaturedata of 2000 and 2001 show that the variation inmaximum temperature is bimodal as characterized bytwo peaks, one in May and the other in October. Thesoil temperature rises from an average of 20°C inJanuary to a maximum of 33.9°C in April or May at 5cm soil depth. The difference in soil temperature at 5,10, and 20 cm depth is generally less than 1°C exceptin April and May. The variation in soil temperature

follows bimodal trend and conform to variation in airtemperature.

The semi-arid tropics of Madhya Pradesh isendowed with plenty of solar radiation round the yearwith net radiation ranging from 5.9 MJ m-2 in Januaryto 11.5 MJ m-2 during May–June. The solar radiationincreases from January and attains a peak levelsometimes during late April and May. After May, thesolar radiation starts declining and reaches a minimumin rainy season that is characterized by heavy clouds.After withdrawal of monsoon, the sky clears and bothsolar and net radiation rise slowly. Soil being black incolor absorbs considerable quantity of solar radiation.

The analysis of natural resource endowments in theregion demonstrated that this ecoregion bears potentiallyproductive environment for soybean-based croppingsystems, nevertheless yields of crops are poor. Safedrainage of excess rainwater from crop field and itsstorage in water harvesting pond and use of integratedSWNM approaches hold the key for enhancing andsustaining the soybean-based cropping system. Withthese in view, comprehensive field experiments wereplanned to address specific issues related to:• Intensification of soybean-based systems through

supplemental irrigation using harvested rainwater;• Integrated nutrient and water management options

for soybean-based systems; and• Assessment of effects of waterlogging on soybean

plant growth and soil processes.

Figure 2. Variation of soil moisture constraintsin soil profile.

Table 1. Details of weather and soilcharacteristics at IISS, Bhopal.

Particulars Value

Air temperatureAverage maximum daily 40.7°C (May)Average minimum daily 10.4°C (Jan)

Relative humidityMean maximum monthly 83% (Aug)Mean minimum monthly 25% (Apr)

Wind speedMean maximum monthly 13.2 km h-1 (Jul)Mean minimum monthly 4.3 km h-1 (Nov)

Pan evaporationAverage maximum daily 16.7 mm day-1

Average minimum daily 3.5 mm day-1

Soil textureSand 15.4%Silt 26.6%Clay 58.0%

Soil moistureLiquid limit 48.62%Plastic limit 23.94%Saturation 62.23%0.3 bar 28.40%15 bar 19.34%Hydraulic conductivity 0.06 m day-1

Infiltration rate 0.24 m day-1

Chemical propertiespH 7.74Electrical conductivity 0.11 dS m-1

Organic carbon 0.45%Calcium carbonate 2.83%Cation exchange capacity 49 Cmol kg-1

68

Nutrient and Supplemental IrrigationManagement in Soybean-WheatSystemAlleviation of waterlogging through safe drainage ofexcess rainwater from field and its storage in waterharvesting pond along with integrated SWNMapproach could enhance and sustain the productivity ofsoybean-wheat system. To validate this hypothesis,field experiments were conducted during rainy season2000 and winter season 2000/01 with the followingtreatments:

• Landform– BBF– FOG

• Supplemental irrigation treatment– Soybean: Rainfed– Wheat :

(1) Pre-sowing irrigation plus one irrigation atcrown root initiation (CRI) and one atflowering stage (three supplementalirrigations): I1

(2) Pre-sowing irrigation, one irrigation atCRI, maximum tillering, and floweringstage (four supplemental irrigations): I2

• Nutrient treatment in soybean (see treatments intabular form). Wheat is grown with 120 N:60P2O5:40 K2O (kg ha-1).

Soybean (rainy season 2000)

During the rainy season in 2000, 798 mm (30% belownormal) rainfall was received. Soybean yield washigher in BBF than in FOG landform (Table 2). Higheryield in BBF landform could be due to better soilconditions (aeration) and less cone penetrationresistance (CPR) in upper layers of the soil in BBFthan in FOG (Fig. 3). The less CPR in BBF wasbecause of the loose soil and higher moisture in upper

N P2O5

Treatment (kg ha-1) (kg ha-1)

N0P 0 60

NP0 30 0

NP 30 60

N50f+50iP 15 kg ha-1 through 60farmyard manure (FYM)and 15 kg ha-1 throughinorganic fertilizer

N50g+50iP 15 kg ha-1 through 60Gliricidia and 15 kg ha-1

through inorganic fertilizer

N50f+50gP 15 kg ha-1 through FYM 60and 15 kg ha-1 throughGliricidia

Table 2. Seed yield, biological yield, and harvest index of soybean in different nutrient and landtreatments1.

Seed yield (kg ha-1) Biological yield (kg ha-1) Harvest index

Treatment BBF FOG BBF FOG BBF FOG

N0P 1505 1440 4160 4185 0.3618 0.3438

NP0 1570 1470 4260 4250 0.3686 0.3447

NP 1590 1480 4380 4315 0.3625 0.3455

N50f+50iP 1590 1500 4390 4360 0.3636 0.3441

N50g+50iP 1600 1510 4430 4370 0.3621 0.3481

N50f+50gP 1620 1510 4460 4370 0.3641 0.3456Mean 1580 1485 4340 4310 0.3618 0.3438

NS NS NS NS NS NS1. See text for treatment details; NS = Not significant.

69

soil layers. Application of nitrogen (N) andphosphorus (P) had no effect on soybean yields, whichcould be due to moisture stress during pod-fillingstage. Runoff was 13% of the seasonal rainfall (798mm) and soil loss was 3.3 t ha-1 in BBF system.

Nodulation status

Number of nodules formed due to various integratednutrient treatments was not markedly different exceptin NP0 of BBF system and N0P and NP0 of FOGlandform. The development of nodules wassignificantly affected by the lack of sufficient amountof available P in the soil. The dry biomass of noduleswas 23% higher in BBF (109 mg plant-1) than in FOG(88 mg plant-1). The integrated nutrient treatmentssignificantly influenced the synthesis of dry biomassof nodules. Overall, the dry biomass of nodules wasgreater in the treatments with the application of (FYMand/or Gliricidia) as a component of fertilization.Nodule dry biomass was maximum in N50f+50gPtreatment; 174 mg plant-1 was observed in BBF and155 mg plant-1 in FOG. The replacement of inorganicN either fully or 50% by organic material (FYM and/or Gliricidia) increased nodule dry biomass over thatof NP treatment from 35 to 131% under different landtreatments.

Nitrogenase activity

The influence of integrated nutrient (inorganic andorganic) treatments on nitrogenase activity of nodules

was significant. Nitrogenase activity was lowest inNP0 (495 and 213 µmoles C2H4 h

-1 plant-1 in BBF andFOG, respectively). Replacement of inorganic N,either fully or 50% with organic material significantlyimproved the nitrogenase activity over the NPtreatment. The highest nitrogenase activity of 2081µmoles C2H4 h

-1 plant-1 in N50f+50gP in BBF system and1857 µmoles C2H4 h

-1 plant-1 in N50g+50iP in FOG systemwas registered.

Microbial biomass C and N

Microbial biomass carbon (C) and N in soybean wassignificantly influenced by integrated nutrient treatmentsin BBF and FOG landforms. In BBF system themicrobial biomass C and N was greater by approximately33% than in FOG system. The lowest microbial biomassC and N was registered in NP0 treatment where fertilizerP was not applied. Nitrogen supply through organicmaterial (FYM and/or Gliricidia) improved the synthesisof microbial biomass C and N. This is because appliedorganic material acted as substrate for microbes andresulted in intense microbial activity and accumulation ofnutrients.

Wheat (winter season 2000/01)

The data on yield, biological yield, and harvest indexof wheat are summarized in Table 3. The influence ofnutrient and supplemental irrigation treatments onyield was not significant because irrigation in thewatershed was not available after pre-sowingirrigation. The biological yield and harvest index alsowere not affected significantly. However, the grainyield, biological yield, and harvest index in general inBBF system were higher than in FOG landformbecause of the better moisture status in BBF than inFOG landform.

The moisture content at 0–15 and 15–30 cm depthswas high up to 80 days after sowing (DAS) beyondwhich the difference in moisture decreased, whereasthe moisture content at 45–60 cm was higher in BBFup to 100 DAS (Fig. 4). The 30–45 cm soil layershowed no pattern. The moisture content at 60–75 cmand 75–90 cm soil layers was higher in BBF than FOGafter 80 DAS (Fig. 4) and the moisture content keptdecreasing in FOG. The soil profile moisture storagechange (moisture use) was also higher, in general, inBBF than FOG (Fig. 5).

Figure 3. Variation of cone penetration resistance(CPR) in broad-bed and furrow (BBF) and

flat on grade (FOG) land treatments on20 September 2000 in Bhopal.

70

P uptake by soybean and wheat

All the nutrient input treatments supplying P alone ortogether with N through inorganic and/or organic sourcescaused a significant increase in P uptake by soybean andwheat over the treatment without P, i.e., NP0 in both BBFand FOG landforms (Table 4). The P uptake differencesamong the P supplying treatments were, however, notsignificant. The mean values of P uptake (across nutrienttreatments) by soybean as well as wheat were relativelyhigher in BBF than in FOG landforms.

Table 3. Grain yield, biomass yield, and harvest index of wheat in different landform, irrigation (I), andnutrient management (NM) treatments in 2000/011.

BBF FOG

Treatment I1 I2 Mean I1 I2 Mean

Grain yield (GY) (kg ha-1)N0P 1643 1628 1636 992 972 982NP0 1663 1642 1652 1021 985 1003NP 1879 1671 1775 1116 1096 1106N50f+50iP 1691 1768 1730 1052 1219 1135N50g+50iP 1609 1842 1726 1050 1011 1031N50f+50gP 1846 1830 1838 1040 1146 1093

Mean 1722 1786 1045 1071Biomass yield (BY) (kg ha-1)N0P 4285 4148 4217 2884 2530 2707NP0 4000 3704 3852 2716 2637 2677NP 4237 4074 4156 2749 2801 2775N50f+50iP 4089 4741 4415 2746 3241 2993N50g+50iP 3704 4370 4037 2831 2874 2852N50f+50gP 4181 4148 4165 2680 2944 2812

Mean 4083 4198 2768 2838Harvest index (HI)N0P 0.38 0.40 0.39 0.34 0.38 0.36NP0 0.42 0.43 0.42 0.38 0.38 0.38NP 0.44 0.40 0.42 0.41 0.39 0.40N50f+50iP 0.42 0.37 0.39 0.39 0.38 0.38N50g+50iP 0.44 0.42 0.43 0.37 0.36 0.36N50f+50gP 0.44 0.43 0.43 0.39 0.38 0.39

Mean 0.42 0.41 0.38 0.38CD (5%) I NM I × NM I NM I × NMGY NS2 NS NS NS NS NSBY NS NS NS NS NS NSHI NS NS NS NS NS NS

1. See text for treatment details. Post-sowing irrigation was not possible due to non-availability of irrigation water in watershed pond.2. NS = Not significant.

Phosphorus availability in soil at the end of onecycle of soybean-wheat rotation

The nutrient input treatments supplying P alone ortogether with N applied through inorganic and/ororganic sources improved significantly the Pavailability in soil over the treatment without P (NP0)in BBF as well as FOG systems. Relatively greater Pavailability in the soil was recorded in the plots treatedwith inorganic plus organic (FYM, Gliricidia) sourcesof nutrients. The P availability was highest in the plots

71

Figure 4. Seasonal variation of moisture content at various soil depths in two landforms.

72

receiving N50i+50fP in BBF (19.7 mg kg-1) as well as inFOG (17.1 mg kg-1) systems. Irrespective of thenutrient treatment, BBF maintained relatively greaterP availability than FOG system, with the average soiltest P values being 16.1 and 13.8 mg kg-1 soil,respectively.

N uptake by soybean, soil organic carbon, andavailable N in soil

The BBF treatment recorded relatively higher Nuptake compared to FOG treatment (Table 5).However, the available N in the soil after harvest ofsoybean was not significantly different in FOG andBBF. Further, the contents of organic C in BBF andFOG treatments were similar. Overall, it was found

Figure 5. Seasonal variation in soil profileFigure 5. Seasonal variation in soil profileFigure 5. Seasonal variation in soil profileFigure 5. Seasonal variation in soil profile

0

1

2

3

4

5

21 40 55 70 84 101 137

Days after sowing

Moi

stur

est

orag

ech

ange

(cm

)

BBF FOG

Figure 5. Seasonal variation in soil profile moisture storage change (0–90 cm depth).

Table 4. Phosphorus uptake (kg P ha-1) bysoybean and wheat in different land and nutrienttreatments in 2000/011.

Soybean Wheat

Treatment BBF FOG BBF FOG

N0P 15.5 15.3 8.4 5.2NP0 13.3 12.7 6.8 4.2NP 15.2 15.1 8.4 5.1N50i+50fP 16.2 15.4 7.7 5.0N50i+50gP 16.2 15.5 7.2 5.0N50f+50gP 15.5 15.8 8.1 5.0 Mean 15.4 15.0 7.8 4.9 CD (5%) 1.41 1.21 1.14 0.731. See text for treatment details.

that both the nutrient management and land treatmentsdid not show any significant effect on soil organic Cand available N status.

The BBF treatment increased the total N uptake bywheat crop compared to FOG. On the other hand, areverse trend was observed in the case of available Nin the soil after harvest of wheat. However, the organiccontent was not affected by either the land treatment ornutrient management practice. As in soybean, thedifferences in the N uptake and available N contentwere not significant in wheat.

Microbial biomass C and N

Under FOG, microbial biomass C and N in wheatrhizosphere in N50f+50iP, N50g+50iP, and N50f+50gPtreatments did not differ significantly from that of NPtreatment. However, under BBF microbial biomass Cwas significantly greater in N50f+50iP and N50f+50gP ascompared to NP treatment. Microbial biomass C andN was higher by 15% in BBF than FOG. This may bedue to the better soil aeration and tilth which providedcongenial soil conditions for development and growthof microorganisms.


Runoff and soil loss

The total amount of runoff water from the FOG plot(131.6 mm) during the rainy season was higher thanthat of BBF plot (74.8 mm). Of the total rainfall of 708mm, 10.6% and 18.6% were lost through runoff fromthe BBF and FOG plots respectively. Seasonal soilloss from the FOG plot was 698 kg ha-1.

73

Bulk density

Bulk density of the surface soil (0–7.5 cm) was less inBBF as compared to the FOG treatment. Thisdifference in bulk density was not tangible in the7.5–15 cm layer. Likewise average bulk density(0–7.5 cm) of the treatments with organic amendmentslike FYM or Gliricidia was lower than inorganicallyfertilized plots.

Soil moisture stock

Soil profile moisture stock for both BBF and FOGland treatments increased (Fig. 6) up to 30 days aftersowing of soybean because of heavy monsoon rainfall(246 mm) in that period (Fig. 7). After that moisturestock decreased up to 45 DAS. Another profilerecharge period was observed between 45 and 60DAS. After 60 DAS moisture content of the profiledecreased continuously due to absence of anysignificant rainfall event in that period (Fig. 7). Thecrop suffered from moisture stress due to earlywithdrawal of monsoon during the pod-filling stage.BBF plots retained more moisture than the FOG plotsup to 48 DAS and in the latter period this difference wasnot clear (Fig. 6). More moisture content in the BBFplots may be due to less loss of rainfall through runofffrom BBF plots (10.6%) than from FOG plots (18.6%).

Nodulation status

Landform had a marked effect on nodulation status. InBBF, number of nodules and its dry biomass was

Table 5. Total N uptake by soybean, soil organic carbon, and available N in soil in different nutrientmanagement and land treatments1.

N uptake Soil organic C Available N after(kg ha-1) after soybean (%) soybean (kg ha-1)

Treatment BBF FOG BBF FOG BBF FOG

N0P 128.3 125.4 0.43 0.44 211.8 217.3NP0 139.3 133.5 0.44 0.44 217.3 221.7NP 143.5 136.3 0.46 0.47 234.7 236.7N50f+50fP 142.8 139.1 0.45 0.45 221.3 224.4N50g+50gP 143.8 139.3 0.47 0.48 219.7 221.7N50f+50gP 145.7 139.8 0.46 0.43 231.3 225.7 Mean 140.6 135.6 0.45 0.45 222.7 224.6

CD (5%) NS NS NS NS NS NS1. See text for treatment details.

Figure 6. Temporal variation of soil profilemoisture stock (0–90 cm) as affected by

land treatment during kharif 2001.

Figure 7. Rainfall during crop growth period inkharif 2001.

greater by 42% and 56% respectively, as compared toFOG. Nodule formation in the presence of organicsdid not vary significantly, except in N50f+50iP in FOG.However, integrated nutrient management treatments,viz., N50g+50iP and N50f+50gP of FOG and N50f+50gP ofBBF significantly improved the synthesis of dry

74

biomass in nodules over the NP treatment ofrespective landforms. The synthesis of dry biomass innodules was significantly reduced in the absence of P.

Seed yield

The variation in seed yield, total biomass, and numberof pods per plant of soybean due to nutrientmanagement was not significant in the 2001 rainyseason (Table 6). In both land treatments, the cropgrew well because of good distribution of rainfall upto 60 DAS. The low yield in general was due towithdrawal of rain during and beyond pod-fillingstage. Though the biomass production was adequate,the number of pods per plant remained low and variedbetween 29 and 38 in both land treatments. Yield was

Table 6. Seed yield, total biomass yield, andpodding of soybean as influenced by landformand nutrient management in 20011.

Treatment BBF FOG

Seed yield (kg ha-1)N0P 1088 1538NP0 1019 1582NP 1058 1644N50f+N50i P 1092 1524N50g+N50i P 1160 1525N50f+N50gP 1047 1644

CD (P = 0.05) NS2 NSTotal biomass yield (kg ha-1)N0P 3298 4702NP0 3291 4424NP 3397 4814N50f+N50iP 3132 4985N50g+N50iP 3509 4524N50f+N50gP 3478 4472

CD (P = 0.05) NS NSNumber of pods plant-1

N0P 31.1 32.2NP0 28.6 28.3NP 31.3 29.1N50f+N50i P 32.7 34.2N50g+N50i P 37.6 28.8N50f+N50gP 37.2 37.7

CD (P = 0.05) NS NS1. See text for treatment details.2. NS = Not significant.

less in BBF than FOG. This may be attributed to thefact that expected advantage in BBF in terms of highmoisture storage and avoidance of water congestioncould not be realized because of well distribution ofrain up to 60 DAS. Less plant population in BBFcompared to FOG could not compensate in terms ofseed yield per unit area as expected in BBF becausethe crop experienced long period of water deficitcondition during and beyond pod formation stage.

Irrigation and Nutrient Managementin Soybean-based Cropping SystemsField experiments consisting of the following treatmentswith three replications were conducted during the winterseason 1999/2000 and rainy season 2000:• Production system:

– Soybean-chickpea– Soybean-linseed– Soybean-wheat– Soybean + Gliricidia-wheat + Gliricidia

• Irrigation:– Soybean: Rainfed– Winter season crops: (i) Pre-sowing; (ii) Pre-

sowing + one irrigation at the most critical stageof crop

• Nutrient management:– Soybean: FYM at 4 t ha-1 + recommended dose

of NPK– Winter season crops: (i) No fertilizer; (ii) 50%

of recommended NPK; (iii) 100% ofrecommended dose of NPK:Soybean 30 N:60 P2O5:30 K2O kg ha-1

(JS 335)Wheat 120 N:60 P2O5:40 K2O kg ha-1

(Sujata)Chickpea 20 N:40 P2O5:20 K2O kg ha-1

(Ujjain-21)Linseed 60 N:30 P2O5:30 K2O kg ha-1

(RS 52)


The residual effect of fertilizer treatments imposed inthe previous season crops (winter season 1999/2000)on the following soybean crop (rainy season 2000)was studied. The seed yield, biological yield, andharvest index show that the crop performance was not

75

significantly different due to the previous treatmentimposed in winter season crops because soybean wasgrown under optimum nutrient conditions (Table 7).

The main effect of nutrients and interaction effectof cropping systems and nutrients on nodule numberof soybean were not significant. However, the cropgrown during the previous season significantlyinfluenced the nodule number. Nodule number and drybiomass in soybean grown after linseed weresignificantly less when compared to other croppingsystems. Highest number of nodules (140 plant-1) anddry biomass (272 mg plant-1) were registered insoybean-wheat (Gliricidia alley) system receiving50% NPK during previous season. Also, nitrogenaseactivity was highest (893 µmoles C2H4 h

-1 plant-1) inthis treatment. Nitrogenase activity was reduced insoybean plants grown after linseed as well as chickpeaas compared to wheat and wheat-Gliricidia alley.

Cropping systems (winter season 2000/01)

Yield data showed increasing trends in all systemswith increasing nutrient doses and irrigation levels(Table 8). As the crop yields of different crops cannotbe compared, the yields were transformed to wheatequivalent yield (WEY). The interaction betweencropping system and irrigation was significant. Themaximum WEY was in soybean-chickpea system withtwo irrigation levels followed by one irrigation. Thedata also indicated that even with one irrigation theWEY of soybean-chickpea system was significantlyhigher than other systems.

The interaction between cropping system and NPKwas also significant. The maximum WEY was insoybean-chickpea system (4181 kg ha-1) at 100% NPKwhich was significantly higher than WEY (3658 kgha-1) at 50% NPK. The WEY in soybean-chickpea

Table 7. Residual effect of fertilizer treatments in previous cropping systems on performance of soybean,rainy season 20001.

Treatment in Soybean-wheatwinter season Soybean-wheat Soybean-linseed Soybean-chickpea (Gliricidia alley)

Seed yield (kg ha-1)0% NPK 1510 1463 1428 149350% NPK 1487 1468 1501 1473100% NPK 1514 1474 1515 1519

Mean 1504 1468 1481 1495CD (5%) CS NM CS × NM

NS NS NSBiomass yield (kg ha-1)0% NPK 4465 4350 4423 447550% NPK 4431 4385 4397 4454100% NPK 4389 4395 4444 4452

Mean 4428 4377 4421 4460CD (5%) CS NM CS × NM

NS NS NSHarvest index0% NPK 0.338 0.336 0.323 0.33450% NPK 0.336 0.335 0.342 0.331100% NPK 0.346 0.336 0.341 0.342

Mean 0.340 0.336 0.335 0.336CD (5%) CS NM CS × NM

NS NS NS1. CS = Cropping system; NM = Nutrient management; NS = Not significant.

76

system even at 50% NPK was significantly higher thansoybean-wheat and soybean-wheat (Gliricidia alley)system.

Higher irrigation and nutrient levels favored higherbiomass C and N in all systems. Microbial biomass Cand N was maximum in soybean-wheat (Gliricidiaalley) and soybean-wheat systems followed bysoybean-linseed and soybean-chickpea. Microbialbiomass C and N in soybean-linseed was noticeablyhigh at two irrigation levels.

The P uptake by wheat, chickpea, linseed, andwheat (Gliricidia alley) was relatively higher with twoirrigations than with single irrigation. The irrigationeffect on P uptake was, however, significant only inwheat. In contrast, the fertilizer rate showed asignificant effect on P uptake in all the crops. Withapplication of 100% recommended fertilizer dose, themean P uptake increased by 5.53 kg ha-1 in wheat, 8.51kg ha-1 in chickpea, 2.35 kg ha-1 in linseed, and 5.84 kgha-1 over the P uptake in the respective control (nofertilizer). The relative magnitude of P uptake indifferent crops was in the order: chickpea > wheat =wheat (Gliricidia alley) > linseed.

In all the soybean-based cropping systems, theeffect of irrigation (to rabi crops) levels on available Pstatus of soil was not significant. Nevertheless, theavailable P in soil under different cropping systems

was slightly lower with two irrigations than with singleirrigation, possibly due to relatively greater P removalby the crops irrigated twice. The available P in soilincreased significantly with an increase in the rate offertilizer application in all the cropping systems.Irrespective of the fertilizer rate, the P availability wasrelatively greater in soybean-chickpea and soybean/Gliricidia-wheat/Gliricidia systems than in soybean-wheat and soybean-linseed systems. The greater Pavailability in soil in soybean-chickpea system mayperhaps be due to higher mobilization of native soil Pas the root exudates of chickpea are known to solubilizethe Ca-P, a predominant P fraction in Vertisols. Insoybean/Gliricidia-wheat/Gliricidia system, decom-position of large amount of Gliricidia leaf fall andloppings may have helped in mobilization of native soilP, thus resulting in increased P availability. Nitrogenuptake increased significantly with increase in the levelof N and irrigation. In wheat, plot with two irrigationsrecorded almost two-fold N uptake compared to thatwith one irrigation. On the other hand, irrigation did notshow any positive effect on N uptake by linseed.

With the increase in the level of N both theavailable N and the organic C content in the soilincreased after harvest of all the four crops. However,there was no significant effect of irrigation on organicC and available N in the soil.

Table 8. Seed yield (kg ha-1) of rabi crops as influenced by cropping systems, irrigation (I), and nutrientsduring 2000/011.

I1 I2

Rabi crop 0% 50% 100% 0% 50% 100%(cropping system) NPK NPK NPK Mean NPK NPK NPK Mean

Wheat 1104 1409 2106 1540 1975 2533 3363 2624(Soybean-wheat +Gliricidia alley)

Wheat 1172 1369 2172 1571 2069 2633 3325 2676(Soybean-wheat)

Chickpea 1338 1829 2012 1726 1538 2223 2619 2126(Soybean-chickpea)

Linseed 725 926 1031 894 923 1083 1136 1047(Soybean-linseed)1. I1 = one pre-sowing irrigation; I2 = I1 + one post-sowing irrigation at maximum tillering of wheat, and flowering of chickpea and linseed.

77


The residual effects of various treatments imposedduring the previous season on performance of soybeangrown with recommended dose of fertilizer werestudied. In soybean-Gliricidia alley system, loppingsof Gliricidia were added at the time of soybeanflowering to these treatments.

Nodule formation and dry biomass in soybeangrown after linseed and chickpea were significantlyreduced. In soybean-wheat (Gliricidia alley) croppingsystem, the nodule dry biomass in soybean (0.28 gplant-1) reduced significantly in comparison to solesoybean (0.38 g plant-1) after wheat. Residual effect ofthe previous crop did not significantly influencebiomass, pod, and seed yields. The low yield ingeneral was due to withdrawal of rain during andbeyond pod-filling stage.

Impact of Waterlogging on Growthand Yield of SoybeanTemporary waterlogging is a common feature ofVertisols during the rainy season. Annual rainfall onthese soils ranges from 750 to 1500 mm in MadhyaPradesh. About 80% of the rainfall received duringfour months (June–September). Rainfall during thesemonths exceed potential evapotranspiration (PET),causing excess water situation leading to temporarywaterlogging and anaerobic conditions in these soils.Soybean is grown in diverse environments fromtropical to cool temperature zones, and in both rainfedand irrigated conditions. In Madhya Pradesh, soybeanis grown as a major kharif (rainfed) crop on Vertisols.Approximately 75% of the total production of soybeanin India is from Madhya Pradesh. Temporarywaterlogging because of poor drainage is known toreduce soybean plant growth and N uptake and yield.To quantify the effects of waterlogging on soil, plantprocesses, and yield of soybean during different cropgrowth stages, a field experiment was conductedduring rainy season in 2000 including treatments withdifferent days of waterlogging at vegetative andreproductive stages of soybean crop.

Nodulation (number of nodules and dry biomass)was not affected significantly by waterloggingtreatments. Shoot and root dry biomass productiondecreased significantly due to waterlogging for

various periods. The impact was severe in cyclicwaterlogging treatment, which produced lowest shootand root dry biomass. Six days of continuouswaterlogging during vegetative stage significantlyreduced the shoot/root ratio and indicated enhancedroot growth as compared to shoot growth.

Two days of waterlogging during vegetative stageand 4 days of waterlogging both during vegetative andreproductive stages increased the seed yieldsignificantly over control. The waterloggingtreatments, namely, 3 cycles of waterlogging duringthe vegetative/reproductive stage, and 6 days ofwaterlogging both during the vegetative andreproductive stages, reduced seed yield, although thedecrease was not significant.

Based on the results of rainy season 2000, thetreatments were modified and a field experiment wasconducted during the rainy season of 2001. Tentreatments were laid out in a randomized block designwith three replications (Table 9). Observations wererecorded next day after withdrawal of waterloggingduring different stages of crop growth.

Table 9. Details of waterlogging treatmentsduring rainy season 2001.

Period of waterlogging (days)during growing season of soybean JS 335

7 days after Vegetative Reproductiveemergence (V1) stage (V2) stage (R)

4V1 4V2 4R8V1 8V2 8R

12V1 12V2 12RControl1

1. No waterlogging during crop growing season.

Soybean yield was significantly influenced bywaterlogging (Table 10). The number of pods rangedbetween 27 and 45 plant-1 and in 4V2 treatment, thedecrease in number of pods over control wassignificant. Highest number of pods was recorded in4R treatment. Straw yield was significantly reduced inmost treatments as compared to control. Seed yieldwas significantly reduced in all treatments except in8V1 and 4R treatments.

78

Acknowledgment

We gratefully acknowledge the guidance and technicalhelp provided by Dr S P Wani, Dr Piara Singh, and

Table 10. Effect of waterlogging on soybean yield during rainy season 20011.

Number of Seed yield Straw yield 100-seedTreatment pods plant-1 (g m-2) (g m-2) mass (g)

4V1 39 117.1 382 6.938V1 31 169.7 397 7.57

12V1 28 155.8 437 6.374V2 27 127.5 342 7.488V2 35 90.9 297 6.22

12V2 32 75.1 219 6.294R 45 196.1 379 6.928R 39 156.5 287 7.29

12R 38 88.9 212 6.39Control 37 184.5 401 7.74

CD (5%) 9 15.4 33 NS2

1. See text for treatment details.2. NS = Not significant.

Mr P Pathak from ICRISAT. We also thank ICRISATfor providing financial support through the AsianDevelopment Bank supported project (RETA 5812)for undertaking these studies.

79

1. International Crops Research Institute for the Semi-Arid Tropics (ICRISAT), Patancheru 502 324, Andhra Pradesh, India.2. Andhra Pradesh Rural Livelihoods Programme (APRLP), Rajendranagar, Hyderabad 500 030, Andhra Pradesh, India.3. District Water Management Agency (DWMA), Nampally, Hyderabad, Andhra Pradesh, India.

Minimizing Land Degradation and Sustaining Productivity by IntegratedWatershed Management: Adarsha Watershed, Kothapally, India

S P Wani1, T K Sreedevi2, P Pathak1, T J Rego1, G V Ranga Rao1,L S Jangawad1, G Pardhasaradhi1, and Shailaja R Iyer3

Abstract

Land degradation is a major threat for sustainable crop production in large areas of the semi-arid tropics(SAT). The Asian Development Bank gave financial assistance to evaluate the on-station watershed workof ICRISAT in on-farm situations. Adarsha watershed at Kothapally in Andhra Pradesh (AP), India is oneof the benchmark sites where the evaluation was carried out. Instead of traditional structure drivenapproach, a new idea of integrated watershed approach was followed wherein various components ofimproved crop production were evaluated on a few selected individual farmers’ fields in addition tocommunity-based soil and water conservation activities. A new implementation arrangement calledconsortium approach wherein all the stakeholders, ICRISAT, DWMA, CRIDA, NGO, and farmers, plannedand implemented various activities in a participatory manner was tried. This work has attracted not onlythe attention of AP Government but also many development agencies like DFID throughout the world. TheAP Government is scaling-up this work in five districts through APRLP. It is one of the successful modulesof watershed development. A grant by Sir Dorabji Tata Trust has been approved to replicate this work inCentral and Northwest India. The success story of this work with details of various activities and theoutputs is given in this paper.

Land degradation is a serious problem throughout theworld, threatening economic and physical survival ofmankind. Key issues on land degradation includeescalating soil erosion, declining soil fertility,salinization, soil compaction, agrochemical pollution,and desertification. The result is a decline in theproductive capacity of land. Existing estimates of thecurrent global severity of the problem (Scherr andYadav 1996) indicate that except for forest andwoodland, the proportion of the land that is degradedis estimated to be more extensive in Africa and Asia.Oldeman (1994) assessed that globally, about 15% ofthe land is severely degraded. Water erosion wasestimated at 56%, wind erosion at 28%, chemicaldegradation at 12%, and physical degradation at 4%.Asia’s degradation is specifically attributed todeforestation with overgrazing and agriculturalactivities contributing as major factors. There is about

17% cumulative productivity loss between 1945 and1990 as a result of land degradation (Crosson 1994).Lal (1995) estimated that the average yield reductiondue to soil erosion is about 6%, ranging from 2 to40%. The International Water Management Institute(IWMI) estimates show that 25% of the world’spopulation and 33% of the developing countrypopulation live in regions that will experience severewater scarcity by 2025. One billion of the world’spoorest people living in the semi-arid tropics (SAT)(Ryan and Spencer 2001) will be affected by waterscarcity (Seckler et al. 1998). The poverty of Asia’spoor is both a cause and a consequence of acceleratingsoil degradation and declining agriculturalproductivity. Poverty reduction is thus the majorchallenge for those responsible for policy and decisionmaking on the protection and sustainable use of landresources in Asia.

80

Poverty and Land DegradationWhenever adverse changes occur in the world, it isusually the poor who suffer most. This situation arisesfrom the very definition of the poor – those who lackadequate access to the basic necessities of life and theresources needed to obtain them. Because of landshortage, accentuated by degradation, the options forpoor will be limited. Production will begin to fall andthere will be an immediate attempt by the farmers inincreasing the inputs to the crop and this non-sustainable management will lead to furtherdegradation. So, it is poverty along with increasedpopulation that plays the greatest part in the casualnexus of land degradation and food insecurity in thedeveloping world.

Erosion: On-site and Off-site ImpactsErosion is the most important factor that degradessoils globally. It is a process where wind and waterfacilitate the movement of topsoil from one place toanother. Soil erosion has been occurring for some 450million years, but the problem has been acceleratedmore recently. As discussed above, this is a result ofmankind’s actions, such as over-grazing or unsuitablecultivation practices which make the land vulnerableduring times of erosive rainfall or windstorms. Soilerosion occurs both incrementally, as a result of manysmall rainfall events, and more dramatically as a resultof large but relatively rare storms. The most seriouson-site impact due to erosion is decreased agriculturalproductivity as seen in several developing countries inAsia.

For sustainable management of natural resourcessuch as water, soil, vegetation, and biota, watershed is alogical unit. Integrated watershed managementapproach covers wide-ranging aspects like health of theland (such as farming systems), agroforestry,infrastructure development, soil and waterconservation, and community participation. Integratedwatershed management is defined as an integration oftechnologies within the natural boundaries of a drainagearea for optimum development of land, water, and plantresources to meet the basic needs of the people in asustainable manner. Watershed management solutionsmust address the problem of rural poverty, protect the

natural resources, and rehabilitate degraded areas,particularly those that pose hazards to human life andwelfare. The approach improves the overall conditionof land resources and also the living conditions of thepeople involved.

New Integrated WatershedManagement Consortium ModelA new consortium model for efficient management ofnatural resources in the SAT has emerged from thelessons learned from long-term watershed-basedresearch of the International Crops Research Institutefor the Semi-Arid Tropics (ICRISAT) and nationalagricultural research system (NARS) partners (Waniet al. 2001). The important components of the newintegrated watershed management model are:• Farmer participatory approach through coopera-

tion model and not through contractual model.• Use of new science tools for management and

monitoring of watersheds.• Link on-station and on-farm watersheds.• A holistic system’s approach to improve

livelihoods of people and not merely conservationof soil and water.

• A consortium of institutions for technicalbackstopping of the on-farm watersheds.


• Minimize free supply of inputs for undertakingtechnology evaluation by the farmers.

• Low-cost soil and water conservation measuresand structures.

• Amalgamation of traditional knowledge and newknowledge for efficient management of naturalresources.

• Individual farmer-based conservation measuresfor increasing productivity of individual farmsalong with community-based soil and waterconservation measures.


• Empowerment of community individuals andstrengthening of village institutions for managingnatural watersheds.

81

About the ProjectThe project “Improving Management of NaturalResources for Sustainable Rainfed Agriculture” isfunded by the Asian Development Bank (ADB) and wasestablished in 1999 in an effort to improve the naturalresource base and to have sustained increase in foodproduction by SAT farmers. The present project involveswatershed research in three countries (India, Thailand,and Vietnam), both on-station and on-farm. The Adarshawatershed at Kothapally in Ranga Reddy district ofAndhra Pradesh is one of the three on-farm benchmarkwatersheds in India. The details of the project activitiesand results of the Adarsha watershed are described.

Process of SelectionICRISAT and the District Water ManagementAagency (DWMA) [earlier Drought Prone AreaProgramme (DPAP)], Government of Andhra Pradeshas well as M Venkatarangaiya Foundation (MVF), anon-governmental organization (NGO), togethersurveyed three watersheds in Andhra Pradesh andselected Adarsha watershed as one of the on-farmbenchmark sites for the ADB-assisted project. In thiswatershed the total irrigable area was less and therewas more dryland (80%). Not a single waterharvesting structure for human and animal use existedat the time of survey in 1998, i.e., at the start of thisproject. A large area is under rainfed farming in thevillage. As there were no interventions made toconserve soil and water, this watershed was selected toencompass the concept of convergence in thewatershed through consortium approach of managingand developing watersheds (Wani et al. 2001).Adarsha watershed was selected after a meeting ofvillagers in “Gram Sabha”, where the villagers cameforward to participate in the proposed watershedactivities. The objective was to improve rainfedagricultural production through integrated watersheddevelopment and reduce poverty of the farmersthrough increased systems productivity on sustainablebasis while minimizing land degradation.

Consortium Partners• ICRISAT• Central Research Institute for Dryland Agriculture

(CRIDA)

• DWMA, Government of Andhra Pradesh• MVF• National Remote Sensing Agency (NRSA)• Farmers (Watershed Association, Watershed

Committee, and self-help groups)

Developmental ActorsDifferent committees and groups were formed in thevillage and leaders were selected by the villagersthemselves. The leaders were involved in the planningof watershed development activities from the initialstage (e.g., selection of the water harvesting sites),implementation of the activities, execution andassessment of all the developmental activities withinthe watershed. The various committees formed in thewatershed are:• Watershed Committee: The committee consists of

a president, secretary, and 8 members representingdifferent sections of the community.

• Watershed Association: The working committeeconsists of a chairman, a secretary, 8 committeemembers, and 270 members; i.e., farmers in thevillage.

• Women self-help groups – Vermicomposting: Tengroups were formed with 15 members each. Thesegroups took up vermicomposting as an enterprisein the village.

• User groups: For water harvesting structures.• Self-help groups: To undertake watershed

development activities.

Approach

• Convergence of various activities in the watershed.• No private contractors were involved in the

watershed development activities.• Inputs for technology evaluation were not free but

were supplied at a minimum subsidy.• Farmers conducted on-farm trials with technical

support from ICRISAT and other researchinstitutes in the consortium.

• Empowerment of farmers was through training andworkshops.

• Availability of inputs and necessary machinerywas ensured.

• The NGO’s strength for social mobilization washarnessed.

82

• Monetary disbursements were by watershedcommittees and not through the NGO/projectimplementing agency.

• Social auditing was done by the villagers.

The Initial Situation – Baseline SurveyAt the outset of the project, a baseline data survey wascarried out, which provided the necessary informationon the existing resource-base and conditions of thevillage for monitoring and evaluation later.

Location

Adarsha watershed is located at longitude 78°5’ to78°8’ E and latitude 17°21’ to 17°24’ N falling inSurvey of India toposheet No. 56 K13 in the village ofKothapally, Shankarpally Mandal in Ranga Reddydistrict of Andhra Pradesh (Fig. 1). The total area ofthe watershed is 465 ha of which 430 ha is cultivatedland.

Physiography

Vegetation

Main rainy season crops grown are sorghum, maize,cotton, sunflower, mung bean, and pigeonpea. In thepostrainy season sorghum, sunflower, vegetables, andchickpea are grown. Wheat and rice are alsocultivated.

Climate

The annual rainfall in Kothapally is about 800 mmreceived mainly during June to October (85%). About25–30% of the rainfall is lost as runoff carrying awaythe fertile topsoil.

Soils

The landscape of the watershed is made up of Vertisolsand associated Vertic soils (90% of the area); Alfisols(10% of the area) are also present. Soil depth asperceived by the farmers and verified by the scientiststhrough random samplings in the watershed is about30–90 cm.

Social structure

The village consists of 274 households with the meanfamily size being seven. The total population is 1492,of which 54% belongs to backward communities, 15%to minorities, 20% to scheduled castes, and 9% toother castes. Beteille (1974) states that literacy andeducation may be unevenly distributed in an agrariansociety and the data in Kothapally supports thisstatement with regard to inequalities between sexesand between castes. In Adarsha watershed, 40% of theland belongs to small holding farmers (0.01 to 1.00ha), 40% to medium holding farmers (1.00 to 2.00 ha),and about 20% of the area to large holding farmers(>2.00 ha).

Groundwater table

The average depth of the 56 wells surveyed was 7.35m (range 2–18.65 m). The variation in thegroundwater table level and the amount of waterharvested is based on the cropping patterns and otherfactors such as soil type, crops grown, topography,runoff, and geological factors of the area.

Crop productivities

The productivity of rice ranged between 0.27 and 2.4 tha-1 for small landholders while for large landholdersit was much less and varied from 0.19 to 0.9 t ha-1. Theaverage productivity in small, medium, and largelandholdings was 1.1, 1.2, and 0.6 t ha-1 respectively.The same trend was observed for pulses also. The crop

Figure 1. Location of Kothapally village inShankarpally Mandal, Ranga Reddy district,

Andhra Pradesh, India.

83

productivities of cotton were 0.9, 0.6, and 0.3 t ha-1 forsmall, medium, and large landholders (Table 1).

Landholding size and use of inputs

Diammonium phosphate and urea

The majority of farmers use fertilizers. The amount ofdiammonium phosphate (DAP) and urea used declinessharply as land size increases.

Potash and super phosphate

These fertilizers are only applied to paddy by farmersin Kothapally. The amount of potash and superphosphate applied declines with increasing land size.In Kothapally watershed in general there is a rapiddecline in usage of fertilizer with increase inlandholdings of around 1–2 ha. As land size increasesin Adarsha watershed the amount of fertilizer applieddecreases.

Farmyard manure and compost

In the Adarsha watershed, the amount of farmyardmanure (FYM) applied per hectare differs among thesmall landholdings. The most significant anomaly isthat for a plot of 5 ha, nearly 6 t ha-1 of FYM is applied,and for a plot of about 4 ha approximately 1.5 t ha-1 ofFYM is applied.

Weedicide and insecticide

Weedicide and insecticide are applied in variousdoses. The micro-watershed shows a sharp decline inweedicide and pesticide usage by farmers owning upto 0.4 ha, and a gradual decline with increasing landsize.

Constraints

After the baseline survey, it was concluded thatKothapally village is characterized by variousconstraints such as:• Low level of literacy• Less proportion of irrigated area (20%) and higher

dryland area (80%)• Inverse relationship between land size and

productivity• Diversity in cropping systems between rainy and

postrainy seasons• Scarcity of labor• Low crop productivity• No water harvesting/storage structures• Less use of fertilizers• Low adoption of pest management practices• Income generating activities are not taken up by

women/villagers

Detailed characterization of soil samples

Soils of Kothapally watershed are of 4 series withvarying depths of 0–40, 0–70, 0–90, and 0–120 cm.The soil series of 0–40 and 0–70 cm depth aredeveloped on basaltic parent material having 1–3%gentle slope. These soils are shallow, well drained withmoderate erosion. These soils have very dark grayishbrown surface; subsurface horizons are clayeythroughout the profile. These soils are suitable forgrowing sorghum, soybean, and black gram. Soilseries of 0–90 and 0–120 cm depth are deep,moderately well drained, flat lands with gentle slope(0–1%). These soils have very dark brown surfacehorizon and very dark grayish brown to dark yellowishbrown subsurface horizons which are clayeythroughout the profile. The soils are developed fromalluvium parent material suitable for long-durationcrops like cotton, pigeonpea, turmeric, etc.

Table 1. Crop productivities (t ha-1) in Adarsha watershed, Kothapally in 1998.

Land- Black Otherholders Rice Turmeric Sorghum Pigeonpea gram Cotton Beans Tomato crops

Small 2.83 2.10 1.47 0.19 0.83 0.21 0.79 – 0.33Medium 3.09 2.75 1.19 0.15 0.57 1.43 1.37 0.81 0.74Large 1.66 1.23 0.54 0.13 0.25 0.67 0.19 0.75 1.33

84

Soil samples from the watershed to a depth of 1 mare characterized in terms of their physical, chemical,and biological parameters. Surface soil pH in bothmedium and shallow soils was around 8.3. Soil pHincreased with soil depth. The organic carbon (C) andtotal nitrogen (N) were more in medium-deep soilsthan in shallow soils. The organic C content of soilsdecreased from 5.7 g kg-1 to 1.0 g kg-1 in shallow soilsand from 6.3 g kg-1 to 3.4 g kg-1 in medium deep soilsin top 15 cm layer compared to 60–90 cm soil depth(Table 2). Similar trends were also observed for totalN content. Available phosphorus (P) as estimated byOlsen’s method was very low (1.4 to 2.2 mg kg-1 soil)in top 15 cm layer and decreased with increasing soildepth. The micronutrients like zinc (Zn), boron (B),

and sulfur (S) were found to be lower than their criticallimits. Fine sand and coarse sand were more in shallowsoils while silt and clay were more in medium-deepsoils (Table 3). Soil moisture content at wilting pointvaried from 21 to 27%.

Soil biological activity parameters such asmicrobial biomass, soil respiration, dehydrogenase,alkaline and acid-phosphatase activities are the directmeasures that indicate the soil health. These biologicalproperties are directly associated with transformationsof various elements in soil which are needed for plantgrowth. Soil biological parameters varied significantlyfor shallow and medium-deep soils in the watershed.Like organic C and total N contents from microbialbiomass C and N soil respiration and other biological

Table 2. Analysis of pre-sowing soil samples collected from Adarsha watershed, Kothapally, May 1999.

Properties Land depth 0–151 15–30 30–60 60–90 Mean SE±

pH Shallow 8.34 8.46 8.76 8.86 8.61Medium 8.27 8.30 8.34 8.40 8.33SE ± 0.04 0.034Mean 8.30 8.38 8.55 8.63SE ± 0.02

EC Shallow 0.20 0.17 0.23 0.40 0.25(m mhos cm-1) Medium 0.19 0.17 0.17 0.16 0.17

SE ± 0.011 0.008Mean 0.19 0.17 0.20 0.28SE ± 0.007

Olsen-P Shallow 1.44 0.67 0.53 0.10 0.68(mg kg-1 soil) Medium 2.20 1.05 0.43 0.31 1.00

SE ± 0.245 0.182Mean 1.82 0.86 0.48 0.20SE ± 1.34

Organic Shallow 5.7 3.7 1.0 1.0 2.9C (g kg-1 soil) Medium 6.3 6.1 4.9 3.4 5.2

SE ± 0.60 0.52Mean 6.0 4.9 2.9 2.2SE ± 0.24

Total N Shallow 639 445 193 172 362(mg kg-1 soil) Medium 647 606 483 315 513

SE± 49.1 43.5Mean 643 526 338 244SE± 18.6

1. Soil depth (cm).

85

parameters decreased with increasing soil depth in theprofile (Table 4).

On-farm Trials and Farmers’ParticipationSeveral farmers in the watershed are coming forwardto take up on-farm trials in their fields with technicalbackstopping from ICRISAT. The number of farmersparticipating in these trials increased since start of theproject. Overall, 137 and 138 farmer participatorytrials were conducted in 2000 and 2001 respectively toevaluate improved management options. The areaunder on-farm trials in 2001 season was substantiallyincreased to 108 ha as compared to that of 2000 (81.9ha) and 1999 (36.8 ha) seasons.

Soil and Water Conservation ActivitiesAn urgent need to conserve water and soil in thewatershed is felt after a thorough analysis of the

transect walk conducted. To control erosion andrestore productivity of degraded soils in this area,several soil and water conservation activities weretaken up to conserve the harvested water and increasethe productivity of the crops. These activities areimportant in maintaining, improving, and enhancingproductivity of the crops. Widespread adoption ofimproved practices is essential for controllingdesertification and restoration of degraded soils.Engineering techniques of erosion control and runoffmanagement can be made more effective when used inconjunction with biological control measures such asvegetative barriers, grassed waterways, etc. InAdarsha watershed in Kothapally, several soil andwater conservation activities along with biologicalcontrol measures were taken up both at farm and atcommunity levels.

Ex-situ conservationExcess water is drained away from the fields safelythrough grassed waterways. A total of 21 potential

Table 3. Texture analysis of pre-sowing soil samples collected from Adarsha watershed, Kothapally, May1999.


Coarse sand (%) Shallow 13.9 22.9 41.1 43.7 30.4Medium 7.6 7.8 10.4 25.9 12.9SE ± 3.98 2.75Mean 10.7 15.3 25.7 34.8SE ± 2.35

Fine sand (%) Shallow 9.4 13.3 15.6 16.2 13.6Medium 5.7 5.8 6.9 11.4 7.4SE ± 2.05 1.37Mean 7.6 9.6 11.2 13.8SE ± 1.24

Silt (%) Shallow 21.5 17.3 18.1 19.2 19.0Medium 25.0 22.3 20.4 16.7 21.1SE ± 1.50 1.03Mean 23.3 19.8 19.2 17.9SE ± 0.89

Clay (%) Shallow 55.2 41.3 31.6 24.7 40.7Medium 61.7 64.4 62.3 48.2 59.2SE ± 5.46 3.74Mean 58.4 52.9 46.9 41.5SE ± 3.25

1. Soil depth (cm).

86

sites for water storage structures were identified by thevillage committees and scientists’ team and 10structures were completed; 270 sites for gully controlstructures were identified and 70 structures werecompleted. Also, 40 ha for field bunding was proposedand completed and 10 gabion structures wereproposed and one structure was completed.

In situ conservationShaping of the land reduces runoff. The land is maderough by broad-bed and furrow (BBF) landformtreatment. The beds are prepared at 0.4 to 0.6%gradient. The BBF method helps to reduce runoff andconserves more water in the soil profile and alsodrains excess water safely away from the crops. This

method is being adopted by the farmers in Adarshawatershed with technical backstopping fromICRISAT. Contour planting on flat (flat on grade)landform is also adopted by some farmers. Bullock-drawn tropicultor, developed by ICRISAT, is used bythe farmers for planting, sowing, fertilizer application,and intercultivation. Planting of Gliricidia is done byfarmers. About 30000 and 16000 Gliricidia plantswere planted in 1999/2000 and 2000/01 respectively,on field bunds by the farmers for stabilizing the bundsto conserve the rainwater and soil. In addition theseplants generate N-rich organic matter for fieldapplication for augmenting N supply for crop growth.This would reduce the dependence on mineralfertilizer N.

Table 4. Soil biological properties of pre-sowing soil samples collected from Adarsha watershed,Kothapally, May 1999.


Soil respiration Shallow 126 107 52 44 82(mg C kg-1 soil 10d-1) Medium 157 112 96 75 110

SE ± 6.0 4.4Mean 142 110 74 59SE ± 3.3

Mineral N (NH4+NO3) Shallow 10.3 7.1 5.8 4.2 6.8(mg N kg-1 soil) Medium 11.7 10.1 5.7 4.9 8.1

SE ± 1.35 1.04Mean 11.0 8.6 5.8 4.5SE± 0.71

Net ‘N’ mineralization Shallow 1.14 0.97 0.28 0.15 0.63(mg N kg-1 soil 10d-1) Medium 2.05 1.12 1.08 0.57 1.21

SE ± 0.77 0.43Mean 1.59 1.04 0.68 0.36SE ± 0.52

Microbial biomass carbon Shallow 288 214 123 62 172(mg C kg-1 soil) Medium 267 191 160 109 182

SE ± 16.9 11.2Mean 278 203 141 85SE ± 10.3

Microbial biomass nitrogen Shallow 45.5 33.9 19.5 9.8 27.2(mg N kg-1 soil) Medium 42.3 30.2 25.2 17.2 28.8

SE ± 2.67 1.77Mean 43.9 32.0 22.4 13.5SE ± 1.63

1. Soil depth (cm).

87

Wasteland development

Common wasteland treatment has been initiated in 1ha land and contour trenches (10 m width and 0.3 mheight) were laid out. Custard apple plantation wasundertaken through local people on wastelands on thetrenches during 2000 and 2001. In all 300 custardapple plants were planted. This will give additionalincome to the villagers as they can market the fruits inthe adjacent cities. The wasteland boundaries wereplanted with Gliricidia plants at 0.5 m spacing to serveas live fence and also as a source of N-rich organicmatter through loppings.

Avenue plantations

Avenue plantation was also taken up in the village as apart of the afforestation program in the village. Treeplantation along the roads, field bunds, and nalas wasundertaken. Teak plantation (2500 trees) in privatefields was also undertaken.

Integrated Nutrient Management

Vegetative bunds

In addition to grass planting, Gliricidia was planted onfield bunds and used to conserve moisture and supplyN to the crop through biologically fixed N byincorporation of loppings into the soil. This reducesthe usage of fertilizers. During 1999–2001 farmersplanted Gliricidia plants on their field bunds.

Nutrient budgeting and balancedfertilization trials

In the watershed, 15 farmers are following theimproved soil, water, and nutrient (SWNM)management options along with conventionalpractices. Balanced nutrient doses were used forsustaining productivity in these watersheds.Rhizobium inoculation of pigeonpea and soybeanseeds was done to increase biological nitrogen fixation(BNF). Crop responses were positive to specificnutrient amendments. Based on soil analysis, B and Sapplications were done at Kothapally and increasedyields were observed. Higher grain yields wereobtained with improved practices and this indicates a

considerable scope for savings on N fertilizer.Quantification of BNF using N-difference method isbeing done using non-fixing crop (maize and sesame)varieties of matching duration with groundnut andsoybean in farmers’ fields.

The nutrient uptake by maize/pigeonpea intercropsystem was more in the improved systems ascompared to that of flat landform treatment. The N-difference and 15N isotope dilution methods were usedto quantify BNF contributions of legumes using non-fixing control plants. Similarly, for the sole maize cropuptake of nutrients was more in BBF system than theflat landform. The nutrient balances based on theavailable data sets showed that in this watershed all thesystems are depleting potassium (K) from soils andmore P is applied than removed by the crops. Nutrientremoval was also more in BBF than in the flatlandform treatment. Higher negative N balance inmaize/pigeonpea in BBF system (–55 kg N ha-1) showsthat the crop extracted more N from the soil whengrown on BBF system than on flat system (–48 kg Nha-1) (Table 5).

In situ generation of organic matter forgreen manuring

Leguminous green manures such as Gliricidia areimportant in maintaining soil and crop productivity.Decomposition and nutrient release of Gliricidialoppings occur at a faster rate due to low C:N ratio.Most of the nutrients especially N and K are releasedwithin 5–10 days of decomposition. Decomposingleaf prunings of Gliricidia are better and rapid sourceof nutrients. Forty-six thousand Gliricidia plants wereplanted during 1999–2001 by farmers on their fieldbunds at Kothapally.

Vermicomposting Boosts IncomesEarthworms are used in vermicomposting as they arevoracious eaters and can transform organic wastes intocompost in a short span. Compost which is processedby earthworms makes good organic fertilizer as itcontains auxins, a growth promoter for plants and alsosome natural antibiotics along with plant nutrients.Vermicomposting is a cost-effective pollutionabatement technology. At Kothapally, 52 womenfarmers were identified for vermicomposting units. Of

88

the ten existing women groups, five groups wereformed and trained in vermicomposting techniques.The groups started the units with the available organicwastes, cow dung, etc. These women self-help groupshave taken up vermicomposting as a micro-enterpriseto generate income.

Method of vermicomposting

Agricultural residues like sorghum straw, paddy straw,dry leaves, pigeonpea stalk, groundnut husk, wheathusk, weeds like Parthenium, and agricultural wastes(e.g., animal manures); dairy and poultry wastes; foodindustry wastes; municipal solid wastes; biogas-sludge; and bagasse from sugarcane industry can beused as raw material for vermicomposting. Thecomposting is done in cement rings or 1.5 m3 tanks.Dry organic wastes, dung slurry, rock phosphate,earthworms, and water are mixed (10:3:0.4:100–150:1). The bottom of the tank is filled up with drymaterial like coconut husk or a polythene sheet isspread and on this 15–20 cm of organic wastes is filledas a first layer, rock phosphate as the second layer, anddung slurry as third layer. More layers are filled oneabove the other in the tank. The top layer is plasteredwith mud slurry to prevent moisture loss. This is left todecompose for 15 days to dissipate the heat generatedduring initial decomposition. Earthworms are releasedinto the compost through the cracks after 15 days. Tomaintain adequate moisture, water should be sprinkledon the vermicompost tank intermittently. The compostwill be ready within 6–8 weeks. The vermicompost isheaped in a cone shape. The earthworms move to the

bottom out of the compost heap and these can becollected and used again.

Response of tomato to vermicompostapplication

In 2001, a demonstration plot was initiated in thevillage with a plot size of 300 m2. Vermicompost wasapplied to the standing crop of tomato at 3–5 t ha-1.The productivity (5.8 and 4.8 t ha-1) of tomatoes wassignificantly higher in plots with 3 and 5 t ha-1

vermicompost when compared with plots withconventional compost (3.5 t ha-1 yield). The wormcastings in the vermicompost have nutrients that are97% utilizable to the plants.

Integrated Pest ManagementIntegrated pest management (IPM) is the coordinateduse of pest and environmental information to designand implement pest control measures that areeconomically, environmentally, and socially sound. Itpromotes prevention over remediation and advocatesintegration of at least two or more strategies to achievelong-term solutions. IPM uses methods such as crop orsite scoring, pest trapping, pest tolerance cropvarieties, weather monitoring, cultural controls,biological controls, and precise timing and applicationof pesticide treatments, only when needed. Completedependency on chemical control for the past threedecades led to unsatisfactory pest management alongwith environmental degradation. ICRISAT along withnational agricultural research and extension systems

Table 5. Nutrient budgeting in farmers’ fields in Adarsha watershed at Kothapally, 1999–2000.

Total inputs Total outputs BalanceCropping system/Landform N P K N P K N P K

Maize/pigeonpeaBBF 28.3 16.4 17.1 84.5 10.6 57.6 –55 +6 –40Flat 32.2 13.8 21.2 80.2 8.8 49.7 –48 +5 –29

Sole maizeBBF 20.5 10.0 0.0 74.8 14.1 70.6 –55 –4 –70Flat 9.0 10.0 0.0 32.7 7.3 35.9 –24 +3 –35

Sole sorghumFlat 18.3 9.9 11.0 41.8 9.7 64.3 –24 +0.2 –53

89

(NARES), NGOs, and farmers conducted research inthe watershed to identify environmentally sound andeconomically viable plant protection technologieswhich reduce yield losses and improve farmers’income. Farm surveys and participatory ruralappraisals identified the non-availability of IPMcomponents such as plant-based products, nuclearpolyhedrosis virus (NPV), pheromones, and pesttolerant varieties. The farmers harvested six-foldincreased yields through better management of pestsby controlling them with neem seed extract. There was6–100% reduction in pesticide usage. After thoroughevaluation of the existing pest management options, acomprehensive IPM package for chickpea andpigeonpea was developed and evaluated in farmerparticipatory approach mode. Revitalizing theeffective indigenous methods like shaking of podborers from the pigeonpea crop and use of neem forpest management was done in both the watersheds.These indigenous methods are effective, cheaper, andenvironment-friendly. Installation of pheromone trapsfor pest monitoring was done every year. Bird percheswere also installed in the fields for birds to rest andfeed on the Spodoptera larvae.

Crop surveys

Crop surveys were carried out to know the plantprotection practices followed by farmers inKothapally. All the farmers interviewed indicated useof chemical pesticides against insect pests. Theyindicated Helicoverpa as the key pest on severalcrops. Endosulfan, cypermethrin, fenvalerate, mono-crotophos, and quinalpos were the commonly usedchemicals across the farming community. Precautionswere not taken while spraying. This preliminarysurvey clearly brought about several inappropriateways of chemical usage, which need to be addressed inthe coming years.

Pest control

Cotton

Cotton crop was sown in the first fortnight of June withthe onset of monsoon. Initially farmers could protecttheir crop by 3–4 chemical sprays against suckingpests like jassids, aphids, and whiteflies. Helicoverpa

population was controlled by Helicoverpa NPV(HNPV), which kept the population below theeconomic injury level.

Pigeonpea

Pigeonpea crop was sown as both sole and intercropwith maize or sorghum. Helicoverpa was the keyconstraint to pigeonpea production. The adultpopulation of Helicoverpa was monitored usingpheromone traps. The farmers applied neem spraysand HNPV sprays followed by manual shaking. Nochemical sprays were used. These farms had lowerpod borer damage and higher yields when comparedwith fields where IPM practices were not followed.

Chickpea

Observations of egg and larval populations indicatedthe onset of pest infestation, particularly Helicoverpaand farmers applied HNPV in their fields. The farmersobtained three-fold more yield (780 kg ha-1) thanyields obtained by farmers (250 kg ha-1) who did notadopt IPM in their fields. The increased yields are dueto IPM as well as the use of the variety ICCV 37supplied by ICRISAT.

Monitoring Helicoverpa by pheromonetraps

Population of adult Helicoverpa was monitored inKothapally village from 2000 by using pheromonetraps with the pheromone lures obtained from theNatural Resources Institute (NRI), UK.

Village-level HNPV production

Among various options, the availability of goodquality HNPV was considered a prime component forspread of IPM. This project quickly identified andinitiated village-level production to cater to the needsof farmers. Many farmers and extension workers fromthis village were given training on HNPV production,storage, and usage on different crops. The villagersquickly adopted the technology and produced 2000larval equivalents (LE) of virus and used on cotton,pigeonpea, and chickpea crops. Besides the village-level production, 11650 LE HNPV was supplied to the

90

farmers through ICRISAT to cover cotton, pigeonpea,and chickpea crops.

The project has given high priority for trainingvillage scouts in identifying various pests and theirnatural enemies in different crops before the croppingseason, and assisted them in monitoring throughoutthe crop period. A slide show emphasizing croppingsystems, various pests and diseases and theirmanagement was organized for the whole villageincluding children. Video shows on correct use ofplant protection emphasizing the importance of IPMwere displayed in the village twice during the season.Farmers were trained on HNPV production atICRISAT, Patancheru and were assisted to take upvillage-level HNPV production. Extension handoutson packages of practices for chickpea and pigeonpeacrops in local language were distributed.

Future of IPM at Kothapally

In the coming years, ICRISAT will be involved indevelopment of technologies for high quality insectpathogens to strengthen the existing IPM activities(viral and fungal pathogens). Basic research needs tobe conducted on the insect host plant interaction andcultural operations on pests and on natural enemies.Potential plant products that are safe and effective inpest management should be identified and developed.Insecticidal resistance in both pests as well as naturalenemies should be monitored. Village-based orregional-based IPM approach should be developedrather than pest-wise or crop-wise approach. Training

clients (researchers, extension workers, NGOs, andfarmers) at all levels in IPM concepts is needed.

MonitoringTo evaluate the impact of watershed managementcontinuous monitoring of all the parameters is done.An automatic weather station was installed tocontinuously monitor the weather parameters (Figs. 2and 3; Table 6). To monitor the groundwater levels 64open wells in the watershed were geo-referenced andregular monitoring of water level and quality was done(Fig. 4). Runoff, and soil and nutrient losses aremonitored using automatic water level recorders andsediment samplers (Fig. 5; Table 7).

Quantification of BNF in farmers’ fields wascarried out using N difference method and 15N isotopedilution method. Pheromone traps were installed tomonitor Helicoverpa populations. Changes incropping intensity, greenery, water bodies, andgroundwater levels were monitored. Geographicalinformation system (GIS) maps indicating soil types,soil depths, and crops grown during rainy andpostrainy seasons have been prepared. Cropproductivities were recorded for each crop every year.

Impact

Improved greenery

The normalized difference vegetation index (NDVI)has been used to monitor the impact of the

Figure 2. Average weekly rainfall recorded at Shankarpally Mandal, Andhra Pradesh.

91

Table 6. Monthly weather data recorded at Adarsha watershed, Kothapally, 1999–2001.

Rainfall Max. temp. Min. temp. Solar radiationMonth (mm) (°C) (°C) (MJ m-2)

19996 54.50 32.49 22.02 16.817 139.24 30.56 20.99 17.368 150.59 29.06 20.46 15.849 115.05 29.11 20.36 14.6410 50.90 30.49 18.34 15.8111 0.00 29.59 12.57 17.0812 0.00 28.07 9.54 15.5620004 4.09 41.56 23.60 22.835 138.00 37.99 23.54 22.256 165.30 31.75 22.26 15.157 132.29 29.96 21.71 15.088 460.09 30.26 21.88 14.049 103.69 32.14 20.86 18.0310 12.40 34.35 19.78 17.5520011 12.70 32.62 15.08 16.092 1.30 30.99 14.84 12.643 4.80 39.09 20.50 20.294 27.70 39.24 22.48 20.225 12.20 41.15 25.85 22.376 112.29 33.89 22.60 17.207 19.60 31.97 22.65 14.86

Figure 3. Rainfall recorded at Adarsha watershed, Kothapally, 2001.

92

Figure 4. Location map showing open wells inAdarsha watershed, Kothapally.

Table 7. Annual rainfall, runoff, and peak runoff rate at Adarsha watershed, Kothapally, 2001.

Description Treated watershed Untreated watershed

Annual rainfall (mm) 612 612Runoff (mm) 22 31Peak runoff rate (m3 s-1 ha-1) 0.027 0.022

implementation of action plan. An increase in thevegetation cover which reflects an improvement in thevegetation cover was observed. The spatial extent ofmoderately dense vegetation cover which was 129 hain 1996 has increased to 152 ha by 2000.

Increased groundwater levels

The groundwater levels and other related observations(pumping hours, area irrigated from each well anddistance between the well and check-dam) fromwatersheds were collected. At Kothapally watershed,throughout the season higher groundwater levels wererecorded from the well near the major check-damcompared to water levels in wells away from thecheck-dam (Fig. 6). This clearly shows theeffectiveness of the improved watershed technologiesin increasing the groundwater recharge therebyimproving the availability of water for agricultural andother uses.

Figure 5. Runoff from two sub-watersheds atKothapally, 2000.

Improved productivities and incomes forfarmers

At Kothapally, farmers evaluated improved cropmanagement practices along with improved landmanagement practices such as sowing on BBFlandform and flat sowing on contour; and usingimproved bullock-drawn tropicultor for sowing andinterculture operations. Farmers obtained two-foldincrease in the yields in 1999 (3.3 t ha-1) and three-foldincrease in 2000 (4.2 t ha-1) as compared to the yieldsof sole maize (1.5 t ha-1) in 1998 (Table 8). Inintercropped maize with pigeonpea, improvedpractices gave a four-fold increase in maize yield (2.7t ha-1) compared with farmers’ practices where theyields were 0.7 t ha-1. In sole sorghum the improvedpractices adopted increased yields three-fold withinone year. In 1999/2000, farmers achieved highestsystem productivity, total income, and profit fromimproved maize-pigeonpea and improved sorghum/

93

Figure 6. Effect of check-dam on groundwater recharge in Adarsha watershed,Kothapally during 2000.

Table 8. Average crop yield (kg ha-1) with improved technologies in Adarsha watershed, Kothapally,1999–2001.

Crop 1998 Baseline yield 1999 2000 2001

Sole maize 1500 3250 3750 3300

Intercropped maize - 2700 2790 2800(Farmers’ practice) 700 1600 1600

Intercrop pigeonpea 190 640 940 -(Farmers’ practice) 200 180 -

Sole sorghum 1070 3050 3170 2600

Intercrop sorghum - 1770 1940 2200

pigeonpea intercropping systems. Along with thehighest system productivity the cost-benefit ratio ofthe improved systems was greater (1:2.47) comparedto the farmers’ traditional cotton-based systems (Wani2000). In 2000/01, several farmers evaluated BBF andflat landform treatments for shallow and medium deepblack soils using different crop combinations. Farmersharvested 250 kg more pigeonpea and 50 kg moremaize per hectare using BBF on medium-deep soilsthan the flat landform treatment. Furthermore, even onflat landform treatment farmers harvested 3.6 t ha-1

maize and pigeonpea using improved managementoptions compared to 1.72 t ha-1 maize and pigeonpeausing normal cultivation practices.

Similar benefits from improved BBF landform andalso improved management options were reported by

the farmers in shallow soils and with other croppingsystems. The rainfall during 1999 in this area was 559mm, which is 30% below normal rainfall, and in 2000the rainfall was 958 mm, which is 31% above normal.In spite of this variation in the rainfall received in 1999and 2000, the productivity of the crops marked asustainable increase during 1999/2000 and 2000/01.

Of all the cropping systems studied in Adarshawatershed, maize/chickpea and maize/pigeonpeaproved to be more beneficial to farmers (Table 9).Farmers could gain about Rs 19590 and Rs 17802with these systems respectively. Sorghum, chickpea,and pigeonpea sole cropping systems also provedbeneficial, whereas sorghum, maize, and chickpeatraditional systems were significantly less beneficial tothe farmers.

94

Human Resource DevelopmentFarmers are exposed to new methods and technologiesfor managing natural resources through training and fieldvisits to on-station and other on-farm watersheds.Farmers and landless families were trained andencouraged to undertake income generating activities inthe watershed, which can help sustain the productivity atwatershed level. Various training sessions were held forfarmers on improved management options like providingtraining on farm implements, IPM, and integratednutrient management options. Along with the farmers,watershed committee members and agriculture andextension officials were trained at ICRISAT on differentaspects of integrated watershed management. Researchscholars and apprentices from various universities ofIndia, Thailand, Vietnam, and New Zealand conductedresearch on integrated watershed management. Specialemphasis was laid to educate women farmers andincrease awareness on new management options. Morewomen were trained in vermicomposting technology atKothapally. Educated youth were trained in skilledactivities like NPV production and vermicomposting,which helped them in generating income (Table 10).

Technology Imbibing into OtherWatershedsAround ten watershed farmers from Nawabpet(Yellakonda watershed, Sainnaguda watershed,Lingampally watershed, Maitaphkanguda watershed,and Gullaguda watershed) and Adilabad adopted theimproved practices which proved to be beneficial inAdarsha watershed, Kothapally and they are in theprocess of evolution. Farmers adopted BBFlandform in their fields. Use of tropicultor forsowing, fertilizer application, and intercultivationactivities impressed them very much and they boughttropicultors for their respective villages. Improvedcropping systems like sorghum/pigeonpea, maize/pigeonpea, sole sorghum, chickpea and maizecropping systems were taken up in about 206 ha inthese watersheds. Improved soil and waterconservation measures have been initiated in thesewatersheds. Farmers are found to be keenlyinterested in adopting Gliricidia plantations,vermicomposting, and HNPV production in theirrespective villages.

Table 9. Benefit-cost ratio of different cropping systems at Adarsha watershed, Kothapally, 2001.

Total Total Totalproductivity cost income Profit Benefit-cost

Cropping system (kg ha-1) (Rs ha-1) (Rs ha-1) (Rs ha-1) ratio

ImprovedMaize/chickpea 4700 6883 26473 19590 1:2.85Maize/pigeonpea 3753 6342 24144 17802 1:2.81Maize 3000 4150 12260 8110 1:1.96Sorghum 3000 3850 13860 10010 1:2.60Chickpea 850 5250 18000 12750 1:2.43Pigeonpea 1090 4890 17120 12230 1:2.50

TraditionalMaize/chickpea 2750 5915 16650 10735 1:1.82Maize/pigeonpea 1715 4452 12769 8317 1:1.87Sorghum/pigeonpea 1116 4050 11610 7560 1:1.87Cotton 1163 16990 26748 9758 1:0.57Maize 1600 3360 7500 4140 1:1.23Chickpea - 4260 11600 7340 1:1.72Sorghum 1011 3050 7055 4005 1:1.31

95

Why is Adarsha Watershed a ModelWatershed?The Adarsha watershed is said to be a modelwatershed as all the activities are through communityinitiatives and the strength lies in local participation ofpeople, especially through women empowerment. Theproject improved the livelihoods of the poor byincreasing the farm productivities, farm incomes,groundwater levels, and improving greenery. Thecapacity of local governments and community-basedorganizations has been enhanced through watershedmanagement and decision-making processes. Thisproject is aware of the need to involve local residentsand community-based organizations, given thatresidents possess unique, first-hand knowledge aboutlocal resources and environmental threats.

ConclusionOn-farm trials were conducted by ICRISAT in 1980sand the results on station were replicated in farmers’fields. But even after 15 years in the same village, theimproved practices were not adopted by the farmers ofthe village; they went back to their traditionalpractices. The researchers found the loopholes for lowadoption of the technology package. A new model ofintegrated watershed management was developed byICRISAT with the lessons learned on farm.Contractual mode of farmer participation did notachieve good results, so a higher degree of farmerparticipation through consultative and cooperativemode was initiated and found to be successful in thewatershed. Gender issues were considered high

priority. As women are the key players in developmentof the society, keen interest was taken to empowerwomen in various income generating activities likevermicomposting and HNPV production within thevillage. On-farm trials were conducted in farmers’fields by providing them only with technicalbackstopping; no subsidies were given. Socialauditing was done by the villagers themselves. Tosustain the productivity in the SAT, a holistic approachof integrated watershed management still needs to bescaled up through appropriate policy and otherinstitutional support and the on-site and off-siteimpacts need to be studied.

AcknowledgmentsThe financial assistance provided by the ADB throughRETA 5812 is gratefully acknowledged. The efforts ofscientists of DWMA, NRSA, and MVF are greatlyappreciated. The contribution from farmers who arethe important partners is greatly acknowledged.

ReferencesBeteille, A. 1974. Studies in agrarian social structure.Oxford University Press.

Crosson. 1994. Degradation of resources as a threat tosustainable agriculture. Presented at the First WorldCongress of Professionals in Agronomy 5–8 Sep 1994,Santiago, Chile.

Lal, R. 1995. Global soil erosion by water and Cdynamics. Pages 131–141 in Soils and global change (Lal,R., Kimble, J., Levine, E., and Stewart, B.A., eds.). BocaRaton, Florida, USA: CRC/Lewis Publishers.

Table 10 Various training programs conducted on integrated watershed management at Kothapally,2000–01.

Stakeholders Number of training courses Participants

Farmers 5 675Agricultural officials 3 80Government officials/nodal officers 3 60Non-governmental organization/

Project implementation agencies 1 30Research scholars/Apprentices from

various countries - 14Visitors - 800

96

Oldeman, L.R. 1994. The global extent of soildegradation. In Soil resilience and sustainable land use(Greenland, D.J., and Szabolcs, I., eds.). Oxon,Wallingford, UK: CAB International.

Ryan, J.G., and Spencer, D.C. 2001. Challenges andopportunities – shaping the future of the semi-arid tropicsand their implications. Patancheru 502 324, AndhraPradesh, India: International Crops Research Institute forthe Semi-Arid Tropics.

Scherr, S.J., and Yadav, S. 1996. Land degradation in thedeveloping world: Implications for food agriculture andthe environment to 2020. Food, Agriculture andEnvironment. Discussion Paper 14. Washington, DC,USA: International Food Policy Research Institute.

Seckler, D., Amerasinghe, U., Molden, D., De Silva, R.,and Barker, R. 1998. World water demand and supply

1990 to 2025: Scenarios and issues. Research Report 19.Colombo, Sri Lanka: International Water ManagementInstitute.

Wani, S.P. 2000. Integrated watershed management forsustaining crop productivity and reduced soil erosion inAsia. Presented at MSEC Meeting, 5–11 November 2000,Semarang, Indonesia.

Wani, S.P., Sreedevi, T.K., Pathak, P., Piara Singh, andSingh, H.P. 2001. Integrated watershed managementthrough a consortium approach for sustaining productivityof rainfed areas: Adarsha watershed, Kothapally, AP – acase study. Presented at the Brainstorming Workshop onPolicy and Institutional Options for SustainableManagement of Watersheds, 1–2 November 2001,ICRISAT, Patancheru, India.

97

Conjuctive Use of Water Resource Technology and Extension in ImprovingProductivity of Rainfed Farming: An Experience

at Lalatora, Madhya Pradesh, India

B R Patil1, A B Pande1, S Rao1, S K Dixit1, P Pathak2, T J Rego2, and S P Wani2

Abstract

Although Vertisols and associated soils in Lalatora, Madhya Pradesh, India are a productive group ofsoils, these soils have several constraints to high agricultural productivity. Low cropping intensity,waterlogging during the rainy season, and lack of farmers’ inability to use modern technology are themajor constraints. On-farm studies made during 1999–2001 showed that use of improved varieties ofcrops such as soybean and chickpea along with improved water harvesting and nutrient management(including application of micronutrients) and integrated pest management significantly increased cropproductivity and income of the farmers. There is a need to further test the package of practices forimproving productivity and farm income.

tions for sustainable use and management of naturalresources. A consortium model for management anddevelopment of integrated participatory watershed isevaluated at Lalatora watershed in Vidisha district,Madhya Pradesh, India. The project aims to developand apply environment-friendly technological optionsto increase the productivity in this area of medium tohigh water-holding capacity soils.

Area, Climate, and PhysiographyLalatora village of Vidisha district of Madhya Pradeshis located in an extensive watershed totaling about10525 ha, lying between latitude 28°8’3” and 24°16’ Nand longitude 77°20’45” and 77°30’15” E at a heightof 415 m above mean sea level. The Lalatorawatershed is a micro-watershed within the Lateri-1milli and is located in the northwest corner of Vidishadistrict. Madhya Pradesh is the largest state in Indiaand extends into three agro-ecological zones (7, 8, and9), which are characterized as having 120 to 180 dayslength of growing period; soils are largely Vertisolswhere the parent material is mainly alluvial. Thecatchment area represents the four major rivers, viz.,Yamuna, Ganga, Narmada, and Godavari. The state is

1. BAIF Development Research Foundation, IIIrd floor, Indra Complex, Manialpur, Baroda 390 004, Gujarat, India.2. International Crops Research Institute for the Semi-Arid Tropics (ICRISAT), Patancheru 502 324, Andhra Pradesh, India.

The concept of sustainable development cannot bediscussed only in the light of economic development.The social and environmental changes and the peopleshould be considered. Sustainable socioeconomicdevelopment is a challenge and an opportunity. Alongwith economic development, sustainable socialdevelopment has become an integral part of the entiredevelopment process as it aims at “the improvement ofthe well-being of the people”. The success and failureof the development program and technologyinterventions largely depend upon the actors involvedin the process. The major stakeholders in the processof development are those who make use of naturalresources. The participation of these users is veryimportant for the management of natural resources aswell as for technology adoption and success. A processoriented step-by-step approach is very effective forsustainability. Farming systems research (FSR) is asuccessful example of this approach, which has beentried out in many developing countries. Integration ofthe roles of scientist, extensionist, and farmers is thekey to success in FSR approach.

The project has adopted a holistic process, i.e.,FSR approach, to improve the productivity of soybean-based rainfed farming with technological interven-

98

divided into six physiographic regions; the district ofVidisha is located in the Vindhya Plateau Zone.

Surface hydrology

Two digital runoff recorders along with automaticpumping sediment samplers were installed at theLalatora watershed to monitor runoff and soil erosionfrom untreated and treated sub-watersheds. Tomonitor the outflow from entire Lalatora watershed,one digital runoff recorder was installed at the majordrain. There is a significant reduction in runoff fromthe treated sub-watershed compared to untreated sub-watershed (Table 1). In 1999, the significant reductionin runoff from treated sub-watershed (24% less thanthe untreated sub-watershed) was observed. During2001, large reduction in runoff volume (81% lesscompared to untreated sub-watershed) was recorded.The difference in the runoff between the treated anduntreated sub-watersheds was greater during 2001compared to 1999. This is mainly due to the fact thatstarting from 1999, more area was brought underimproved technologies including check-dams andstructures in the treated sub-watershed. The peakrunoff rates from the treated sub-watershed were alsosignificantly lower compared to untreated sub-watershed (Table 1). During 1999, the peak runoff ratein the treated watershed was only one-third that ofuntreated watershed. During 2001, the peak runoffrate in the treated watershed was 30% lower thanuntreated watershed. During three years (1999–2001),the highest peak runoff rate of 0.218 m3 s-1 ha-1 wasrecorded from the untreated watershed.

During all the three years (1999–2001), few majorstorms contributed 50–75% of the seasonal runoff

(Fig. 1). During 1999, one single storm on 5September resulted in more than 50% seasonal runofffrom both treated and untreated sub-watersheds.Similar trend was seen during 2000 and 2001. Theeffectiveness of treated watershed in controllingrunoff from small and medium storms is shown inFigure 1 for 2001. During 2001, runoff from all thesmall and medium storms were totally controlled inthe treated watershed except for one large storm on 12July 2001.

Groundwater hydrology

At Lalatora, 12 open wells were monitored atfortnightly intervals to record the groundwaterfluctuations. The mean water level in open wellsbefore watershed development was about 6.5–9.5 m.The water level in open wells increased substantiallyin subsequent years after implementing watersheddevelopment work, particularly construction of thecheck-dams and other water harvesting structures.During 2000, the mean water level in the wells nearcheck-dams was consistently around 1.5–2.0 m up toOctober, whereas the water level in the wells locatedaway (about 1000 m) from the check-dams was atabout 8.5 m throughout the year (Fig. 2). Even duringrainy season the wells away from check-dams did notshow significant increase in water level, while thewater levels in the open wells near the check-dams hadsignificant increase particularly during the rainy season.

The Lateri block is considered the mostunderdeveloped area within the district of Vidisha, withvery limited irrigation and no major or medium-scaleindustry. The average rainfall is 1000 mm. The soils ofthe area range between medium black to red soils.

Table 1. Seasonal rainfall, runoff, and peak runoff rate from two sub-watersheds at Lalatora, 1999–2001.

Runoff from twosub-watersheds (mm) Peak runoff rate (m3 s-1 ha-1)Rainfall

Year (mm) Untreated1 Treated2 Untreated1 Treated2

1999 1203 296 224 0.218 0.0652000 932 234 NR3 0.019 NR3

2001 1002 290 55 0.040 0.0271. Untreated = Large watershed with most of the areas untreated.2. Treated = Small watershed with major areas treated with improved technologies including hydraulic structure.3. NR = Not recorded.

99

Agriculture is the main occupation in the block, butemployment is seasonal due to lesser crop intensitybecause of low irrigation availability. About 20% of thepopulation migrates seasonally. The postrainy season(rabi) is the main cropping season with 35,000 ha sownarea while 10,000 ha is sown during the rainy season(kharif). Double cropping is undertaken in only 3750 ha(Rangnekar 1999). An automatic weather station wasestablished in the watershed. Daily weather readings arecollected and are maintained.

Soils

The Lalatora watershed in particular is spread on theDeccan Trap basalt where the parent material ismainly alluvial. The physiography of the area is levelto very gently sloping land where certain pocketstowards the north of the area are highly gulliedcreating a certain amount of relief, which may createfurther problems of management. Five soil series havebeen identified. These are Vertisols characterized bygray, very deep, dark grayish brown to olive brown

with a clayey surface horizon and calcareous Bhorizon. The predominant clay mineral ismontmorillonite. These soils have greater microporevolume due to high amount of very fine clay present inthe soil (NBSSLUP 2000). About 60% soils aremedium (30 to 60 cm depth) while 20% are deep (>60cm depth) and 20% are shallow (<30 cm depth).

Socioeconomic StatusA baseline socioeconomic and natural resourceinventory survey was conducted through a stratifieddetailed household survey and participatory ruralappraisal (PRA) methods. Primary data have beencollected from 396 households from 7 villagescovering top, middle, and lower toposequencepositions in Lalatora watershed. The data have beencollected using an interview schedule prepared bytrained investigators. The sample for the study ispresented in Table 2. Low literacy rate, poorresources, and inadequate extension are the mainreasons for backwardness of the community.

Figure 1. Daily runoff events from the treated and untreated sub-watersheds during 1999 and2001 at Lalatora.

100

Figure 2. Groundwater levels in open wells at Lalatora watershed, 1998–2001.

101

Family information

Ninety-four percent of the households are male-headed and 6% are headed by females, mostlywidows. Main source of income for 56% of thehouseholds was from agriculture, 28% from non-agricultural labor, 14% from agricultural labor, andaround 2% from the services sector. The occupationalcategorization of individual family members ispresented in Table 3. Lalatora has 56% small farmers(57), 29% medium farmers (30), and 15% largefarmers (15). Literacy status revealed that 52% areilliterate, 29% reach up to primary education, 12%complete secondary education, and 7% complete highschool education.

Social structure

The average size of the household is 9.4 persons ofwhich 5.3 are male and 4.1 are female. The family sizeof the landless laborers is much smaller at 5.5 personsper household, perhaps due to lower income andassets. The availability of labor is seasonal and agreater family size would require them to migrate forwork. The recent study (Vadivelu et al. 2001) revealedthat in most of the cases people with some

landholdings enter into share cropping contracts andthese people ‘crowd-out’ the share croppers from theshare cropping market. The obvious reason is that thelandholders would assume to have a better knowledgeof agricultural operations and would also be in aposition to repay the borrowed money from the landlord(through growing wheat from their own land).

Land-use PatternAbout 60% soils are black soils, medium in depth (30 to60 cm depth), 20% are deep black soils, more than 60cm deep, and 20% are shallow soils (30 cm depth). Inthe Lateri block 53% of the land is used for agriculture,17% is classified as wasteland, 25% as forest land, 4%as grazing land, and 1% as fallow land. The averagelandholding of the farmers from the study area is around4 ha. The average landholding size is 4 ha, with 2.73 haof dryland and 1.27 ha of irrigable land.

The primary data collected revealed that onlysoybean is grown in large areas (Rangnekar 1999).The other major rainy season crops are sorghum andmaize, while pulses are grown in smaller areas.Soybean was the crop preferred by 74% of the farmersin kharif followed by rice (21%), sorghum (3%), and

Table 2. Sample area (ha) for the study at Lalatora.

Marginal Small Medium Large LandlessRegion farmers farmers farmers farmers laborers Total

Upper area 15 15 18 14 9 71Middle area 16 28 30 16 10 100Lower area 17 27 29 34 23 130Lalatora micro-watershed 23 20 14 12 26 95

Total 71 90 91 76 68 396

Table 3. Occupation categorization (%) of family members1.

Work Migration Unemployment

Agri + Agri + For ForGender Agri NFW NFW Trade Trade Agri NFW Always Sometimes

Male 28.8 9.8 12.2 19.3 15.7 2.7 - 2.1 9.2Female 24.5 13.3 30.3 8.5 6.9 2.7 0.5 2.7 10.6 Overall 27.4 11.1 18.7 15.5 12.6 2.7 N 2.3 9.71. Agri = Agriculture; NFW = Non-farm work; N = Negligible.

102

maize (2%). In the rabi season, wheat was preferred by46% of the farmers followed by chickpea (43%), rice(6%), and lentil (5%).

During the rainy season, out of an average 2.73 haof dryland, only 0.26 ha is cultivated. During the rainyseason, 88% of land in Lalatora is kept fallow by thefarmers. The fallow lands are largely rainfed landswithout any source of irrigation and farmerstraditionally grow postrainy season crops on storedsoil moisture and use supplemental irrigation obtainedon payment from the neighbors. Soybean is the onlycrop cultivated during this season with an averagegrain yield of 0.95 t ha-1, whereas in the postrainyseason, 71% area is under wheat and 24% underchickpea.

Relationship between Soil, Rainfall,and Cropping PatternThe soils have a higher clay content characterized bygreater water-holding capacity with poor drainage andproblems due to waterlogging. The average annualrainfall is about 1000 mm with high intensity ofrainfall resulting in runoff and soil erosion. Theproblem in the rainy season is ponding of water causedby high rainfall intensity.

Farmers prefer to grow the rainy season crop(soybean) under irrigation. Also, the delayed harvestof soybean does not affect the growth of the postrainyseason crop chickpea and/or wheat. In drylands,farmers prefer to leave the plots fallow as sequentialcropping is risky. A study by Pandey (1986) indicatedthat in only 9 out of 29 years, the residual moisture issufficient for sequential cropping. Another study byRosenzweig and Binswanger (1993), using data over10-year period, indicated that risk-coping mechanisms(post-ante consumption smoothing mechanisms arestronger) in wealthier farmers are higher and theygenerally tend to take the risks.

The average yield of all the crops in Lateri blockexcept soybean is lower than the state average. It ispertinent to note that Madhya Pradesh is the largestsoybean growing state in India where more than5 million ha produce about 4 million t of soybean witha production of 750–1062 kg ha-1, which amplysupports the project objective of studyingsustainability of agriculture in the rainfed tropics.

On-farm Participatory TrialsMethodology

Detailed soil survey and soil analysis was carried out.Geohydrological study of soil losses and water runoffare being recorded to monitor the effect of water andsoil conservation work carried out in the watershed.Automatic weather station is commissioned to recordthe microclimatic data. Participatory approach wasadopted in water and soil conservation activity bymotivating and mobilizing people’s participation inwatershed development. Medium and small farmers,who were willing to participate in the on-farmparticipatory trials (OFPTs), were selected. The plotsize was 0.15 to 0.25 ha. All the farmers involved inthe trials were given training on the recommendedpractices and record keeping.

Government partnership

The Government of Madhya Pradesh has supportedthe management of natural resources through thewatershed development programs under the RajivGandhi Watershed mission. The major activitiescarried out under this mission were water harvesting,soil conservation, improving vegetative cover, andempowering community with emphasis on weakersection of the society.

Study on existing farming system

The information was gathered by using the followingtools to understand the existing agricultural systemincluding soil, crop husbandry, and socioeconomicaspects of the families:• Baseline survey: socioeconomic survey; and crop

production data• Group discussions• Transect walk and PRA• Secondary data collection

Constraints analysis

In the soybean-based farming system in Lalatorawatershed area, the following important constraintswere identified jointly with farmers:• Soybean cultivation has low productivity ranging

from 900 to 1200 kg ha-1.

103

• Soils were deficient in boron (B) and sulfur (S).• Waterlogged conditions prevailed many a times if

not continuously in the growing season.• Risk of erratic rainfall and insufficient moisture in

the soils.• Market fluctuation affected the price realization by

farmers.

Selection of interventions

The following interventions were selected to improveproductivity of soybean and chickpea to suit localconditions and resources:• Evaluation of best-bet options for soybean-based

systems.• Evaluation of response to micronutrients (B as

borax at 10 kg ha-1, S as gypsum at 200 kg ha-1, andcombination of both).

• Broad-bed and furrow (BBF) system (1 m broad-bed and 30 cm wide furrow) was formed withtractor-drawn implement.

• Introduction of improved seeds of soybean (JS335) and chickpea (ICCV 2, ICCV 10, and ICCC37) varieties.

• Seed treatment with Rhizobium (or phosphate-solubilizing microorganisms) culture to enhancenutrition and thiram for disease control.

• Use of nuclear polyhedrosis virus (NPV) forcontrolling Helicoverpa (pod borer) attack onchickpea.

• Water and soil conservation measures.

On-farm trials: The approach

The approach of farmer-managed trials was adopted. Thetrials were basically managed by farmers with technicaladvice provided by scientists and non-governmentalorganization (NGO) staff for planning, conducting, andmonitoring the trials. This approach was advantageous asscientists received feedback and accordingly user-friendly interventions were devised. There was no changein other agronomic practices, which are generallyadopted by the farmers. The guiding principles adoptedfor on-farm trials (OFTs) were:• Farmers were provided scientific information

about the constraints and possible solutions forincreasing productivity.

• Farmers were involved from the beginning and theOFTs were planned in partnership with thefarmers.

• Inputs for OFTs were provided at subsidized ratesinitially.

• Trained NGO staff collected records of inputs andoutputs for OFTs along with neighboring non-trialfarmers as controls.

• Farmers evaluated the OFTs in their own way.• In 1998 rabi season, one on-farm trial was

conducted with chickpea variety ICCC 37 forevaluation by farmers and seed multiplication.

• In 1999 kharif, OFTs were conducted on soybeanJS 335 along with best-bet options put together byscientists from Jawaharlal Nehru Krishi VishwaVidyalaya (JNKVV), International CropsResearch Institute for the Semi-Arid Tropics(ICRISAT), Central Research Institute for DrylandAgriculture (CRIDA), and Indian Institute of SoilScience (IISS). A multi-pronged integrated pestmanagement (IPM) strategy was also adoptedusing physical method (plowing in summer),chemical method, and biological method (use ofpheromone traps).

• In 1999/2000 rabi season, one trial was conductedon chickpea using 3 varieties (ICCV 2, ICCC 10,ICCC 37), combined with seed treatment andreduced tillage.

• In 2000 kharif, OFTs were conducted on soybeanusing improved variety JS 335, BBF,micronutrient amendments (B and S), and nutrientbudgeting.

• In 2000/01 rabi season residual effects of B and Samendments on wheat and chickpea yields wereevaluated. Nutrient budgeting trials for soybean-based systems were continued.

• In 2001 kharif the following OFTs were conductedon soybean:– Cultivation on fallow land using short-duration

soybean variety Samrat.– Micronutrient trials using B and S.– Landform trials comparing BBF.– Mixed cropping with maize.– Introduction of new crops (pigeonpea,

groundnut, and maize).

104

Performance of On-farmParticipatory TrialsDuring 1998/99 rabi season interactions with farmersat Lalatora revealed that most farmers were usingtraditional chickpea variety Katila. They were verykeen to evaluate improved chickpea varietiesdeveloped by ICRISAT. Five farmers volunteered toevaluate ICCC 37 and breeders’ seed of ICCC 37 wasprovided to these farmers. The farmers had agreed toreturn double the quantity of seeds provided as theircontribution for building seed bank at Lalatora. On anaverage, farmers harvested 20% more grain yield ofICCC 37 with maximum yield of 1680 kg ha-1 asagainst the yield of 950 kg ha-1 from the traditionalvariety (Table 4). The farmers were happy with thebold shiny seed and high yield. During 1999/2000 rabiseason more farmers came forward to evaluateimproved chickpea varieties and also reduced tillagefor chickpea crop production.

Use of improved chickpea varieties ICCV 10,ICCV 2, and ICCV 37 resulted in higher yields (957 to1471 kg ha-1) over the local variety (923 kg ha-1). Thevariety ICCV 10 was found more promising than othervarieties evaluated. The combinations of improved

variety with seed treatment and reduced tillageexhibited better production than improved varietyalone (Table 4). Profitability of improved practicesover traditional practices was higher by Rs 6421 ha-1.Improved variety with seed treatment yielded 59%higher yield than the local variety. The gross profit wasalso significantly higher.

Breeders’ seeds were provided initially to thefarmers for evaluating the performance of chickpeavarieties. The farmers were helped in roguing the oddplants in the chickpea fields under evaluation. Fourwomen self-help groups (SHGs) consisting of 40members purchased seeds and with the help of BAIFstaff stored treated seeds in government godowns untilplanting of the next rabi crop. Through this mode, thefour SHGs sold 1200 kg of quality seeds in the villageand earned the profits. In addition, individual farmersused their own seeds of improved varieties and alsoprovided on cost to farmers of nearby watersheds.

Evaluation of Best-bet OptionDuring 1999 rainy season, the scientists from JNKVV,CRIDA, and ICRISAT put together a best-betcombination option for soybean-based systems. This

Table 4. Performance of farmer participatory evaluation of chickpea varieties at Lalatora1.

No. of Area covered Yield Increase (%)Variety farmers (ha) (kg ha-1) over control

1998ICCC 37 5 1.8 Av. 1150

Max. 16801999ICCV 10 14 3.75 Av. 1471 59

Max. 2500

ICCV 2 5 1.25 Av. 1280 38Max. 1600

ICCC 37 25 6.25 Av. 957 4Max. 1700

Control 14 4 Av. 923(traditional farms)1. Interventions included improved variety, seed treatment, non-nodulating seed, and reduced tillage; no intervention in control.

105

option consisted of use of improved variety ofsoybean (JS 335), seed treatment with thiram alongwith Rhizobium and phosphate solubilizingmicroorganisms, application of diammoniumphosphate (DAP) at 50 kg ha-1, and IPM. In the firstyear, 27 farmers evaluated this option for soybeancovering 40 ha (Table 5). Average increase in soybeanyield was 34% above the baseline/control plot yield of950 kg ha-1. The range of soybean yields with best-betoption varied from 0.81 to 1.06 t ha-1. Detailedanalysis of benefit-cost ratio for the farmers whoevaluated this option was worked out and the net profitwas Rs 5575 ha-1 (Vadivelu et al. 2001).

Response to micronutrient amendments

In 1999 season, soil samples from Lalatora werecollected and analyzed. The analysis revealed thatthese soils were deficient in B and S. These resultswere shared with the farmers and the implicationswere explained. Remedial measures were alsosuggested. During 2000 rainy season 13 farmers cameforward to evaluate effects of B, S, and B+Samendments (10 kg borax = 1 kg B ha-1, 200 kggypsum = 30 kg S ha-1, and 1 kg B + 30 kg S ha-1). Thefarmers evaluated these treatments in strips,monitored the plant growth during the season, andharvested separately.

The results showed that B, S, and B+S treatmentssignificantly increased soybean yields (1710 to 1780

kg ha-1) and the yields were 34–39% higher than thebest-bet option treatment (Table 5), which served ascontrol treatment without B, S, and B+S amendments.Farmers found these trials educative and they werehappy to harvest increased yields of soybean, whichalso resulted in increased income.

During the postrainy season these plots weremaintained and six farmers planted wheat followingnormal practices to evaluate the residual benefits of B,S, and B+S amendments. Residual benefits of B and Samendments for soybean in kharif season increasedwheat yields by 30 to 39% over the untreated controlplot yields. Total systems productivity increased by 63to 73% due to single dose of 1 kg B and 30 kg Sapplication. The economic analysis of these trialsshowed that farmers’ net incomes increased by Rs8190 to 8850 ha-1 due to B and S amendments,respectively.

The economic analysis of the OFT in 2001 showedthat intervention of combined application of B and Sgave maximum benefit amounting to Rs 26454,followed by only B (Rs 26609) and S application (Rs25955). All these three interventions proved to bebeneficial to the farmers with 1.8 benefit-cost ratio ascompared to control with traditional practices (1.3)and gave almost 49% higher benefits to the farmers.These amendments not only increased the incomes,but also improved the water-use efficiency throughincreased productivity. The farmers were impressedwith these results. Besides Lalatora farmers, other

Table 5. Performance of soybean variety in on-farm trials at Lalatora.

No. of Area covered Yield Increase (%) overVariety farmers (ha) (kg ha-1) Interventions control farm

1999JS 335 27 40.5 1275 Improved seed; 34

seed treatment andapplication of DAP

Control 14 6 950 No intervention(traditional)

2000JS 335 13 6 1730 Boron application 35JS 335 13 6 1710 Sulfur application 34JS 335 13 6 1780 Micronutrients 39JS 335 6 3 1500 Broad-bed and furrow 17Control (JS 335) 13 6 1280 Best-bet option

106

farmers from the neighboring watersheds ofAnandpur, Lateri, Khairkhali, and Mahauti alsoindented in advance their B and S requirements withthe BAIF staff.

During 2001 kharif season, 12 farmers furthercontinued evaluation of B and S amendments on thesame plots. The results revealed that the response to Band S amendments was 10 to 16%, which was almosthalf that of the response observed in 2000 kharifseason. The probable reasons for the reduced benefitsdue to B and S amendments could be: (i) occurrence of45 days drought (dry spell) during reproductive stageof soybean, (ii) application of B and S after sowing ofthe crop due to continued rains in the beginning ofseason, and (iii) improved availability of B and S fromthe previous amendments.

Evaluation of broad-bed and furrowlandform treatment

Three farmers evaluated BBF landform for soybeanproduction and observed 17% increase in soybeanyields over their normal practice. Considering the factthat the black soils of this region are prone towaterlogging, BBF technology was suggested forevaluation. The BBF practice proved better (1500 kgha-1) than the traditional cultivation practices of thefarmers (Table 5), which involved planting on flatsurface. There was initial resistance to BBF due tosigns of low and slow germination as a result of poorsowing practices. This lacuna has been rectified withprovision of modified seed drills for BBF operationsand adequate field training to the farmers. With furtherimprovements in planting on BBF, during 2001 kharifseason, 11 farmers evaluated BBF landform treatment.During the drought season also, BBF increasedsoybean yields by 16% over and above the yields ofbest-bet option. However, farmers feel and perceivethat they lose some land in furrows and more croplines could be sown in the field. Availability ofequipment for BBF preparations and planting areperceived as constraints by the farmers.

Technology ExchangeUltimate success of any technology depends on itsacceptance by the farmers. If the technology is noteconomically viable, socially acceptable, and not

user-friendly, it would be rejected by the farmers. Thiswas ascertained by the feedback received from thefarmers. Based on the encouraging demand-drivenexperience, extension of technology disseminationwas initiated by organizing:• field days, field exposure visits, and scientist visits;• increasing awareness through enhanced interactions

with extension officials;• promoting seed banks to ensure timely seed supply

through SHGs (4 SHGs procured and supplied1.2 t seeds to 12 farmers, in addition to projectsupply and farmers’ own seeds);

• field demonstrations;• providing literature about improved options;• linkages with other agencies (swayam sidha,

market access); and• government organization, NGO, and farmer visits.

These activities have proved to be very effective indissemination of the technologies. The impact is clearlyseen by increased demand for improved varieties inlarge numbers from non-operational project areas.Another interesting feature was attracting financialsupport from the Government of Madhya Pradesh forundertaking physical structures for water conservation.

Adoption and Farmers’ Perceptionson InterventionsAdoption status and farmers’ opinion on variousinterventions carried out during the project arereported in Table 6. The most important constraint ofwaterlogging in the rainy season requires adequatedrainage systems. Eighty-one per cent of therespondents categorize the adoption as ‘partial’. Mostof the innovations that were identified as non-adoptionare either due to lack of technical knowledge or theexpensive nature of the operation. One hypothesis isthat the problem lies more in the nature of theperceived higher cost, which the farmer is not willingto invest. This calls for a properly designed programwhich will provide sufficient subsidy with reasonablecontribution from the farmers. Farmer interest in theinterventions is given below:• All farmers appreciated:

– Introduction of new crop variety;– Seed treatment;– Use of biofertilizer.

107

• Most farmers appreciated:– Mixed cropping;– Micronutrient application.

• The practice of reduced tillage was appreciated byless than half of the participating farmers.

• BBF landform was appreciated by least number offarmers.

Summary and ConclusionThe major constraint in the watershed area is lowcropping intensity, as majority of the dryland is leftfallow during the rainy season. Waterlogging and soilerosion are the other constraints. Lack of initiative bythe farmers is due to their perception that higher costsare involved in undertaking these investments andthey expect the Government to take a lead role. Theyields of most of the crops are lower than the stateaverage, except soybean. In general the yields are lessthan 1 t ha-1; however, there is huge variation in theyields over years. The education level is poorparticularly in the women’s group. The infrastructuralfacilities in terms of electricity, roads, andtelecommunications are also poor.

Farmers were happy to evaluate improved options,which increased their farm incomes. Along with theimproved crops/varieties, farmers appreciated muchthe technologies to harvest rainwater through farmponds, best-bet options, amendments with selectednutrients, use of NPV, and the concepts of village-based seed banks and vermicomposting.

All the participating farmers opined that theintroduction of improved varieties of chickpea and theimproved practices have given higher yields than localpractices and local variety and they were satisfied with

the interventions. All the interventions suggested inthe project were user-friendly, except BBF in whichthe availability of implement and skill to use were theconstraints encountered by the farmers. Availability ofthe improved variety, quality seeds, and credit were otherconstraints faced by farmers. Work load and itsdistribution should be analyzed further as some practicesmay involve more women participation or vice versa.

AcknowledgmentsWe acknowledge the help of the scientists for theirtechnical support, and the BAIF staff, particularlyMr Somnath Roy and Mr R K Verma, who helped 12farmers to undertake evaluations. Our sincere thanksare due to the farmers who actively took part in OFPTs.The financial support from the Asian DevelopmentBank (ADB) is gratefully acknowledged.

References

NBSSLUP. 2000. Detailed soil survey of a part of Lalatorawatershed. Nagpur 440 010, Maharashtra, India:NBSSLUP.

Pandey, S. 1986. Economics of water harvesting andsupplementary irrigation in the semi-arid tropical region inIndia: a systems approach. Armidale, Australia: Universityof New England. 312 pp.

Rangnekar, D.V. 1999. Lateri – Another challenge forBAIF for sustainable, integrated rural development. Pages62–69 in Improving management of natural resources forsustainable rainfed agriculture in India. Proceedings of thepre-launch workshop, 16–18 Dec 1998, ICRISAT Center,Patancheru, India. Patancheru 502 324, Andhra Pradesh,India: International Crops Research Institute for the Semi-Arid Tropics.

Table 6. Adoption of soil and water conservation (SWC) measures.

Adoption status1 Causes for non-adoption1

SWC Not Partly Fully Technical Verymeasures adopted adopted adopted Ignorance constraints expensive Inconvenient

Land leveling 2.08 91.67 6.25 NR2 45.37 46.15 8.48Waterway 9.38 81.24 9.38 NR 45.37 54.63 NRFarm pond 34.49 51.72 13.79 NR 47.12 52.88 NR1. Data indicate respondents (%).2. NR = No reply.

108

Rosenzweig and Binswanger. 1993. Wealth, weather riskand the composition and profitability of agriculturalinvestments. Economic Journal 103:56–78.

Vadivelu, G.A., Wani, S.P., Bhole, L.M., Pathak, P., andPande, A.B. 2001. An empirical analysis of the

relationship between land size, ownership, andproductivity – new evidence from the semi-arid tropicalregion in Madhya Pradesh, India. Natural ResourceManagement Program Report no. 4. Patancheru 502 324,Andhra Padesh, India: International Crops ResearchInstitute for the Semi-Arid Tropics. 50 pp.

109

Efficient Management of Natural Resources: A Way to Sustain FoodSecurity in the Rainfed Sloping Lands of Northern Vietnam

T D Long1, A Ramakrishna2, H M Tam1, N V Viet1, and N T Chinh1

Abstract

Rainfed sloping lands occupy one-third area of northern Vietnam and are threatened with destruction ofnatural resource base due to improper land use practices. Current production practices are exacerbatingsoil loss and destruction of the natural habitat as the soils are deeply weathered, poor in nutrients, andhighly vulnerable to erosion. These ecosystems have much lower carrying capacity and respond to cropintensification by rapid decline in productivity, even total collapse if not managed properly.

Remoteness and inaccessibility, low biological productivity, environmental degradation, disease andhealth problems, population increase, and lack of a development paradigm tailored to the specialconditions are the key constraints for development. Sustainable farming on these lands in the perspectiveof a seriously deteriorated ecology and environment is not an easy task. Proper understanding ofconstraints and development of appropriate technologies with focus on soil, water, and nutrientmanagement help optimize food production and combat resource degradation. Research and applicationof watershed based integrated natural resource management technologies offered excellent opportunitiesfor crop diversification to meet market orientation, sustaining food production at higher levels, improvingsoil health, recharge of aquifers, and enhancing household incomes for better rural livelihoods in thesloping land ecoregions of northern Vietnam.

In Vietnam, uplands cover three-fourth of the territoryand shelter one-third of the population (28 out of 84million people) of the entire nation. The uplands are afragile environment characterized by sloping landsthat are prone to erosion, with low natural soil fertilityand declining forest cover. They are currentlythreatened with ecological degradation, which isalready severe in some areas. Substantial areas ofcultivated land have been seriously affected by soilerosion and land degradation, resulting from improperland use practices. More than 11 million ha (33%) isbarren land as a result of deforestation andinappropriate land use.

As increasing population expand to steeper, morefragile areas in the uplands, more catchments areaffected by severe soil erosion, declining soilproductivity, and environmental degradation.Watershed land degradation now poses a threat to theeconomies of Vietnam, and to the livelihoods of the

ever-growing population that depend on theseresources.

On-site soil loss reduces soil fertility in terms ofchemical, physical, and biological degradation. Thesesoil changes will in due course reduce crop yields andhence income and household food security. The off-site effects of soil erosion often have broadereconomic and environmental implications includingsedimentation, flooding, and reduced water qualityresulting in poor living conditions of the people.

The Integrated Watershed Development Program,a new paradigm for research, has been promoted underthe Asian Development Bank (ADB)/InternationalCrops Research Institute for the Semi-Arid Tropics(ICRISAT) initiative since 1999 to address the aboveconstraints. This program focuses on:• Simultaneous development of land, water, and

biomass resources in the light of the symbioticrelationship among them.

1. Vietnam Agricultural Science Institute (VASI), Hanoi, Vietnam.2. International Crops Research Institute for the Semi-Arid Tropics (ICRISAT), Patancheru 502 324, Andhra Pradesh, India.

110

• Integrated farming systems approach.• Meeting food, fodder, and fuel requirements of the

human and livestock population that depend onthese resources.

• Ensuring environmental sustainability along witheconomic viability by promoting low-costtechnologies.

• Improving land productivity by promotingimproved agronomic practices and input use.

• Releasing population pressure on land by creatingnon-farm employment.

• Development of local institutions for futuremanagement through participatory approach.The central thrust of our research is to enhance

productivity of land and water resources on the basisof a scientifically defined watershed that connotes ageographical unit rather than economic administrativeunits (like household or village). It is also ensured thatwhole range of stakeholders, from land users to policymakers, are involved in the generation and promotionof improved land use practices.

The approach we have taken was to ensuremaximum participation of farmers in planning andexecution of all our activities. All the watershedinterventions, viz., introduction of new crops andcropping systems, soil and water conservation,integrated nutrient management (INM), integratedpest management (IPM), etc. are thoroughly discussedand decided by the farmers. Researchers andextension workers aid in decision-making process andfacilitate agreed activities by providing technicalsupport.

Micro-watershed is used as a demonstration blockfor appreciating the benefits in terms of reduced runoffand soil loss through scientific measurements.Farmers in rest of the watershed are evaluatingimproved soil, water, and nutrient managementoptions and cropping systems along with IPM andintegrated disease management (IDM) for efficientuse of natural resources and sustainable productivitygains. Studies on nutrient budgeting and response tomicronutrients are being conducted with closecooperation and involvement of farmers.

The partnership research at the benchmarkwatershed was conducted under three sub-projects andthe results are presented.

Sub-project 1: Socioeconomic Surveys

Purpose

Target the group of farmers for exchange of improvedwatershed natural resources management (NRM)technologies and monitor the impact of interventionsmade in the watershed.

Objectives

• Conduct surveys and enlist present practices.• Quantify economic benefit and household income

due to improved technologies.• Assess adoption pattern and constraints to the

adoption of watershed technologies.

Survey

Secondary data and survey of the target zone providebasis for the choice of representative “benchmarkresearch location”. More detailed information on thebenchmark area will be needed in order to defineresearch priorities and can be collected by means ofrapid rural appraisal (RRA) or participatory ruralappraisal (PRA) by the on-farm research team. Thesurvey consisting of direct observations and interviewsbring to life the problems faced as well as theopportunities, which exist for improvement. The RRAor PRA survey technique is a good tool for developingthe necessary insights into how integrated systemoperates. The technique lays increased emphasis onfarmer participation in the collection and interpretationof information over the diagnostic surveys.

We conducted a socioeconomic survey using aschedule carved out from the schedules used by theAsian Grain Legumes On-farm Research (AGLOR)and the Jawaharlal Nehru Krishi Vishwa Vidyalaya(JNKVV), India. The survey covered the watershed asa whole for general description of macro-economics,population, infrastructure, institutions, and otheraspects. But most of the work concentrated at villageand household levels. The main areas covered were:• Demography, manpower, skills, and education.• Rural sociology with emphasis on traditions and

values, constraints to development, land tenure,participation, power structure, and leadership.

111

• Agro-economy: commodities, yields, farmingsystems, inputs, technology, income, and cash flow.

• Household survey of selected target groups.• Institutions and basic social services.

Information on physical (rainfall, temperature,solar radiation, topography, and soil) and biological(natural vegetation, pests, and diseases) elements wereobtained to determine what crops can be grown in anarea, given a suitable human environment (economic,institutional, and social elements). A checklist to keeptrack of the topics for discussion and exploration wasdeveloped. For recording physical information onindividual fields, a simple data sheet was used.

Natural and socioeconomic conditions

Out of 1522 ha, although 53% area is suitable foragricultural purposes only 28% is being cultivated.Again most of these lands were brought recently underarable cropping. The average family size is small with58% of the population in the age group of 17 to 55years. Since majority of the population is young andengaged in agricultural production, adoption of labor-intensive new production technologies and farmingsystems should not pose any problem.

Cropping patterns and land use

Major crops in terms of cropped area are maize (83%),sugarcane (8%), legumes (13%), and watermelon(6%). Groundnut was grown in the past but is notcultivated now due to severe problem of pod rot.Cereal monocropping (maize-maize) is predominantand occupies 77% of the cultivated area followed bywatermelon-maize cropping system (11%). Cereal-

legume cropping is practiced in only 2–3% cultivatedarea.

Input usage

High quantity of inorganic fertilizers is used (Table 1).Usage of organic manure (39–46 t ha-1) is limited towatermelon and sugarcane crops. Insecticide usage islimited to sugarcane alone.

Crop yields

The average yields are low to moderate (maize 0.9–7 tha-1; watermelon 10–36 t ha-1; and mung bean 0.3–1.2t ha-1) with low benefit-cost ratio (maize 0.4,watermelon 0.7, and mung bean 0.9) (Table 2).Discussions with the farmers revealed that productionpotential is high if appropriate crops and productiontechnologies are used. Improved seed and culturalpractices are being adopted only in maize.

Constraints to production

The survey has brought out the following importantconstraints faced by the farmers in the benchmarklandscape watershed.

Farmer perceived

• Lack of water for crop intensification (97.9%).• Unavailability of credit and complicated loan

procedures (91.8%).• Fertilizers are expensive (83.7%).

Table 1. Input usage in various crops.

Particulars Maize Watermelon Sugarcane Mung bean Cowpea Rice

Seed1 (kg ha-1) 23 1 - 22 23 100Urea (kg ha-1) 444 561 670 12 Nil 220Super phosphate (kg ha-1) 525 579 554 500 500 500Murate of potash (kg ha-1) 136 127 1467 Nil Nil 85Manure (t ha-1) Nil 46 39 Nil Nil 10Labor (person-days) 198 552 414 190 215 2001. Seed price (VND per kg): Maize 18100, watermelon 554700, mung bean 11180, cowpea 14000, and rice 2500 (US$ 1 = 14000 VND).

Data for sugarcane not available.

112

• Lack of capital to purchase inputs (80%).• Lack of knowledge on plant protection and

improved production practices (79.6%).• Monopoly of market forces (75.5%).• Non-availability of market facilities (71.4%).• Lack of extension services and demonstration of

new technologies (71.4%).• Unavailability of farmyard manure (FYM) (67.3%).

Researcher perceived

• Soil erosion.• Inappropriate soil, water, and nutrient management

practices.• Improper land use planning.• Natural resource base degradation.

Constraints and opportunities

We examined the constraints (in the farming systemsand the environment) that limit the systemsproductivity and made an attempt to focus onopportunities that increase systems productivity. Anumber of specific challenges were identified thatneed to be addressed for development to be carried outsuccessfully in the sloping ecoregions of northernVietnam. A distinction was made between theconstraints that in principle can be addressed directlyby the research team (‘addressable’) and those thatcannot (‘non-addressable’). A priority list ofconstraints and opportunities identified is provided.

Constraints

• Physical constraints: broken terrain, steep slopes,and poor soils.

• Environmental constraints: deforestation, landdegradation, moisture stress during critical stagesof crop growth, and low biological productivity.

• Infrastructure constraints: inadequate communica-tion, transportation, and production infrastructure;and unskilled agricultural force.

• Economic constraints: subsistence orientation; andinadequate development of market and trade.

• Cultural constraints: low levels of education andknowledge; and persistence of traditional patternof behavior.

• Intellectual constraints: inadequate scientificknowledge of the sloping land ecoregions; andlack of suitable strategies to guide developmentand planning.

Opportunities

• The benchmark watershed has good potential forintroduction of new crops and cropping systemssince the current cropping systems are givingmeager income and mining the soil fertility withassociated erosion of natural resource base.

• Identification and/or introduction of appropriatetechnologies with focus on soil, water, and nutrientmanagement at micro-level in a watershed contextwill help in optimizing food production andarresting further erosion of natural resource base.

• Farmers are currently relying on high doses ofinorganic fertilizers with little or no application oforganic fertilizers. Good scope exists forintroduction of appropriate INM practices.

• Most farmers are unaware of improved productiontechnologies. There is a need to demonstrate newcrops/cultivars, integrated pest and diseasemanagement technologies, improved crop produc-

Table 2. Yield and output of crops grown in Thanh Ha State Farm, Vietnam.

Yield (t ha-1) OutputPrice1

Crop Range Average (VND kg-1) Average (VND ha-1) CV (%)

Maize 0.9–7.0 3.4 1742 5923176 46.4Watermelon 10.0–36.0 17.8 1582 28166666 48.0Sugarcane 20.0–83.0 58.3 134 7839999 49.3Mung bean 0.3–1.2 0.7 7600 5320000 21.8Cowpea 0.6–1.2 0.8 5400 4320000 27.5Rice 3.0–6.1 3.2 1900 6080000 15.61. US$ 1 = 14000 VND.

113

tion practices, and systems diversification forhigher productivity and household incomes.

• A paradigm tailored to the special conditions of thesloping land ecoregions needs to be developed.Farmers themselves are strongly aware of some of

the constraints, while the team members perceivedother constraints as important. The decision on whichconstraints to tackle first may be influenced by thisdifference in perception. For example, the researchersconsidered soil erosion hazard as the number oneproblem, while farmers did not regard it as being quiteserious. Erosion hazard may be seen as a ‘strategic’problem, i.e., one which is likely to increase in thefuture unless measures are taken immediately. Tobuild up credibility, the team however, decided to firstaddress those constraints, which the farmers considerurgent, even if they are not most important inresearchers’ point of view.

From constraints to solutions

The following process was followed for analysis ofconstraints and project planning:1. Analyze the causes underlying the major constraints.2. Examine whether there is sufficient evidence for

these causes, if not take up diagnostic research tofind answers.

3. Analyze whether a constraint or cause can betackled directly by on-farm testing with availabletechnology or whether technology must bedeveloped.

4. Choose specific, well-defined technologies for on-farm testing.Choosing the most appropriate technology always

requires a good knowledge of both the target systemand range of available technological options.Knowledge of the target system and the farmingenvironment was obtained from the diagnostic surveyand through collection of information. Knowledgeabout the technology was obtained by means ofsystematic search for information from experts,literature, and existing databases. The examples ofgroundnut and soil fertility are given in Table 3.

Farmers’ involvement in the choice ofinnovations

The research team, after carrying out the ex anteanalysis of possible innovations, met the cooperatingfarmers and discussed the proposed innovations andsolicited farmers’ inputs. The average landholding inVietnam is very small (1000 m2 upland or 600 m2 ricefield) and production losses if any due to improperpractices advocated need to be compensated. The

Table 3. Prioritization of constraints, likely causes, and research activity in Thanh Ha State Farm, Vietnam.

Technology testing AdditionalConstraint Cause On-farm On-station diagnostic studies

Failure of High disease Introduction of high- Screening of Quantify fungusgroundnut pressure. yielding, disease resistant potential buildup and diseasedue to pod rot cultivars. cultivars. relationships.

Introduction of appropriate Identify hot spotsIPM technologies. and abandon fungus-

infested fields.

Declining Continuous maize Integration of legumes. Screening Characterization ofsoil fertility monocropping. potential soil resource.and crop legumes.productivity

Reduced fallow Introduction of integrated Seed multipli-and soil erosion. land, water, and nutrient cation.

management technologies.

114

approach adopted therefore, is to encourage maximumparticipation of farmers in planning and execution ofall the activities. All the watershed interventions, viz.,introduction of new crops and cropping systems, soil andwater conservation, INM, IPM, etc. are thoroughlydiscussed and decided by the farmers. Researchers andextension workers aid in decision-making process andfacilitate agreed activities by providing technical support.

Micro-watershed is used as a demonstration blockfor appreciating the benefits in terms of reduced runoffand soil loss through scientific measurements.Farmers in rest of the watershed evaluate improvedsoil, water, and nutrient management options andcropping systems along with IPM and IDM forefficient use of natural resources and sustainableproductivity gains. Studies on nutrient budgeting andmicronutrient requirements for different systems areunderway with close cooperation and involvement offarmers.

Conclusions

The socioeconomic surveys helped us to identify thebenchmark research location, which is therepresentative of a well-defined target zone. Targetzones were delineated within the project’s mandatedregion on the basis of similarities in climate, soilclasses, dominant cropping systems, etc. Similarzones would be expected to face similar constraints toagricultural production, and to have similaropportunities to overcome them. The workinghypothesis is that the performance of innovations willbe similar across the target zone, and the chances thatfarmers will then adopt them will also be similar.

Sub-project II: Ecoregional Databases

Purpose

Assemble all available databases on soil, climate, crops,and input use for the targeted rainfed production systemapplicable in the ecoregion of northern Vietnam.

Objective

• Study spatial distribution of the constraints,analyze yield gaps, and examine opportunities forcrop intensification in the target ecoregion(s).

Characterization of environment

Climate

The climate in the landscape watershed is monsoonalwith hot, wet summers (April–August) and cool,cloudy, moist winters (December–February). The totalannual rainfall is 1500–2000 mm. The averagetemperature is 25°C, with an average maximum of35°C (in August) and an average minimum of 12°C (inJanuary). The southwest monsoon occurs from May toOctober, bringing heavy rainfall. Temperatures arehigh. November to May is the dry season with a periodof prolonged cloudiness, high humidity, and light rain.Weekly weather data for the past 15 years from thenearby weather station was collected (Fig. 1). All themeteorological parameters are being collected atregular intervals. Rainfall variability is measured withautomatic rain gauge located in the watershed.

All available databases on soil, climate, crops, andinput use for the targeted rainfed production of

Figure 1. Climate of Hoa Binh Province, 1984–98.

115

northern Vietnam (7 provinces) were assembled tostudy spatial distribution of the constraints, analyzeyield gaps, examine opportunities for cropintensification, and scale-up improved integratedwatershed technologies.

Soils

The soil was sampled up to a depth of 1.5 m fordetailed biological, chemical, and physicalcharacterization and also based on the toposequence.Some salient observations are:• Soil is medium loamy in texture, acidic in nature

with very poor organic matter, medium inpotassium, and very low in phosphorus (P) content.

• Soils need organic and inorganic supplements andparticularly P fertilizer for good productivity. It isbetter to use thermo phosphate than superphosphate in these soils.

• Soil is rich in microbial population with largebiodiversity and has good ability to developbiological activities with cultivation.

Yield gap analysis

We undertook the yield gap analyses for importantcrops grown in the target ecoregions of northernVietnam using conventional approach of yieldoptimization trials conducted on research stations bycomparing with the farmers’ field plot yields. Data fromexperiment station and fields of 10 farmers each in 5districts of a province for three years (1997–99) in fiveprovinces for maize, groundnut, and soybean werecollected. Yield gap analysis was undertaken for eachprovince by computing the difference between averagefarmers’ yields from 5 districts in a province and thecrop yield from the experiment station in that province.The results showed that for maize mean yield fromstation trials was 4.9 t ha-1 and farmers’ average yieldwas 2.8 t ha-1 indicating a yield gap of 2.1 t ha-1 (Table4). The yield gap for maize in 5 provinces variedbetween 1.7 and 2.5 t ha-1. The average yield gap was2.1 t ha-1 and it varied between 1.1 and 3.6 t ha-1 indifferent provinces. The coefficient of variation forgroundnut yield was 41.7% which indicated largevariation in yield gap within the provinces. For soybeanmean yield gap was 1.8 t ha-1 and within the provinces itvaried from 1.6 to 2.0 t ha-1 with a coefficient of

variation of 7.7%. Daily climate data for the period1982 to 1997 was collected from Phu Tho, Vinh Phuc,Hoa Binh, Nam Ha, Nam Dinh, Ninh Binh, and Ha Tayprovinces to undertake yield gap analysis using theCROPGRO model and to examine the opportunities forcrop intensification in the target ecoregion.

Sub-project III: On-farm ResearchPurpose

Test, validate, and evaluate suitable natural resourcemanagement technologies in partnership with thefarmers in farmers’ fields in the benchmark watershed.

Objectives

• Introduce improved soil, water, nutrient, and pestmanagement technologies for sustained increasesin agricultural productivity.

• Reduce soil degradation and increase rainwateruse efficiency through increased soil moisture,runoff water harvesting, and increasedgroundwater recharging.

• Evaluate suitable cropping systems based on theagroecological potential of the region to increasefarm income.

Soil and water conservationWe have undertaken various measures for increasedwater and soil conservation in the benchmarkwatershed such as:• Landform treatments (ridge and furrow, contour

planting).

Table 4. Maize yield gap between experimentalplot yields and farmers’ field yields.

Experimental Farmers’ YieldProvince yield (t ha-1) yield (t ha-1) gap (t ha-1)

Hoa Binh 4.7 2.8 1.9Ha Tay 5.4 3.1 2.3Phu Tho 4.2 2.5 1.7Vinh Phuc 5.5 3.0 2.5Ninh Binh 4.6 2.8 1.8Mean 4.9 2.8 2.1SD 0.5 0.2 0.3CV (%) 10.2 7.3 15.1

116

• Grassed waterways and drainage channels.• Field bunding.• Biological and mechanical barriers across the

slope on contours.• Trenches and silt traps to reduce the rainwater

velocity and increase opportunity time forinfiltration.

• Percolation tanks (40 m3 capacity) to store excesswater.

• Planting of Gliricidia sepium on the propertybunds and contours.

Groundwater monitoring

Groundwater level in 10 open wells (8 inside and 2outside the watershed) on the toposequence (top,middle, and lower part of the landscape watershed)was monitored at fortnightly intervals to observe waterfluctuations and water yield to quantify the influenceof improved soil and water conservation practicesundertaken in the benchmark watershed. There wasabout 2.5–3 m increase in the water level in thebenchmark watershed wells compared to those outsidethe watershed (Fig. 2). In addition, groundwater levelfluctuations were less pronounced with stable wateryield particularly in the dry season.

Moisture conservation

In northern Vietnam, the important productionconstraints for the groundnut crop are low temperatureat maturity in autumn-winter, and low temperature atgermination and moisture stress at maturity in springseason. Trials were therefore, initiated to evaluate theeffect of straw and polyethylene mulch on soilmoisture, temperature, and pod yields of groundnut.Polyethylene mulch increased soil temperature by 2–3°C in autumn-winter and 1–2°C in spring at 10 cmdepth with associated conservation of soil moisture inthe entire soil profile (Fig. 3). Increase in soiltemperature in spring promoted early (about 2–3 days)and better germination with good seedling vigor whilein winter, good pod development and early maturitywas noticed.

Application of polyethylene mulch resulted indoubling the groundnut yield (1.5 t ha-1) than thecontrol (0.7 t ha-1) treatment in autumn-winter seasonin 2000. The straw mulch treatment, which is

environment-friendly and also economical, increasedgroundnut yields by 71% over the non-mulch controltreatment (1.2 t ha-1). Both the mulch treatmentsincreased number of pods plant-1, pod weight, andbiomass. However, in spring 2001, significantlyhigher yields (3.23 t ha-1) were recorded inpolyethylene mulch treatment than in control (2.74 tha-1). The beneficial effects of straw mulch appearedto be masked by the increased incidence of fungaldisease.

Land degradation

To quantify the effect of land degradation in terms ofreduced productivity, we studied the effect of fieldlocation on a toposequence in the watershed on cropproductivity. Soil samples up to a depth of 105 cmwere collected for detailed biophysical and chemicalcharacterization for identifying the suitable indicatorsof land degradation.

Soil biological activity parameters such asmicrobial biomass, soil respiration, dehydrogenase,alkaline and acid-phosphatase activities are the directmeasures that indicate the soil health. These biologicalproperties are directly associated with transformationsof various elements in soil which are needed for plantgrowth. Soil biological parameters varied significantlyon a toposequence. Biomass carbon (C) andrespiration values for top 10 cm samples from top ofthe toposequence were similar to the values of the 10–20 cm samples from the middle of toposequence.Detailed analysis of parameters will enable us to relatethe indicators of soil degradation with the cropproductivity for estimating the productivity losses dueto degradation.

Variation in biological soil quality attributes alongthe toposequence and soil depths were studied indetail (Table 5). The results indicated a widevariation for all the parameters along the location ona toposequence. The organic C content was high(8517–9633 mg C kg-1 soil). Similarly, soilrespiration also showed the same trend. Further,analysis of results reveal that the samples from top ofthe toposequence showed more soil C, microbialbiomass C and nitrogen (N), and respiration than thesamples from middle and lower positions on atoposequence. These results point out that as thefarmers grow fruit trees on top of the toposequence

117

Figure 3. Changes in soil temperature and moisture in mulch and normal cultivation of groundnut.

Figure 2. Groundwater level in the open wells along the toposequence in the benchmark watershed.(Note: 0 = Ground level.)

118

the soils were not degraded on top, whereas, theagricultural systems followed on middle and lowerpositions of a toposequence which are cultivatedhave caused degradation. The results also suggest adirect relationship between rainfall and soil organicC content. Further analyses of these data sets willreveal detailed understanding of relationshipsbetween environmental management factors and landdegradation.

Runoff and soil loss

The two micro-watersheds were equipped withdigital recorders to monitor runoff and sedimentsamplers to measure soil loss and nutrient loss inrunoff as well as in soil sediment. In 2000, annualrainfall of 1349 mm and runoff of 29.5% rainfall wasrecorded. Total soil loss from the developedwatershed in 2000 was 6.8 t ha-1.

Nutrient managementImproved nutrient management in maize

Farmers in the benchmark watershed over the yearshave increased the quantity of nitrogenous fertilizer(600 to 750 kg urea ha-1) in maize crop to maintain theyields. This has resulted in increased incidence ofpests and diseases and decline in household incomes.Improved nutrient management practice (180 N:90P2O5:90 K2O; 10 t FYM; and 400 kg lime) wascompared with farmers’ practice (275–300 N:80P2O5:45 K2O) in maize to wean the farmers away fromhigh dependence on inorganic fertilizers, encouragebalanced fertilization, and reduce cost of cultivation.Higher grain yields were obtained with improvedpractice in all the three years and the results haveclearly indicated good scope for savings of 95 to 120kg ha-1 of N fertilizer (Table 6).

Evaluation of micronutrient requirements

Soil analysis indicated that the soils in the benchmarkwatershed are deficient in micronutrients like boron,zinc, sulfur, and molybdenum. Trials were initiated ongroundnut and soybean to quantify advantages ofmicronutrient application with and without Rhizobiuminoculation. Micronutrient application resulted in27% higher pod yield over farmers’ practice (2.75 tha-1). The results, however, indicated limited scope forreduction of N fertilizer and urgent need to identifyappropriate Rhizobium strains and/or effectiveapplication methods for added advantage.

Green manure, compost, and mulching

Farmers have planted 40,000 Gliricidia saplings onfarm bunds and near the mechanical structures in thebenchmark watershed. The growth was very fast with

Table 5. Variation in soil biological propertiesalong the toposequence at 0–105 cm soil depth.

Soil property Lower Middle Top

Microbial biomass C 108 112 125(mg kg-1)

Microbial biomass N 11 10 16(mg kg 1)

Mineral N 19 18 12(mg kg-1)

Net N mineralization 9 8 10(mg kg-1 soil 10d-1)

Organic C 8517 8233 9633(mg kg-1)

Table 6. Influence of improved nutrient management practices on grain and biomass yields (t ha-1) of maize.

1999 Autumn 2000 Spring 2000 Autumn 2001 Spring

Practice Grain Biomass Grain Biomass Grain Biomass Grain Biomass

Farmers’ practice 4.50 8.84 5.32 11.80 4.90 8.80 5.57 9.58Improved practice 4.81 9.67 5.70 11.85 5.37 11.82 6.62 11.26 SE± 0.12 0.20 0.13 0.16 0.12 0.21 0.25 0.28 CV (%) 9.00 17.00 10.00 5.70 9.80 10.70 5.00 9.00

119

ample biomass production. In Vietnam, Gliricidia canbe cut 5 times in a year with an estimated biomassproduction of 25–50 t ha-1. This rich organic mattercontaining 3–5% N can meet 100 to 200 kg Nrequirement of crops when applied in situ.

Due to limited turn around time, farmers in ThanhHa watershed used to burn maize straw after springseason to get the fields quickly prepared for autumn-winter cropping. Demonstrations were held todiscourage straw burning and farmers were giventraining on minimum cultivation, mulching, in situ soilincorporation, and composting.

Integrated pest and disease management

Integrated pest and disease management studies wereconducted during 1999 to 2001 with 7 farmer groupsin the benchmark watershed. Regular monitoring ofdisease and insect buildup, appropriate selection ofvariety (resistant), seed treatment, cultural practices(lime application, land configuration, timely sowing,Rhizobium inoculation, etc.), and need-basedchemical application are some of the measures thatwere followed and compared with farmers’ practice toquantify the benefits and economic suitability.

Bacterial wilt, collar rot, and root rot are importantdiseases in groundnut, while stem borers in maize andMaruca testulalis and Helicoverpa armigera in mungbean and soybean appear to be most devastating pests.Manganese toxicity (appears like virus infection) iswidespread in spring season. Mulching reducedsoilborne fungi significantly but led to increasedincidence of leaf spots due to luxuriant crop growth.Increased incidence of rust, damping-off, and early

leaf spot was noticed on top of the toposequence ingroundnut while late leaf spot and root rot were moresevere in the middle part of the toposequence.

Introduction, evaluation, and seedmultiplication of novel crops

A decade of renovation policy (Doi moi)implementation has changed the agriculturalproduction model in Vietnam from communityoriented to household based. Abolition of governmentmanaged and sponsored production system andestablishment of new system has led to scarcity ofimproved varieties. It has, therefore, become veryimportant to take up varietal improvement and seedmultiplication to ensure that appropriate varieties areidentified/developed and produced in sufficientquantities to meet the production demands.

More than 75% area was under maize (springmaize-autumn maize) before our intervention in thewatershed. This has resulted in decline in soil fertilityand increase in input costs over the years. Trials toevaluate novel crops (soybean, groundnut, and mungbean) and improved cultivars were taken up under theaegis of the National Legume Research andDevelopment Program and funding support from theAustralian Centre for International AgriculturalResearch (ACIAR). Technical backstopping and seedmultiplication of these remunerative crops resulted inreduction in maize area by about half in the benchmarkwatershed (Fig. 4). Many farmers are interested ingroundnut cultivation in large areas but non-availability of seed and seed storage facilities are themajor constraints.

Figure 4. Proportion of land under various crops in Thanh Ha watershed.

120

Large-scale on-farm demonstrations

On-farm demonstrations were conducted with mungbean, soybean, groundnut, and watermelon. Improvedmung bean cultivar T 135 was assessed for itssuitability as a catch crop in watermelon-maizecropping system. T 135 produced 1.12–1.24 t ha-1 anddid not pose any problem for normal maizecultivation.

An early-maturing soybean cultivar TN 12 (70–75days) was introduced to increase cropping intensityand system productivity. The new cropping systemallows four crops, i.e., watermelon-soybean-groundnut-sweet potato/vegetables in one year, ifwater is made available during the dry season.Improved groundnut cultivars LO 2 and MD 7 withhigh yield potential (3–4 t ha-1) and tolerance tobacterial wilt and pod rot were introduced. Farmerswere highly impressed with the crop and showedinterest in planting in large areas.

Improved production practices

Farmers in the watershed were following traditionalcultural practices due to technological inaccessibility.Improved production practices (integrated nutrient,pest, and disease management; and agronomy) werecompared with farmers’ practice in maize. Improvedproduction practices increased maize grain yields by8–18% over the years with 38% reduction in Nfertilizer (90–120 kg ha-1) usage. Improved productionpractices were also introduced in soybean, groundnut,and mung bean.

Land use planning for increased householdincomes

Unlike other Asian countries, the landholdings ofVietnamese farmers are very small. The averagefamily holding in drylands is around 0.5 to 1 ha. It is,therefore, important that the farm is utilized in mostprudent way for higher household income and foodsecurity. Efforts have been made to identifyappropriate crops and crop combinations in variousseasons for enhanced household income. For example,maize, groundnut, and soybean combination gavehigher income in spring while the traditional maizemonocropping system was not at all economical (Fig.5). Also, crop performance differed significantlyacross seasons.

Soils in the sloping lands were highly vulnerable toerosion when cleared of vegetative cover and weresubjected to various forms of land degradation. Lossof humus rich topsoil left behind the subsoil devoid ofvital plant nutrients leading to rampant infertility andpoor water-holding capacity. It is, therefore, importantto identify crops that not only perform well on thesesoils but help improve soil health over the years. Tofind out the influence of land degradation on cropproductivity and profitability we have delineated thegrain yields of soybean, groundnut, mung bean, andmaize based on the location on the toposequence in thewatershed (Figs. 6 and 7).

In general, higher grain yields and farm incomeswere obtained in the lower and middle part of thetoposequence compared to that of top due to lessdegradation and better soil fertility. Farmers areincurring higher expenditure due to increasedfertilizer usage on top of the toposequence. Groundnutcan be grown successfully on top, middle, and lowerpart of the toposequence while mung bean andsoybean need high level of management on top of thetoposequence for obtaining good yields. This kind ofinformation would assist in appropriate land useplanning and development of targeted nutrientmanagement technologies for system resilience andincreased household incomes.

Technology Exchange and HumanResource DevelopmentHuman resource development and technologyexchange were important components of the projectand consistent efforts have been made to provide onthe job training to large number of national staff anddisseminate integrated watershed technologies widelyas natural resource management is still an unexploredresearch area in Vietnam.

On the job in-country training

Several in-country training courses were organized forresearchers, extension workers, and farmers during thelast three years with the help of faculty from bothnational and international organizations. Theseinclude (i) recent concepts in integrated participatorywatershed management; (ii) improved soil and watermanagement in watershed context; (iii) integrated pest

121

Figure 6. Influence of toposequence on crop productivity in Thanh Ha.

Figure 7. Influence of toposequence on crop profitability in Thanh Ha.

Figure 5. Influence of land use planning on household income, Thanh Ha, 2000.(Note: M = maize, GN = groundnut, SB = soybean, and MB = mung bean.)

122

and disease management; (iv) improved productionpractices of groundnut, soybean, and mung bean;(v) INM; (vi) Rhizobium inoculation; and(vii) Gliricidia sepium nursery management. Also,many national program scientists received informaltraining in design and development of efficient landand water management technologies at watershed scale.

Regional training courses and travelingworkshops

Scientists of the Vietnam Agricultural ScienceInstitute (VASI) benefited greatly from these activitiesas they were provided excellent opportunities to visitother benchmark watersheds of the project, interactwith peers, and exchange views and experiences.

Training at ICRISAT

Three young and two senior researchers from VASIvisited ICRISAT and received training on integratedwatershed management, data collection, andmodeling.

Public awareness programs

Farmers’ days were conducted in each croppingseason in the landscape watershed and all the farmersin the Thanh Ha State Farm were invited to getaquainted with different components and technologiesof integrated watershed. Field days proved to be veryefficient in getting the message across as severalprovincial and district authorities, technology

exchange departments, research managers, and policymakers were invited and were highly impressed withthe technological innovations. Videos and brochureson improved production practices and watersheddevelopment were prepared for use by extensionagencies for wider dissemination and adoption oftechnologies.

Looking AheadWatershed based integrated natural resourcemanagement technologies provided an excellentopportunity for efficient management of rainwater,i.e., controlling runoff to reduce erosion, increaseinfiltration, enhance moisture levels in the soil,canalizing and harvesting surplus water for life savingirrigation and summer cropping, and recharginggroundwater for sustaining the production at higherlevels to meet the growing food demands in theuplands. Crop diversification with legumes andoilseed crops helped in improvement of soil health andopened up new opportunities for enhanced householdincomes and rural employment in the hithertomalnourished and poverty-stricken sloping lands. Inthe coming years the strategic goal should be to scale-up improved technologies to other sites in the targetecoregion to capitalize the benefits achieved in thebenchmark watershed and consolidate the gains ofimproved soil, water, and nutrient managementtechnologies and cropping systems through wideradoption for sustaining agricultural production in thesloping land ecoregions of northern Vietnam.

123

Improving Management of Natural Resources for Sustainable RainfedAgriculture in Northeastern Thailand

Narongsak Senanarong1, Precha Cheuwchoom1, Chanvid Lusanandana1,Somchai Tangpoonpol1, Suppachai Atichart1, Arun Pongkanjana2,

Pranee Srihaban2, Thommasak Singhapong2, Aran Patanothai3,Banyong Toomsan3, Sawaeng Ruaysoongnern3, and T J Rego4

Abstract

Lands of northeastern Thailand are sloping and fragile. A large proportion of these soils are degraded dueto soil erosion. Degraded soils are one of the major constraints for agricultural production in this region.ICRISAT in collaboration with the Department of Agriculture, Land Development Department, and KhonKaen University, Thailand started a project in 1999 with the financial support from the Asian DevelopmentBank to improve the management of natural resources for sustainable rainfed agriculture throughintegrated watershed approach. This paper summarizes all the research work carried out for three yearsduring the project period. This includes selection of benchmark site in the ecoregion, baseline surveys,establishment of monitoring devices and various interventions in cropping systems, land and watermanagement and fertility management areas, and human resource development. The initial results ofresearch indicate a reduction in soil erosion and improvement of crop yields due to various interventions.There is sufficient scope to scale up this work in the ecoregion. The details of various activities undertakenand the outputs are presented in the paper.

1. Department of Agriculture (DOA), Bangkok, Thailand.2. Land Development Department (LDD), O/o Agricultural Research and Development (OARD), Region 3, Khon Kaen, Thailand.3. Khon Kaen University (KKU), Khon Kaen, Thailand.4. International Crops Research Institute for the Semi-Arid Tropics (ICRISAT), Patancheru 502 324, Andhra Pradesh, India.

Agricultural production in northeast (NE) Thailandcompared to other regions is diverse but mainlydependent on rainfed production and constrained dueto moderate to low seasonal rainfall, lack of waterduring the dry season, undulating terrain, and poorsoils. In NE Thailand only 8% area is irrigated andremaining 92% is either rainfed or partially irrigatedwith the water harvested from higher slopes. Besides,in most part of the northeastern region theunderground water is mostly saline because of theunderlying rock salt geological formations.

Soils are mostly degraded primarily due to thepredominance of fragile shallow and eroded soils.The common land use practices are mainly in the formof shifting cultivation; however, farmers draw adistinction between shifting cultivation and the morecommon practice of “land rotation farming”. Under

such a practice, land is fallowed for 3–5 years for soilfertility regeneration. With increasing populationlandholdings are becoming smaller and smallerresulting in intensifying crop production to fulfill thedemands of food. Alongside, however, the soilerosion has also increased and serious soildegradation is taking place. Due to frequent plowingof land, bush regrowth gets reduced so that there islittle vegetative cover to protect the land. Therefore,land degradation by soil erosion has lately increased,mainly due to the adoption of inappropriate newproduction technologies.

Significantly, more than 95% lowland paddy in NEThailand is also grown under rainfed agriculture. Thecropping systems in rainfed agriculture include maizeas a cash crop on the high slopes and upland rice onthe lower slopes; tree crops and fruit trees are usually

124

grown close to supplementary water resources on thelower slopes. Sometimes, legumes and cereals arerotated with upland rice and maize, according to soilfertility and economic returns. Cassava is anothermajor crop.

The Environment and NaturalResource Degradation

The basic constraints of rainfed agriculture impactseveral soil degradation processes individually orinteractively at different levels of land use hierarchy.At the highest level of toposequential hierarchy, soilerosion from steep slopes is extremely severe becausethe land preparation is done along the slope bytractors. This practice substantially increases soilerosion. Forest fires, particularly during the extendeddry periods, and unavailability of water during dryperiods constrain establishing plantations on steepslopes. In the mid-slopes, the second level oftoposequential hierarchy, some soil conservationpractices can be applied to reduce soil erosion byflexible land preparation, introduction of smallerplots (<1 ha), which shorten slope lengths, andgrowing alternative crops. For the relatively flatundulating area, the third level of toposequentialhierarchy, soil erosion is low due to the presence oftrees and more care taken by farmers to conservelands. Therefore, classification of land types isneeded for matching appropriate technology optionsto combat soil erosion. Moreover, where the soilerosion is not severe, nutrient depletion may well bethe cause of decline in productivity in low inputintensified systems (Table 1). In the intensifiedproduction systems, soil acidification is taking placeeither in the subsoil or in the topsoil or in both. Allthese problems have resulted in decline in crop

productivity, and have eventually restrictedintroduction of alternative crops. Some off-siteeffects are unabated soil erosion, which has led topublic health problems of poor quality water, siltationof reservoirs resulting in a decline of water and fishresources, and related environmental issues. Otheroff-site effects of land degradation include furtherencroachment and clearing of forest for new fertileland by the farmers thus setting in motion acontinuous cycle of decline in reserve forestresources. This is a vicious problem for sustainingnatural resources in the rainfed agricultural areas.

To reduce the negative effects of conventionalfarming practices on the degradation of land, Thaiinstitutions and the International Crops ResearchInstitute for the Semi-Arid Tropics (ICRISAT), withthe assistance from the Asian Development Bank(ADB), had initiated a pilot rainfed agricultureproject in 1999 in NE Thailand. The objectives of theproject are to:• Evaluate the degree and potential of land

degradation in NE Thailand.• Screen and identify appropriate existing

technologies to control soil degradation.• Evaluate improved conservation-effective land

and water management technologies to rehabilitatedegraded soils.

• Field test new technologies to sustain productivityand minimize land degradation.

Selection of Benchmark Watershed SiteIn March 1999, a team of scientists from theDepartment of Agriculture (DOA), LandDevelopment Department (LDD), and Khon KaenUniversity (KKU) in Thailand and ICRISATidentified Tad Fa watershed for on-farm benchmarkresearch and development. The site is situated about

Table 1. Nutrient loss (t yr-1) in different regions of Thailand.

Region Nitrogen Phosphorus Potassium

Northern 38,288 4,467 75,588Northeastern 18,896 1,212 91,644Eastern 17,890 1,074 30,860Southern 17,310 453 13,254Source: Land Development Department.

125

150 km northwest of Khon Kaen. It is at a junction ofthree big watersheds namely Mae Khong in thenortheast, Chi in the east, and Pasak in the southwest(Fig. 1). Tad Fa watershed represents the “ecoregion”covered by these three watersheds which cover47,000, 49,480, and 15,780 km2 respectively. Tad Fawatershed has 2,500 ha land, which is a part of “NamChern” sub-watershed (2,920 km2) in the Chiwatershed. The Tad Fa watershed falls in twoprovinces. The eastern part of river Tad Fa is in KhonKaen Province which has nearly 700 ha while thewestern side is in Petchabun Province. All theresearch and development work was carried out in theeastern part of Tad Fa watershed called Huay Ladcovering 200 ha cultivated land spread in twovillages, Ban Tad Fa and Ban Dong Sakran.

Research and DevelopmentThe following research and development activitieswere undertaken:• Collection and analysis of ecoregional database to

identify constraints in the ecoregion and yield gap

analysis to find out the scope for yield improvementin the ecoregion.

• Participatory rural appraisal (PRA) of Tad Fa(eastern part) watershed to identify and prioritizethe constraints for enhanced crop production onsustainable basis while conserving the naturalresources.

• Strategic research to overcome nutrient constraintsas well as to quantify land degradation.

• On-farm development and evaluation of varioustechnologies through farmer participatory approach.

• Continuous monitoring and evaluation ofimproved options.

• Human resource development of the nationalagricultural research system (NARS) researchersand farmers through various training programs,workshops, and field days.

Ecoregional Database AnalysisThe report contains information on the agroecologyof three main watersheds (Mae Khong, Chi, andPasak) surrounding Tad Fa. Biophysical and

Figure 1. Tad Fa wataershed in NE Thailand.

126

socioeconomic data were collected from secondarysources. Tad Fa watershed is tropical (26–28ºC). Theannual rainfall is 1200–1400 mm and evaporation is1900–2000 mm. Topographically it has sloping-upland complexes; soils are mostly Ustults and theland use is mostly comprised of forestry, agroforestry,horticulture, and field crops.

Physical constraints

No rainfall in the dry season (November to March) isa major constraint. A less important constraint is thehigh relative humidity in the wet season (June toOctober) which encourages pests and diseases indryland crops like maize.

Relief is a major constraint in hilly and mountainousterrain. The steep and uneven slopes make cultivationdifficult and result in rapid runoff of rainfall,accompanied by sheet and gully erosion. Flooding oflowland is also a major constraint resulting in lowyields during intensive rains.

Low soil fertility affects large areas in highlandsand strongly leached soils on slightly higher terrain inlowlands. Shallow soil depth and lateritic gravelaggravate the fertility problem. Loss of appliednutrients occurs during the wet season, especially onsteep slopes. Shallow soils have reduced water-holding capacity in the soil profile, limit rootingdepth, and increase erosion hazard.

Technological constraints

A large number of technological innovations areavailable to overcome physical constraints likeirrigation, drainage, flood control, systems of highlandagriculture and forest conservation, application of

fertilizers and pesticides, weed control, and seedsupply. But these technologies should be modified asper the characteristics of the location and problem.

Institutional constraints

Since Thailand has a well developed research,training, extension, and agricultural credit systemthere are minor institutional constraints. However,farmers’ groups or cooperatives to manage naturalresources are not prevalent.

Socioeconomic constraints

Farmers are economically poor and education in thefamily in this region is quite low. Thai farmers arequite hard working and adaptable. Since 1975 evenpopulation growth has been checked which may result instopping further fragmentation of land. The governmenthas provided infrastructural support and even guaranteesminimum farm-gate price for certain crops.

Yield Gap AnalysisFive major crops (rice, maize, soybean, groundnut,and sunflower) of NE Thailand were chosen for thestudy. Experiment station yields, representativecountry yields, and regional yields have beencompared with the crop yields harvested by farmersin NE Thailand. The average yield of rice in NEThailand is 1.8 t ha-1 compared to experiment stationyield of 3.4 t ha-1 (Table 2). The yield gap is 1.7 t ha-1;however, it is only 0.2 t ha-1 when compared with thecountry’s average yield. The average yield of maize inNE Thailand is 2.5 t ha-1 while experimental stationyield is 4.7 t ha-1, and the country’s average yield is

Table 2. Average crop productivity of five crops in different regions of Thailand in 1998.

Yield (t ha-1)

Region Rice Maize Soybean Groundnut Sunflower

Northeastern 1.8 2.5 1.2 1.3 1.5Northern 2.8 2.9 1.2 1.5 1.5Central Plain 2.9 3.1 1.2 1.5 1.5Southern 2.1 2.6 1.3 1.1 – Country 2.0 2.8 2.8 1.4 1.5

127

2.8 t ha-1. The NE region is further subdivided intohighland, upland, and lowland areas. Maize yield is1.5 t ha-1 in highlands, 2.4 t ha-1 in uplands, and 3.5 tha-1 in lowlands. Even though water may not belimiting for the rainy season maize, it clearly indicatesthe degradation of soil in the highlands and uplands.

Participatory Rural Appraisal of TadFa WatershedA PRA was conducted in 1999 in the eastern part ofTad Fa watershed, which was further divided intothree parts based on three streams namely Samtada,Lad, and Wang Deun Ha. In addition to these threeagricultural areas, two additonal forest watershedswere identified, one in the north and another in thesouth of these agricultural watershed areas. The PRAon socioeconomic aspects was conducted in all thethree agricultural watershed areas. The objectives ofthis survey were:• To understand the existing socioeconomic situations

in Tad Fa watershed in order to plan a researchprogram for sustaining agricultural production.

• To select a catchment which is representative ofthe agroecology where research and developmentwork will be carried out.The survey indicated that there are three regions,

based on soil quality, in the watershed. The middleportion is the most fertile while the regions to the northas well as to the south are less fertile. The soil depthranges from 0.5 m to 2 m. The soil is sandy loam at thesurface and is clayey loam to loam at subsurface. Thereare nearly 80 farm ponds in eastern Tad Fa, of whichonly 4 store water throughout the dry season, whileothers dry in summer because of very porous subsoiland high seepage losses. Farmers have planted fruittrees only around their houses and not on steep slopesas desired (and recommended) by the government.

In Tad Fa watershed upland rice is mainly grownfor home consumption. Maize is the main cash crop.Ginger has been tried since two years by few farmersbut is very risky due to disease problems and pricefluctuations. Soybean is not grown because of highvegetative growth and very poor grain yields. Rice(local variety) is grown only in poor soils or lessfertile patches since more vegetative growth has beenobserved in the fertile lands. A very small amount of

urea is mixed with the rice seeds at sowing. Onlymaize and ginger crops are fertilized. Rice is plantedin June and harvested in October. About 2.5 to 3.0 tha-1 grain yield is obtained. Often, maize is growntwice a year depending on the onset of monsoon. Thefirst crop is grown during March to July and secondcrop during July to November. Farmers apply about150 kg ha-1 of NPK (15:15:15) fertilizer and harvest 3to 3.5 t ha-1 of grain yield. Ginger is grown in March/April to December. A heavy dose of NPK fertilizer at600 kg ha-1 is applied.

Farmers have identified the following constraints:land tenure, lack of capital, lack of water resources,costly agricultural inputs, price fluctuation, lack ofgovernment support, lack of transport facilities, soilerosion, forest fires, and labor shortage. Since theseare displaced farmers, they do not have much capitalto invest. But highest priority is given to childreneducation. There is only one primary school in thevillage and children have to go to other villages forhigh school education. Second priority is given forhousing. As most of the farmers have poor temporaryhouses, they wish to build new houses.

Farmers give third priority for agriculture.Fortunately, the land is reasonably fertile. Rice isgrown as a subsistence crop without much fertilizerapplication. Maize, which is grown as a cash crop isfertilized. However, farmers have to invest a sizeablecapital on labor because household labor is scarce; alloperations like land preparation, seeding, weeding,and harvesting are given on contract to serviceproviders. Few farmers have tried the risky gingercrop with huge investment and most of them havesuffered heavy losses.

Overcoming Nutrient ConstraintMost of the farmers apply chemical fertilizers to theircash crops to harvest good yields. Chemicalfertilizers are one of the costliest inputs and there wasa need to search other alternatives or supplementsources to overcome nutrient constraints. There is notmuch scope to use farmyard manure as farm animalshave been replaced by farm machines for draftpurposes. The use of legumes in the cropping systemwould certainly help to reduce the amount ofchemical nitrogen (N) fertilizer. Legumes were

128

evaluated to quantify N2 fixation and the benefits oflegumes using 15N abundance method and 15N isotopedilution method on farmers’ fields at Ban Koke Monlocated near Ban Tad Fa.

Based on the N difference method N2 fixationvaried from 20 to 104 kg N ha-1 and net N benefit tothe succeeding crop was estimated at 2 to 51 kg ha-1.Following legumes, a maize crop was grown with 40kg N ha-1 along with the organic matter from thelegume residues. Grain yield of succeeding maizecrop was significantly (P ≤0.05) higher by 27 to 34%in treatments following black gram, rice bean, andsunnhemp over the yield of maize crop (Table 3).Although N2 fixation was highest in sword bean(104 kg N ha-1) the benefits were not translated interms of increased maize yields. These resultsdemonstrated that it is not only the quantity of N2

fixed that determines the benefit to the succeedingcrop but also the quality of organic matter and Nrelease pattern from the legume residue. However, inthe long term, sword bean could play an importantrole for sustaining land productivity.

Growing black gram, rice bean, and sunnheamp inthe system would help in reducing N requirement forthe succeeding maize crop. The actual realizedbenefits from legumes in terms of increased N uptakeby the succeeding maize crop varied from 5.3 to 19.3kg N ha-1 whereas the expected benefits from legumesthrough biological nitrogen fixation (BNF) and soil Nsparing effect over a maize crop varied from 15 to 64kg N ha-1 (Table 4). These results revealed that forquick benefits for succeeding maize crop farmerswould be benefited by growing legumes such as ricebean, sunnhemp, and black gram.

Table 3. Dry matter (kg ha-1) of maize grown after five different crops at Ban Koke Mon in the rainy season2000.

Treatment Stover Cob Seed1 Total

Rice bean 7069 816 4541 a 12425Sunnhemp 6634 786 4720 a 12141Sword bean 6689 659 3642 b 10991Black gram 6786 875 4488 a 12149Maize2 5560 697 3525 b 9781

F test NS3 NS * NSCV (%) 14.57 14.41 13.36 13.13

1. Figures followed by the same letter do not differ significantly at P ≤0.05; * = Significant at P <0.05.2. The maize crop received nitrogen (N) from legume crop residue plus 40 kg N ha-1 in the form of chemical fertilizer.3. NS = Not significant.

Table 4. Nitrogen benefit realized from legumes in maize-based systems.

N benefit ExpectedNet N Total N uptake realized from benefit from

N fixed by benefit by succeeding legume over BNF + N savinglegume expected1 maize maize2 benefit

Crop (kg ha-1) (kg ha-1) (kg ha-1) (kg ha-1) (kg ha-1)

Rice bean 20 2 75.9 19.1 15Sunnhemp 90 31 76.1 19.3 44Sword bean 104 51 62.1 5.3 64Black gram 27 8 68.9 12.1 21Maize - –13 56.8 - -1. N2 fixed – Seed N.2. Total N uptake by succeeding maize – Total N uptake by maize grown after maize.

129

Quantification of Land DegradationIn NE Thailand, types of land degradation (e.g.,biological and chemical) are not fully studied. Tostudy the effect of land degradation on cropproductivity, sites in the toposequence were identifiedand crop yields were monitored during 1999, 2000,and 2001 (Table 5). Soil samples at these spots up to110 cm depth were collected and analyzed forphysical, chemical, and biological properties (Table 6).The maize grain yield data clearly indicated the loss ofproductivity on steep slopes and on moderate slopeswhen compared to mild slopes. The clay and organicmatter content at these spots indicated that precious clayand organic matter have been eroded from the steepslopes. Most of these changes have occurred in thetopsoil layers which are very important for cropproduction. A new methodology to quantify landdegradation called Soil Fingerprint method is being used.

Fingerprint technique (FPT) is an approach formonitoring land degradation as it impacts soil quality.FPT primarily utilizes characteristics of landform andsome soil physical properties as a basic guideline forcomparable paired sites. Usually the comparison isbased on pairing of the virtually similar location withdifferent land use system, namely forest as control (lessdisturbed) site and agriculture plot as degraded site.Thereafter, further soil chemical analysis is done toidentify soil profile similarities and depths ofundisturbed horizon using both general soil chemicalproperties and soil charge fingerprinting characteristicsof each horizon. Once the profiles are consideredcomparable, the discrepancies in soil depth andphysical properties of comparable profiles wereconsidered as soil loss through erosion and soilphysical degradation, respectively. FPT-basedinformation could be used as a good first

Table 5. Maize grain yield (t ha-1) across topo-sequence in NE Thailand during 1999 to 20011.

Toposequence 1999 2000 2001

Steep (>15%) 3.1 (3) 4.5 (4) 2.1Moderate (5–15%) 3.6 (6) 4.8 (5) 2.9Mild (2–5%) 4.1 (2) 5.3 (4) 3.41. Figures in parentheses refer to the number of farmer fields at

each slope.

approximation to estimate loss of soil organic matter,clay, mineral bases, and soil fertility under differentland uses. These soil parameters are important for theevaluation of soil and land degradation, which is amajor constraint to the sustainability of agriculture inthe tropics.

The watershed was surveyed to identify suitablesites for soil degradation study using FPT. From thelandscape layout, it was found that at least 7 siteswere considered physically suitable. Land use historywas further investigated using a rapid rural appraisal(RRA) technique. From the 7 sites, only 5 transectswere found suitable. These transects were used as thefinal study sites. Soil physical analysis was completedfor all samples. In general, the charge fingerprintsindicated that soils under 30 cm depth have similarcharge fingerprinting.

On-farm Development and Evaluationof Various Technologies throughFarmer Participatory ApproachOut of 700 ha of land in the eastern part of Tad Fa, weselected the middle portion of the watershed calledHuay Lad, which had about 200 ha of land undercultivation; also the two villages, Ban Tad Fa and BanDong Sakran, were located in this area. Most of thefarmers from Ban Tad Fa village had land in thenorthern Huay Samtada.

We concentrated in the Huay Lad area (DongSakran village) for research and development work.There were 49 farm ponds in the Huay Lad. Weidentified two micro-watersheds for our research.One micro-watershed was “traditional”, which hadmoderate slope and nearly 70% land had fruit treesand in the remaining area other annual crops likemaize and upland rice were grown. The other micro-watershed had moderate as well as steep lands andmostly annual crops like maize and upland rice weregrown. All the interventions were carried out in thismicro-watershed. In almost 80% of this micro-watershed, “hillside ditches” were dug for soilconservation on contour. Vetiver and maize were plantedalong the side of “hillside ditches”. It was suggested thatall farmers should plant crops like maize along thecontour instead of usual up and down the slope. Oneautomatic runoff and sediment sampling system wasinstalled at the lowest point of each micro-watershed to

130

monitor runoff and soil loss. The area of the traditionalmicro-watershed was 17.8 ha with 4 farmers while thatof the improved micro-watershed was 12 ha with 5farmers. Two farmers had land in both micro-watersheds. An automatic weather station was installednear the research area to monitor rainfall, temperature,

sunshine, humidity, wind velocity, and soil temperatureon a continuous basis at fixed intervals of time.

Soil survey of the entire Huay Lad agricultural landwas done and detailed soil map and land use map wasprepared. Majority of the soil was silty clay loam with avery small fraction of clay loam. Almost all the clay loam

Table 6. Biological and chemical properties of soil samples from toposequence in Ban Tad Fa watershedin NE Thailand.

Topsequence 0–101 10–20 20–30 30–50 50–70 70–90 90–110

Organic C (g kg-1 soil)Top 28 27 26 14 13 9 7Middle 31 29 26 18 12 10 10Lower 40 34 29 20 35 20 19 LSD = 1.15Total N (mg kg-1 soil)Top 2073 2085 1956 1755 1324 1249 1092Middle 1967 1771 1785 1376 1178 1352 1012Lower 2336 2287 1971 1563 2345 1630 1462 LSD = 621.2Net “N” mineralization (mg kg-1 soil 10d-1)Top 11.89 10.03 6.80 5.52 2.30 1.97 1.47Middle 14.22 11.16 8.93 6.07 3.84 3.75 3.04Lower 15.11 14.49 12.72 9.04 5.73 4.53 4.70 LSD = 6.034Biomass C (mg kg-1 soil)Top 366 304 275 258 178 149 133Middle 362 300 240 206 173 124 100Lower 384 328 276 213 128 145 112 LSD = 86.3

Clay content (g kg-1 soil)Top 330 350 380 330 330 0 0Middle 390 380 430 420 370 230 0Lower 450 450 450 490 550 550 590 LSD = 2.4Fine sand (g kg-1 soil)Top 90 70 70 180 140 0 0Middle 80 80 80 130 120 160 0Lower 60 60 60 70 60 60 40 LSD = 1.6Gravel (g kg-1 soil)Top 190 150 130 100 140 0 0Middle 130 120 100 80 250 150 0Lower 140 140 130 110 120 100 90 LSD = 1.91. Soil depth in cm.

131

had 2–5 and 5–12% slope while a small proportion ofsilty clay loam had 2–5% slope and the rest had 5–12,12–20, and even 20–35% slope. There were 13 distinctsoil series and their variants in Huay Lad.

Detailed baseline survey of all the 10 householdsinvolved in the micro-watershed was done. This willbe used to measure the impact of interventions. Thesurvey covered size of family, age, education, sourceof income, size of landholding, land use, crops grown,agricultural implements, animals reared, and financialstatus of farmers. Even though all farmers recognizedsoil erosion as a cause for land degradation none ofthem have seriously followed any measures to checksoil erosion. Since the history of cultivation of theselands is only about 80 years, the soils are still rich inorganic matter and support reasonable cropproduction. But farmers apply chemical fertilizer totheir maize crop, which is a major cash crop and isgrown on moderate and steep slopes. Upland rice ismostly grown on mild slopes without much fertilizerssince these soils are not eroded. Even though plantingof fruit trees and other trees on steep slopes isrecommended, farmers have planted fruit trees on mildslopes where they have some water sources to irrigatethese trees during dry periods of establishment. Theresults of on-farm research are given below:• Comparison of groundnut yields grown in the two

rainy seasons:Early season (first crop) gave higher yield of freshpods (3.7 t ha-1) when compared to second season yieldof 2.5 t ha-1. These low yields in both the seasons maybe due to phosphorus (P) deficiency in the soil.

• Evaluation of the performance of soybean in thebenchmark watershed area:Five soybean varieties were grown. The grainyields were low ranging from 510 to 875 kg ha-1. InThailand yields less than 1.2 t ha-1 are considereduneconomic. The experimental fields had less than5 ppm of P as against the threshold level of 15ppm. Thus the performance of soybean was poorperhaps due to acute P deficiency. Weedcompetition was also severe, as farmers do notadequately weed the crop. At present soybean cropdoes not hold promise in Tad Fa watershed.

• Productivity of upland rice at two toposequentialslopes:Rice is a popular crop among all farmers and isgrown in the lower part of the toposequence. Often

farmers grow only traditional varieties and mainlyfor home consumption on 0.5 to 1.0 ha of land.Nearly 75% of farmers grow “Lao Taek”. Mildslope (2–5%) and very mild slope (<2%) did notaffect the rice yield significantly; the mild slopeproduced 3.5 t ha-1 when compared to 3.4 t ha-1 invery mild slope area.

• Evaluation of the productivity of relay croppedrice bean at different sites in the toposequence:Rice bean is a popular legume crop grown insloping land ecologies of NE Thailand. In about40% of the maize growing area rice bean is relayplanted. It is sown in the standing crop of maize atthe flowering time. Since it is sown without anyland preparation (unlike sequential planting) relayplanting is a soil conservation efficient system. Insteeper slope (>15%), the yield (970 kg ha-1) is 25–30% less compared to moderate slope (5–15%)(1270 kg ha-1) or mild slope (2–5%) (1360 kg ha-1).Poor soil as well as less amount of soil moisturemay be responsible for low yields at steep slopes.This system has to be popularized with most of themaize farmers who sometimes try a second crop ofmaize, which suffers due to terminal drought orsometimes they are not able to plant the secondcrop due to late onset of monsoon and late plantingof first maize crop.

• Response of maize to fertilizer at upper and lowerslope of toposequence:Most farmers apply fertilizers at 150 kg ha-1 ofNPK (16-20-0). To see the maize performancewithout fertilizers at the upper and lower portionof toposequence, a small trial was carried out indifferent farmers’ fields. At the upper site maizeproduced 2.6 t ha-1 and at the lower site it produced4.1 t ha-1 of grain without application of anyfertilizer while the yields were 4.1 and 6.7 t ha-1

respectively with fertilizer. This again reveals thedegradation of soil at the upper end oftoposequence and indicates that in terms ofeconomic losses land degradation has resultedequivalent to 2 t of maize grains ha-1 yr-1. Suchdifferences are variable in some years.

• Minimum tillage and contour cultivation:In order to reduce tillage on steep slopes which maytrigger enhanced soil erosion, hand dibbling wastried on steep slopes and tractor planting on contouron moderate and mild slopes. Minimum tillage was

132

as effective as tractor planting. Grain yield of maizewas 4.1 t ha-1 on steep slope, 4.5 kg ha-1 on moderateslope, and 4.9 kg ha-1 on mild slope.

• Maize relay cropping with legumes:Rice bean was relay cropped with maize on twoslopes. On moderate slope, yield was 980 kg ha-1

while on mild slopes 1060 kg ha-1 was recorded.This small increase on mild slopes may be due tohigher moisture retention at lower ends. In anothertrial a new legume crop, black gram, was grown asrelay crop with maize. Compared to rice bean itproduced only 290 kg ha-1 while the former gave810 kg ha-1. Black gram was shaded by maize andalso was severely suppressed by weeds. Manypests attacked the foliage as well as the pods. Blackgram may not be suitable as relay crop with maize.

• Fruit tree planting and intercropping:Our continued efforts resulted in planting of 625fruit tree seedlings of longan, litchi, and mango in4 ha in the watershed. Few farmers grew maize asan intercrop between the fruit trees providingshade for trees as well as helped in suppressingfast-growing Mimosa weeds. Farmers harvestedmaize grains providing direct income until thetrees started producing fruits. In this processfarmers need not suffer for a few years till the treesstart bearing fruits. But some technologies shouldbe improved to reduce competition to fruit trees bymaize and Mimosa.

• Community-based banana dryer:The farmers of Tad Fa get very low returns fortheir banana crop. The scientists of DOA analyzedthis problem and came up with a practical solutionof value addition to their product and also toincrease the shelf life by drying the ripe bananaand then selling. This is possible in summermonths. In rainy season they cannot do itefficiently and the product may get spoiled. Toovercome this problem, with the help of KKUengineers they developed solar dryers as there isno electricity in the village. On a trial basis DOAplans to install community-based solar dryers witha very nominal charge for maintenance. If thisintervention succeeds, then the farmers will notonly get additional income but also the fundsgenerated for the community will help to buyadditional dryers.

• Minimum tillage cultivation for maize:As mentioned earlier, farmers plow the land deepwith hired tractors. Due to our persuasion a fewfarmers followed minimum cultivation. The grainyield of maize cultivar CD-DK 888 was 4020kg ha-1 while that of Suwan 1 was 2100 kg ha-1.These yields are fairly good. Also, the cost ofcultivation was less and minimum tillage preventssoil erosion from the sloping lands.

• Sequential cropping of groundnut:Groundnut was planted in July as a sequential cropafter maize. The variety Khon Kaen 5 produced4980 kg ha-1 fresh pods while Native (red seed)produced 4540 kg ha-1 fresh pods.

Human Resources DevelopmentThe following activities were undertaken during theproject period to develop human resources of partnerNARS:• A training course on database management system

for sustainable rainfed agriculture was held atICRISAT, Patancheru, India from 15 November to3 December 1999. One Thai researcher attendedthe training.

• A traveling workshop was organized from 27August to 12 September 2000. Two Thairesearchers participated and traveled to all theproject sites in India, Thailand, and Vietnam.

• A training workshop on “Impacts of Variability ofNatural Resources on the Performance ofCommunity Scale Watershed” was held from 16 to29 November 2000 at ICRISAT and two Thairesearchers participated.

• A training workshop on “On-farm ParticipatoryResearch Methodology” was held at Khon Kaenfrom 26 to 31 July 2001. Two scientists, each fromDOA and LDD, attended this workshop.

• A training program on “Farmer ParticipatoryResearch” was conducted by the faculty of KKUfor the staff of DOA and LDD involved in the TadFa watershed. This training was held in twosessions. The first one was held at Nam Nao from10 to 12 May 2001 and a total of 18 staff fromDOA and LDD attended. The second one was heldfrom 18 to 20 October 2001 at ChumPhe and 22members attended.

133

Lessons Learned and Future Strategy• Land degradation in general and soil erosion in

particular has been perceived as a major constraintfor sustainable agricultural production in the studyarea not only by the agricultural scientists but alsoby the farmers.

• Since the cultivation history is only about 80 years,the soils are reasonably productive at present butpoor productivity has been noticed both on steepand moderate slopes of the toposequence. Farmersapply fertilizers to cash crops on these sites.

• Land degradation in the watershed on atoposequense is evident and clearly demonstratedby the varying yields of maize from 300 to 3000 kgha-1 during different cropping seasons dependingon rainfall. Farmers recognize these effects and areready to work along with scientists to developmeasures for soil and water conservation.

• Farmers have recognized water conservation as animportant aspect and are willing to contributeresources (cash and in kind) for undertakingrainwater harvesting structures.

• Due to lack of capital as well as the fear ofevacuation as they do not have permanent tenure,farmers had not taken up major soil and waterconservation and management practices.

• Since the government has recognized the settlementas villages in 1998 the farmers are optimistic ofownership of their lands and started heeding theadvice of scientists and policy makers to plant morefruit trees in the agricultural land. Looking at theresponse of the farmers since two years we arehopeful that fruit tree planting will cover a largearea in the near future, which not only will reducesoil erosion but also will improve the economicstatus of the farmers in a sustainable way.

• Since the underground parent material is veryporous, almost all the dug tanks do not hold waterfor summer. We may have to look either forimprovement of these structures or for possibilitiesof using groundwater and sharing the same for thedrier months by the community.

• There is a need to evaluate the soil conservationpractices like hillside ditches, vetiver planting, andcontour planting of crops.

• The annual cash crops are very much influencedby markets. Crops like ginger and pineapple havedisappeared from this area. Maize crop was beingcultivated as a cash crop for a long-time on steepand moderate slopes. Since one year the prices ofmaize are falling down. Many farmers have startedgrowing vegetables as cash crops. We believe fruittrees will stabilize the situation in the future.

• Finally, this watershed approach with more focuson soil conservation and rainwater harvest shouldbe extended to other areas in the ecoregion withemphasis on both community as well as individualfarmers.

Acknowledgments

We sincerely thank Dr Montien Somabhi and MrPanya Ekmahachai of DOA, Dr Manu Srikhajonand Mr Rungroj Puengpan of LDD, and Dr S MVirmani (formerly at ICRISAT) who were involvedin the initial planning and execution (under theleadership of Dr Narongsak Senanarong of DOAand country coordinator). We also thank Dr S PWani, Project Coordinator and Mr P Pathak andother staff of ICRISAT whose active supporthelped to run this project successfully and toachieve the planned outputs.

134

Improving Management of Natural Resources for Sustainable RainfedAgriculture in Ringnodia Micro-watershed

R A Sharma1, O P Verma1, Y M Kool1, M C Chaurasia1, G P Saraf1,R S Nema1, and Y S Chauhan2

Abstract

The current productivity of rainfed lands in Madhya Pradesh, India is about 1.0 t ha-1 although there isscope to obtain >3 t ha-1. To assess and evaluate the potential of improved soil, water, and nutrientmanagement options through integrated watershed management at Ringnodia in Indore in westernMadhya Pradesh, a micro-watershed of 390 ha was delineated. Soybean is a major crop during the rainyseason and yield of <1 t ha-1 is obtained in the micro-watershed. Landholdings in the watershed aregenerally small. The input use is low with little soil and water conservation measures in vogue amongfarmers. About 30–40% of the total rainfall is lost through runoff, carrying productive soils and nutrientswhile crops experienced drought stress in the rainy as well as postrainy seasons. With a critical advisorysupport from scientists, the watershed farmers could augment water storage capacity in the villagethrough construction of percolation/storage tanks and renovation of existing ponds. For safe disposal ofwater from the watershed, waterways were developed and wire mesh bound boulder structures wereconstructed to reduce soil loss and runoff. These water storage structures could store up to 30 ha-m waterrepresenting about 70% of total runoff from 100 ha cultivated area and thus reduce runoff and soil losses.This increased groundwater recharge, which manifested in increased water table in most wells includingthe abandoned ones.

The scenario analysis suggested various cropping options for enhanced yield with limited irrigation(soybean-wheat) or under rainfed conditions (pigeonpea/sorghum intercrop). Sorghum/pigeonpeaintercrop was, however, less popular amongst the farmers. The introduction of extra-short-durationpigeonpea opened avenues for diversification and its adoption is likely to increase. Under rainfedconditions, double cropping could be practiced in two out of three postrainy seasons. Soybean yieldsincreased marginally by gypsum application and also by planting on mini-ridges. The medium-durationchickpea cultivar JG 218 gave higher yield than short-duration cultivars ICCV 2 and ICCC 37 indicatingsufficient moisture for the traditional types. Pests were the major yield reducers in soybean and adoption ofintegrated pest management options nearly tripled soybean yield.

In another micro-watershed at the College of Agriculture, Indore interaction between land and waterconservation measures and efficient cropping systems was examined. Soybean/pigeonpea strip crop andsoybean-wheat systems were more productive than soybean-chickpea and soybean-linseed systems.Chickpea and wheat could easily be established with minimum tillage when planted in moist seed zone at15 cm depth after the harvest of soybean.

1. Jawharlal Nehru Krishi Vishwa Vidyalaya (JNKVV), College of Agriculture, Indore 452 001, Madhya Pradesh, India.2. International Crops Research Institute for the Semi-Arid Tropics (ICRISAT), Patancheru 502 324, Andhra Pradesh, India.

The state of Madhya Pradesh comprises about 14% ofthe total rainfed area of India. Soybean-basedproduction systems dominate the agriculturalscenario and have appreciably improved the

economic status of farmers of rainfed areas of thestate. Soybean is grown in about 4 million harepresenting 75% of total soybean area in India. It fitswell in the double cropping pattern with wheat and

135

other rainfed crops. The introduction of soybean inthe state in the 1970s has substantially minimized theincidence of rainy season-fallows and thus directlycontributed to sustainability of rainfed agriculture.Since 1987, when the soybean variety JS 335 wasreleased by Jawaharlal Nehru Krishi VishwaVidyalaya (JNKVV), Indore, Madhya Pradesh andimproved production technology was available,soybean area increased fourfold and productionfivefold. Soybean has greater tolerance towaterlogging than most other crops including cerealsgrown earlier in the region, which perhaps is thesingle most important reason for its widespread andrapid adoption.

Although the production of soybean is increasingdue to substantial increases in area under the crop andits yield, there is still a large yield gap of about 1.2 tha-1 that needs to be bridged (Singh et al. 2001).Soybean area in India is likely to increase to 10million ha by 2010, a larger portion of which wouldbe in Madhya Pradesh. With intensification ofsoybean production systems, about 20 million t ofsoybean can be produced. Unfortunately, at thecurrent level of productivity, there is appreciableunderutilization of land, water, nutrients, and climaticresources. Well-documented congruent relationaldata sets are needed on water balances, nutrientturnover, weather and crop development interaction,and seed viability to first understand and to tackle theproblems associated with the low productivity andunsustainability. There are a number ofsocioeconomic and technological constraints that areresponsible for low yields of soybean, resolution ofwhich in integrated fashion can help in bridging thegap between the realizable and realized yields. This ispossible through a holistic management of naturalresources of soil, water, biotic, abiotic, andsocioeconomic constraints.

The research partnership between JNKVV and theInternational Crops Research Institute for the Semi-Arid Tropics (ICRISAT), Patancheru, India primarilyfocused on enhancing productivity of the soybean-based production systems. It also covered otheraspects of enhancing farm productivity andsustainability such as diversification and cropintensification. The partners identified twobenchmark watershed sites, one at Ringnodia villagein Sanwer Tahsil of Indore district for on-farm

strategic and applied research, and another at theCollege of Agriculture, Indore for conducting on-station strategic research.

Ringnodia Micro-watershed

Characterization

Site selection

Ringnodia, a 390-ha micro-watershed, is a part of theNational Watershed Development Project for RainfedAreas (NWDPRA), Solsinda, located at 20 km fromIndore city (22°51’ N and 75°51’ E) on Indore-Ujjainhighway at an altitude of about 540 m above mean sealevel (Fig. 1). The watershed is located in the middleand lower reaches of Khan River. The topography ofthe Ringnodia micro-watershed can be divided intothree sections: a recharge zone of 18.2 ha with a slopeof 8% or more, a transition zone with slope of 2–8%,and a cultivated area of 327 ha with a slope of 2% orless, comprising medium to deep Vertisols.

This watershed was selected because the croppingintensity was very low (<130%), the rainy seasoncrop yield was low (<1 t ha-1), and irrigated area was<30% (mainly through tube wells and open wells).Due to lack of soil and water conservation measuresmost of the runoff water eroded the valuableproductive lands. The watershed had two ponds thatwere in dilapidated state and stored very little waterdue to silting and heavily breached bunds. While lackof water conservation and soil erosion due to highvelocity runoff in the transition zone was a majorproblem in the upper reaches of the watershed,waterlogging was common in the lower reaches of thewatershed. A few main watercourses had developedfrom the several field washes that carried the runoffwater towards the lower reaches, forming deepgullies. In the postrainy season, crops were grown on<30% area and generally suffered from drought.

Weather

The mean annual rainfall at Indore is about 960 mm, amajor portion of which is received between 25 and 41standard meteorological week (SMW) (Fig. 2). Therate of evaporation (mm day-1) increases steadily fromthe 2nd SMW and attains a peak during the 21st SMW,

136

SoilsThe soils of the Ringnodia micro-watershed areshallow to deep black with variable depth (~0.5 to>0.9 m depth). They occur on bare hill slopes to flattopography. The cultivated soils are mostly clay loamin texture with high moisture retention capacity,normal to somewhat alkaline in reaction (soil pH 7.5–9.3), and electrical conductivity (EC) values <1.00 dSm-1 at 25°C for most soils indicating normal soils. Thesoils in general were low to medium in soil fertilitystatus with respect to available nitrogen (N),phosphorus (P), and sulfur (S), while they were highin potassium (K) content.

The deep soils (>0.9 m) hold about 225 mm plantavailable water to a meter depth. The soils ofwatershed are classified into six soil series:

and thereafter, shows a steep decline up to 29th SMW,and a gradual decline up to 37th SMW. The rainfalland evaporation patterns showed that the periodbetween 26th and 37th SMW is surplus water period.Sowing during the rainy season is usually taken upfrom the 23rd SMW.

At Ringnodia, rainfall was near normal in 1999whereas it was 44% less in 2000 and 28% less in 2001(Fig. 3). In 2000, the beginning of the rainy seasonwas considerably delayed and there were very littlerains after 15 August; hence, the postrainy seasoncrops could be grown in <10% area only. Thebeginning of the rainy season was normal in 2001.Further, about 70 mm rainfall in October, after theharvest of soybean, helped in sowing the postrainyseason crops.

Figure 1. Location map of Ringnodia micro-watershed in Indore district of Madhya Pradesh, India.

137

Figure 2. Weather at Indore during 1971–95.

Figure 3. Daily weather during 2000–01 at Ringnodia micro-watershed recordedthrough automatic weather station.

138

Panchdaria, Runija, Kamliakheri, Sarol, Baloda, andMalikheri. The deep black soils belong to Sarol andBaloda series, which comprise members of finemontmorillonitic hyperthermic family of VerticUstochrepts and Pellusterts, respectively. Theshallow soils belong to fine clayey montmorillonitichyperthermic family of Lithic Ustochrepts.

Land use pattern

Soybean or soybean mixed with a small population ofmaize was the dominant cropping system during therainy season. There was very little rainy seasonfallow. This is in contrast to Lalatora watershed inVidisha district of Madhya Pradesh, where soybeanwas generally grown only on irrigable lands asprospects of an irrigated postrainy season crop wereassured (Vadivelu et al. 2001). However, unlike theprevious rainy seasons, in 2001 there wasconsiderable diversification with other crops such assorghum, groundnut, pigeonpea (ICPL 88039, JA 4,and local), vegetables, sole maize, and green foddercrops. There was a noticable increase in sorghum areawhich was attributable to shortage of cereal grain fordomestic consumption as most farmers could not takewheat crop in the preceding postrainy season. This ishow farmers seem to adjust their cropping patterns todrought conditions in the region.

The rains during September/October have a strongbearing on the prospects of postrainy season crops. In2000, there was scanty rainfall (<12 mm) inSeptember/October and hence the area under thepostrainy season crops was drastically reduced. Somefarmers whose tube wells had water managed toestablish chickpea and wheat crops. In 1999 and2001, there were good rains during September andOctober, which facilitated planting of wheat andchickpea as well as potato in some areas.

Socioeconomic profile

The village has a total population of 855 persons,which includes 435 males, 420 females (466 adultsand 389 children). Literacy is about 40%. Agricultureprovides the major source of income for 56% of thevillagers. They also depend on livestock and poultry,and work in the nearby industries. The landholding inthe village varies from >4 ha for the large landholders

constituting about 9% of the total farmers, 2 to 4 hafor medium landholders (26%), and <2 ha for smalllandholders (65%).

The gross annual family income from all sourceswas US$4000 for the large landholders, US$3200 forthe medium landholders, and less than US$2800 forthe small landholders. The medium and smalllandholders also depended on other sources ofincome such as working as farm labor. There were 14tractors, 14 cultivators, 10 disc harrows, 7 seed drills,10 potato planters, 50 sprayers, and 50 threshers inthe village. Livestock was maintained by allcategories of farmers. There were 143 buffaloes, 59cows, 92 oxen, and 60 goats in the village. The largelandholders spent about 32% of income onagricultural inputs and management of crops, whereasmedium and small landholders spent only about 10%.The village had 14 open wells and 19 tube wells, andtwo silted ponds (prior to our intervention) spread inabout 9 ha area.

Major constraints identified throughparticipatory rural appraisal

Through a participatory rural appraisal (PRA) thefollowing constraints were identified:• Poor resource base (poor soil fertility) and lack of

awareness and adoption of improved land andwater management practices.

• Water scarcity as a consequence of poor rainwatermanagement practices.

• Less crop diversification.• Temporary waterlogging on flat lands in lower

reaches during the rainy season.• Soil erosion in the transition zone during the rainy

season. Silt loaded runoff in the initial stages ofseedling establishment adversely affected soybeancrop in the lower reaches.

• Lack of credit to buy quality seeds and other inputs.• Lack of awareness about practicing improved

package of practices, e.g., seed treatment withfungicide and Rhizobium culture, recommendedfertilizer doses, method of fertilizer application,plant protection measures, and weed managementpractices.

• Absentee landlordism: some farmers were livingin Indore city and thus were unable to giveadequate attention to crop management.

139

Development of Ringnodia micro-watershed

Prior to undertaking any development works in thewatershed, most of the runoff originating in thehillock region used to flow through the cultivatedarea. This water flowing at a great velocity carriedconsiderable soil to the crops in the lower reaches.Soybean crop being sensitive to such deposition in theearly stages (Sullivan et al. 2001) was adverselyaffected. The turbid water eventually reached theponds and over the years reduced their capacity tostore water to the extent that these could barely store10% of the total runoff. As the storage capacity ofthese ponds had become very limited, this water usedto eventually flow wildly out of the village and joinmain watercourses to Solsinda and Katkaya Nala(Fig. 4), and eventually to Khan river that joinsKshipra river at Ujjain. In general, the watershed washaving fern type catchments. In this type ofcatchment, runoff is more since discharge is spreadover a longer period.

Land and water conservation measures

To conserve natural resources of water and soil, thefollowing measures were undertaken:• Participatory rural appraisal and topographic survey

of the watershed and its instrumentation forundertaking developmental works were completed.A watershed committee has been formed tomobilize and motivate farmers to enlist theirparticipation in the developmental works.

• Storage structures/percolation tanks wereconstructed in three sections (Table 1) coveringabout 0.3 km in early 2000 for recharginggroundwater and protecting the cultivated areafrom flooding in the lower reaches. Thesestructures received about 80% of the rainfall asrunoff from about 9.5 ha hillock area. It was,however, realized during the rainy season 2000that the storage structures could protect only asmall part of the watershed (<2 ha) as large gullieswere still forming in the cultivated area due torunoff from an adjacent 7.5 ha hillock area. A 0.3-km long diversion bund in an area contiguous tothe storage structures was therefore, constructedprior to the rainy season 2001 to reduce velocity of

runoff from this barren hillock region and safelydispose it through the watercourses.

• Twenty-five loose and wire mesh bound boulderstructures were constructed to reduce velocity ofrunoff and retain silt upstream and therebystabilize watercourses.

• Waterways were improved for safe disposal ofwater from fields.

• A temporary bund was constructed on the mainwaterway to increase discharge to ponds.

• Deepening, shaping, strengthening by pitching,and repair of the breached bunds of village pondswere carried out to increase storage capacity.These ponds can now store about 20–30 ha-mwater representing about 75% of the total runoff ofthe 100-ha catchment in the upper reaches in anormal rainfall year (Fig. 5). The de-silting of theponds served twin purposes: enhancement ofstorage capacity of ponds, and application of thesilt in fields for improving soil fertility.All these works were undertaken in a farmer

participatory mode with financial assistance ofUS$4000 from the District Collector, Indore, underthe XI Finance Commission. These works alsocreated lot of awareness and interest in soil and waterconservation works among farmers of the nearbyvillages and extension agencies operating in the area.

Quantification of the benefits of soil and waterconservation

In the rainy season 2000, the runoff from the treatedand untreated areas was small (<15 mm) and thedifferences were marginal. The construction of a 0.3-km long diversion bund in 2001 resulted in 34%reduction in runoff. Most of the runoff had occurredprior to planting of the rainy season crops.

In 2001, soil loss was high (about 0.9 t ha-1) in theuntreated area compared to 0.1 t ha-1 in the treatedarea. The organic carbon content in the silt lostthrough runoff from the treated and untreated areaswas 0.76% and 0.84% and available N was 0.014%and 0.015%, respectively. The organic carbon lossdue to runoff was 7.6 kg ha-1 from the untreated areaand 1.14 kg ha-1 from the treated area. The N loss dueto runoff was less than 1 kg ha-1 from both areas.

There was no appreciable difference in waterbalance and the water use efficiency between the

140

treated and untreated areas. The water use rangedfrom 329 mm for shallow soils to 400 mm for deepsoils in the treated area. Since there was very littlerunoff during the cropping period (no waterlogging orsilt deposition), the differences due to land and soilmanagement were unlikely to have occurred betweenthe treated and untreated areas. The water useefficiency ranged from 3.5 kg mm-1 ha-1 in shallowsoils in the treated area to 7.6 kg mm-1 ha-1 in mediumsoils in the untreated area.

The storage structures constructed in the transitionzone increased water table of the wells (groundwaterrecharge) (Fig. 6). The difference in the water tableacross wells near to the storage structure and farthestpoint was up to 3 m when measured with respect touniform reference level.

Strategic research on best-bet options

Devising efficient cropping options

The development and application of simulationmodels of crops is well established in studying cropresponse to changes in genotype, cultivar, soil,weather, climatic patterns, and managementpractices. To identify best production system, wehave generated scenarios of crop production using theAPSIM (Agricultural Production Systems Simulator)(McCown et al. 1996). This model allows modelingof crop and pasture production, residue decomposition,and soil water and nutrient flow to be readilyconfigured to simulate various production systems,including crop sequences and intercropping, and soil

Figure 4. Ringnodia micro-watershed map.(Note: The dark area is treated area and gray adjacent area is untreated control, both extending up to pond area.)

sN

141

Figure 5. Water storage in the renovated big pond in Ringnodia micro-watershed afterthe first major runoff event in June 2001.

(Note: This existing tank did not store any water due to breached bund and silting priorto our intervention as shown in the inset.)

Table 1. Salient features of water storage structures constructed in transition zone at Ringnodia micro-watershed during 1999.

Feature Segment 1 Segment 2 Segment 3

Area contributing to pond (ha) 4.5 3.5 1.5Type Earthen bund Earthen bund Earthen bundShape Semi-circular Trapezoidal TrapezoidalLength of bund (m) 110 100 65Surface area (m2) 4400 2500 1625Top width of bund (m) 3.2 3.2 3.2Storage-capacity (maximum) (ha-cm) 140 80 48Excavation/earthwork (m3) 1620 1200 460Excavation (%) 14.7 12 9.6Function Water storage Percolation Percolation

142

Figure 6. Water table fluctuations in terms of reduced level (RL) in the nearest (treated) and farthest(untreated) wells to the water storage structures in Ringnodia micro-watershed.

(Note: The arrow indicates date of construction of storage structures.)

and crop management to be dynamically simulatedusing conditional rules.

The long-term scenarios for the 18 growing seasonsderived using the APSIM model (McCown et al. 1996,Robertson et al. 2001) suggested that among therainfed systems, pigeonpea/sorghum intercrop was themost productive cropping system with >5 t ha-1 totalproduction and least variable (Fig. 7). In the postrainyseason, systems involving partially irrigated wheatwere most productive. Indeed, soybean-wheat rotationis practiced in about 64% of the total arable area in thepostrainy season in the region. Rainfed croppingsystems involving only legumes either as intercrops(soybean/pigeonpea) or sequential crops (soybean-chickpea) were less productive. The introduction ofextra-short-duration pigeonpea as intercrop withsoybean was not very productive, probably because ofits matching life cycle duration with soybean. Aprovision of limited supplemental irrigation tochickpea could further boost the prospects of onlylegume-based systems. However, a legume-cerealsystem should be more appropriate, cost effective, andhighly remunerative diversified farming system as itcan derive maximum benefit of N-fixing capability oflegumes and can meet the farmer’s varied requirementsof food and fodder.

Diversification with pigeonpea

Very few farmers in the micro-watershed grewpigeonpea as an intercrop with soybean. Thus,pigeonpea observation trials were laid out tointroduce pigeonpea (variety JA 4) as a sole crop inthe prevailing cropping system at two locations. Also,three trials of soybean/pigeonpea intercroppingsystem were laid out. The 1999 rainy season wasslightly adverse for soybean crop as there occurred along dry spell after sowing and heavy rains during thereproductive period causing a severe setback to itsproductivity. Pigeonpea crop grown either alone or asan intercrop with soybean could resist theseaberrations (Table 2). However, due to operationalinconvenience in planting (sowing in different rowsposed difficulty with existing planters) and harvesting(due to crops of dissimilar maturity period), farmersdid not favor this system. Further, with medium-duration pigeonpea, there was little prospect of asecond crop in the postrainy season. It gave <0.3 t ha-1

during 2000 due to terminal drought. In 2001 none ofthe farmers grew soybean/pigeonpea intercrop.

As medium-duration pigeonpea suffers fromterminal drought and does not allow double cropping,extra-short-duration pigeonpea genotype ICPL 88039that matures in 120 days was introduced in the

143

Figure 7. Long-term scenarios (mean of 18 years from 1984–2001) for different cropping system optionsand coefficient of variation (CV) for Ringnodia micro-watershed.

(Note: Soy = soybean, ESDP = extra-short-duration pigeonpea, MDPP = medium-duration pigeonpea,CP = chickpea, Wh = wheat, - = sequential cropping, / = intercropping.)

Table 2. Productivity of soybean and pigeonpea as sole and intercrops in farmers’ fields in Ringnodiamicro-watershed, rainy season 1999.

Yield (kg ha-1) Gross returns2

Treatment Location 1 Location 2 Location 3 Mean LER1 (Rs ha-1)

Sole soybean 960 780 765 835 1.0 6680Soybean/pigeonpea 1.7 18285Soybean 782 636 652 690Pigeonpea 940 826 786 851Sole pigeonpea 1330 1140 1239 1.0 185851. Land equivalent ratio.2. Calculated on the basis of price in January 2000.

144

watershed with 24 farmers growing the crop in about0.05-ha plots. The extra-short-duration pigeonpeaescaped terminal drought on both shallow (0.62 t ha-1)and deep soils (1.03 t ha-1) and gave higher yield thanmedium-duration cultivars. There are indications thatthe adoption of extra-short-duration pigeonpea mayincrease in near future as not only farmers ofRingnodia but also farmers in adjacent villages whobought seed from Ringnodia farmers chose tocontinue growing the crop in 2001.

Land configuration

Waterlogging is a major constraint, particularly onflat lands in normal to high rainfall years when thereare heavy silt loaded inflows from upper reaches. Weevaluated broad-bed and furrows (BBF) landformand found that in spite of below normal rainfall in 2000season soybean yield from flat landform and BBF wassame. However, farmers were reluctant to adopt thismeasure due to operational difficulties in planting,intercultural operations, and reduced plant population.

Alternatively 5 to 10 cm high ‘mini-ridges’ wereevaluated. These were made using a small duck-footshaped spade between two tines. While making theridges, this spade removed pre-emergence weeds andmade the soil loose thereby ensuring good plantstand. In 2001, although yield of soybean obtained on

mini-ridges was 8.5% greater than the adjacent flatbeds, the difference was statistically not significant,probably due to below normal rainfall.

Integrated nutrient management

Five farmers evaluated recommended doses of fertilizerwith soybean and found no increase in yields withfertilizer or Rhizobium application suggesting that Navailability did not limit soybean yield in this watershed.During 2001, 12 farmers evaluated response of soybeanto 30 kg S application through 200 kg ha-1 gypsum.Application of gypsum increased soybean yields by 70kg ha-1 (5.4% more) over the non-gypsum plots.

Integrated pest management

Insect pest damage was one of the major bioticconstraints for the yield gap between the potential andthe realized yield in the watershed. Although farmershave access to insecticides they do not sprayrecommended quantities in time. They required anexposure to integrated pest management (IPM)options so that they could control the pests withoutdepending on high doses of insecticides. Acombination of insecticides and herbal preparationslike Neemol and cow urine were found to be effectivein reducing pest population (Tables 3 and 4). The

Table 3. Girdle beetle damage and semilooper larvae on soybean in farmers’ fields at Ringnodia micro-watershed during the rainy season 2000/011.

Girdle beetle Semilooperdamaged plants2 larvae2

Seed yieldTreatment Before After Before After (t ha-1)

T1 1.10 0.37 0.34 0.28 1.56T2 0.77 0.57 0.38 0.26 1.24T3 1.08 0.78 0.48 0.41 1.04T4 1.88 1.80 0.47 0.70 0.54 SEm 0.06 0.04 0.06 CD at 5% 0.17 0.10 0.191. Data is mean of values obtained from seven farmers’ fields.

T1 = Two sprays of Quinalphos 25 EC, 2 ml L-1 at 1.0 L ha-1; T2 = One spray of Quinalphos 25 EC, 2 ml L-1 at 1.0 L ha-1 followedby Neemol spray (20 days after Quinalphos spray); T3 = Two sprays of Neemol at 5 ml L-1 (20 days interval) at 2.5 L ha-1; T4 =Absolute control.

2. Number m-1 row length at 72 h before and after spray.

145

yield increase was up to three times in the IPM plots.The increase in yield was largely due to effectivecontrol of girdle beetle and semilooper larvae.

Fruit trees and Gliricidia plantation

In a quest for diversification of income sources,planting of several fruit trees such as jack fruit,papaya, custard apple, and jamun was encouraged.Each farmer was given two seedlings of each type thatwere planted in the backyard or on field boundaries.The mean survival of different fruit tree speciesvaried from 40 to 100%. Jamun and custard applewere best established. A low rainfall in 2000 affectedthe survival of other tree species. Thus, there appearsto be a scope for introducing selected fruit trees anddiversifying farmers’ income sources. Gliricidiaseedlings were distributed among farmers forgenerating foliage to improve soil N and organicmatter. In 2000, Gliricidia seedlings failed to surviveas they were planted on diversion bund and hillock

area where little individual care could be providedand moisture stress was acute.

Technology exchange

The adoption of BBF technology was very low in theRingnodia micro-watershed as probably farmers werenot fully aware of its benefits. To familiarize with theBBF system five farmers of Ringnodia village weretaken to a neighboring benchmark site in Lalatora duringMay 2001. They learned about BBF on deep black soilsand its advantages. They also exchanged their viewsabout the features of short-duration soybean Samratwhich could be used for double cropping.

Impact on crop productivity

Impact assessment was done in 2001. The data basedon sample farmers of Ringnodia micro-watershed,revealed the overall impact of the interventions madewith respect to soil and water management (Table 5).

Table 4. Effect of integrated pest management on pest control and soybean yield.

Girdle beetle damage1 (%) Grain yieldTreatment 33 DAE 45 DAE Cumulative (t ha-1)

Monocrotophos at 30 & 45 DAE (1 L ha-1) 10.6 39.0 24.8 1.71Neemol at 30 & 45 DAE (1.5 L ha-1) 15.4 53.1 34.6 1.56Cow urine (50 ml L-1) 15.2 61.2 38.2 1.40Control 21.0 76.0 48.5 1.11 SEm 0.85 2.63 1.57 0.0771. DAE = Days after emergence of soybean crop.

Table 5. Increased productivity (t ha-1) of important crops grown in Ringnodia micro-watershed basedon sample survey.

1998/99 2000/01

Landholders Soybean Wheat Chickpea Potato Soybean Wheat Chickpea Potato

Small 0.70 1.60 0.50 15.54 0.85 1.68 0.71 18.90Medium 0.75 2.40 0.60 16.10 0.92 2.51 0.79 19.80Large 0.80 2.30 0.60 17.00 0.96 2.47 0.80 21.40

Increase (%) 21 6 36 17

146

College of Agriculture Micro-watershed

Background and treatment details

The productivity of soybean-based cropping systemsin Madhya Pradesh is constrained by prolonged wetspells during the months of July and August and dryspells during the reproductive stage of soybean.Inadequate and imbalanced nutrition managementpractices followed by farmers also contribute to lowand unstable productivity of soybean. Otherproduction constraints include weeds, insect pests,and diseases. The strategic research work at theCollege of Agriculture micro-watershed in Indore wastaken up to sustain productivity of soybean-basedcropping systems to conserve and optimally utilizenatural resources, and to determine soybean cultivarparameters for modeling.

A 2-ha micro-watershed was delineated in the ‘C’block of the College of Agriculture farm (22°43’ Nand 76°54’ E, about 540 m above mean sea level),Indore. The soils of the watershed are Vertisolsbelonging to Sarol series with medium fertility andlow organic matter and are more than a meter deep.The plant available water-holding capacity was about230 mm to a meter depth.

There were two main plots accommodating twoland configuration treatments, flat (control) and BBF,and four subplot treatments of cropping systems,soybean (JS 335)-chickpea (JG 218); soybean (JS335)-linseed (ILS 252); soybean (JS 335)-wheat(Sujata); and soybean (JS 335)/pigeonpea (JA 4)(4:2) strip cropping system. The treatments were laidout in a split plot design. The gross subplot size was40 m × 15 m. The planting dates were: 4 July 1999, 6July 2000, and 22 June 2001 for the rainy seasoncrops, soybean and pigeonpea; 26 October 1999 and10 October 2001 for the postrainy season crops,linseed, chickpea, and wheat. The harvesting dateswere October 21 and 25, in 1999 and 2000,respectively, for soybean; 3 January 2000 forpigeonpea; 28 February 2000 for chickpea and linseed;and 23 March 2000 for wheat. There were no postrainyseason crops in 2000 due to insufficient moisture insurface layers and the postrainy season crops of 2001were planted on 6–10 October. The pigeonpea cropplanted in 2000 failed due to terminal drought.

Observed and simulated performance

Soybean yields were high (2 t ha-1) in 2001, but in theprevious two seasons they were very low. Theobserved yield was compared with APSIM simulatedyield. In 2001, the observed yield and otherparameters were very similar to simulated yield(Table 6). However, in the previous two seasons, theobserved yield was significantly less than thesimulated yield (Fig. 8). This low yield could beattributed to various factors, such as weed incidenceand damage by stray cattle. There was only a marginaladvantage of about 60 kg ha-1 yield of soybean due toplanting on BBF, which was statistically notsignificant. This could be because waterlogging didnot occur in any of the seasons.

Table 6. Observed versus APSIM predictedparameters of the phenological stages, yield, andtotal dry matter of soybean variety JS 335 sownon 22 June 2001.

Parameter Observed Predicted

Days to flowering 42 48Days to maturity 100 99Yield (t ha-1) 1.89 1.94Total dry matter (t ha-1) 5.45 5.19Harvest index (%) 34.7 37.4

Figure 8. Simulated and observed soybean yields atthe College of Agriculture micro-watershed, Indore.

147

Among the component crops, yield of pigeonpeaand wheat was better than that of chickpea andlinseed. In 2001, the postrainy season crops of wheat,linseed, and chickpea were planted under zero tillageand rainfed conditions as per the treatments. Theestablishment and growth of chickpea was better thanwheat whereas linseed failed to establish.

Water balance

The field water balance of the College of Agriculturemicro-watershed was computed for 2000 and 2001rainy season considering all components. In 2000 and2001, soil moisture at sowing was 340 and 201 mm,and at harvest was 237 and 275 mm, and rainfall was231 and 375 mm, respectively. The runoff wasnegligible in both the years. The water use by thesoybean crop grown on flat and BBF systems was 333mm in 2000 and 283 mm in 2001 with no appreciabledifference between them. The water use efficiency forsoybean under different cropping systems was 1.81–2.2 kg mm-1 ha-1 for sole crop in 2000, and 5.25–5.51kg mm-1 ha-1 for intercrop to 5.70–7.37 kg mm-1 ha-1

for sole crop in 2001 regardless of land treatments.

Lessons Learned from On-farm andOn-station Strategic Research Workand Way ForwardAs is apparent from the production statistics ofMadhya Pradesh, soybean yield has doubled duringthe past 11 years; area has increased four times andtotal production five times. The increase in area underthe crop has occurred partly through reduction inrainy season fallows, and partly by substitution ofcrops such as sorghum. While a sizable area underrainy season fallows still exists in the state (Vadiveluet al. 2001), the scope of increasing productivityexists only through increased crop yields andcropping intensity in the postrainy season. Using soiland water conservation measures and appropriatecrop and nutrient management practice in Ringnodiamicro-watershed, yields as high as 2 t ha-1 can berealized with ease in a normal rainfall year.

The Ringnodia micro-watershed presented anopportunity to explore the possibility of increasingcrop yields as well as cropping intensity throughfarmers’ participation, which was low to begin with

and the soil and water resources in the village werepoorly managed. It was possible to increase the waterstorage potential up to 70% of the total runoffpotential. Also, the potential for soil erosion andwaterlogging was substantially reduced. Farmers hadgood insight about the problems, but lacked initiativeand resources to undertake massive developmentworks. Their cooperation, catalyzed by our intervention,made a difference to the village resources in severalways as mentioned earlier.

Farmers were eager to construct more waterstorage structures as they have now realized the valueof water storage without which their crops wouldsuffer, wells would go dry, and even drinking waterwould become scarce. The open and tube wells couldprovide better insurance against drought provided theground recharge could be enhanced. Farmers wereinterested in finding an alternative to soybean asprices have been low over the years.

Tractor mounted seed drill designed to sow onBBF was not found to be suitable for soybean as it hadpoor depth control, but was found to be appropriatefor chickpea sowing as it could sow deep in the moistzone. Trees provide insurance against aberrantweather. Plantation efforts did not progress due tolimited rainfall. However, custard apple and jamunhad good survival ability. As some common land wasavailable in the village there may be scope to developthese trees as a common property resource.

AcknowledgmentsWe are thankful to Dr S P Wani (ICRISAT), Dr R KGupta (JNKVV), and Mr P Pathak (ICRISAT) fortheir valuable suggestions to undertake this research.Financial assistance provided by ICRISAT throughAsian Development Bank project (RETA 5812) isgratefully acknowledged.

ReferencesMcCown, R.L., Hammer, G.L., Hargreaves, J.N.G.,Holzworth, D.P., and Freebairn, D.M. 1996. APSIM: anovel software system for model development, modeltesting and simulation in agricultural systems research.Agricultural Systems 50:255–271.

Robertson, M.J., Carberry, P.S., Chauhan, Y.S.,Rangnathan, R., and O’Leary, G.J. 2001. Predictinggrowth and development of pigeonpea: simulation model.Field Crops Research 71:195–210.

148

Singh, P., Vijaya, D., Chinh, N.T., Aroon Pongkanjana,Prasad, K.S., Srinivas, K., and Wani, S.P. 2001. Potentialproductivity and yield gap of selected crops in the rainfedregions of India, Thailand, and Vietnam. Natural ResourceManagement Program Report no. 5. Patancheru 502 324,Andhra Pradesh, India: International Crops ResearchInstitute for the Semi-Arid Tropics. 52 pp.

Sullivan, M., Vantoai, T., Fausely, N., Beuerlein, J.,Parkinson, R., and Soboyejo, A. 2001. Evaluating on-farmflooding impacts on soybean. Crop Science 41:93–100.

Vadivelu, G.A., Wani, S.P., Bhole, L.M., Pathak, P., andPande, A.B. 2001. An empirical analysis of therelationship between land size, ownership, and soybeanproductivity – new evidence from the semi-arid tropicalregion of Madhya Pradesh, India. Natural ResourceManagement Program Report no. 4. Patancheru 502 324,Andhra Pradesh, India: International Crops ResearchInstitute for the Semi-Arid Tropics. 50 pp.

149

Use of Satellite Data for Watershed Management and Impact Assessment

R S Dwivedi1, K V Ramana1, S P Wani2, and P Pathak2

Abstract

Over-exploitation of natural resources for meeting the increasing demand for food, fuel, and fiber of theever growing population has led to environmental degradation and calls for their optimal utilization basedon their potential and limitations. Information on the nature, extent, and spatial distribution of naturalresources is essential. Spaceborne multispectral measurements made at regular intervals hold immensepotential of providing such information in a timely and cost-effective manner, and facilitate studyingdynamic phenomenon. The geographic information system (GIS) provides an ideal environment forintegration of information on natural resources with the ancillary information for generating derivativeinformation which is useful in decision making. The study was taken up to generate the action plan for landand water resources development and to monitor the progress of its implementation in the Adarshawatershed, Kothapally, Ranga Reddy district, Andhra Pradesh, India. The approach involves generationof thematic maps on various natural resources through a systematic visual interpretation of satellite data,integration of such data with the ancillary information and generation of action plan in the GISenvironment, and monitoring vegetation development as a sequel to implementation of action plan bygenerating Normalized Difference Vegetation Index (NDVI) from the Indian Remote Sensing Satellite (IRS-1C/-1D) Linear Imaging Self-scanning Sensor (LISS-III) data. Soil erosion by water is the major landdegradation process operating in the watershed. There has been an improvement in the vegetation coverowing to implementation of various soil and water conservation measures, which is reflected in the NDVIimages of pre- and post-implementation periods.

Soil erosion by water and wind is the major landdegradation process in the arid and semi-arid regionsof the world. Globally, about 1.965 billion ha of landis subjected to some kind of degradation. Of this,1.094 billion ha of land is subjected to soil erosion bywater and 549 million ha of land to soil erosion bywind. On an average 25 billion tons of topsoil fromcroplands is being washed into oceans. In India alone,out of 329 million ha geographical area, 150 millionha land is affected by wind and water erosion (GOI1976). Annually about 6000 million tons of soil is lostthrough soil erosion by water (Das 1985). Also,shifting cultivation, waterlogging, and salinizationand/or alkalization have affected an estimated 4.36million ha, 6 million ha, and 7.16 million ha of landrespectively (GOI 1976). Frequent floods anddrought further compound the problem. Soildegradation contributes to an increase in atmosphericcarbon dioxide through rapid decomposition of

organic matter. In addition, rapid industrialization anddeforestation have led to building up of greenhousegases in the atmosphere resulting in global warming.Degradation of vegetation by deforestation for timberand fuel wood, shifting cultivation, and occasionallyforest fire is a very serious environmental problem.Biodiversity conservation is equally important for thesustainability of vegetation. Optimal utilization ofnatural resources based on their limitations andpotential is, therefore, a prerequisite for sustainedagricultural production.

The Role of Remote SensingFor optimal utilization of natural resources,information on their nature, extent, and spatialdistribution is a prerequisite. Until the 1920s, suchinformation had been collected by conventionalsurveys, which are labor-intensive, cost-prohibitive,

1. National Remote Sensing Agency (NRSA), Department of Space, Balanagar, Hyderabad, India.2. International Crops Research Institute for the Semi-Arid Tropics (ICRISAT), Patancheru 502 324, Andhra Pradesh, India.

150

and impractical in the inhospitable terrain. During the1920s and early 1970s, aerial photographs were usedfor deriving information on various natural resourcesincluding lands subject to degradation by variousprocesses (Bushnell 1929, USDA 1951, Howard1965, Iyer et al. 1975). Since the launch of the EarthResources Technology Satellite (ERTS-1), laterrenamed as Landsat-1, in 1972, followed by Landsat-2, -3, -4, and -5, SPOT-1, -2, -3, and -4, and the IndianRemote Sensing Satellites (IRS-1A, -1B, -1C, and-1D) with Linear Imaging Self-scanning Sensors(LISS-I, -II, and -III), spaceborne multispectral datacollected in the optical region of the electromagneticspectrum have been extensively used in conjunctionwith the aerial photographs and other relevantinformation supported by ground truth, for derivinginformation on geological, geomorphological, andhydro-geomorphological features (Rao et al. 1996a,Reddy et al. 1996); soil resources (Singh and Dwivedi1986); land use/land cover (Landgrebe 1979,Raghavaswamy et al. 1992, Rao et al. 1996b); forestresources (Dodge and Bryant 1976, Unni 1992, Royet al. 1996); surface water resources (Thiruvengadachariet al. 1996); and degraded or wastelands (FAO 1978,Karale et al. 1988, Nagaraja et al. 1992, Dwivedi etal. 1997a, 1997b). Futhermore, spaceborne multi-spectral data have been operationally used forintegrated assessment of natural resources andsubsequent generation of action plans for land andwater resources development and for assessment ofthe impact of their implementation.

Biomass has been used as a surrogate measure toevaluate the impact of the implementation of actionplan for land and water resources development. Highabsorption of incident sunlight in the visible red(600–700 nm) portion and strong reflectance in thenear-infrared (750–1350 nm) portion of the electro-magnetic spectrum has been used to derive vegetation

indices, which indicate the abundance and conditionof biomass. The index is typically a sum, difference,ratio, or other linear combination of reflectance factoror radiance observations from two or morewavelength intervals. The vegetation indices thusdeveloped are highly correlated with the vegetationdensity or cover; photosynthetically active biomass(Tucker 1979, Wiegand and Richardson 1984); leafarea index (Wiegand et al. 1979); green leaf density(Tucker et al. 1985); photosynthesis rate (Sellers1987); and amount of photosynthetically active tissue(Wiegand and Richardson 1987). Landsat-TM datahave been used for deriving various vegetationindices which in turn were used to assess the impactof soil conservation measures in the treatedwatersheds (NRSA 1996, 1999). The study reportedhere was taken up to (i) generate the action plan forsustainable development of land and water resources,and (ii) assess the impact of the action plan in theAdarsha watershed using IRS-1B/-1C and -1DLISS-II and -III data (see Table 1).

Test SiteWith an area of 1083 ha, Adarsha watershed inKothapally is bound by geo-coordinates 17°21’ to17°24’ N and 78°5’ to 78°8’ E and forms part ofShankarpally mandal (an administrative unit) ofRanga Reddy district, Andhra Pradesh, India.Vertisols and associated Vertic soils occupy 90% ofthe watershed area. However, Alfisols do occur to anextent of 10% of the watershed area. The main kharif(rainy season) crops grown are sorghum, maize,cotton, sunflower, mung bean, and pigeonpea. Duringrabi (postrainy season) wheat, rice, sorghum,sunflower, vegetables, and chickpea are grown. Themean annual rainfall is about 800 mm, which isreceived mainly during June to October.

Table 1. The details of remote sensing data used.

Satellite/sensor Path/row nos. Date of pass

IRS-1B LISS-II 26–56 25-11-1996

IRS-1C LISS-III 99–60 01-04-1996

IRS-1D LISS-III 99–60 02-04-2000 & 29-11-1999

IRS-1D PAN 99–60 01-12-1999

151

DatabaseWe have used the Indian Remote Sensing Satellite (IRS-1B/-1C and -1D) Linear Imaging Self-scanning Sensor(LISS-II and -III) and Panchromatic sensor (PAN) datafor deriving information on various natural resourcesand for generation of action plans for land and waterresources development (Table 1). In addition, Survey ofIndia topographical maps at 1:50,000 scale, andpublished soils and other resources maps and reportswere also used as collateral information.

MethodologyThe methodology involves database preparation,generation of thematic maps on natural resources, andtheir integration with the socioeconomic data toarrive at a locale-specific prescription for land andwater resources development. The schematic diagramof the approach is given in Figure 1. Once action planis implemented, the next logical step is to assess itsimpact on environment and the beneficiaries.

Figure 1. Schematic diagram of the approach.

152

Database preparation

The first step in generating the multi-sensor data setsis the geo-referencing of the image to a common mapgrid. When merging higher resolution data with thelower resolution images, usually high resolutionimage (here PAN data with 5.8 m spatial resolution) isused as a reference for respective enhancement of thelower resolution (LISS-III data with 23.5 m spatialresolution) data (Cliché et al. 1985). To begin with,the Survey of India topographical maps at 1:50,000scale were scanned on a Contex FSS-800 scanner at300 dots per square inch (dpi). The digital LISS-IIIwas later co-registered to digital, topographicdatabase on a Silicon Graphics Octane work stationusing 20 tie points (ground control point) and image-to-image registration algorithm. The IRS-1D PANdigital data was subsequently co-registered to LISS-III data following similar approach. Subsequently, theLISS-III data was resampled to 6 m pixel dimensionusing nearest neighborhood algorithms for furtherprocessing. The IRS-LISS-II data was also digitallyco-registered to IRS-1C LISS-III data and resampledto 24 m pixel dimension. The three bands, namely0.52–0.59 µm, 0.62–0.68 µm, and 0.77–0.86 µm ofLISS-III data were digitally merged with PAN usingBrovey transformation algorithm. The Broveytransformation is a formula-based process that worksby dividing the band to display in a given color by thesum of all the color layers, i.e., red, green, and blueand then multiplying by the intensity layer.

Generation of thematic maps

Thematic maps on hydrogeomorphological conditions,soil resources, and present land use/land cover havebeen generated through a systematic visualinterpretation of IRS-1B/-1C/-1D LISS-II and -IIIdata in conjunction with the collateral information inthe form of published maps, reports, wisdom of thelocal people, etc. supported by ground truth. Theinformation derived on the lithology of the area andgeomorphic and structural features in conjunctionwith recharge condition and precipitation was used toinfer groundwater potential of each lithological unit.Soil resource maps of the area have been prepared bydelineating sub-divisions within each geomorphic

unit based on erosion status, land use/land cover, andimage elements, namely color, texture, shape, pattern,and association. Soil composition of eachgeomorphic unit was defined by studying soil profilesin the field and classifying them based onmorphological characteristics and chemical analysesdata (USDA 1975, 1998).

In addition, derivative maps, namely landcapability and land irrigability maps were generatedbased on information on soils and terrain conditionsaccording to criteria from the All India Soil and LandUse Survey Organization (All India Soil and LandUse Survey 1970). Land capability classification is aninterpretative grouping of soils mainly based on:(i) inherent soil characteristics, (ii) external landfeatures, and (iii) environmental factors. Thegroupings enable one to get a picture of (i) thehazards of the soils to various factors which cause soildamage and deterioration or lowering in fertility, and(ii) its potentiality for production. The interpretationof soil and land conditions for irrigation is concernedprimarily with predicting the behavior of soils underthe greatly altered water regime brought about by theintroduction of irrigation. For arriving at landirrigability classes, soil characteristics, namely,effective soil depth, texture of the surface soil,permeability, water-holding capacity, coursefragments, salinity and/alkalinity, presence of hardpan in the surface, topography, and surface and sub-surface drainage are considered.

Land use/land cover maps have been preparedusing monsoon (kharif) and winter (rabi) cropgrowing seasons and summer period satellite data fordelineating single-cropped and double-cropped areasapart from other land use and land cover categories.Furthermore, micro-watersheds and water bodieshave been delineated and the drainage networks havealso been mapped. Slope maps showing various slopecategories have been prepared based on contourinformation available at 1:50,000 scale topographicalsheets. Rainfall data were analyzed to study therainfall distribution pattern in time and space.Demographic and socioeconomic data were analyzedto generate information on population density,literacy status, economic backwardness, and theavailability of basic amenities.

153

Generation of action plan

The generation of an action plan essentially involvesa careful study of thematic maps on land and waterresources, both individually as well as incombination, to identify various land and waterresources regions or Composite Land DevelopmentUnits (CLDU) and their spatial distribution, potentialand limitations for sustained agriculture and otheruses, and development of an integration key. It wasachieved by scanning the thematic maps on aCONTEX FSS 800 black and white scanner at 400dpi. It was followed by vectorization, projection toreal world coordinates, editing map compilation andunionizing the thematic boundaries in a geographicinformation system (GIS) domain using ARC/INFOversion-7 software. Each CLDU was studiedcarefully and a specific land use and soil and waterconservation practice was suggested based on itssustainability. Subsequently, taking landform as abase an integration key in terms of potential/limitations of soils, present land use/land cover, andgroundwater potential, and suggested alternate landuse/action plan was developed.

Implementation of action plan

The action plan and/alternate land use practices anddrought-proofing activities emerging from this approachhave been implemented by the district/mandalauthorities using the state-of-the-art technology foreach action item to fully exploit the contemporarydevelopments in agriculture, science, and technology.

Impact assessment

Since vegetation condition is the reflection of soilsand hydrological conditions which are altered in theevent of implementation of suggested action plan, ithas been taken as a surrogate parameter forassessment of the impact of such treatment in thewatershed. The Normalized Difference VegetationIndex (NDVI) values from near infrared (NIR) andred (R) band responses in the IRS-1B/-1C/-1D LISS-II and -III data were generated on a Silicon Graphicswork station as follows:

NIR (DN) – R (DN)NDVI (output DN) =

NIR (DN) + R (DN)

DN represents digital number in respective spectralbands. The equation produces NDVI values in therange of –1.0 to 1.0, where negative values generallyrepresent clouds, snow, water, and other non-vegetated surfaces, and positive values representvegetated surfaces.

Results

Natural resources

Lithologically, the watershed comprises of basalt andlaterites. The moderately dissected plateau which isinterspersed with structural valleys constitutes themajor landform. While the undissected plateau haspoor groundwater potential, the dissected plateau haspoor to moderate potential. Structural valley has goodgroundwater potential depending on the nature of thefracture. Whereas Vertic Haplaquepts have developedover structural valleys the dissected plateau supportthe development of shallow soils namely LithicUstochrepts and Lithic Ustorthents. VerticUstochrepts, however, do occur in local depressionswithin the dissected plateau. The watershed is mainlyused for raising both kharif and rabi crops. A fewpockets of land, however, is wasteland mostly in theform of land with/without scrub. The land under kharifcrops constitute the major land use and land covercategory followed by double cropped land (Fig. 2).

Action plan

Since the watershed very often experiences drought,apart from alternate land use based on potential andlimitations of natural resources, various drought-proofing measures such as vegetative barriers,contour bunding, stone check-dams, irrigation watermanagement, horticulture, groundwater developmentwith conservation measures, and fodder andsilvipasture in marginal lands have been undertaken.The suggested optimal land use practices areintensive agriculture, intercropping system, improvedland configuration, agro-horticulture, horticulturewith groundwater development, and silvipasture.

Implementation of action plan

Various soil and water and conservation measures,e.g., broad-bed and furrows, contour planting,

154

waterways and drainage channels, field bunding,wasteland development, storage of excess waterthrough construction of check-dams, dug out ponds,gabion structures, gully plugging, and increasedcropping intensity have been undertaken in thewatershed. In addition, integrated nutrient and pestmanagement trials have also been conducted.

Impact assessment

Soon after implementation of the suggested actionplan, the area undergoes transformation, which ismonitored regularly. Such an exercise not only helpsin studying the impact of the program, but alsoenables resorting to mid-course corrections, ifrequired. Parameters included under monitoringactivities are land use/land cover, extent of irrigatedarea, vegetation density and condition, fluctuation of

groundwater level, well density and yield, croppingpattern and crop yield, occurrence of hazards, andsocioeconomic conditions. Land use/land coverparameters include: changes in the number and aerialextent of surface water bodies, spatial extent of forestand other plantations, wastelands, and cropped area.

As mentioned earlier, NDVI has been used tomonitor the impact of the implementation of actionplan. A close look at the NDVI images of 1996 and2000 reveals an increase in the vegetation coverwhich is reflected in improvement in the vegetationcover (Fig. 3). The changes in the vegetation covercan be seen in the satellite image as variations in thered-colored patches, and in the NDVI images aschanges in yellow and pink colors. The spatial extentof moderately dense vegetation cover which was 129ha in 1996 has risen to 152 ha in 2000. Though thesatellite data used in the study depicts the terrain

Figure 2. LISS-III and PAN merged image and land use map of Adarsha watershed, Kothapally, India.

155

conditions during 1996, implementation activitiesstarted only in 1998. It is, therefore, obvious that itwill take considerable time for detectable changes inthe terrain and vegetation conditions.

ConclusionsThe study vividly demonstrates the potential ofspaceborne multispectral data in deriving informationon natural resources. The GIS provides an idealenvironment for integration of data on naturalresources with the ancillary information andfacilitates generation of action plan for developmentof land and water resources. After implementation ofaction plan, multi-temporal satellite data help inmonitoring its success and progress. The change invegetation cover in the Adarsha watershed as a resultof adopting soil conservation measures during 1996to 2000, is an indicator of the success ofimplementation of such action plans. High spatialresolution panchromatic and multispectral data fromIKONOS-II and the future earth observation missionssuch as Resourcesat-1, Cartosat-1 and -2, Quick Bird,Almaz-1B, etc. may further enhance our capability ofgenerating farm-level action plan for land and waterresources development, and to study the success andprogress of the implementation of such action plans.

Acknowledgments

We would like to place on record our sincere thanks toDr D P Rao, Director, National Remote SensingAgency (NRSA), Department of Space, Governmentof India, Hyderabad, India for evincing keen interest,providing necessary guidance and facilities, andactive involvement in this study. Thanks are also dueto Prof. S K Bhan, Deputy Director (Applications),NRSA for his timely and kind advice at various stagesof the study. Thematic and action plan maps of thewatershed provided by Dr A Perumal, Head,Integrated Survey Division, NRSA have been quiteuseful in finalizing the manuscript.

ReferencesAll India Soil and Land Use Survey. 1970. Soil surveymanual. New Delhi, India: All India Soil and Land UseSurvey, Ministry of Agriculture, Government of India.Bushnell, T.M. 1929. Aerial photographs of Jennigscounty. Proceedings of Indiana Academy of Science39:229–230.Cliché, G., Bonn, F., and Teillet, P.M. 1985. Integration ofthe SPOT panchromatic channel into its multispectralmode for image sharpness. Photogrammetric Engineeringand Remote Sensing 51(3):311–316.Das, D.C. 1985. Problem of soil erosion and landdegradation in India. Presented at the National Seminar on

Figure 3. Satellite and NDVI images of Adarsha watershed, Kothapply, India.

156

Soil Conservation and Watershed Management, 17–18 Sep1985, New Delhi, India.

Dodge, A.G., and Bryant, E.S. 1976. Forest type mappingwith satellite data. Journal of Forestry 74(8):526–531.

Dwivedi, R.S., Kumar, A.B., and Tiwari, K.K. 1997a.The utility of multi-sensor data for mapping eroded lands.International Journal of Remote Sensing 18(11):2303–2318.

Dwivedi, R.S., Ravi Sankar, T., Venkataratnam, L.,Karale, R.L., Gawande, S.P., Rao, K.V.S.,Senchaudhary, Bhaumik, K.R., and Mukharjee, K.K.1997b. The inventory and monitoring of eroded landsusing remote sensing data. International Journal of RemoteSensing 18(1):107–119.

FAO (Food and Agriculture Organization of the UnitedNations). 1978. The application of Landsat imagery soildegradation mapping at 1:50000 scale. FAO SummaryReport. Rome, Italy: FAO.

GOI (Government of India). 1976. Report of the NationalCommission on Agriculture. Parts V, IX, and AbridgedReport, Ministry of Agriculture and Irrigation. New Delhi,India: GOI.

Howard, J.A. 1965. Small scale photographs and landresources in Nyamweziland, East Africa. PhotogrammetricEngineering 32:287–293.

Iyer, H.S., Singh, A.N., and Ram Kumar. 1975. Problemarea inventory of parts of Hoshiarpur district throughphoto-interpretation. Journal of the Indian Society ofRemote Sensing 11:79–85.

Karale, R.L., Saini, K.M., and Narula, K.K. 1988.Mapping and monitoring ravines using remotely-senseddata. Journal of Soil Water Conservation, India 32:75.

Landgrebe, D. 1979. Monitoring the earth’s resourcesfrom space – Can you really identify crops by satellites?Presented at the National Computer Conference, New YorkCity, USA, June 7, 1979.

Nagaraja, R., Gautam, N.C., Rao, D.P., andRaghavaswamy, V. 1992. Application of Indian RemoteSensing Satellite (IRS) data for wasteland mapping. Pages170–174 in Natural resources management: A newperspective (Karale, R.L., ed.). Bangalore, India: NationalNatural Resources Management System (NNRMS),Department of Space.

NRSA (National Remote Sensing Agency). 1996.Evaluation study of treated sub-watersheds in Kundah,Lower Bhawani and Tungabhadra catchments usingremote sensing techniques. Project Report. Hyderabad,India: NRSA.

NRSA (National Remote Sensing Agency). 1999.Monitoring the impact of soil conservation measures usingremote sensing. Project Report. Hyderabad, India: NRSA.

Raghavaswamy, V., Gautam, N.C., Rao, D.P., andNagaraja, R. 1992. Land use/land cover mapping andmonitoring urban sprawl using IRS data - A few examples.Pages 120–127 in Natural resources management: A newperspective (Karale, R.L., ed.). Bangalore, India: NationalNatural Resources Management System (NNRMS),Department of Space.

Rao, D.P., Bhattacharya, A., and Reddy, P.R. 1996a. Useof IRS-1C data for geological studies. Current Science(Special Section: IRS-1C) 70(7):619–623.

Rao, D.P., Gautam, N.C., Nagaraja, R., Ram Mohan, P.1996b. IRS-1C applications in land use mapping andplanning. Current Science (Special Section: IRS-1C)70(7):575–581.

Reddy, P.R., Vinod Kumar, K., and Seshadri, K. 1996.Use of IRS-1C data in ground water studies. CurrentScience (Special Section: IRS-1C) 70(7):600–605.

Roy, P.S., Dutt, C.B.S., Jadhav, R.N., Rangnath, B.K.,Murthy, M.S.R., Gharai, B., Lakshmi, V.U., Kandya,A.K., and Thakker, P.S. 1996. IRS-1C data utilization forforestry applications. Current Science (Special Section:IRS-1C) 70(7):606–613.

Sellers, P.J. 1987. Canopy reflectance, photosynthesis andtranspiration: The role of biophysics in the linearity of theirinterdependence. Remote Sensing of Environment21:143–183.

Singh, A.N., and Dwivedi, R.S. 1986. The utility ofLandsat-MSS as an integral part of database for small scalesoil mapping. International Journal of Remote Sensing7(9):1099–1108.

Thiruvengadachari, S., Jonna, S., Hakeem, K.A., Raju,P.V., Bhanumurthy, V., Rao, G.S., Paul, P.R., Shankar,E.S., and Jayaseelan, A.T. 1996. Improved watermanagement: IRS-1C contribution. Current Science(Special Section: IRS-1C) 70(7):589–599.

Tucker, C.J. 1979. Red and photographic infrared linearcombinations for monitoring vegetation. Remote Sensingof Environment 8:127–150.

Tucker, C.J., Townshend, J.R.G., and Goff, T.E.. 1985.African land cover classification using satellite data.Science 227:369–375.

Unni, N.V.M. 1992. Forestry and ecology applications ofIRS-1A data. Pages 108–119 in Natural resourcesmanagement: A new perspective (Karale, R.L., ed.).Bangalore, India: National Natural ResourcesManagement System (NNRMS), Department of Space.

157

USDA (United States Department of Agriculture). 1951.Soil survey manual. Agriculture Hand Book No. 18.Washington, DC, USA: USDA.

USDA (United States Department of Agriculture). 1975.Soil taxonomy - A basic system of soil classification formaking and interpreting soil surveys. Agriculture HandBook No. 436. Washington, DC, USA: USDA.

USDA (United States Department of Agriculture). 1998.Key to soil taxonomy. Washington, DC, USA: USDA.

Wiegand, C.L., and Richardson, A.J. 1984. Leaf area,light interception, and yield estimates from spectralcomponent analysis. Agronomy Journal 82:623–629.

Wiegand, C.L., and Richardson, A.J. 1987. Spectralcomponent analysis. Rationale and results for three crops.International Journal of Remote Sensing 8:1011–1032.

Wiegand, C.L., Richardson, A.J., and Kanemasu, E.T.1979. Leaf area index estimates for wheat fromLANDSAST and their implications for evapotranspirationand crop modeling. Agronomy Journal 71:336–342.

158

Statistical Analysis of Long Series Rainfall Data: A RegionalStudy in Southeast Asia

J P Bricquet1, A Boonsaner2, T Phommassack3, and T D Toan4

Abstract

Many developing countries have long series of rainfall data but these have not been analyzed for theirtrends and probable occurrence. A study of rainfall data at Hoa Binh in Vietnam (41 years), LuangPrabang in Laos (51 years), and Phrae in northern Thailand (28 years) was undertaken by theManagement of Soil Erosion Consortium project. The initial analysis of rainfall data suggests that the datais homogeneous and indicates that there is a small decrease in annual rainfall in all 3 locations. The paperpresents summary statistics and frequency analysis of annual rainfall and maximum daily rainfallrespectively for the locations. The authors suggest that detailed and in-depth analysis of more rainfallstations including these 3 stations should be carried out for further study to incorporate the probabilisticinformation in decision making.

1. International Water Management Institute (IWMI), Southeast Asia Regional Office, Bangkok, Thailand.2. Royal Forest Department, Bangkok, Thailand.3. National Agronomy and Forestry Research Institute, Vientiane, Laos.4. National Institute for Soils and Fertilizers (NISF), Hanoi, Vietnam.

Management Institute (IWMI), it was possible to gainaccess to three recorded series in this region.

Location of the StudyThe study was conducted for three stations relativelyclose to the MSEC project sites, namely Hoa Binh inVietnam (20°49’ N, 105°20’ E), Luang Prabang inLaos (19°53’ N, 102°08’ E), and Phrae in northernThailand (18°08’ N, 100°10’ E) (Fig. 1). The periodof record is 41 years for Hoa Binh, 51 for LuangPrabang, and 28 for Phrae. For all stations, dailyrainfall data are available.

Results of Analysis

Homogeneity of the data

The homogeneity of hydrological or meteorologicaldata is the requirement for a valid statisticalapplication. The most commonly used informationabout non-climatic influences comes from records ofstation movement, changes in instrumentation,problems with instrumentation, sensor calibration,

The occurrence of many extreme events in hydrologycannot be forecast on the basis of deterministicinformation with sufficient skill and lead time as thosedecisions which are sensitive to their occurrence. Insuch cases, a probabilistic approach is required toincorporate the effects of such phenomena intodecisions. If the occurrence can be assumed to beindependent of time, then frequency analysis can beused to describe the likelihood of any one or acombination of events over the time horizon of adecision (WMO 1983). Interpretation of precipitationhas two major purposes. The first purpose is toevaluate the observations that sample a precipitationevent or series of events. The evaluation of theobserved sample includes consideration of extraneousinfluences, such as deficient or changing gaugeexposure, and interpretation of the effects of physicalenvironment, such as physiography. The otherpurpose is to describe the event in a form appropriatefor display, subsequent analysis, or other applications.

Little is known about trends in rainfall in theSoutheast Asia region. A study conducted by Mantonet al. (2001) showed a gap in the Indo-Chinesepeninsula. With the Management of Soil ErosionConsortium (MSEC) led by the International Water

159

changes in surrounding environmental characteristics,observation practices, and other similar features(Guttman 1998). The double mass curve analysisintroduced by Kohler (1949) is a graphical method ofidentifying and adjusting inconsistencies in a stationrecord by comparing its time trend with those of otherstations. Changes in slope of double-mass curve maybe caused by changes in exposure or location of gauge,change in procedure in collecting and processing data,

etc. (WMO 1994). The data collected at all the siteswithin the region should be highly correlated, havesimilar variability, and differ only by scaling factorsand random sampling variability. As shown in Figure2, there is no change in the slope of the curve for bothstations. So, we can consider all series ashomogeneous. With very close correlation coefficient,it is also possible, if needed, to extend the series of HoaBinh and Phrae but only for annual or monthly values.

Figure 1. Location of the three experimental stations (H) in Southeast Asia.

Figure 2. Double mass curve.

160

Annual rainfall analysis

Climatic normal is defined by the WorldMeteorological Organization (WMO) as “periodaverages of a climatic element such as temperature orprecipitation computed for a uniform and relativelylong period comprising at least three consecutive ten-year periods” (WMO 1983). Manton et al. (2001)assumes that the annual total rainfall had generallydecreased between 1961 and 1998. The number ofrainy days (with at least 2 mm of rainfall) has alsodecreased significantly in Southeast Asia. Looking atthe variations of the annual total rainfall (Fig. 3) of the

three stations, it is difficult to detect a trend of anincrease or decrease in the annual rainfall. However,a small decrease may be detected but cannot bequantified.

The 1957 value of Luang Prabang seems very low.Checking the record, we cannot reject this very smallamount of rainfall (511.1 mm). We can assume thesame comment for the 1993 value for Phrae (635.9mm). Standard statistical results are presented inTable 1. Annual rainfall usually follows a Gaussstatistical distribution. Accordingly, the differentreturn period values can be calculated with theNormal law (Table 2). The results indicated that all

Figure 3. Variation of the annual rainfall at three experimental stations.

Table 1. Summary statistics of the annual rainfall of three stations.

Standard StandardLocation Average Variance deviation Minimum Maximum kurtosis

Hoa Binh 1856.7 127290 356.8 1085.2 2671.7 –0.473Luang Prabang 1261.2 65058 255.1 511.1 1827.5 0.817Phrae 1084.6 34107 184.7 635.9 1461.3 0.273

Table 2. Frequency analysis of the annual rainfall (mm) of three stations.

Location 0.01 0.1 0.5 0.9 0.99

Hoa Binh 1064.2 1420.8 1856.7 2292.7 2649.3Luang Prabang 710.7 958.4 1261.2 1563.9 1811.6Phrae 654.9 848.3 1084.6 1321.0 1514.3

161

stations had already reached annual values close tothe millennium frequency (0.99) for the maximumvalues recorded. For the minimum values, all stationshad already reached centennial frequency (0.01).

Maximum daily rainfall

Figure 4 presents the distribution of the maximum dailyrainfall for the three stations. As usual for this type ofdata, we do not have a Gauss distribution and the datacan be adjusted with a Log Normal distribution (Fig.5). Standard statistics are given in Table 3.

The calculations of return period values are donefollowing a Pearson 3 law (Table 4). The highestobserved value for Hoa Binh station (416.4 mm) isprobably a millennium occurrence. Also, all stationshave reached maximum daily values equal to orhigher than the centennial calculation. The samecomment can be made for the lower values. Should

we see here an increase in heavy precipitation eventslike in the United States (Karl and Knight 1998)?

ConclusionsThis study is just a starting point for a deeperinvestigation on the climate in the region. As we candetect a decrease of the annual total rainfall for thethree stations, we should continue this study withmore stations from the region including South China.A limitation of such studies is the low spatial densityof stations with homogeneous data.

Acknowledgments

This study is a contribution to MSEC led by IWMIand funded by the Asian Development Bank (RETA5803). Special thanks to the National Departmentsof Meteorology of the different countries whichprovided the data.

Table 3. Summary statistics of the maximum daily rainfall of three stations.

Standard StandardLocation Average Variance deviation Minimum Maximum kurtosis

Hoa Binh 156.1 4293.2 65.5 58.3 416.4 6.620Luang Prabang 83.2 852.8 29.2 27.7 180.7 3.615Phrae 90.2 1278.8 35.7 43.3 218.2 7.062

Figure 4. Frequency of maximum daily rainfall at three stations.

162

Figure 5. Adjustment of maximum daily rainfall at Luang Prabang.

Table 4. Frequency analysis of the maximum daily rainfall of three stations.

Location 0.01 0.1 0.5 0.9 0.99

Hoa Binh 64.4 90.2 142.6 238.6 378.9Luang Prabang 34.7 50.7 79.2 121.0 168.1Phrae 48.4 58.0 81.3 132.4 222.7

ReferencesGuttman, N.B. 1998. Homogeneity, data adjustments andclimatic normals. In Abstracts, 7th International Meeting onStatistical Climatology, Whistler, Canada, May 25–29,1998.

Karl, T.R., and Knight, R.W. 1998. Secular trends ofprecipitation amount, frequency, and intensity in theUnited States. Bulletin of the American MeteorologicalSociety 79:1107–1119.

Kohler, M.A. 1949. Double-mass analysis for testing theconsistency of records for making adjustments. Bulletin ofthe American Meteorological Society 30:188–189.

Manton, M.J., Della-Marta, P.M., Haylock, M.R.,Hennessy, K.J., Nicholls, N., Chambers, L.E., Collins,

D.A., Daw, G., Finet, A., Gunawan, D., Inape, K., Isobe,H., Kestin, T.S., Lefale, P., Leyu, C.H., Lin T.,Maitrepierre, L., Outprasitwong, N., Page, C.M.,Pahalad, J., Plummer, N., Salinger, M.J., Suppiah, R.,Tran, V.L., Trewin, B., Tibig, I., and Yee, D. 2001. Trendsin extreme daily rainfall and temperature in Southeast Asiaand the South Pacific. International Journal of Climatology21:269–284.

WMO (World Meteorological Organization). 1983. Guideto climatological practices. WMO-No. 100. Switzerland:WMO.

WMO (World Meteorological Organization). 1994. Guideto hydrological practices. WMO-No. 168. Switzerland:WMO.

1

0.8

0.6

0.4

0.2

0

Cum

ulat

ive

prob

abilit

y

1 40 80 120 160 200Maximum daily value (mm)

163

Factorial Analysis of Runoff and Sediment Yield fromCatchments in Southeast Asia

T Phommassack1, F Agus2, A Boonsaner3, J P Bricquet4, A Chantavongsa5, V Chaplot6,R O Ilao7, J L Janeau4, P Marchand6, T D Toan8, and C Valentin6

Abstract

The objective of this study was to evaluate the impacts of environmental factors (e.g., climate, topography,and land use) on runoff and sediment yield from 5 catchments and 21 sub-catchments located in fivecountries (Indonesia, Laos, Philippines, Thailand, and Vietnam) of Southeast Asia. Slope parameters wereestimated using Arc-View software after catchment delineation with BASIN3 software. Land use wascharacterized from field surveys. Topographic attributes were derived from digital elevation models with a10-m mesh. Precipitation amounts were determined using both automatic and manual rain gauges.Erosion and runoff were recorded from each catchment and sub-catchments during 2001 both manuallyusing staff gauges and automatically using automatic water level recorders. Bed load sediments (BL), i.e.,the sediments trapped in the weirs, were collected and weighed after each main rainfall event.

Mean catchment area was 40.1 ha (min. 0.9 ha, max. 290 ha). The annual precipitation amount (P)ranged from 938 to 3840 mm; the precipitation ratio (Pr), i.e., the ratio between the minimum monthlyprecipitation (Pn) and the maximum monthly precipitation (Px), ranged from 0.03 to 0.31; and the slopegradient (S) from 7.7° to 31°. Land use variables included the areal percentages of annual crops (C), cropsassociated with conservation practices (Cp), fallows or pastures (Fa), and forests (Fo). Teak andeucalyptus tree plantations, and orchards were placed in a single category (O). Mean runoff coefficient (R)was 22% (ranging from 0.4 to 48%); mean BL was 3 t ha-1 (0.01–20 t ha-1). Mean suspended loadsediments (SL) was 1.76 t ha-1 (0.04–6.37 t ha-1) and mean sediment concentration (SC) was 1.63 g L-1

(0.33–3.50 g L-1).Annual runoff coefficient was predicted (R²=0.73) using a stepwise regression model combining the

impact of Px, S, and C. Similarly, SL was predicted (R²=0.75) by C and Pr; C seemed to have the maininfluence on BL (R²=0.41). No significant relation was found between runoff coefficient or sediment yieldparameters and the catchment area. Also, there was no relation between runoff and sediment yield and Foor Cp. This may be due to the absence of these two types of land use in some of the tested catchments. Theseresults clearly illustrate the impact of annual crops on runoff and sediment yield under sloping landconditions.

1. National Agriculture and Forestry Research Institute (NAFRI), Vientiane, Laos.2. Center for Soil and Agroclimate Research (CSAR), Bogor, Indonsia.3. Mae Yom Research Station, Royal Forest Department (RFD), Bangkok, Thailand.4. International Water Management Institute (IWMI), Southeast Asia Regional Office, Bangkok, Thailand.5. Soil Survey and Land Classification Center (SSLCC), Vientiane, Laos.6. IWMI-IRD, Vientiane, Laos.7. Philippine Council for Agriculture, Forestry and Natural Resources Research and Development (PCARRD), Los Banos, Laguna,

Philippines.8. National Institute for Soils and Fertilizers (NISF), Hanoi, Vietnam.

Water availability, water quality, and sedimentdelivery have become a vital issue for food security,human health, and environment. In particular, most

concerns stem from the rapid changes in land usepatterns caused by demographic, economic, political,and/or cultural transitions (e.g., Ingram et al. 1996).

164

Inappropriate land use is one of the major causes ofdecreasing water supply, and accelerated soil erosionand nutrient loss, particularly in areas of recent landcover changes such as tropical regions. Theconversion of tropical rainforests to pastures orcultivated land commonly results in a reduction ofsurface soil porosity. Increased erosion removes thetop soil where the major part of organic matter andnutrients are present. In addition to a decreased on-site productivity, these processes lead to off-siteconsequences including flooding, decrease ofgroundwater recharge, pollution by nutrients, heavymetals, and pesticides, siltation, and eutrophication ofreservoirs (IGBP 1995).

Despite the crucial need for a sound assessment ofthese processes, available data remained scarce andwas based on a single process observed at a specificscale (e.g., soil loss from erosion plots). In the slopinglands of Southeast Asia, land use changes are veryrapid due to strong demographic, economic, andpolitic drivers. In many locations, the pristine foresthas been cleared for slash and burn cultivation or formore intensified systems based on the use ofpesticides, fertilizers, and machinery. At the onset ofthe rainy season, the tilled soil left bare tends to crustand generates runoff, which favors gully erosion.Conversely, appropriate land use can lead to soil andwater conservation.

To tackle these issues, the Management of SoilErosion Consortium (MSEC) initiated a researchproject in six countries of Asia with the support of theAsian Development Bank (ADB). It aims atdeveloping, adapting, and disseminating appropriatetools and methodologies to reduce on-site and off-siteeffects of erosion on land and water resources(Maglinao and Penning de Vries 2003). Eachparticipating country has selected a catchment ofabout 100 ha and delineated four to five sub-catchments within the catchment. These catchmentsare adequately equipped to monitor runoff andsediment yield and are representative of theprevailing biophysical and socioeconomicconditions. The objectives of this study are: (i) toassess runoff and sediment yield annual budget fromthe data collected in 2001 in 5 of these catchmentsand 21 sub-catchments; (ii) to assess the impact ofland use, climate, topography on runoff, bed loadsediments (BL) and suspended load sediments (SL)

annual budgets; and (iii) to predict annual runoffcoefficient and sediment yield using a statisticalmodel.

Data AcquisitionLand use was assessed from detailed field surveys.Land use types included: forest (Fo), annual crops(C), and fallows or pastures (Fa). Crops associatedwith conservation practices (Cp) were mainly coffeeand agroforestry techniques with annual crops. Teakand eucalyptus tree plantations, and orchards wereplaced in a single category (O). Topographic featuresof catchments were derived from Digital ElevationModel (DEM) with a 10-m mesh. In most cases,DEMs have been constructed at 10-m resolution byinterpolation from digitized contours with a 5-minterval of 1:50,000 topographic map. Some accurateDEMs also have been established by interpolationfrom field spot heights using theodolithe (e.g., Laos).Spline Interpolations were performed using theinterpolation tool of Arc-View software. Catchmentswere then automatically delineated from DEMs andweirs location within UTM system using BASIN3(EPA 2000). Automatic delineation of catchmentswas used to avoid inaccuracies associated with expertjudgment. The BASIN3 delineation tool allows amanual delineation of sub-catchment boundaries anduses the multiple flow direction algorithm of Quinn etal. (1991). After removal of the sinkholes from theDEM map grid, a flow direction was automaticallyestimated to determine a stream network using athreshold area of 0.5 ha. Topographic characteristicsas slope angle were estimated from DEMs of eachcatchment using Arc-View software.

Runoff and sediment yield were monitored from the26 catchments (5 main catchments and 21 sub-catchments) in five countries (Indonesia, Laos,Philippines, Thailand, Vietnam) throughout the rainyseason of 2001. The area of the catchments and sub-catchments ranged from 0.9 to 290 ha (Table 1). Thedetailed description of the biophysical andsocioeconomic conditions in these catchments are givenin the individual country reports. Manual rain gaugeswith a daily time step and automatic weather stationswith a 6-minute data acquisition were installed in eachmain catchment. Weather and water-level data weredownloaded every week.

165

Runoff and erosion data were collected at theoutlet of each catchment where hydrologic stationshave been constructed and instrumented. Erosion andrunoff in each catchment were recorded throughout2001 both manually using staff gauges andautomatically using the automatic water levelrecorders. The location of these structures andequipment were reported using a geographicpositioning system (GPS). Water level in the weirs

was automatically recorded at a time step lower than10 minutes. Water samples were collected during themain rainfall event to assess the sedimentconcentration. The time interval for water samplingdiffered among sites. Samples were collected at timeintervals from 2 minutes to 1 hour depending on waterdischarge peaks. Sediment concentration has beenassessed in Indonesia, Laos, and Thailand catchmentsonly. Bed load sediments, i.e., the sediments trapped

Table 1. Main topographic and land use factors of the 5 catchments and 26 sub-catchments in Southeast Asia1.

Sub- Study Surfcatchment name (ha) S (°) C (%) Fa (%) Cp (%) O (%) Fo (%)

LaosS0 L0 1.2 25.0 0.0 69.0 0.0 31.0 0.0S1 L1 19.6 29.0 9.2 76.0 0.0 0.8 14.0S2 L2 32.8 27.0 1.7 19.7 0.0 7.0 11.6S3 L3 51.4 25.0 18.6 61.2 0.0 9.5 10.2S4 L4 60.2 28.0 2.3 52.7 0.0 9.9 35.1S5 L5 63.0 17.0 0.0 55.7 0.0 31.0 13.4

VietnamW1 V0 4.6 28.0 80.0 0.0 20.0 0.0 0.0W2 V1 9.4 29.0 0.0 5.0 95.0 0.0 0.0W3 V2 6.2 27.0 0.0 35.0 65.0 0.0 0.0W4 V3 11.7 31.0 0.0 60.0 40.0 0.0 0.0MW V4 59.5 25.0 28.0 22.0 20.0 30.0 0.0

IndonesiaBabon Ib 290.0 10.0 0.0 0.0 0.0 100.0 0.0Sill Is 150.0 8.0 10.0 0.0 90.0 0.0 0.0Tegalan It 3.2 14.0 50.0 0.0 0.0 50.0 0.0Rambutan Ir 2.0 12.0 0.0 0.0 0.0 100.0 0.0Kalisidi Ik 38.5 8.0 0.0 0.0 0.0 100.0 0.0

PhilippinesMain P0 84.5 14.0 23.7 45.0 16.8 0.0 15.4MC1 P1 24.9 14.0 10.0 60.2 12.1 0.0 16.1MC2 P2 17.8 14.0 47.8 39.3 5.6 0.0 11.2MC3 P3 7.9 14.0 15.2 76.0 12.7 0.0 12.7MC4 P4 0.9 25.0 42.6 31.9 0.0 0.0 21.3

ThailandW1 T1 7.5 14.1 30.0 0.0 30.0 30.0 0.0W2 T2 7.8 11.6 80.0 0.0 0.0 1.0 10.0W3 T3 2.2 12.7 3.0 2.0 2.0 92.0 0.0W4 T4 7.5 14.6 60.0 0.0 15.0 15.0 10.0Flume Tf 79.5 11.4 70.0 5.0 12.0 8.0 5.01. Surf is the catchment area; S is the slope angle; areal percentage is denoted as C for annual crops, Fa for fallows or pastures, Cp for

crops with conservation practices, O for orchards, and Fo for forest.

166

in the weirs were collected and weighed after eachmain rainfall event or once at the end of the rainyseason. Runoff and sediment yield data werecomputed over 2001 to obtain yearly means. Runoffwas derived from water depth after calibration curveswere established in the field. Mean annual suspendedsediment concentration was combined with waterfluxes data to assess the annual suspended load, usingdata interpolation between the sampling periods.

Statistical Analysis and ModelingThe relation between environmental factors and runoffcoefficient (R), BL, and SL was first investigated usingcorrelation coefficient. Variance analysis was donebetween runoff and sediment yield budget asdependent variables and environmental factors asindependent variables. These environmental factorsincluded the annual precipitation amount (P), theprecipitation ratio (Pr) between the minimummonthly precipitation (Pn) and the maximum monthlyprecipitation (Px), the slope gradient (S), thecatchment area (Surf), and the areal percentage ofeach land use type. Statistical analysis was carried outwith the Statistica® package for use on a personalcomputer (StatSoft 1995). The statistical significanceof results was evaluated using the P-level,considering values that yield 0.05 (*) and 0.01 (***)as statistically significant and highly significant,respectively.

Forward stepwise regression analyses were used topredict runoff and sediment yield budgets usingenvironmental factors as predictors. Regressionanalyses were performed using the Statistica®

package. Only parameters with statistical significanceat the 0.01 level were considered for computingpredictive equations and reporting results. Dependingon the available data, these analyses have beenconducted for a varying number of catchments andsub-catchments, i.e., n=16, 21, and 11 for R, BL, andSL, respectively.

Stepwise Regsression AnalysisStatistics of environmental variables and sedimentvariables are presented in Tables 2, 3, and 4. Runoffand sediment yield were not significantly related toSurf. Runoff coefficient was highly correlated to O

(R = –0.87) and slope angle S (R = 0.56). BL wassignificantly correlated to C (R = 0.83) while SL wasmainly related to C (R = 0.76) and Pr (R = 0.61).

The results of the stepwise regression analysis aregiven in Table 5. The relationships between the observedand predicted mean annual runoff and between theobserved and predicted mean annual sediment load arepresented (Figs. 1 and 2). Deviations from the regressionline are very low for runoff values lower than 15%. Forhigher runoff values, spot scattering slightly increased,deviations being lower than 30% of runoff value. Lessthan half of the variance of BD is explained by C(R² = 0.41). The regression model for SL included Prand C. It explained a high proportion of the variance ofSL (R2 = 0.75), with an equal distribution of theregression errors.

It is interesting to note that the model deviationerrors could be ascribed to country conditions,suggesting some site specificity. For instance, thePhilippine catchments were systematically situatedunder the regression line for BL lower than 10 t ha-1,the highest loads being highly underestimated (Fig.3). BL of Vietnam catchments were underestimatedfor low erosion rates and over estimated for highvalues. This regression model seems to best predictBL for Indonesian and Laotian catchments.

DiscussionUnder analogous circumstances, runoff coefficientand sediment yield usually decrease when thecatchment area increases. Despite values rangingfrom 0.9 to 290 ha, such relation could not beestablished here, presumably because of theoverriding influence of land use. As observed also inWest Africa, the runoff coefficient increases due to astronger development of surface crusts even whenannual rainfall decreases (Valentin 1996). The arealpercentages of fallow, forest, or crops associated withconservation practices did not statistically influencerunoff production or sediment yield, probablybecause their effects were counterbalanced by otherfactors as slope angle for runoff and the arealpercentage of annual crop for sediment yield.Contrary to what have been observed on micro-plotsin the Thai catchment, runoff at the catchment scaleincreases with mean slope angle. The twoobservations are consistent because the water flux

167

Table 2. Main statistics for environmental factors1.

Surf P Pm Pr S C Fa Cp O FoDescription (ha) (mm) (mm) (%) (°) (%) (%) (%) (%) (%)

Mean 40.15 2019 466 16 18.8 22.4 27.5 16.8 24.0 7.2Minimum 0.94 1385 275 3 8.0 0.0 0.0 0.0 0.0 0.0Maximum 290.00 3840 672 31 31.0 80.0 76.0 95.0 100.0 35.1SD 62.08 938 151 9 7.7 26.8 28.4 27.1 34.8 8.91. Surf is the catchment area; P is the annual precipitation amount; Pm is the maximum monthly precipitation; Pr is the ratio between the

minimum monthly precipitation (Pn) and Pm; S is the slope angle; the areal percentages are denoted as C for annual crops, Fa forfallows or pastures, Cp for crops with conservation practices, O for orchards, and Fo for forest.

Table 3. Main statistics for runoff and sediment yield variables1.

R BL SL SCDescription (%) (t ha-1) (t ha-1) (g L-1)

Mean 22.3 3.01 1.76 1.63Minimum 0.4 0.01 0.04 0.33Maximum 48.0 20.02 6.37 3.50SD 17.8 4.92 1.99 1.041. R = runoff coefficient; BL = bed load sediments; SL = suspended load sediments; and SC = sediment concentration.

Table 4. Correlation coefficients between environmental factors, runoff, and sediment yield variables1.

Surf P Pm Pr S C Fa Cp O FoDescription (ha) (mm) (mm) (%) (°) (%) (%) (%) (%) (%)

R 0.04 –0.95* –0.91* 0.95* 0.55* 0.25 0.53 0.21 –0.87* 0.39BL –0.29 –0.16 0.01 0.22 –0.37 0.83* –0.4 –0.07 –0.27 0.04SL –0.08 –0.54 –0.3 0.61* –0.28 0.76* –0.37 0.43 –0.47 0.21SC 0.33 0.35 0.2 –0.39 0.23 –0.04 0.18 –0.45 –0.05 0.211. Surf is the catchment area; P is the annual precipitation amount; Pm is the maximum monthly precipitation; Pr is the ratio between the

minimum monthly precipitation (Pn) and Pm; S is the slope angle; the areal percentages are denoted as C for annual crops, Fa forfallows or pastures, Cp for crops with conservation practices, O for orchards, and Fo for forest.* = Significant at P = 0.05.

Table 5. Regression analysis between runoff coefficient, sediment yield factors, and the most relevantenvironmental factors.

Regression equations No. of catchments R²

Runoff coefficient (%)Rest = 24.1 – 0.48 Px + 0.42 S + 0.42 C 16 0.73

Bed load sediment (t ha-1)BLest = 0.28 + 0.648 C 21 0.41

Suspended load sediment (t ha-1)SLest = –0.79 + 0.638 C + 0.503 Pr 11 0.75

168

Figure 1. Observed mean annual runoff as a function of predicted runoff for a data set of16 catchments in Southeast Asia.

(Note: Y=X and 95% confidence lines are presented.)

Figure 2. Observed mean annual suspended sediment load as a function of predictedsuspended sediment load for a data set of 11 catchments in Southeast Asia.

(Note: Y=X and 95% confidence lines are presented.)

169

recorded from a steep slope catchment can result fromthe combination not only of surface runoff water butalso of water that has been infiltrated in the hillslopeand ex-filtrated in the bottom as suggested bypreliminary isotopic assessment in Laos.

The deviation errors from the model for bed load,which are specific to each country, can reflect theimpact of other factors such as surface stoniness, soilresistance to shear stress, and mean depth of soil.Thinner soil with low soil water storage may affectsediment losses (Burt 2001). Under these slopingland conditions, the percentage of annual crops, notassociated with conservation practices, appears as themain factor controlling sediment yield, both in terms

of suspended load and bed load, irrespective of thesize of the catchment. These results indicate thatannual crop proportion seems to be the key parameterfor bed load and suspended load production. Moresurprisingly, no significant relation was foundbetween runoff coefficient or sediment yield variablesand the areal percentages of land occupied by forestor crops associated with conservation practices. Thismay be due to auto-correlation among the variousland use types. This study should be conducted over alonger period to validate and refine these preliminarystatistical analyses and to test the influence of otherland use types on runoff and sediment yield from thesame catchments.

Figure 3. Observed mean annual bed load as a function of predicted bed loadfor a data set of 21 catchments in Southeast Asia.

[Note: Y=X and 95% confidence lines are presented. Relation between estimated valuesand catchment named after country: Indonesia (It, Ib, Ir, Ik), Laos (L0 L1, L3, L4),

Philippines (P0 to P4), Thailand (T1, T2, T4), and Vietnam (V0 to V4).]

22

20

18

16

14

12

10

8

6

4

2

0

–2

Predicted bed load (t ha-1)

Obs

erve

d be

d lo

ad (t

ha-1

)

–2 0 2 4 6 8 10 12 14 16 18 20 22Regression95% confidence

170

ConclusionsThis study leads to three main conclusions:• The MSEC network of catchments and sub-

catchments provides an invaluable tool to test theimpact of land use changes on runoff productionand sediment yield on various biophysical andsocioeconomic conditions.

• The annual runoff coefficient increases due tosurface layer crusting even with decreasing annualrainfall and increasing slope angle. Thus moreattention should be paid to water conservation inthe drier catchments with very steep slopes (e.g.,Laos) than in wetter countries with gentler slopes(e.g., Indonesia).

• The areal percentage of the catchment cultivatedwith annual crops is the best predictor of sedimentyield.

ReferencesBurt, T.P. 2001. Integrated management of sensitivecatchment systems. Catena 42:275–290.

EPA. 2000. http://www.epa.gov/ost/ftp/basins/system/BASINS3.

IGBP (International Geosphere-Biosphere Programme).1995. Land-use and land cover change. Science ResearchPlan, Stockholm. 123 pp.

Ingram, J., Lee, J., and Valentin, C. 1996. The GCTE soilerosion network: a multidisciplinary research program.Journal of Soil and Water Conservation 51(5):377–380.

Maglinao, A.R., and Penning de Vries, F. 2003. TheManagement of Soil Erosion Consortium (MSEC): Linkingland and water management for sustainable uplanddevelopment in Asia. Pages 31–41 in Integrated watershedmanagement for land and water conservation andsustainable agricultural production in Asia: proceedings ofthe ADB-ICRISAT-IWMI Project Review and PlanningMeeting, 10–14 December 2001, Hanoi, Vietnam (Wani,S.P., Maglinao, A.R., Ramakrishna, A., and Rego, T.J., eds.).Patancheru 502 324, Andhra Pradesh, India: InternationalCrops Research Institute for the Semi-Arid Tropics.

Quinn, P., Beven, K., Chevallier, P., and Planchon, O.1991. The prediction of hillslope flow paths for distributedhydrological modelling using digital terrain models.Hydrological Processes 5:59–80.

StatSoft. 1995. Statistica for windows. Computer programmanual. Tulsa, Oklahama, USA: StatSoft Inc.

Valentin, C. 1996. Soil erosion under global change.Pages 317–338 in Global change and terrestrial ecosystems(Walker, B.H., and Steffen, W.L., eds.). InternationalGeosphere-Biosphere Programme Book Series, No. 2.Cambridge, UK: Cambridge University Press.

171

Land Use Characteristics and Hydrologic Behaviorof Different Catchments in Asia

M G Villano1 and R O Ilao2

Abstract

Development and management of natural resources is crucial for increasing and sustaining the systemproductivity and maintaining ecological balance. There is lack of information in most of the less developedcountries to know the effect of land use on hydrological behavior. Such information is useful forformulation and implementation of the appropriate natural resource management policies. The paperpresents the hydrological effects of various land uses in some catchments in Asia. This activity is a part ofthe on-going catchment study of the Management of Soil Erosion Consortium (MSEC)-International WaterManagement Institute and identifies indicators/criteria such as peak runoff, total runoff, duration ofrunoff, time of concentration of runoff, sediment load, and groundwater levels to evaluate impact ofchanges in land uses. A set of recommendations is drawn for collection of reliable hydrologic data formeaningful analysis and practical utility of such data to planners and policy makers.

Quantitative data on the impact of different land useson the behavior of catchments are indispensableinformation not only for extension workers to advisefarmers on conservation farming techniques, but alsofor planners and decision-makers to formulateappropriate and implementable policies forsustainable development and management of thenatural resources. Such information is also veryuseful to engineers for the proper design of watercontrol structures for irrigation, flood control, andsoil and water conservation. Among others, the moresignificant environmental impacts of land use andcover type modifications are hydrologic in nature.Specifically, these include variation in runoff, soilerosion, and sediment transport, and groundwaterquantities. In most of the less-developed countries,however, there is practically a dearth of factual dataon the effects of land use on the hydrologic behaviorof a catchment. Indeed, the on-going catchment studyof the Management of Soil Erosion Consortium(MSEC)-International Water Management Institute(IWMI) is a move towards addressing such data gaps.

The objectives of this paper are to: (i) identifycriteria and indicators for evaluating the relationships

1. University of the Philippines Los Banos (UPLB), Laguna, Philippines.2. Philipines Council for Agriculture, Forestry and Natural Resources Research and Development (PCARRD), Los Banos, Laguna,

Philippines.

of land use characteristics and hydrologic responsesin a catchment level; (ii) present MSEC methodologyof hydrologic field data monitoring and preliminaryanalysis of available results; and (iii) make relevantrecommendations.

The Hydrologic Cycle

The hydrologic cycle or water cycle may bevisualized as starting with the upward movement ofwater in the form of vapor (caused by radiation fromthe sun) from bodies of water (e.g., ocean, lakes,reservoirs, and rivers) and moist soil by evaporationand from vegetation by transpiration (Fig. 1). Suchvapor is transported by air mass, forms clouds, andmay fall on the land surface vegetation and into thebodies of water including the ocean as precipitation.Precipitation reaching land surfaces is disposed off indifferent ways. Some infiltrate below the ground andothers move down the land slopes as surface runoff oroverland flow. As surface runoff moves over the landsurface, some of it may still infiltrate below theground and the rest may reach depressions, gullies,streams, and rivers and finally leads to the lake or

172

Figu

re 1

. The

hyd

rolo

gie

cycl

e.

173

ocean. Surface runoff added to the stream flow isreferred to as direct runoff.

The portion of precipitation that infiltrates belowground surface enhances the soil water. If the infiltratingwater encounters an impermeable soil layer, it may seeplaterally and will return to the surface at some locationaway from the point of entry. This component of runoffis known variously as interflow, storm seepage, or quick-return flow. As more fraction of precipitation infiltratesbelow the ground surface, the soil water furtherincreases and reaches the maximum water retentioncapacity of the soil. Beyond this capacity, succeedinginfiltration water may percolate down the soil profile andinto the groundwater (aquifers). This could increase theunconfined aquifer’s water level or piezometric waterlevel of confined aquifers. Some of this move asgroundwater flow towards the stream as its base flow.Confined aquifer is the main source of free-flowing well.A perched aquifer is a layer of semi-permeable soilformation above the normal water table. It couldintercept and store infiltrating water and with favorableconditions it may release water out to the mountain sidesas spring water. Such water could contribute to the baseflow of a stream.

Hydrologic Effects of Land Use andCover Type ModificationsAs presented by David (1984), typical land uses may becategorized as forest land, grassland, arable land, andurban and industrial lands. In a given catchment, thechange of existing land use to another or the conversionof a significant portion of it to other land uses may resultin positive or negative effects. For example, conversionof a significant forest area to arable land could acceleratesoil erosion leading to decreased crop productivity level,increased sediment material downstream, and decreasedfresh water supply availability. On the other hand, theintroduction of soil conservation measures (e.g., contourcultivation, hedgerows, terracing, and check-dams) inavailable lands may control soil erosion and runoff, andincrease dependable water supply.

Effects of land uses on infiltration andpercolation

It is through infiltration and percolation thatgroundwater in aquifers is recharged leading to

sustained freshwater supply and enhancedenvironment in general. Of the above mentioned landuses, forest cover has the greatest effect on enhancingthe infiltration and percolation rates of the soil. Theincrease in infiltration is mainly due to the addition ofhigh organic matter of decayed litter that falls on theground surface while the improvement in percolationrates of the soil profile is affected by the decayed deep-penetrating tree roots and burrowing earthworms.While well-managed grassland also improves surfacesoil infiltration rates, the relatively shallower rootscould hardly improve the percolation rates of lowersoil profile. Generally, infiltration and percolation ratesof arable lands are relatively low, especially if noadequate soil and water conservation is practiced.

Effects of land uses on runoff

Storm runoff hydrograph

Total runoff measured in a stream is the sum ofsurface runoff (or direct runoff), interflow, and baseflow (or groundwater flow). The most dramaticimpacts of land use and cover type are on thevariation with time of these stream flow componentsin a storm hydrograph (Fig. 2). Specifically, therelevant hydrograph components affected by land useare the peak flow, direct runoff volume, and baseflow. For good soil and water conservation, directrunoff should be less but base flow should be high.

The magnitude of surface runoff is highlydependent on the infiltration and percolation rates ofthe soil. The effects of different land uses or covertypes on surface runoff are related to their effects oninfiltration and percolation rates. Generally, forestland will yield the least surface runoff volume andpeak flow, followed by grassland, arable land, andurban and industrial lands in increasing order. Interms of runoff coefficient values (the fraction ofprecipitation that reaches the stream as surface ordirect runoff), estimated runoff-producing potentialsof vegetation and land use common in the Philippinewatersheds are given in Table 1. Other sample valuesare given in Tables 2 and 3.

Forest land with dense trees effectively restrict fastmovement of surface water resulting in longer time topeak than arable lands without soil conservationmeasures. Arable land with hedgerows, terraces,check-dams, and other conservation structures,

174

however, may exhibit longer time to peak thangrasslands or urban and industrial lands. There is alimit to the amount of precipitation that could beallowed to infiltrate and percolate below groundsurface. Thus, extended downpours with intensityvery much greater than the infiltration rate of soiloccurring in forest lands, could still result into verydestructive flash floods downstream.

Annual runoff hydrograph

Annual hydrograph shows the variation of daily,weekly, or decadal mean flows over a year. Based onthe annual hydrographs, streams may be classifiedinto three classes as: (i) perennial, (ii) intermittent,and (iii) ephemeral (Subramanya 1984) (Fig. 3).

A perennial stream always carries some flows.There is considerable base flow (or groundwaterflow) throughout the year as the water table is abovethe stream bed (an effluent stream) even during the

dry season. Normally, the catchment or recharge areaof this stream is of good cover such as primary forestor well managed grasslands.

An intermittent stream has limited contribution fromgroundwater. The stream may be effluent during wetseason, but becomes influent during the dry season.Depending on the catchment’s geological nature, such astream may become perennial under favorable land useand ground cover which could significantly enhanceannual recharge of the groundwater.

A stream which does not have any base flowcontribution is referred to as ephemeral. Its annualhydrograph (Fig. 3) shows series of short-durationspikes marking flash flows in response to storms. Itbecomes dry soon after the end of the storm flow. Thisstream may become intermittent or even perennialwith improved land use and ground cover if the soiland geological characteristics would allow significantsubsurface water recharge and fair to goodgroundwater flow.

Figure 2. Components of a storm hydrograph.

175

Table 1. Estimated crop cover coefficient (C) values for the common cover conditions of Philippinewatersheds.

Cover C value C value (%)

Bare soil 1.0 100Primary forest (with dense undergrowth) 0.001 0.1Second growth forest with good undergrowth and high 0.003 0.3mulch coverSecond growth forest with paths of shrubs and plantation 0.006 0.6crops of 5 years or moreIndustrial tree plantation (ITP)

Benguet pine with high mulch cover 0.007 0.7Mahogany, narra, 3–8 years with good cover crop 0.01–0.05 1–5Mahogany, narra, 8 years or more with good undergrowth 0.01–0.05 1–5Yemane, 8 years or more 0.08 8Mixed stand of ITP plant species, 8 years or more 0.07 7

Agroforestry tree speciesCashew, mango, and jackfruit, less than 3 years, without 0.25 25intercrop and with ring weedingCashew, mango, and jackfruit, 3 to 5 years without 0.15 15intercrop, without ring weedingCashew, mango, and jackfruit with intercrop or 0.08 8native grass undercoverMixed stand of agroforestry species, 5 years of 0.08 8more with good coverCoconut with tree intercrops 0.05–0.10 5–10Coconut with annual crops as intercrop 0.10–0.30 10–30Ipil-ipil, good stand, first year with native grass intercrop 0.20 20Ipil-ipil, good stand, 2 years or more with high mulch cover 0.10 10Ipil-ipil for leaf meal or charcoal 0.30 30

GrasslandsImperata or Themeda grasslands, well established 0.007 0.7and undisturbed, with shrubImperata or Themeda grasslands, slightly grazed, 0.15 15with patches of shrubShrubs with patches of open, disturbed grasslands 0.15 15Well-managed rangeland, slightly grazed cover of 0.30–0.80 30–80slow development, first yearWell-managed rangeland cover of fast development, 0.05–0.10 5–10first year, ungrazed

continued

176

Table 1. continued.

Cover C value C value (%)

Well-managed rangeland, slightly grazed cover of 0.01–0.10 1–10slow development, 2 years or moreWell-managed rangeland cover of fast development, 0.01–0.05 1–5ungrazed, 2 years or moreGrassland, moderately grazed, burned occasionally 0.20–0.40 20–40Overgrazed grasslands, burned regularly 0.40–0.90 40–90

Annual cash cropsMaize, sorghum 0.30–0.60 30–60Rice 0.10–0.20 10–20Groundnut, mung bean, soybean 0.30–0.50 30–50Cotton, tobacco 0.40–0.60 40–60Pineapple 0.20–0.50 20–50Banana 0.10–0.30 10–30Diversified crops 0.20–0.40 20–40New Kaingin areas, diversified crops 0.30 30Old Kaingin areas, diversified crops 0.80 80

OthersBuilt-up rural areas, with home gardens 0.20 20Riverwash 0.50 50

Source: David (1988).

Table 2. Sediment and surface water yields1.

Average Average Runoff Annual sediment yield (t ha-1)Land use or annual annual coefficientcover type rainfall (mm) runoff (mm) (%) Range Mean

Open landCultivated 1321 406 31 8.11–106.40 53.75Pasture (one unit) 1295 381 29 2.94–5.02 3.98

Forest land 14 0.32Abandoned fields 1295 178 10 0.02–1.33 0.25Depleted hardwoods 1295 127 2 0.05–0.79 0.05Pine plantations 1372 25 18 0.00–0.20 0.05Mature pine-hardwoods2 1295 229 0.02–0.10

Gullies3 1346 - - 208.31–986.69 449.721. Data are means of 9 values, 3 replications of each cover for the 3 years, 1959–61 except pine-hardwoods (1960–61).2. These watersheds are on hydrologically shallow soils.3. Average annual and sediment outflow from 7 gullies for the 5 years, 1956–60.Source: Ursic and Dendy (1965).

177

Figure 3. Annual runoff hydrographs.(Source: Subramanya 1984)

178

Effects of land uses on soil erosion andsediment transport

Soil erosion by water may be defined as thedetachment and transport of soil from the land surfaceby rainfall and runoff energy (wind erosion is lesssignificant than water erosion in Asian countries). Thesubsequent transport of detached particles downstreamthrough channels or streams is referred to as sedimenttransport. Sediment discharge, sediment load, andsediment yield refer to the amount of sedimentdelivered downstream per unit area per unit time bythe catchment through its channels or stream network.The magnitude of soil erosion is dependent on theinteractive effects of climate, soil characteristics,geology, topography, surface cover, and land use. Fora given catchment of known soil types, geology, andtopography, however, the rate of soil erosion may beexpected to vary with the interaction of precipitationwith land use and cover type.

Different ground cover types and land uses vary inthe degree of protecting the soil against the detachingand transporting power of raindrops and surfacerunoff. Thus, lands fully covered with forest trees andgrasses are ably shielded against raindrop erosionwhile arable lands with bare areas are vulnerable to it.As the kinetic energy or erosive power of surfacerunoff is also dependent on its mass, land use or covertype which enhances infiltration and percolation andtends to yield lesser runoff; hence, less soil erosion.

The effect of land use is on the production ofsediment material through erosion and its efficacy inintercepting or trapping such sediment material. Thearable lands are expected to yield more sedimentmaterials compared to forest and grasslands.

Criteria and Indicators for EvaluatingHydrologic Impacts of Land Use andCover Type Modification

The more significant hydrologic impacts of land useand cover type that need quantification are runoff,groundwater recharge, soil erosion, and sedimenttransport. For a particular catchment of known soiltype, geology, and topography, the variations in thesehydrologic processes are mainly dependent on theinteractions of other hydrologic processes, namely,precipitation, infiltration, and percolation.

The best way of characterizing land use effect onrunoff behavior is to quantify the more relevantcomponents of storm hydrographs of a catchment. Ofspecific interest are the peak runoff, time to peak,direct runoff volume, and base flow volumeincrement. The first three are more relevant to theirimpacts on flood hazards, soil erosion, and sedimenttransport (Table 4). Base flow is the available streamflow during non-rainy days. Another indicator of therunoff-producing effect of land use is the runoffcoefficient (ratio of direct runoff to precipitation) ofthe catchment. Indicators of the effects of land use ongroundwater recharge may include measurement ofchanges in spring water discharge, depth of staticwater table of unconfined aquifers and depth of staticpiezometric water level in confined aquifers. Annualstream hydrographs could also indicate possible long-term changes in groundwater storage.

On a catchment level, the effects of the differentcombinations of land uses and cover types on soilerosion and sediment transport may be assessed bymeasuring the sediment yield of the stream flow.

Table 3. Barlow’s runoff coefficient Kb in different catchments (for use in Uttar Pradesh, India).

Kb1 (%)

Class Description of catchment Season 1 Season 2 Season 3

A Flat, cultivated, and absorbent soils 7 10 15B Flat, partly cultivated, stiff soils 12 15 18C Average catchment 16 20 32D Hills and plains with little cultivation 28 35 60E Very hilly, steep, and hardly any cultivation 36 45 811. Season 1: Light rain, no heavy downpour; Season 2: Average or varying rainfall, no continuous downpour; and Season 3: Continuous

downpour.Source: Subramanya (1984).

179

However, the total stream sediment load measured atthe gauging station could also include eroded soilsfrom gullies, stream banks, and landslides. If suchadded sediments are significant in magnitude,appropriate corrections on the total sediment yieldmay have to be done before computing the soilerosion rate (sheet and rill or interrill).

MSEC Methodology of HydrologicData Monitoring

Hydrologic data collection andmeasurement

The hydrologic data being monitored are listed ineach of the MSEC catchments in Table 5. Figure 4shows a typical instrumentation layout. Regular fieldmonitoring of the possible change in the land uses andcover types is being done in each catchment. Specificfarm activities are also noted including soil and waterconservation measures being practiced.

Data analysis approaches

Analysis and interpretation of field data may be donein two ways. One is to find effects of land usemodifications on selected hydrologic criteria of agiven catchment of known soil, geologic formation,and topographic characteristics. It may takeconsiderable time before statistically significantchange on some hydrologic indicators could beobserved. It may, for instance, take several yearsbefore a newly-started reforestation project couldeffect significant change on a storm hydrographcomponent such as peak flow, time to peak, and baseflow. The factorial analysis of runoff and sedimentyield from 5 catchments and 21 sub-catchments ofSoutheast Asia conducted by Phommassack et al.(2003) is of this type.

Another way is to determine the significance ofland use and other catchment characteristics such assoil infiltration rate, land slope, and rainfall intensityon the variations of hydrologic criteria (such as runoffcoefficient, peak runoff, base flow, and sediment

Table 4. Criteria and indicators for evaluating hydrologic effects of land use and cover type.

Some quantifiableMajor hydrologic criteria Major impacts onenvironment hydrologic indicators

Runoff Flood water hazards downstream Storm hydrograph peak flow;storm hydrograph time to peak.

Water source for dams/reservoirs Storm hydrograph directrunoff; downstream flooddamages; runoff coefficient.

Available stream water Storm hydrograph base flow;during dry days annual stream flow hydrograph.

Groundwater recharge/flow Fresh water resources availability Change in soil water storage;and sustainability change in spring water discharge

from perched aquifers; change indepth of static water table ofunconfined aquifer; change in depthof static piezometric water level inconfined aquifer; storm hydrographbase flow; annual stream flowhydrograph.

Soil erosion and Potential downstream Stream flow sediment load/sediment transport sedimentation problems yield; sediment delivery ratio.

180

yield). Based on about a year of field observationsfrom MSEC catchments, such analysis tend toindicate that areal percentage of annual crops seem tohave the main influence on bed load sediment.

The following analysis is focused on analysisusing available preliminary observations. While thelimited data may not demonstrate significantrelationships between land use and selectedhydrologic indicators within a catchment, thispresentation mainly intends to demonstrate sampleprocedures in analyzing some of the identifiedhydrologic indicators and possibly establish base datafrom which to compare future observations in thesame catchment.

Initial ResultsInitial hydrologic data of three MSEC-participatingcountries in Asia are given in Table 6.

Initial values of runoff coefficientsAs discussed earlier, runoff coefficient is an indicatorof the surface runoff-producing characteristics of thecatchment. The smaller the value of runoffcoefficient, the better. More fraction of the rainfall

infiltrates below the ground surface thus adding to thesoil water storage and even to the water table throughpercolation.

Storm runoff coefficients range from 0.01 to18.7% for main catchments and from 0.01 to 0.05%for micro-catchments (Table 6). For monthly runoffcoefficients, the values range from 0.55 to 10.20% formain catchments and from 0.01 to 18.89% for micro-catchments. These values are within the ranges ofestimated C values for similar vegetation or land usespresented in Table 1. Serving as base data, theirvariations with time and land use change are to beobserved in succeeding years.

Initial values of storm hydrographbehavior

Only one catchment (Mapawa catchment) hasavailable data for storm hydrograph analysis. Threesample storms were selected from 2000 and four from2001. The hydrograph of a selected storm is shown inFigure 5 while the magnitudes of relevant hydrographcomponents of the seven sample storms are given inTable 7. These data will serve as baseline informationfor comparing hydrograph behavior of storms inthe future as appropriate conservation farming

Table 5. Hydrologic data and measuring instruments in MSEC catchments.

Hydrologic data Measuring instrument per catchment

PrecipitationTotal daily rainfall 5–8 manual rain gauges.Rainfall intensity 1 automatic mini-weather station with automatic rainfall recorder.

Runoff/Stream flow Compound sharp-crested weir (contracted rectangular weir with V-notch). Waterlevel is being measured by a staff gauge and an automatic water level recorder.

Sediment load/yieldBed load Approach channel of the weir serves as bed load sediment tank or interceptor.

Such trapped sediments are considered bed load. After every sediment-producingstorm rainfall, water is drained from the tank and then the bulk sediment volumeis measured. A sample is taken out and oven-dried in the laboratory to get the drymass density. This density multiplied by the bulk volume is the total sediment drymass for the storm.

Suspended load Initially, no sampler was used. Manual scooping was done with can sampler orseries of sampling cans at different levels in the weir approach channel.ICRISAT-design automatic pumping type samplers were delivered to mostcatchments during the last quarter of 2001 but not all were installed.

181

Figure 4. Typical MSEC catchment hydrologic instrumentation for monitoring rainfall,stream flow, and sediment load.

182

Tabl

e 6.

Sum

mar

y of

initi

al h

ydro

logi

c da

ta in

thre

e M

SEC

-par

ticip

atin

g co

untr

ies i

n A

sia,

200

0–01

1 .

Map

awa

(Phi

lippi

nes)

Bab

on (I

ndon

esia

)D

ong

Cao

(Vie

tnam

)R

ambu

tan

Tega

lan

Kal

isid

iH

ydro

logi

c in

dica

tors

(Mai

n)M

ain

MC

1M

C 2

MC

3M

CM

CM

C

Run

off c

oeff

icie

nt (%

)St

orm

runo

ff c

oeffi

cien

t0.

01–0

.80

0.30

–18.

70.

010.

040.

05(A

v. 0

.224

)(A

v. 8

.6)

(1 st

orm

)(1

stor

m)

(1 st

orm

)(3

0 st

orm

s)(4

7 st

orm

s)

Mon

thly

runo

ff co

effic

ient

0.55

–3.1

31.

12–1

0.20

0.65

–4.3

60.

21–1

1.62

1.34

–6.6

50.

01–1

.67

0.03

–18.

89N

A(A

v. 2

.57)

(Av.

5.0

2)Av

. 2.4

8(A

v. 7

.70)

(Av.

3.8

9)(A

v. 0

.24)

(Av.

2.1

4)(N

o. o

f mon

ths w

ith ru

noff

)8

127

54

1111

Stor

m h

ydro

grap

h be

havi

orPe

ak ru

noff

(m3 s

-1)

NA

0.18

6–1.

122

NA

NA

NA

NA

NA

NA

Tim

e to

pea

k (m

in)

10–1

25(7

stor

ms)

Tim

e ba

se (h

)8.

17–1

9.67

(7 st

orm

s)

Sedi

men

t loa

d (t

ha-1

)B

ed lo

ad1.

037

1.58

016

.842

0.56

4

Susp

ende

d lo

ad0.

063

0.57

43.

714

2.44

0

Tota

l soi

l los

s1.

100

0.36

0.08

0.56

0.59

2.15

420

.556

3.00

41.

Mai

n =

Mai

n ca

tchm

ent;

MC

= M

icro

-cat

chm

ent;

NA

= D

ata

not a

vaila

ble.

183

Table 7. Magnitudes of storm hydrograph components for sample storms in Mapawa catchment (mainweir) in the Philippines during 2000 and 2001.

Direct runoffRainfall Direct runoff Peak runoff Time to peakDate of occurrence (mm) m3 mm coefficient (%) (Qp) (m

3 s-1) (tp) (mm)

2000September 24–25 0.55 15October 5–6 0.917 125November 5 0.669 10

2001September 13–14 33.90 4127.40 4.74 14.0 0.826 35September 26–27 35.60 2997.00 3.44 9.7 0.186 105October 4–5 22.00 4563.00 5.24 23.8 1.122 20October 6–7 22.20 4806.90 5.53 24.9 0.927 30

Total/Average 113.70 18.95 16.7

Figure 5. Storm hydrograph on 6–7 October 2001 for Mapawa catchment (main weir), Philippines.

184

technologies start to be widely practiced by farmerswithin the catchment. At present, no significant soiland water conservation practices are followed in thecultivated areas of the Mapawa catchment.

Initial values of sediment yield

Reported sediment yield from one main catchmentand six micro-catchments indicate that bed loadranges from 0.6 to 17.0 t ha-1. Bed load (sedimentstrapped in the weir approach channel) is about19.94% of the total sediment load. Total sedimentload of the stream, i.e., the total soil loss leaving thecatchment ranges from 0.1 to 21 t ha-1. An attempt torelate these soil loss values to the observed land useand field conditions during the periods of sedimentload monitoring was presented by Maglinao andPenning de Vries (2003).

RecommendationsThe on-going MSEC catchment studies woulddefinitely be of great contribution towards generatinginformation to show that appropriate land uses couldsustain the use of our land and water resources andenhance the environment. However, generation ofadequate and reliable information would requirelonger duration of field data observations. Forexample, increasing the aerial extent of forest coverand conservation farming practices could enhancefresh water supply. But this would require sufficientperiod to observe significant increases in the baseflow of a stream or in the water table of aquifers.There is indeed a need to extend the study to gathersufficient data for a more convincing conclusion andrecommendations. Monitoring of hydrologic fielddata is quite costly. Gathering of enormous butunreliable data should therefore be minimized if nottotally avoided. The specific recommendations forimproving the reliability of field data are:• Functionality and accuracy of field instruments

should be ascertained. Preparation of a fieldmanual on the proper observation, operation, andmaintenance of installed instruments should beconsidered. It should be simple enough to be easilyunderstood by field observers. If necessary, it maybe written in the local dialect of the concernedobserver.

• Capability and sustainability of field dataobservers should be ascertained. More than onecapable staff may be needed to minimize data gapswhen an observer is sick or resigns.

• Occasional checking/verification of actual fielddata monitoring to ensure uniformity of procedureand reliability of collected information.

• Use of uniform technical terms and units to avoidconfusion and for easy comparison of results withother publications. A glossary of terms and unitsmay have to be agreed upon.

• Delineation of catchment/micro-catchment divideswith respect to the actual stream flow gaugingstations should be done using topographic map andre-checked in the field. Some catchment areasbeing reported are not yet fixed.

• Inclusion of other indicators of groundwaterrecharge specifically the spring water discharge (ifthere exists some within the catchment area) andwater table/piezometric water level variation in allcatchments if possible. A common methodologyand instrumentation for these indicators isnecessary if they are to be monitored.

ReferencesDavid, W.P. 1984. Environmental effects of watershedmodifications. Presented at the Seminar-Workshop onEconomic Policies for Forest Resources Managementsponsored by the Philippine Institute for DevelopmentStudies held in Calamba, Laguna, Philippines, Feb 17–18,1984.

David, W.P. 1988. Soil and water conservation planning:Policy issues and recommendations. Journal of PhilippineDevelopment XV(1):47–84.

Maglinao, AR., and Penning de Vries, F. 2003. TheManagement of Soil Erosion Consortium (MSEC): Linkingland and water management for sustainable uplanddevelopment in Asia. Pages 31–41 in Integrated watershedmanagement for land and water conservation andsustainable agricultural production in Asia: proceedings ofthe ADB-ICRISAT-IWMI Project Review and PlanningMeeting, 10–14 December 2001, Hanoi, Vietnam (Wani,S.P., Maglinao, A.R., Ramakrishna, A., and Rego, T.J., eds.).Patancheru 502 324, Andhra Pradesh, India: InternationalCrops Research Institute for the Semi-Arid Tropics.

Phommassack, T., Agus, F., Boonsaner, A., Bricquet,J.P., Chantavongsa, A., Chaplot, V., Ilao, R.O., Janeau,J.L., Marchand, P., Toan, T.D., and Valentin, C. 2003.Factorial analysis of runoff and sediment yield from

185

catchments in Southeast Asia. Pages 163–170 in Integratedwatershed management for land and water conservationand sustainable agricultural production in Asia:proceedings of the ADB-ICRISAT-IWMI Project Reviewand Planning Meeting, 10–14 December 2001, Hanoi,Vietnam (Wani, S.P., Maglinao, A.R., Ramakrishna, A.,and Rego, T.J., eds.). Patancheru 502 324, AndhraPradesh, India: International Crops Research Institute forthe Semi-Arid Tropics.

Subramanya, K. 1984. Engineering hydrology. New Delhi,India: Tata McGraw-Hill Publishing Co. Ltd. 316 pp.

Ursic, S.J., and Dendy, F.E. 1965. Sediment yields fromsmall watersheds under various land uses and forest cover.In Proceedings of the Federal Inter-Agency SedimentationConference held at Jackson, Mississippi, USA, 28 Jan–1Feb 1963. Miscellaneous Publication No. 970,Washington, DC, USA: ARS-USDA.

186

Nutrient Loss and On-site Cost of Soil Erosion Under DifferentLand Use Systems in Southeast Asia

F Agus and Sukristiyonubowo1

Abstract

It is generally understood that high slopes, high amount and intensity of rainfall, and intensive cultivationleads to high runoff, soil loss, and nutrient losses. This paper presents the results obtained in nutrient andsoil erosion research conducted by the Management of Soil Erosion Consortium (MSEC) project on 19catchments in Indonesia, Philippines, Laos, and Vietnam. In most of the countries little or no externalnutrients are added to upland crops. Though the nitrogen, phosphorus, and potassium losses per unit areadue to soil erosion are not high, the annual upland food crop farming system is not very profitable andsustainable because of nutrient losses due to leaching and removal of residues. The replacement cost ofnutrient losses is high. The less nutrient exploitative system, addition of external nutrients, and control oferosion are the options to come out of the vicious circle of land degradation, unsustainability, and poverty.

1. Center for Soil and Agroclimate Research (CSAR), Bogor, Indonesia.

High amount and intensity of rainfall, coupled withsteep slopes and intensive cultivation in the uplandsof Southeast Asia has led to high runoff and erosion,and in turn high nutrient content in runoff water andtransported sediments. Nutrient transport out of thecatchment occurs regardless of external nutrientinputs to the system. In Kaligarang watershed of Java,Indonesia, around 20 kg nitrogen (N), 6 kgphosphorus (P), and 9 kg potassium (K) leaveannually a 0.9-ha annual food crop-based catchment.Limited fertilizer application is given to the systembecause of inability of the farmers to purchasefertilizers, insecure tenure, or combination of both(Agus et al. 2001). In this catchment, fertilizerapplication is prioritized for paddy farming and ureais considered by most farmers as the most importantnutrient source. Similar farmers’ practices werereported in Dong Cao catchment in Vietnam (Toan etal. 2001). In Indonesia, farmers use about 50 to 75 kgN ha-1 yr-1 and variable amount of farmyard manure(FYM) for paddy fields. In paddy fields in Vietnam,fertilizer is applied at 72, 25, and 53 kg ha-1 yr-1 of N,P, and K, respectively, besides 5 to 8 t ha-1 yr-1 ofFYM. In the annual upland system, fertilizerapplication is rare among the upland poor. Fertilizerapplication usually goes with other inputs used. For

instance, farmers planting cassava, sweet potatoes,and local variety of upland rice are unlikely to applyany fertilizers, but where farmers adopt improvedvarieties of maize, rice, soybean, etc., fertilizerapplication becomes a more regular practice. Higherfertilizer application is found in vegetable crop-basedfarming system.

Despite the low rate of fertilizer application, soilnutrients are lost from the system through variousprocesses such as plant uptake, leaching, denitrification(for N), and nutrient transport with runoff water andtransported sediment during erosion. Continuousdepletion of soil nutrients leads to decline in soil fertility.By knowing the amount of nutrients lost througherosion, one can plan strategies to rectify degradationand to ensure sustainability of the scarce natural capital.

Nutrient transport through the process of erosionposes a threat to the off-site areas especially if theconcentration is high and the off-site areas are used forother key livelihood activities such as drinking water,fisheries, etc. Therefore, the assessment of catchmentscale nutrient transport is important as a partialassessment of the total nutrient loss in the system. Themore complete method, i.e., nutrient balance analysis,will give a comprehensive assessment but in the firstphase of the Management of Soil Erosion Consortium

187

(MSEC) research, the assessment is concentrated on thelateral loss through erosion. Some countries in theconsortium have planned to conduct a complete nutrientbalance study during the second phase of MSEC.

This paper reviews and discusses nutrient loss andthe costs of these losses in Indonesia, Philippines,Vietnam, and Laos. MSEC Phase 1 uses thecatchment as the scale of analysis because plot-scaleresearch as was common in the past reflects only thesoil and nutrient loss out of the plot and ignores theamount of soil and nutrients transported out of theplots. The weakness of the micro-catchment scalemeasurement, however, is the difficulty of evaluatingthe processes within each segment of the catchment.

Nutrient Loss and Soil ErosionResearch

MSEC research is conducted in 6 countries in Asia,but this paper covers results only from four countries,namely, Indonesia, Philippines, Vietnam, and Laos(the other two countries in the consortium areThailand and Nepal). In each country, catchment-scale research has been adopted.

The biophysical aspect of research was initiatedwith the construction of V-notch weirs and sedimenttraps and, for Indonesia, also Parshall flume in thedifferent catchments. The pairs of sediment traps andV-notch weirs were installed with both AutomaticWater Level Recorder (AWLR) (Orphimides orThalimedes) and manually recorded staff gauges. Thesediment traps are used for collecting bed load.‘Coarse’ soil particles and aggregates (the bed load)eroded from the catchment and collected by the trapswere weighed and the water content was determinedgravimetrically for every eroding rainfall event. Formost countries, nutrient (N, P, K) contents in the bedload samples were determined periodically based oncomposite samples of a few rainfall events. Inaddition, nutrients contained in runoff water were alsodetermined although some countries have yet toconduct the measurement. The on-site costsassociated with nutrient loss was calculated based onthe conversion of elemental nutrients to fertilizermaterial equivalent and then multiplying the valuewith the prevailing fertilizer prices.

Progress

Catchment setting

The catchments vary in size from 0.9 to 285 ha. InIndonesia, the study site is located in Ungaransubdistrict, Central Java Province. The catchment islocated relatively close to the urban developmentarea. Farming constitutes the second or third sourceof income by most households. Annual rainfall of3,800 mm in 2000/01 was above the normal annualaverage of 2,800 mm and considerably higher thanthat in the other Southeast Asian countries. Despitethe high rainfall, the runoff did not exceed 10%indicating the high infiltration capacity of theInceptisols at the site. Soil loss from each sub-catchment, except for the annual crop-based sub-catchment, was less than 2 t ha-1 yr-1. For the Tegalansub-catchment, which is dominated by annual uplandfood crops, soil loss was about 20 t ha-1 yr-1. This highsoil loss can be attributed to the relatively exposedsoil surface to rain drops, little litter cover, intensivetillage, steep slopes, and small catchment size (Table 1)(Agus et al. 2001).

The Philippine catchment is located in MindanaoIsland. The rainfall in 2001 was 2,574 mm. The areaof the sub-catchments ranged from 0.9 to 85 ha. Forthe sub-catchments between 8 and 85 ha, soil loss wasonly up to 1 t ha-1 yr-1. Catchment MC4 with a size of0.9 ha and 40% cultivated or bare land, had an annualsoil loss of 52 t ha-1. This may be attributed to the factthat the bare and cultivated areas were relatively closeto the sediment trap, otherwise this value seems to beextremely high for catchment scale measurement.

The Dong Cao catchment in Vietnam has an annualrainfall of about 2,000 mm. It represents the typicalcultivated mountainous uplands with slopes rangingfrom 15 to 60%. The altitude varies from 125 to 700m above sea level. The main crops are cassava, taro,groundnut, rice, maize, and forest plantation such aseucalyptus, Acacia mangium, cinnamon, etc. Waterfrom streams at the catchment is used for irrigation of10 ha of paddy in Dong Cao village. The sub-catchments range from 4.8 to 96 ha, and the annualsoil loss ranged from 1.6 to 4.4 t ha-1 (Table 1).

The MSEC study site in Laos is located in theLuang Prabang Province. Luang Prabang is

188

Tabl

e 1.

Cha

ract

eris

tics o

f MSE

C c

atch

men

t in

four

cou

ntri

es.

Run

off

Are

aR

ainf

all

coef

ficie

ntSo

il lo

ssSo

il or

der/

Land

use

/D

omin

ant

Cat

chm

ent

(ha)

(mm

)(%

)(t

ha-1 y

r-1)

Subg

roup

farm

ing

syst

emsl

ope

(%)

Ferti

lizer

app

licat

ion

(per

ha)

Indo

nesi

aTe

gala

n1.

138

200.

0520

And

icC

assa

va, m

aize

45–4

715

00 k

g ca

ttle

dung

Eutro

pept

sR

ambu

tan

0.9

3820

0.01

1.7

And

icR

ambu

tan,

22–5

540

6 kg

ure

a; 4

20 k

g TS

P; 6

58 k

g K

Cl1

Dys

trope

pts

shru

bK

alis

idi

20.0

3820

0.09

1.9

And

icR

ambu

tan

22–5

540

6 kg

ure

a; 4

20 k

g TS

P; 6

58 k

g K

Cl1

Dys

trope

pts

Pars

hal F

lum

e28

5.0

3820

0.02

0.04

Typi

cR

ice,

etc

.0–

55Fo

r ric

e on

ly 1

00 k

g ur

eaTr

opaq

uept

sPh

ilipp

ines

MC

124

.925

742.

070.

1Fa

lcat

a,2

bags

(1 b

ag =

50

kg) c

hick

enba

mbo

o,du

ng a

nd 6

bag

s com

plet

egr

assl

and

(98%

),fe

rtiliz

er 1

4-14

-14

vege

tabl

es a

ndro

ot c

rops

(2%

)M

C2

17.9

2659

6.57

0.5

Fore

st, g

rass

land

2 ba

gs c

hick

en d

ung

and

6 ba

gs(8

5%),

crop

land

,co

mpl

ete

ferti

lizer

14-

14-1

4sh

rubs

(10%

)M

C3

8.0

2659

3.63

1.0

Gra

ssla

nd (7

5%),

2 ba

gs c

hick

en d

ung

and

6 ba

gscu

ltiva

ted

(15%

),co

mpl

ete

ferti

lizer

14-

14-1

4se

ttlem

ent (

10%

)M

C4

0.9

2443

-51

.6G

rass

land

,2

bags

chi

cken

dun

g an

d 6

bags

trees

(60%

),co

mpl

ete

ferti

lizer

14-

14-1

4cu

ltiva

ted

and

bare

land

(40%

)W

hole

84.5

2623

4.44

0.45

Gra

ssla

nd,

2 ba

gs c

hick

en d

ung

and

6 ba

gsba

mbo

o,co

mpl

ete

ferti

lizer

14-

14-1

4eu

caly

ptus

,ve

geta

bles

and

root

cro

ps (2

0%)

cont

inue

d

189

Tabl

e 1.

con

tinue

d.

Run

off

Are

aR

ainf

all

coef

ficie

ntSo

il lo

ssSo

il or

der/

Land

use

/D

omin

ant

Cat

chm

ent

(ha)

(mm

)(%

)(t

ha-1 y

r-1)

Subg

roup

farm

ing

syst

emsl

ope

(%)

Ferti

lizer

app

licat

ion

(per

ha)

Vie

tnam

W1

4.8

2035

4.4

Cas

sava

, gra

ss-

W2

9.4

2035

3.9

Cas

sava

,-

Acac

ia m

angi

umW

35.

220

352.

9C

assa

va, t

aro,

-Ac

acia

man

gium

W4

12.4

2035

1.6

Cas

sava

,-

Acac

ia m

angi

um,

seco

ndar

y fo

rest

MW

96.0

2035

1.9

Cas

sava

, arr

owro

ot,

seco

ndar

y fo

rest

-L

aos

S01.

312

290.

01U

ltiso

ls B

ush

fallo

w (6

9%),

25-

Alfi

sols

teak

(31%

)S1

19.6

1229

2.1

Ulti

sols

,B

ush

fallo

w (7

6%),

29-

Alfi

sols

,fo

rest

and

teak

(14%

),En

tisol

s a

nnua

l c

rops

(9%

),pe

renn

ial c

rops

(1%

)S2

35.7

1229

2.3

Alfi

sols

,B

ush

fallo

w (8

0%),

27-

Ulti

sols

,fo

rest

and

teak

(15%

),En

tisol

s a

nnua

l cro

ps (2

%),

pere

nnia

l cro

ps (3

%)

S355

.912

293.

9A

lfiso

ls,

Bus

h fa

llow

(60%

),25

-En

tisol

sfo

rest

and

teak

(10%

),U

ltiso

ls a

nnua

l cro

ps (2

0%),

pere

nnia

l cro

ps (1

0%)

S465

.412

296.

4A

lfiso

ls,

Bus

h fa

llow

(53%

),28

-U

ltiso

ls,

fore

st a

nd te

ak (4

3%),

Entis

ols

ann

ual c

rops

(2%

)pe

renn

ial c

rops

(2%

)1.

Ferti

lizer

app

licat

ion

for r

ambu

tan

(Nep

heliu

m la

ppac

eum

) was

cea

sed

in 1

999

till m

id-2

001

beca

use

of e

ncro

achm

ent b

y vi

llage

rs. I

n N

ovem

ber 2

001

ferti

lizer

app

licat

ion

as w

ell a

s ef

forts

to re

gene

rate

som

e of

the

trees

wer

e re

sum

ed.

190

predominantly mountainous, consisting mostly(99.99%) of hills with steep and very steep slopes (8%to >55%) while the flat and gentle slopes (0 to 2%)represent <1% of the area and lie at the foothills, at thebottom of the valley. Elevation varies from 290 to 2257m above sea level. The most common soil order isUltisols and found on the slopes ranging from 8% to50%.

The province has a wet-dry monsoon tropicalclimate. The dry season (November to March) is coldand mostly dry, while the wet season (April toOctober) is hot and humid. The average annualrainfall is 1403 mm and in 2001 it was 1230 mm. Thesub-catchment of 1.3 ha, planted to teak and coveredby bush yielded only 0.01 t ha-1 yr-1 of sediment. Otherlarger sub-catchments ranging between 20 and 65 haand cultivated to annual upland crops yielded 2.1 to6.4 t ha-1 yr-1 of sediments.

Nutrient loss

Nitrogen

In the 19 catchments and sub-catchments in the fourcountries, there was relatively high amount of Nleaving the catchment (Table 2). High N loss ingeneral was associated with annual upland cropsystems. But in Laos, with relatively low rainfall,perennial tree cover and shifting cultivation system,N loss was also high. In Indonesia, over 21 kg ha-1 yr-1

of N leaves the annual crop-based Tegalan sub-catchment. Nitrogen loss from Kalisidi as high as 9 kgha-1 yr-1 was associated with the high erosion due toencroachment of the rambutan plantation by localvillagers (Agus et al. 2001). The level of 21 kg ha-1

yr-1 of N loss from the Tegalan annual upland systemis considered very high owing to low addition ofnutrients into the system. About 1.5 t ha-1 yr-1 of cattlemanure is added to the sub-catchment and thiscontributes only about 20 kg N with the assumptionthat N concentration in FYM is 13 g kg-1. Other Nsources include plant residue, N2 fixation duringrotation with leguminous crops and, to a limitedextent, N fertilizers. This indicates negative balanceof N in the current system. Unless N fertilizer or sometechniques of soil and nutrient conservation areapplied, the Tegalan system will not be sustainable.

In Philippines, the MC4 sub-catchment thatincludes 40% of cultivated and bare lands yielded as

high as 144 kg ha-1 yr-1 of N. The estimate of N loss inthis case was based on the organic matter content ofthe eroded sediment. Nitrogen content in the organicmatter was assumed as high as 5% (Duque et al.2001). Fertilizer application in the MC4 sub-catchment included 2 bags of chicken dung and 6bags of 14-14-14 (i.e., approximately 300 kg of NPKfertilizer per year). With 14% N content, this amounttranslates to 42 kg N addition per crop per year. Twobags (approximately 100 kg) of chicken dung cancontribute only about 2 kg N and N from fertilizer andchicken dung put together is equivalent to about 44 kgand is far below the N loss of 144 kg ha-1 yr-1

indicating substantial N deficit as the current processcontinues. The extremely high soil N loss could easilybe verified by long-term crop yield and soil total Ndata. Duque et al. (2001) presented soil organicmatter level of 32 g kg-1, which in most soils in thetropics could be considered as moderate level. Shouldthe actual soil and N losses stay at the level as shownin Tables 1 and 2, there will be significant decline incrop yield and soil test levels.

In Vietnam, like in the Philippines and Indonesia,high N loss of 9 to 11 kg ha-1 yr-1 in W1, W2, and W3sub-catchments (Table 2) is associated with cassavaplanting (Table 1). As mentioned by Toan et al. (2001),fertilizer and manure are used mainly for paddy fields,but almost no fertilizer is used for the upland system.Nitrogen loss of around 10 kg ha-1 yr-1 plus the amountof N that is taken out from the catchment with harvest,and other loss through leaching indicates declining soilfertility in the Vietnamese sub-catchments. In sub-catchments W4 and MW, N loss was relatively low (3.5to 4 kg ha-1 yr-1) and this may be associated with thepresence of secondary forest in parts of the catchmentsand with their large size.

Despite relatively low annual rainfall (1,230 mm)during the year and the relatively large size (20–65ha) of sub-catchments S1, S2, S3, and S4 in Laos,relatively high amount (5–16 kg ha-1 yr-1) of N lossoccurred. Furthermore, the dominant (about 80%)land use in the study area was rotating cultivated land(shifting cultivation) which is characterized by lowexternal nutrient inputs. Despite the existence of teakplantation near the trap of catchment S4, the averageN loss was 16 kg ha-1 yr-1 and this again raises concernon the sustainability of the system. Nitrogen loss fromcatchment S0 was negligible (0.03 kg ha-1 yr-1). With

191

the small sub-catchment size, this trend is contrary tothat of the other three countries.

Phosphorus and potassium

Data in Table 2 show that P loss in general was relatedwith N loss. In the Philippines, however, despite highsoil and N losses in catchment MC4, P loss was almostundetectable. Extractable P content in the Philippinessoil was 3 to 13 mg kg-1. For Indonesia, Vietnam, andLaos, P loss, albeit low in absolute value is consideredsignificant as almost no external P was added to thesecatchments. For annual crop-based farming, P isespecially high in the seed which is the part of harvest

almost completely removed from the system. This P off-take indicates unsustainability of the current systems.

Potassium is subject to lateral transport along withsoil mass transport in exchangeable form and/or alongwith runoff water during and following storms. Thus, Kloss through erosion was almost parallel with soil and Nlosses in the four countries. Potassium is rarely appliedby farmers in the upland farming systems, but cropresidue recycling under minimum input agriculturalsystems could reduce the need for K and calcium (Ca)fertilization by 50%, P by 30%, N and magnesium (Mg)by 90% depending on the kinds of successive plants inthe rotation (Wade et al. 1988). Thus the promotion ofcrop residue recycling technique should be intensified.

Table 2. Nutrient loss through erosion and on-site cost at MSEC catchments of different countries1.

Nutrient loss (kg-1 ha-1 yr-1) On-site costCatctment N P K (US$ ha-1 yr-1)

IndonesiaTegalan 21.53 5.82 9.02 20.54Rambutan 0.89 0.89 1.11 2.00Kalisidi 9.24 0.21 5.97 8.01Parshal Flume 0.60 0.00 2.10 1.46

PhilippinesMC1 0.50 0.003 0.50 0.30MC2 2.30 0.0003 0.05 1.03MC3 4.80 0.004 0.19 2.28MC4 144.20 0.079 6.09 67.86Whole 1.30 0.0008 0.15 0.67

VietnamW1 10.79 4.81 4.26 11.99W2 10.83 4.97 2.46 10.20W3 8.73 3.99 2.68 10.83W4 4.03 2.25 1.38 4.41MW 3.55 1.94 2.58 3.96

LaosS0 0.03 0.00 0.00 0.01S1 4.74 0.89 0.82 3.53S2 5.12 0.93 0.79 3.72S3 12.51 1.91 0.76 7.93S4 16.27 2.73 0.98 10.561. In Indonesia and Laos, nutrient concentration was determined from both the bed load and runoff water samples, while in Philippines

and Vietnam it was based on only the bed load.

192

Nutrient loss and gap to sustainability

In general soil fertility declined because of significantnutrient loss through erosion even if very limited orno external nutrients were added by the uplandresource-poor farmers in the studied catchments. Inthe flat area on Oxisols and Ultisols of Sitiung,Sumatra, Indonesia after more than five years of soilfertility management research, Wade et al. (1988)came up with nutrient management recommendationfor continuous farming systems (Table 3). Capabilityof local farmers to purchase inputs was far below therecommended level.

In Laos, where shifting cultivation is mainlypracticed, depending on the length of fallow period, thesystem may be able to rejuvenate. But when the systemintensifies, there will be progressive need for externalinputs. In Indonesia, Vietnam, and Philippines wherefarmers practice continuous farming system, the lowexternal nutrient inputs as given in Table 1 mayprogressively fail to support the current system in thenear future.

The options will be to control erosion significantly,add external nutrients, and/or modify the currentsystems to less exploitative ones such as agroforestryor agro-silvo-pastoral system. The current agriculturalsystem may have been in a vicious cycle of poverty andland degradation. The challenge ahead will be toeradicate poverty and at the same time come up withmore soil conserving farming practices.

On-site cost of soil erosionWith the level of nutrient losses as discussed earlier,the replacement cost of nutrients that should be borneby farmers based on fertilizer prices could be as highas US$ 20 in Indonesia. In the Philippines, it could goas high as US$ 68. This, as has been mentioned informer sections, only takes into account the lossthrough erosion, but not the loss through leaching,off-take with harvest, and other biochemicalprocesses. The problem is more complicated forannual upland food crop farming systems because, asin Indonesia, the replacement cost for nutrient was the

Table 3. Initial rate and annual maintenance rate (in parentheses) of nutrient inputs needed to satisfynutrient requirements of six main food crops grown in clayey Oxisols and Ultisols in Sitiung upland,Sumatra, Indonesia.

Input Rice Soybean Maize Groundnut Mung bean Cowpea

Lime 0.5–1.5 4–5 3–5 3–5 5 1–3(Mg CaCO3 ha-1)1

N (kg N ha-1) 45 0 135 0 0 0(45) (0) (135) (0) (0) (0)

P (kg P ha-1) 20 80 80 80 80 80(20) (20) (20) (20) (20) (20)

K (kg K ha-1)2 60 40 60 60 80 60

Mg (kg Mg ha-1)3 8 24 32 8 nd4 nd

S (kg S ha-1) 0 0 0 0 0 01. Use equation LR = 1.5 [(Al – (RAS × ECEC / 100)] for initial and maintenance rate, where LR is lime requirement in Mg ha-1 of the

CaCO3 equivalent, Al is aluminum concentration in cmolc kg-1, and ECEC (effective cation exchange capacity) is the sum ofexchangeable Ca, Mg, and K plus KCl extractable Al in cmolc kg-1. RAS values of maize, groundnut, rice, soybean, cowpea, and mungbean are 29, 28, 70, 15, 55, and 5%, respectively.

2. These are initial and maintenance rates if crop residue is removed. Return of crop residues would permit these rates to be reduced by50%. Reference point for initial rate is when the soil is degraded, such as after bulldozing land clearing or after several years of severeerosion.

3. Unless dolomitic lime is utilized, maintenance rates are not given.4. nd = Not determined.

Source: Adapted from Wade et al. (1988).

193

highest and the system happened to be the leastprofitable. When family time for farming is takeninto account, the annual upland farmers suffer lossof about US$ 51 annually or a benefit-cost ratio of–0.29. If they replace the lost nutrients they mayget better crop yields, but the ability to spend extraUS$ 20 besides US$ 51 loss is questionable. Inmany cases, however, farmers do not consider theirown labor time and they feel secure if they can getfood from farming.

For the Philippines, revenues generated fromseveral field crops is presented in Table 4. Maizegives the lowest return compared with vegetablecrops. Fertilized and unfertilized maize gives US$ 34and US$ 111 returns, respectively. With various on-farm expenses, it is very unlikely that the farmers canafford to invest for the nutrient replacement which isas high as US$ 68 (Table 4).

Conclusions and Suggestions• Except in Laos, nutrient loss per unit area was

lower as the catchment size increases.• High nutrient loss through erosion is also likely to

happen when the farming system base is annualupland crops.

• Annual upland food crops based system is ingeneral the least profitable and it is the system thatis highly vulnerable to erosion and thus demandinghigh replacement cost of nutrient loss.

• Long-term crop yield and spatially distributed soiltest data should be included as an integral part ofthis nutrient loss study and there is a need for aunified research methodology in the MSECproject to enable better cross country analysis.

ReferencesAgus, F., Sukristiyonubowo, Vadari, T., Hermiyanto, B.,Watung, R.L., and Setiani, C. 2001. Managing soil erosionin Kaligarang catchment of Java, Indonesia. MSEC-IndonesiaCountry Report. Bangkok, Thailand: International WaterManagement Institute, Southeast Asia Regional Office.

Duque (Sr), C.M., Tiongco, L.E., Quita, R.S., Carpina,N.V., Santos, B., Ilao, R.O., and de Guzman, T. 2001.Management of soil erosion consortium: An innovativeapproach to sustainable land management in thePhilippines. MSEC-Philippines Annual Report. Bangkok,Thailand: International Water Management Institute,Southeast Asia Regional Office.

Toan, T.D., Phien, T., Phai, D.D., and Nguyen, L. 2001.Soil erosion and farm productivity under different landuses. MSEC-Vietnam Annual Report. Bangkok, Thailand:International Water Management Institute, Southeast AsiaRegional Office.

Wade, M.K., Gill, D.W., Subagjo, H., Sudjadi, M., andSanchez, P.A. 1988. Overcoming soil fertility constraintsin a transmigration area of Indonesia. TropSoils BulletinNo. 88-01. Raleigh, North Carolina, USA: North CarolinaState University.

Table 4. Average yield of major crops planted within MSEC catchment in the Philippines.

Yield Average price ValueCrop (kg ha-1) (US$ kg-1) (US$ ha-1)

Potato 4700 0.33 1566.67

Sweet pea 1220 1.00 1220.00

Baguio beans 2500 0.18 444.44

Cabbage 3500 0.27 933.33

Wongbok 3200 0.22 711.11

Maize (no fertilizer) 233 0.14 33.66

Maize (with fertilizer) 767 0.14 110.79Source: Adapted from Duque et al. (2001).

194

Soil Erosion and Farm Productivity Under Different Land Uses

T D Toan1, F Agus2, C M Duque3, and A Chanthavongsa4

Abstract

Land use is an important factor of soil erosion. The paper presents initial evaluation of soil erosion andnutrient losses under different land uses, farm productivity, and income for various annual crops on arablelands in four participating countries namely Indonesia, Laos, Philippines, and Vietnam for theManagement of Soil Erosion Consortium project. Initial results indicated that the soil loss was highestunder annual crops followed by perennial crops, grasslands, bush land, and forests. The yields of annualcrops are low and because of intensive cropping and reduced fallow period on arable land the yields havedeclined over a period. Based on the results it is concluded that there is a need for implementation of soiland water conservation measures to conserve natural resources and increase the potential of systemproductivity.

Soil erosion is commonly observed in sloping landsespecially in the tropical zones of Southeast Asia.Farming and other economic activities on slopinglands have become environmentally unsustainablecausing deleterious on-site and off-site effects. In1998, the Management of Soil Erosion Consortium(MSEC) initiated a project aimed at developing andpromoting sustainable and socially acceptablecommunity-based land management options througha participatory and interdisciplinary approach in sixcountries in Asia with support from the AsianDevelopment Bank (ADB). This paper presents aninitial evaluation of soil erosion, land use, and farmproductivity in four participating countries, namelyIndonesia, Laos, Philippines, and Vietnam. Itdiscusses the land management options forsustainable adoption in developing agriculture andforestry on sloping lands.

Approach and Methodology

Site selection and characterization

MSEC uses an integrated, interdisciplinary,participatory, and community-based approach inresearch involving all land users and stakeholders on

a catchment scale. Representative catchments wereselected by using carefully defined criteria andmethodological guidelines developed for the purpose.More detailed characterization was done to establishthe baseline information about the sites. Within thecatchments, several micro-catchments representingvarious land uses were further identified anddelineated to conduct more detailed soil erosion andhydrological studies.

Hydrological and soil erosion monitoring

Hydrological monitoring stations equipped withautomatic water level recorders, manual staff gauges,sediment traps, automatic weather stations, andmanual rain gauges were installed. Data collectionand analysis followed guidelines set up by themembers of the consortium. In addition to thebiophysical data, monitoring of the socioeconomicparameters was also undertaken.

Land management options

On the basis of the information gathered andconsultations with farmers, the best-bet options wereidentified and introduced in the catchments.

1 National Institute for Soils and Fertilizers (NISF), Hanoi, Vietnam.2. Center for Soils and Agroclimate Research (CSAR), Bogor, Indonesia.3. Central Mindanao University (CMU), Musuan, Bukidnon, Philippines.4. Soil Survey and Land Classification Center (SSLCC), Vientiane, Laos.

195

Considering the inputs of the farmers, anyintervention would be better adopted and sustained.But the intervention should address both soilconservation and farm productivity.

Results and Discussion

Catchment profiles

The experimental catchments selected in Indonesia,Laos, Philippines, and Vietnam ranged from 71 to 139ha with four smaller micro-catchments representingdifferent land uses delineated within. The catchmentshave slopes ranging from 12 to 80% and averageannual rainfall ranging from 1,080 to 2,500 mm. In thecatchments in Indonesia, Laos, and Vietnam, waterflows in the creeks throughout the year, while in thecatchment in the Philippines, water flows only duringthe rainy season. The catchments are dominated byannual cash crops with some patches of perennials andare cultivated primarily by ethnic minorities. Ingeneral, the model catchments represent a resourcemanagement domain with biophysical andsocioeconomic characteristics common in the marginalsloping uplands.

Soil erosion and sediment yield

Soil loss in the different micro-catchments in HouayPano catchment in Laos is presented in Table 1. The soilloss under land use with high bush fallow and forest

(Weir 1) was lowest (0.5 t ha-1), while in the micro-catchment with more annual crops, the amount of soilloss was 1 t ha-1. In Mapawa catchment in Bukidnon,Philippines, about 80% of the area is under forest andgrassland in Weir 1, Weir 2, and Weir 3. The amount ofsoil loss ranged from 0.1 to 0.6 t ha-1 (Table 2). In Weir 4,which is mostly cultivated, the soil loss was 14 t ha-1. Soilloss under different land uses at Dong Cao catchment isshown in Table 3. Soil loss was high in Weir 1, Weir 2,and Weir 3 and ranged from 3 to 5 t ha-1. More than 50%of the area of these sub-catchments are cultivated toannual crops such as cassava, taro, and forest trees andthe soil cover was very poor.

Cost of nutrient loss

In Indonesia, the lowest on-site cost of nutrient losswas US$ 0.46 ha-1 yr-1 in Rambutan sub-catchmentfollowed by US$ 1.38 ha-1 yr-1 in Babon catchment.The on-site cost was US$ 3.74 ha-1 yr-1 and US$ 2.84ha-1 yr-1 in Tegalan and Kalisidi sub-catchmentsrespectively. In the Mapawa catchment in thePhilippines the highest on-site cost (680 PhP ha-1 yr-1)was in Weir 4, where more than 40% of the area iscultivated. In Dong Cao catchment in Vietnam thelowest cost was in the main weir at US$ 3.95 ha-1 yr-1.

Crop yield and income

In the Babon catchment in Indonesia, the benefit-costratio of paddy farming, tegalan, and rambutan were

Table 1. Soil loss under land use in the micro-catchments in Houay Pano catchment, Luang Prabang,Laos during May to September 2001.

Micro- Area Average Soil losscatchment (%) slope (°) Land use (t ha-1)

Weir 1 29.3 29 76% bush fallow, 14% forest, 0.59% annual crops, 1% perennial crops

Weir 2 19.8 27 80% bush fallow, 15% forest, 2% annual crops, -3% perennial crops

Weir 3 27.7 25 60% bush fallow, 10% forest, 20% annual crops, 1.010% perennial crops

Weir 4 13.1 28 53% bush fallow, 43% forest, 2% annual crops, 1.92% perennial crops

Weir 5 2.5 17 56% bush fallow, 44% forest 2.0

196

0.91, –0.29, and 0.73 respectively. Rambutan yieldsfruits only seasonally and due to lack of postharvestprocessing equipment, the fruit prices becomeunacceptably low during the peak season. Paddyfarming is economically most profitable.

In the catchment in Laos, farmers practice rice-based cropping system. Crops like maize, cucumber,vegetables, root crops, and chili are commonintercrops wit`h upland rice. Sometimes they are alsogrown in adjacent plots or in separate fields. They can

Table 2. Soil loss under land use in the micro-catchments in Mapawa catchment, Bukidnon, Philippinesduring January to August 2001.

Micro- Area1 Average Soil losscatchment (ha) slope (°) Land use (t ha-1)

Weir 1 24.93 -2 2% vegetable and root crops, 98% falcata, 0.444(19.5) bamboo, eucalyptus, grassland

Weir 2 17.88 - 20% vegetable, root crops, 80% grassland, 0.608(21.2) bamboo, eucalyptus, falcata, settlement

Weir 3 7.96 - 10% settlement, 15% cultivated, 75% grassland 0.132(9.4)

Weir 4 0.94 - 40% cultivated (14% of cultivated area is left bare), 14.326(1.1) 60% grassland, trees

Main Weir 84.5 - 20% vegetable and root crops, 80% grassland, 0.261(100) bamboo, falcata, settlement

1. Figures in parentheses are percentages.2. Slope in micro-catchment not recorded.

Table 3. Soil loss under land use in the different micro-catchments from January to October 2001 inDong Cao catchment, Hoa Binh, Vietnam.

Micro- Area1 Average Soil losscatchment (ha) slope (°) Land use (t ha-1)

Weir 1 4.77 28 67% cassava, 33% natural grass 5.59(5.0)

Weir 2 9.45 26 24% cassava + Acacia mangium, 4.43(9.8) 59% cassava + taro, 17% natural grass

Weir 3 5.19 20 52% cassava + Acacia mangium, 3.49(5.4) 48% cassava + taro + Acacia mangium

Weir 4 12.36 29 26% cassava + Acacia mangium, 1.82(12.9) 0.8% cassava, 74% natural grass

Main Weir 64.23 19 22% cassava + Acacia mangium, 2.80(66.9) 1.14% arrow root, 41% cassava,

2.86% Vemicia montana, 4.86% eucalyptus,4.09% natural grass

1. Figures in parentheses are percentages.

197

also be grown as intercrops with banana. Most uplandrice farmers practice 2–3 years bush fallow rotationwith one year of rice on the same land. The averageyield of upland rice in this area is 1.2 t ha-1, whilemaize and job’s tears yield an average 1.3 t ha-1 and1.4 t ha-1, respectively. Farmers reported that becauseof more intensive cropping and reduced fallowperiod, crop yields have declined 2–3 timescompared to the crop yields 20 years ago.

In the MSEC catchment in the Philippines,unfertilized native maize produced only an averagegrain yield of 233 kg ha-1 in one crop, while fertilizedmaize yielded about 767 kg ha-1 (Table 4). Theaverage production of potato is 4700 kg ha-1 per crop.Leafy vegetables such as cabbage and wongbok arealso grown. In Dong Cao catchment in Vietnam, themain crop is cassava. Arrow root and taro are grownas intercrops with cassava. These crops increasedland cover density, reduced soil erosion, andincreased farmers’ incomes (Table 5).

ConclusionFarming activities in the MSEC catchment enhancessoil erosion. Data on soil loss obtained from the weirsof the four countries indicated that the amount of soilloss mostly depends on land use and land management.There is a great need for adoption of soil and waterconservation measures to reduce soil erosion andincrease crop yields. The farmers realize this and mostof them have indicated their interest to adopt integratedmeasures like contour hedgerow farming, and farmingsystems for crop cover in the rainy season. The amountand size of sediment originating from the catchmentlargely depended on land cover and managementsystems and catchment size. Intensive soil tillage andminimal surface cover for annual crops in thecatchment made it relatively susceptible not only tobed load transport but also to suspended sedimenttransport. The relation between soil erosion and cropyields cannot be formulated because not enough datahave been collected in the catchments studied.

Table 4. Average yields of major crops planted within MSEC catchment in the Philippines.

Average yield Average price ValueCrop (kg ha-1) (Ph. peso kg-1) (Ph. peso ha-1)

Potato 4700 15.00 70,500.00Sweet pea 1220 45.00 54,900.00Baguio beans 2500 8.00 20,000.00Cabbage 3500 12.00 36,000.00Wongbok 3200 10.00 32,000.00Maize (without fertilizer) 233 6.50 1514.50Maize (with fertilizer) 767 6.50 4,985.50

Table 5. Average yield of major crops planted within the Dong Cao catchment in Vietnam during 2000–02.

Average yield Average price ValueCrop (kg ha-1) (Ph. peso kg-1) (Ph. peso ha-1)

Cassava 16,600 383 6,363,760Arrow root 1,350 525 708,750Taro 1,200 1,267 1,520,400

198

Economic Incentives and the Adoption of Soil Conservation Practices

E Biltonen1

Abstract

Soil conservation measures are an important component of farmer participatory and multi-disciplinarycatchment management approach for the management of soil erosion and for improving the agriculturalsystem’s productivity. The degree of adoption of soil conservation measures by farmers varies with theeconomic incentives offered by competing alterations. These alternatives are influenced by socioeconomicfactors, resources, cost-benefit analysis of technologies, time horizon of benefits, discount rate, value offarm labor, farmers’ perceptions, land tenure, and sustainability of the measures.

This paper presents an analysis of the research conducted on catchments in Indonesia, Laos,Philippines, and Vietnam by the Management of Soil Erosion Consortium project. The average householdincome, average expenditure, net farm income, income from cropping, and estimated on-site erosion costwere collected for all catchments. The study indicates that successful implementation of soil conservationmeasures must have significant on-site impact on production and benefits of a practice must substantiallyoutweigh the cost. The measures should offer increased and sustainable benefits with minimal fluctuation.The farmer’s perception on soil erosion as a problem is important. Incentives for adoption of soilconservation practices should be devised so as to increase the likelihood that a farmer voluntarily adoptsthe measures and decrease the chance of conflict between policy makers and farmers.


The Management of Soil Erosion Consortium (MSEC)is one of four consortia within the Soil, Water, andNutrient Management (SWNM) program of theConsultative Group on International AgriculturalResearch. The primary objective of the consortium is thedevelopment of community-based land managementpractices from a catchment perspective. The researchapproach has been designed to follow a participatory,multi-disciplinary, and catchment-based approach. Overthe last several years, MSEC has conducted research onvarious soil conservation practices in a number ofsettings and soil conservation options have beenidentified and verified through MSEC’s work. The nextphase of efforts in realizing the objective of MSECinvolves the dissemination of the accrued knowledgeand the adoption of recommended soil conservationoptions.

The next phase of MSEC work extends the researchfindings beyond the examination of biophysical

feasibility to socioeconomic feasibility. Besidesproductivity impacts, conditions such as society,knowledge, market access, and householddemographics influence what a household perceives tobe in its best interest and, therefore, its productiondecisions (Enters 1998). Of great importance for theadoption of soil conservation practices is the properprovision of incentives. Economists believe strongly inthe idea of incentives as motivators of human behavior(Young 1996). Incentives send signals to peopleregarding the impact that a certain activity will have ontheir well-being. These are especially important inweighing the trade-offs between competing alternatives.

This paper investigates the role that economicincentives may play at the household level in theadoption of soil conservation practices as determinedby MSEC research. The paper then examines researchconducted in the four MSEC-participating countries:Indonesia, Laos, Philippines, and Vietnam.

199

Economic IncentivesAn incentive is “something that encourages ormotivates somebody to do something” (Source:Encarta World English Dictionary, MicrosoftCorporation, 1999). The encouragement is providedby the condition of whether the thing will add to ordetract from one’s well-being. This paper considersincentives as both natural and imposed byinstitutional processes. In economics, an analysis ofexisting and potential incentives is usually conductedin the well-known form of cost-benefit analysis. Animportant characteristic of cost-benefit analysis is theassumption that all costs and benefits (incentives anddisincentives) can be quantified as if they were tradedin a market place. Traditional cost-benefit analysisruns into problems when the costs and benefits arederived from non-market items.

Economists normally assume that a person acts tomaximize their utility or net benefits. However, if aperson is living near or below a subsistence level,then their objective may not be maximization of netbenefits, but maximization of the probability ofsurviving (Miracle 1968). This difference hasimplications for the incentives necessary for theprivate farmers to adopt soil conservation measures.

A Modern Approach to SoilConservationThe traditional approach for the economic analysis ofsoil degradation is to concentrate on either nutrient lossor soil erosion in isolation with the aim of finding astructural solution to the problem. The land husbandryview aims to maintain the productivity of soil bygetting individual farmers to adopt good practices(FAO 2001). The difference is in the concentration onland use practices rather than structures as the solutionto erosion problems. The land husbandry approachdiffers from the traditional approach in that it perceivesthe farmer as a knowledgeable actor who is constrainedby circumstance.

Concerning a household’s decision making, acurrent approach is that the household makes itsdecisions based on a given set of attributes includingsocioeconomic factors, resources, and technologies(FAO 2001). The decisions are influenced by theincentives existing within the specific context of

credit availability, land tenure status, and informationand data available to the farmer. These factors areimpacted by external shocks and policies, which feedinto a household’s perceptions (Fig. 1).

The farmer’s perceptions concern the resourcebase available for productive activities and thebenefits and costs involved with a productive activity.It is then the farmer’s own decisions about productionpractices that ultimately have an impact on the naturalenvironment. A change in the incentives can be usedto signal to the farmer that existing resource use needsto be changed as their relative attractiveness declines.This is achieved when a farmer weighs the incentivesand disincentives of various options. This generalevaluation process has been formalized in cost-benefit analysis.

The Cost-Benefit Analysis of SoilErosion

Cost-benefit analysis is a common senseimplementation of economic theory. It has generallybeen applied to larger projects; however, it can beapplied to any investment activity of any size. Cost-benefit analysis provides a tool to aid decision makersin making informed decisions based on a consistentand logical framework of analysis. This frameworkevaluates the discounted stream of future net benefitsarising from various options.

A crucial element is determining from whose pointof view the results of a cost-benefit analysis will beutilized. For example, a poor upland farmer willprobably not consider the costs that soil erosion fromhis land may impose on an industry several hundredkilometers downstream. However, a national leveldecision maker should consider this cost. Similarly, afarmer would probably not willingly go along with asoil conservation measure that held positive benefitsfor the country, but caused a net loss to the farmer.When analyzing incentives, it is vitally important toconduct the analyses from the point of view of thoseaffected.

For the individual farmer, soil conservation willincrease net benefits from productive activities.However, impact of soil erosion on productivity iscomplex. As a result, complex models have beendeveloped to analyze the benefits and costs. In

200

developing countries, however, the data is often notavailable or a working knowledge of the model islacking. This has created a tradeoff for the analystregarding data availability and model complexity. Asa result, these models have been limited in theirusefulness because of their complexity.

AlternativesCost-benefit analysis seeks to value and rankalternatives. Therefore, the range of impacts of soilerosion will need to be identified. A proper analysis ofalternative soil conservation measures will draw acomparison between situations with the practice and

Figure 1. The role of incentives in farmer decisions on soil productivity.

201

without. This approach will give a clear indication ofwhat would happen with adoption of the soilconservation practice compared to what will happenif the soil conservation practice is not adopted.

Time horizon

The time horizon is also important to cost-benefitanalysis. Proper time horizon should be equal to theexpected life of the conservation measure. If it is to bean indefinitely sustained practice then it ispermissible to use a time horizon of 50 years as theeffect of even a low discount rate will reduce netbenefits to negligible levels after 50 years.

Enters (1998) has warned against simply assumingonly a short-term planning horizon in poor farmerdecision making. Instead, the nature of a farmer’sneeds should be assessed. For instance, havingenough food to eat would be a short-term need, whilehaving a child educated could be considered a long-term need. The implication is that an analysis ofincentives will need to consider the impact that theincentive will have on the target group’s needs.

Discount rate

Understanding the discount rate is crucial to analyzeincentives for the private investor. The discount ratereflects the relative attractiveness of immediateconsumption versus consumption postponed until afuture time. Determining the appropriate discountrate is a much debated issue in the field of economics.In simple terms, the discount rate is the premium onemust receive to be induced to forego immediateconsumption.

It is generally assumed that the more desperatepeople become, the higher their discount rates will be.The higher a farmer’s discount rate the lower is thepresent value of the damage done by future soilerosion. Therefore, the benefits arising from theadoption of soil conservation measures should be high.

As standard practice, the Asian Development Bank(ADB) uses a discount rate of 10 to 12% as a criterionfor judging a given project (ADB 1997). However,the precise calculation of specific discount rates isvaried and contentious. Consequently, someresearchers recommend using a range of discountrates due to the difficulties in determining the

“correct” discount rate. A decision maker with a firmgrasp of the concept of the discount rate can make abetter informed decision.

Value of labor

Another factor for any valuation method is the valueof a farmer’s own labor. This is the opportunity costof using his time to undertake the productive activityor soil conservation measure. There are two elementsto this value. There is that part of labor that is used formanual labor, which is usually equated with theunskilled labor wage rate. An often overlooked aspectis the farmer’s management labor. This laborcomponent involves the utilization of a farmer’s skillin choosing crops, overseeing production, choosingwhen to harvest, etc. Management skills are normallymore highly valued than manual labor skills. Enters(1998) has pointed out that the labor cost associatedwith various soil conservation measures is often thebiggest obstacle to their adoption.

Evaluation criteria

Evaluation criteria typically involve the net presentvalue, internal rate of return, and benefit-cost ratio.These criteria have advantages and disadvantages(ADB 1997). The net present value is the discountednet benefits summed over a given time horizon. Thisvalue should be greater than zero for an option to beconsidered. The internal rate of return is the rate ofreturn a given alternative will yield considering thediscounted stream of costs and benefits. It should begreater than some benchmark rate in order to beconsidered. The benefit-cost ratio is the ratio ofbenefits over costs. If this ratio is greater than 1, thenbenefits outweigh costs and the alternative can beconsidered acceptable.

Valuation TechniquesThe two techniques for valuation of soil erosion thathave been commonly used in the literature are:(1) replacement cost method; and (2) change inproductivity method (Enters 1998). The replacementcost approach (which was utilized in several of thecase studies) takes the value of soil erosion to be theequivalent cost of replacing the lost nutrients. This

202

estimated cost not only includes the cost of inorganicfertilizers, but also transportation and labor requiredto replace the lost nutrients. Weaknesses with thisapproach include the fact that estimated replacementquantities may have little or nothing to do withproduction requirements. Furthermore, as erosioncontinues the estimated annual costs will demonstratethe perverse trend of declining. An additionaldifficulty with the replacement cost approach is thatfarmers may not understand the concept very well(Enters 1998).

The change in productivity approach measures thecost of soil erosion as the decline in the value of soil-dependent productive activities. This approach has aneasily understood logic, particularly in analyzingfarmer responses. A key problem with this approachlies in isolating the impact that soil erosion has onchanges in production. The farming practice causingsoil erosion needs to be compared with a soilconserving productive practice, which means thatchanges in the cost structure also need to beaccounted for in the analysis. The change inproductivity approach has the most intuitive appealfor encouraging farmer participation.

Sustainability of Soil ConservationMeasuresAnother important area of concern is the sustainabilityof a soil conservation measure. Essentially, a soilconservation measure will need to be funded in order tobe maintained. This can be in the form of farmer’s owncontribution, subsidies, credit provision, farmer’svolunteered labor, or some combination of these.Ideally, the enabling environment is such that farmersare willing to undertake the conservation practice withno additional input from the government. If subsidiesare required as initial startup funding, then either thegovernment must adopt a long-term commitment toprovide the subsidies (and related compliancemonitoring) or devise a procedure for phasing out thesubsidies.

Case StudiesThis section examines the research conducted inIndonesia, Laos, Philippines, and Vietnam so that ananalysis can be made regarding possible incentives.

Information is obtained from progress reports for thefour countries.

Indonesia

The Government of Indonesia is quite conscious ofthe fact that farmer profits will largely determine thesustainability of any soil conservation program. TheMSEC site in Indonesia is the Kaligarang catchmentin Java. The population growth in Indonesia willincrease food demand by 30% by 2025 (Agus andSukristiyonubowo 2000). Efforts by the governmentto increase food production are being hampered bycontinued soil erosion that diminishes the capacity togrow adequate food supplies. Currently, farmers havebeen unwilling to adopt soil conservation measuresciting the lack of any short-term enhancement ofprofits. Furthermore, the temporary nature ofexternally provided incentives has limited thesuccessful adoption of soil conservation measures ona sustainable basis (Huszar and Pasaribu 1994, Agusand Sukristiyonubowo 2001).

Average farm income from both on- and off-farmsources was reported as US$ 373 (2,980,400 Rp).Farm plot sizes are small (0.05 to 0.25 ha) with asingle farmer often farming multiple plots. Whilemost farmers hold the perception that soil fertility islow, they also lack the ability to buy adequateamounts of fertilizers. Off-farm employment isundertaken by approximately one-third of the villagemembers in addition to farming activities. Low farmincome is attributed to small farm size, low soilfertility, low value crop cultivation, and low inputtechnology implementation. Currently, farmingaccounts for less than half of a family’s annualincome, but farming is still perceived as an importantactivity for ensuring food security and as a source ofincome (Agus and Sukristiyonubowo 2000).

Some farmers have responded to market incentivesfor the crops they have chosen to grow. Experience inthe study site demonstrates those farmers are willingto switch to alternative technologies that theyperceive are appropriate. The choice of a best-betoption was based on the understanding that a farmer’sadoption of a conservation measure will bedetermined by the measure’s contribution to ahousehold’s income. The annual replacement cost oflost nutrients was estimated at US$ 3.74 ha-1 (Agus

203

and Sukristiyonubowo 2001). This amount is equal toabout 1.5% of the average annual total income.However, the per-household cost of soil erosion willbe smaller if farm size is less than 1 ha.

A benefit-cost ratio was reported for threedifferent production activities: paddy, annual uplandcrops, and rambutan plantation. The ratiosrepresented the profitability of three crop options andwere computed as profit over expenditure.Profitability is concerned with the size of the profitrelative to the utilized resources value. The presentedratios indicated that paddy would produce a relativelyhigher profit for the required expenditures than theother two crops. However, a rambutan plantation,while requiring higher expenditure, would produce agreater profit. For analysis of incentives, the costrequirements for each crop and the land constraintsfacing a farmer are relevant. In this case, if land size isa constraint then choosing the crop that generates thegreatest profit would be the objective, even though itsprofitability is lower than paddy.

LaosForests in the northern and central regions of Laoshave suffered severe encroachment by farmers.Forested area has been reduced in these areas to only30% of total area. The mountainous topography ofLaos exacerbates the effect of deforestation on soilerosion rates. These rates were estimated at 30 to 150t ha-1 yr-1 (Soil Survey and Land Classification Center2001). Soil erosion has been positively identified as aconstraint to the sustainable development ofagriculture on steep sloping land areas.

The MSEC study area in Laos is the Houay Panocatchment located in Luang Prabang Province. Thearea has a population density of 23 persons km-2 andan annual growth rate of 2.3%. In the studycatchment, approximately 97% of the population isengaged in farming as the primary source of incomeand livelihood and the average household is made upof 6.2 persons. Within the MSEC study watershed,shifting agriculture accounts for 80% of the classifiedland use. Furthermore, the problems of land qualityand weeds have tended to worsen as farming isintensified. Thus, the labor requirements for weedcontrol have increased. Additionally, the increasedpressure on land have reduced the soil fertilityresulting in lower yields.

Protection measures were initiated under the 1994Land and Forest Allocation Scheme. Land useallocations were determined based both onproduction and protection of the soil. However, landremains the property of the state. Land is leased tohouseholds for farming on a long-term basis with theaverage landholding being 3 ha in the MSEC site.

Farmers in the study area earn about 70% of theirincome from farming activities. However, most of theagricultural production is kept for homeconsumption. Most of the inputs for agriculturalactivities are retained from the previous harvest orfrom an extension worker. The average annualincome from on-farm sources in the study area wasUS$ 207 (1,572,000 kip). However, 61% of thehouseholds earned less than US$ 132 (1,000,000 kip)per year from farming activities indicating a degree ofincome inequality within the catchment. Sixty-threepercent of farmers surveyed invested less than US$ 4(30,000 kip) in agriculture. No estimates of the cost ofsoil erosion were given.

PhilippinesFood and fiber production by subsistence farmers inthe Philippines is increasingly threatened by soilerosion. The topography and rainfall patternsexacerbate the soil erosion problems on slopinglands, which are compounded by cultivationpractices. The study site chosen for the MSECresearch is the Mapawa catchment located nearLantapan, Bukidnon. Farming is the majoroccupation in the catchment area. Most of the land inthe catchment area is classified as private.

Farmers in the catchment area practice maizemonoculture (42%) as well as mixed cropping (38%).Only 50% of surveyed farmers use fertilizer with 30%applying complete fertilizer alone and 20% applyingcomplete fertilizer and chicken dung. The farmers areprimarily landowners with a small fraction farming astenants. Among the farmers, 32% have adopted somesort of soil conservation practice. The rest have notadopted a soil conservation practice, primarily citingthe high labor cost involved with establishment. Thismay be due to the fact that soil erosion is notperceived as a serious problem by the farmers(Caprina and Duque 2000).

The Philippine study site conducted both on- andoff-site analyses of the cost of soil erosion. On-site

204

costs were estimated using the replacement costmethod, while off-site costs were estimated as theproportion of dredging costs of an irrigation schemethat could be attributed to sedimentation from theMSEC study catchment. The on-site cost of soilerosion in the study site was estimated at US$ 0.20 to0.52 ha-1. No income data was reported forcomparison.

Vietnam

In Vietnam soil erosion has occurred on 13 million haof land, which represents 40% of the total land area(Toan and Phien 2001). The MSEC study site is DongCao catchment located in Ha Tay Province inNorthern Vietnam. Water exiting the catchment isused to irrigate 10 ha of paddy. Access to the siterequires the use of a 4-wheel vehicle. Current soilnutrient conditions in the catchment are poor,highlighting the need for soil conservation measuresespecially for improving soil quality.

All farmers in the study site are full-time. Typicalupland farming patterns involve no fertilizer withintercropped rotation or fallow periods. Paddycropping involves only low rates of fertilizerapplication. Cassava is the main crop and is usuallygrown as monoculture or in rotation with taro. In2001, the farmers adjusted their cropping patterns toallow for both higher food production and highercover density to protect against rainy season erosion.Off-farm income was an insignificant source ofincome. The average cultivated area was 2.14 ha, ofwhich 1.5 ha area is upland and 0.29 ha is paddy land.This total area was divided into seven separate plots.Due to several plots, farmers generally did not applyfertilizers. Reported yields were low in comparison tothe Red River Delta (except for the Chinese ricevariety Q5 which yielded 6.5 t ha-1). The average netincome for households in the Dong Cao catchmentwas about US$ 139 (VND 2,100,000).

Soil erosion costs were estimated using thereplacement cost method. The cost estimates in thesub-catchments ranged from US$ 4 to 12 ha-1

depending upon nutrient loss. For the wholecatchment, the per-hectare cost was estimated at US$4 because most of the soil erosion remained within thecatchment. This represents a cost equal to 2.8% of theaverage annual income, although the cost is notincurred in financial terms.

Lessons LearnedThere are many lessons to be drawn from the workalready done in the MSEC study sites. These lessonsgive a good indication of the current status within thecatchments and guidance in the implementation of anenabling incentive system. These lessons also giveguidance as to where further work could be conducted.

There appears to be wide recognition andpromotion of the idea that soil conservation measures,if they are to be sustainably adopted, must hold appealto the individual farmers. Successful implementationof soil conservation options must have significant on-site impact on production. Soil conservation musthold short-term benefits, as well as long-term benefitsfor the farmers. Thus, the benefits of a practiceadopted must outweigh the costs for the farmer.

In all four studies summarized here, it has beendemonstrated that for most people within thewatershed, farming is the primary occupation andsource of income. In Indonesia off-farm activities inthe catchment hold greater opportunity for increasedincomes. Thus incentives that impact the financialstructure of farming operations can have seriousimplications for household incomes.

The Philippine study reported that 38% of farmershad already adopted a soil conservation practicebefore the MSEC study began. It would be interestingto know which crops the farmers adopted, what aretheir sources of income, and what their personalreasons are for adopting the soil conservationmeasures. This could make for an interesting casestudy on natural versus state-provided incentives.

Indications of incomes in the study areas werepresented, although some manipulations werenecessary to get income values in comparable forms(Table 1). The presented values try to represent incomefor average farms in the study catchments and are notnormalized on a per-hectare basis. Additionally,expenditures and net income values do not benefit fromdetailed crop budgets or household expendituresurveys in all cases. The savings rate is probably notvery high. Thus, it may not be practical or sociallyacceptable to apply “Polluter pays” fines on the uplandfarmers. This is further demonstrated by the data inTable 2 which compares estimated replacement costsof soil erosion for the four countries to total householdincome. Table 3 compares the estimated costs of soilerosion to income from cropping activities. In both

205

cases, it can be seen that the cost of soil erosion is asmall fraction of income, either total or fromcropping. The exception is the high cost of erosion inthe Mapawa catchment MC4. Compared to the highcost of labor for erosion control measures, the cost oferosion is relatively small. It would seem unlikely thatfarmers will be interested in controlling erosion basedon cost alone.

Analysis of data from Indonesia indicate that abenefit-cost ratio of 1 is required for benefits to atleast equal costs. Only paddy and rambutan plantationbut not upland crop options satisfy the minimumrequirement of a benefit-cost ratio of 1 (Table 4);paddy yields a higher benefit-cost ratio. The benefit-cost ratio in this case is a revenue-to-cost ratio, which

gives an indication of relative profitability. Rambutanplantation, however, yields a larger profit than paddy.This is indicated by the higher net present value.Therefore, if land is a constraint and sufficient capitalis available, a rambutan plantation would be thepreferred option.

A proper cost-benefit analysis would need toconsider alternatives based on with- and without-conservation measure scenarios. A 20-year timehorizon was assumed for conducting a quick analysis.The ADB-recommended 12% discount rate wasutilized. It was assumed that without soilconservation, yields would decline by 3% a year.Adoption of soil conservation measures would reducethe yield decline to 1% per year but add 5% to annual

Table 1. Income comparison of four case studies.

Average Percentage ofhousehold Average Net farm Income from total income

income expenditure1 income cropping earned fromCountry Catchment (US$) (US$) (US$) (US$) cropping

Indonesia Babon 373 358 15 183 49

Laos Houay Pano 296 13 282 207 70

Philippines Mapawa2 29 0 29 29 100 Mapawa3 726 196 530 726 100

Vietnam Dong Cao 774 628 147 458 591. The value for Laos includes only farm expenditure and that for Philippines includes only fertilizer expenditure.2. Maize monoculture without fertilizer.3. Potato/maize mixed cropping assuming 0.5-ha farm plot.

Table 2. Total household income and estimated cost of soil erosion.

Average Estimated on-site On-site erosion cost ashousehold income erosion cost percentage of

Country Catchment (US$) (US$ ha-1) household income1

Indonesia Babon 373 0.46–3.74 0–1

Laos Houay Pano 296 – –

Philippines Mapawa2 29 0.05–52.28 0–90 Mapawa3 726 0–4

Vietnam Dong Cao 774 3.96–11.99 1–31. Cost adjusted to assumed farm size.2. Maize monoculture without fertilizer.3. Potato/maize mixed cropping assuming 0.5-ha farm plot.

206

costs (Obviously, these assumptions are developedfor demonstration purposes only and results shouldnot be considered reliable). Results of the informalanalysis are presented in Table 5. Both the benefit-cost ratio and the net present value are higher for thesoil conserving option. This indicates that it ispreferable to adopt soil conservation measuresaccording to economic criteria. Positive net benefitsfrom soil conservation would not occur until the thirdyear after adoption. A poor farmer may find this waitunacceptable (indicating a higher discount rate) andprefer not to adopt the conservation measure.

Conclusions and RecommendationsThis paper has outlined some of the concepts oneconomic incentives and linked them with four casestudies conducted under the MSEC program. It hasbeen shown that estimated costs of soil erosion bythemselves will probably be insufficient to motivatefarmers to adopt conservation measures. Instead, it

will be necessary for farmers to understand the long-term impacts and adopt the perception thatcontrolling these impacts are in their best interest.The Philippines offers an interesting opportunity tostudy a situation where some farmers have adoptedsoil conservation measures, while others havehesitated out of concern for the high labor cost. Thevoluntary adoption is perhaps linked to what appearsto be a higher dependence on cropping activities and amore desperate income situation. Identification of thereasons behind this behavior and quantification of thedeterminants would help in planning the use ofincentives.

Regarding incentives, it is useful to keep in mindthat ultimately it is the farmer who adopts the soilconservation measures. If the farmer does not viewsoil erosion as a problem, then they will not beconcerned with protecting against it. The farmer canbe a valuable asset in planning and implementing soilconservation measures. Gaining a farmer’s input canhelp scientists and policy makers determine whether a

Table 3. Cropping, income, and estimated cost of soil erosion.

Income Estimated on-site On-site erosion cost as afrom cropping erosion cost percentage of income

Country Catchment (US$) (US$ ha-1) from cropping1

Indonesia Babon 183 0.46–3.74 0–2

Laos Houay Pano 207 – –

Philippines Mapawa2 29 0.05–52.28 0–90 Mapawa3 726 0–4

Vietnam Dong Cao 458 3.96–11.99 2–61. Cost adjusted to assumed farm size.2. Maize monoculture without fertilizer.3. Potato/maize mixed cropping assuming 0.5-ha farm plot.

Table 4. Cost-benefit analysis of three options for soil conservation.

Net presentCost Revenue Profit Benefit-cost value1

Crop US$) (US$) (US$) ratio (US$)

Paddy 150 286 137 1.91 1236Annual upland crops 177 125 –51 0.71 –463Rambutan plantation 250 433 183 1.73 16531. Discount rate equal to 12%.

207

lack of adoption is due to lack of knowledge aboutsoil erosion or lack of concern for damages. A lack ofknowledge about the damage of soil erosion and thebenefits of conservation can be rectified through theuse of extension agents and training programs.However, if the farmer is knowledgeable and does notperceive erosion as a serious problem then othermeans will need to be determined in order to fosteradoption.

It is necessary to determine the degree to whichfarm production is used for income generation orhome consumption. The more dependent a householdis on farm production for home consumption orincome generation the more vulnerable the farmerwill be to fluctuations in production. It would beuseful to demonstrate to the farmer the reducedvulnerability to production fluctuations or decliningtrends that soil conservation offers.

Farmers are also concerned with costs ofproduction. It will be important to determine the coststructure of production. If the cost (including labor) ishigh relative to production and income values, then it isless likely a farmer will voluntarily adopt the soilconservation measures. Farmers are also concernedwith benefits, normally in the form of higherproduction values and income. However, the reductionin risk/vulnerability to production declines is also animportant factor to be considered. Any conservationmeasure must be able to clearly demonstrate significantbenefits and low relative costs.

Cost-benefit analysis, when properly applied, canaid in demonstrating the net benefits of a particularsoil conservation option. Cost-benefit analysisconsiders the projected changes in crop productivity,price, costs of soil conservation, and laborrequirements over time. In comparisons between

existing and experimental farm plots (yield gapanalysis), it will be crucial to determine the cost oflabor (especially management) to derive an accurateidea of the true incremental benefit.

Policy makers interested in seeing the adoption ofsoil conservation measures must strive to provide anenabling environment. The sustainability of anyconservation measure will rely on the commitment ofthe farmer, scientists, and the policy makers. It may beuseful to use farmer participation not only to provideinput, but also to provide output. Also, the farmersshould be allowed to choose the options that appeal tothem.

Finally, in examining incentives and whether toimplement new ones or take advantage of naturalones, the analyst should think from the perspective ofthe farmer. That is, given the existing state of the farmhousehold income generating activities, what set ofconditions are needed to encourage adoption of soilconservation measures? The importance of farmers’perceptions cannot be emphasized strongly enough.In this way, incentives can be devised that appeal tothe farmer leading to a higher chance of voluntaryadoption of soil conservation measures and lesschance of conflicts between policy makers andfarmers. Successful incentives must be orientedtoward the problems of farmers.

ReferencesADB (Asian Development Bank). 1997. Guidelines for theeconomic analysis of projects. Manila, Philippines: ADB.

Agus, F., and Sukristiyonubowo. 2000. Catchment approachto managing soil erosion in Kaligarang of Java, Indonesia. InSoil erosion management research in Asian catchments:Methodological approaches and initial results. Semarang,Indonesia: International Water Management Institute.

Table 5. Demonstration of cost-benefit analysis results for paddy option.

Benefit-cost Net presentDescription ratio value (US$)

Soil erosive practice 1.41 765

Soil conserving practice 1.64 957

Incremental benefit of soil conservation 192

208

Agus, F., and Sukristiyonubowo. 2001. Country report:Managing soil erosion in Kaligarang catchment of Java,Indonesia. Bogor, Indonesia: Center for Soil andAgroclimate Research and Development.

Caprina, N.V., and Duque, C.M. 2000. Management ofSoil Erosion Consortium (MSEC): An innovative approachto sustainable land management in the Philippines. In Soilerosion management research in Asian catchments:Methodological approaches and initial results. Semarang,Indonesia: International Water Management Institute.

Enters, T. 1998. Methods for the economic assessment of theon- and off-site impacts of soil erosion. Bangkok, Thailand:International Board for Soil Research and Management.

FAO (Food and Agriculture Organization of the UnitedNations). 2001. The economics of soil productivity in sub-Saharan Africa. Rome, Italy: FAO.

Huszar, P.C., and Pasaribu, H.S. 1994. The sustainabilityof Indonesia’s upland conservation projects. Bulletin ofIndonesian Economic Studies 30(1):105–122.

Miracle, M.P. 1968. Subsistence agriculture: Analyticalproblems and alternative concepts. American Journal ofAgricultural Economics 50(May):292–310.

Soil Survey and Land Classification Center. 2001. Aninnovative approach to sustainable land management inLao. Semi-annual progress report. Vientiane, Lao PDR:Ministry of Agriculture and Forestry.

Toan, T.D., and Phien, T. 2001. Soil erosion and farmproductivity under different land uses. Hanoi, Vietnam:National Institute for Soils and Fertilizers.

Young, R.A. 1996. Water economics. In Water resourceshandbook (Mays, L.W., ed.). New York, USA: McGraw-Hill.

209

Consortium Approach in Soil Management Research:IWMI’s Experience in Southeast Asia

A R Maglinao and F Penning de Vries1

Abstract

Land and water degradation problems are major constraints in improving living conditions of people inmarginal and sloping uplands of Asia. Past research and development efforts have not been successful tosolve these problems. A new approach of networking involving all the stakeholders was tried by IBSRAMwith the financial assistance from the Asian Development Bank (ADB) and Swiss Agency for DevelopmentCooperation. This project namely ASIALAND Network to manage sloping lands was started in 1988 andcontinued up to 2001. Based on the success of their project a new approach called Management of SoilErosion Consortium (MSEC) started in 1996 and ADB funded it in 1998 for 4 years. In these twoapproaches all stakeholders like researchers of NARES, international agricultural research centers,advanced research institutions, farmers, and NGOs worked in a participatory manner with a common goaland shared information among the stakeholders. The details of organization, working, and achievementsare covered in this paper.


Under this scenario, the International Board for SoilResearch and Management (IBSRAM), since itsinception, had employed the networking strategy in theimplementation of its activities on soil managementresearch. The regional networks had involved thenational agricultural research and extension systems(NARES) in parts of Africa, Asia, and the SouthPacific, addressing soil management problems inVertisols, sloping lands, and acid soils. Thismechanism has been continued by the InternationalWater Management Institute (IWMI), which took overIBSRAM’s programs when it ceased to exist, and anIWMI regional office for Southeast Asia (IWMI-SEA)was established in Bangkok, Thailand on 2 April 2001.

This paper highlights the experiences of IWMI inemploying the networking and consortium approach inthe implementation of its research programs on themanagement of sloping lands in Asia. It is based onlessons learned from the activities of the ASIALANDNetwork on the Management of Sloping Lands and theManagement of Soil Erosion Consortium (MSEC).

Past research and development efforts have not beenable to provide sustainable solution to land and waterdegradation problems, and soil erosion has remaineda major constraint in improving the living conditionsof the people in the marginal and sloping uplands inAsia. While sustainable land and water managementis primarily the activity and responsibility of thefarmers and other major users of the land, it alsorequires support from research and developmentagencies and organizations working at various levels.To be efficiently conducted, programs on soilmanagement require a good knowledge of soils andtheir environment, of the farmers and their practices,as well as of the latest developments in research onthe subject. Individual efforts in soil management aretime consuming and costly. The use of existingknowledge, the sharing of new findings by nationaland international institutions working on the samesubject, and the coordination of these efforts arebelieved to be the most cost-effective ways to tacklethis problem.

210

Organization of the Network andConsortiumNetworking hopes to help the different partners,especially the NARES cooperators to conduct theinvestigations (by designing methodologies,disseminating information, and coordinatingprograms) necessary for the practical validation ofimproved soil management and land use practices(IBSRAM 1985). The organization of the regionalnetwork program consists basically of three mainbodies (Fig. 1). The cooperators initiate andundertake the soil management program through:(i) simple participation in the different programactivities, mainly by sharing information; (ii) activeparticipation both by having an accepted program andby participating in all the various program activities;(iii) basic participation by having an approvedprogram, some basic research related to theobjectives of the network, and also participation in allthe program activities; and (iv) support participationby the international and other research agencies byundertaking some part of the basic research related to

the objectives of the network, either alone or inconjunction with other cooperators.

IWMI, through a Coordinator and backed by thenetwork steering committee (NSC), catalyzes,coordinates, and assists the NARES partners inconducting their activities. It provides assistance inthe preparation and in the presentation of projects todonor agencies. The Coordinator serves as the linkbetween the different partners and IWMI. Also, thecoordinator helps strengthen the nationalcooperators’ program by regular visits andconsultations and by backstopping the network orconsortium activities. This was the mechanismfollowed by the ASIALAND Network on theManagement of Sloping Lands, which is one of theearlier networks established. Considering the lessonslearned from the earlier regional network projects,particularly the ASIALAND network on slopinglands, and other studies on soil erosion, MSEC wasorganized to implement soil erosion managementresearch. The major difference from the ASIALANDnetwork is that it works on the catchment level ofstudy.

Figure 1. The organization of the regional network program.

NATIONALCOOPERATORS

IWMIcoordinator

NSC

DONORS

Internationaland otherresearchagencies

211

The consortium employed an organizationalmodel that engages scientists and research institutionsto tackle a common goal. The model allowsparticipation of those who can contribute, exploitssynergies, and is mutually beneficial. Researchplanning is undertaken through consultation amongconcerned NARES, international agriculturalresearch centers (IARCs), non-governmentalorganizations (NGOs), and farmers. A facilitator, asteering committee, and an annual assembly areessential to ensure the effective operation of theconsortium. NARES play the central role in theconsortia and in participatory research, but with abroad responsibility for underpinning applied andstrategic research. Another major attraction of theconsortium model is the capacity to draw the interestof advanced research institutes (ARIs), which cancontribute significantly through strategic and basicresearch. In turn, the consortium provides a carefullyconsidered development context for their morestrategic interests. The whole idea of the program is totake a bottom up approach in research planning withiterative discussions between farmers, NARES,IARCs, and ARIs in the definition and implementationof the research process. In both instances, the donorsplay a crucial role. They provide funds for programcoordination and in most part, support the activities ofthe individual country participants.

Implementing the Networkand Consortium Programs

The ASIALAND network

The ASIALAND network was established in 1988with initial funding from the Asian DevelopmentBank (ADB) and continued with the support from theSwiss Agency for Development Cooperation (SDC).Presently, the participating countries include China,Indonesia, Laos, Malaysia, Philippines, Thailand, andVietnam. Its main objective is to assist the nationalagricultural research systems (NARS) in conductingresearch and related activities that promotesustainable land management on sloping lands inAsia. The specific objectives are to:• Assist NARS in validating or testing improved soil

management technologies;

• Facilitate the exchange of research information onsoil management among agricultural scientists inthe region through meetings, workshops, andpublications;

• Strengthen the research capacity of the NARSparticipating in the network; and

• Evaluate and select cost-effective and farmer-acceptable options for a more sustainableagricultural production in sloping lands.The network realizes that technology transfer

should be demand-driven, and these needs shouldcome from the farmers themselves. Researchers canact as facilitators for farmer-clients to identify theirneeds (Sajjapongse and Maglinao 1998). To addressits objectives, the network has chosen a stepwiseapproach in identifying appropriate soil conservationtechnologies and extending them to the farmers. It isenvisaged that implementation should be carried outstep by step in phases. In Phases 1 and 2, the projectfocused mainly on research. Various improvedtechnologies were developed and validated underfarmers’ conditions. These technologies include alleycropping, grass strip barriers, hillside ditch,agroforestry, strip cropping, legume cover crop, andcrop residue management.

In Phase 3, a different approach was introduced totake into account the farmers’ socioeconomic context.Instead of researchers doing experiments in farmers’fields, researchers acted as facilitators. In addition,extension services started to be involved.Technologies were chosen by the farmer and testedagainst their common practices in their fields. InPhase 4, the most promising technologies weredisseminated, and pilot conservation farming villageswere established. Phase 5 of the project started inSeptember 2001 and is aimed at enhancing wideradoption through a more intensive and extensiveextension campaign. At the end of this phase, it isexpected that at the national level, a larger network ofinstitutions would be created. Furthermore, thenetwork itself will be strengthened to pave the way forsustaining their activities even after the internationalleadership ceases.

The MSEC project

The MSEC was established to address the inadequateunderstanding of the socioeconomic and policy

212

factors underlying land degradation andimprovement. The fragmented research on soil, water,and nutrient management with little coordination atthe national, regional, and international levels hasalso resulted in very insignificant impact of soilmanagement technologies. MSEC is one of the fourconsortia established through the Soil, Water, andNutrient Management (SWNM) initiative of theConsultative Group on International AgriculturalResearch (CGIAR). The participating countries in thecurrent project are India, Indonesia, Laos, Nepal,Philippines, Thailand, and Vietnam.

Consultations among various stakeholders in theestablishment of the consortium started in 1996 andsupport for the operation of the project in theparticipating countries began in late 1998 with fundsprovided by ADB. The objectives of the consortiumare to:• Develop sustainable and acceptable community-

based land management systems that are suitablefor the entire catchment;

• Quantify and evaluate the biophysical, envi-ronmental, and socioeconomic effects of soilerosion, both on- and off-site;

• Generate reliable information and prepare scien-tifically-based guidelines for the improvement ofcatchment management policies; and

• Enhance NARES capacity in research on integratedcatchment management and soil erosion control.Project implementation follows an interdisciplinary,

participatory, and community-based approach. Thisensures that all stakeholder groups in the landscapeaffected by soil erosion, including farmers and policymakers, benefit from the knowledge generated,recognize the scope and severity of the problem, andmake appropriate decisions about investments andland use policy in the sloping land areas. It startedwith the selection of representative catchments inparticipating countries by an interdisciplinary teamusing carefully defined criteria and methodologicalguidelines (IBSRAM 1997). Visits and dialogueswith local institutions, scientists, and farmers werefacilitated by the NARES.

After finally selecting the model catchments, moredetailed characterization was done to establish thebaseline information about the sites. Different toolsand techniques for conducting both biophysical andsocioeconomic surveys were employed (Maglinao et

al. 2001). Several microcatchments representingvarious land uses were further identified anddelineated to conduct more detailed soil erosion andhydrological studies. Hydrological monitoringstations equipped with automatic water levelrecorders, manual staff gauges, sediment traps,automatic weather instrumentation, and manual raingauges were installed. Data collection, monitoring,and analysis followed the agreed upon protocol.

To provide guidelines for and direction of theimplementation of the consortium activities, theconsortium Steering Committee was formed. Withinthe countries, collaboration among relevant partnershas likewise evolved (Table 1). The organization ofthese teams from different institutions and disciplinesis expected to enhance the participatory,interdisciplinary, and inter-institutional mechanismthat the consortium is implementing. Generally, thisarrangement is committed through formal agreementssigned between and among institutions. Furtherstrengthening the institutional linkages and sharing ofinformation among the consortium partners is theconduct of the consortium annual assembly. Thisprovides the opportunity to discuss issues concerningthe implementation of the approach. The first fouryears of the project will be completed in August 2002and a follow-up phase has been prepared. Theconsortium is envisioned to function for at least 10–12 years.

Project ExperiencesThe experience of the ASIALAND network and theMSEC project in Southeast Asia could be valuablelessons that may be considered in future projects thatintend to do similar research and extension approaches.The following sections highlight these experiences inrelation to project design, project implementation, andproject outputs. Some insights on the participatory andconsortium approach are also given.

Project design

Research and development work on land managementand ultimate extension and adoption of their resultsnormally require a longer time frame. TheASIALAND network project has been carried out inphases for more than 12 years. This has caused

213

anxiety to the cooperators every time the currentphase was about to end. As such, there were outputsthat should have been delivered very much earlier hadthere been assurance of a much longer-termcontinuity. The network was not planned to be a long-term project right from the beginning (Santoso et al.2000). Therefore, from phase to phase, all thecollaborating countries with the leadership of thenetwork coordinator have to prepare new proposalsfor the succeeding phases. Fortunately, SDC hasrecognized the accomplishments of the project andcontinued to provide support for a longer period.

In the case of MSEC, a long-term program hasbeen developed although funding is only assured forlimited duration. In any case, both projects weredeveloped and designed considering the increasingawareness of the effects of soil erosion and landdegradation on agricultural productivity and theenvironment and the advances in research on thesubject. The objectives were identified based on thecommon observation that adoption of land management

technologies is generally insignificant. The objectiveshave been set to address the important issues asperceived by the major stakeholders and the designprocess was carried out with their active involvement.Because of the involvement of a broad range ofstakeholders, the project design process has taken noless than two years from conceptualization andconsultations to the start of implementation.

As the project looks both at research and researchmethodology, further refinement of the design itself isalso carried out. In the ASIALAND network, this wasdone in the preparation of the succeeding phases. InMSEC, the design intended to integrate biophysicaland socioeconomic aspects, but it was observed that inthe early stages, there was a very strong focus on thebiophysical aspects and therefore further discussion onhow the socioeconomic and institutional researchcould be addressed was recommended. Thisconsideration becomes very important when looking atthe off-site impacts of soil erosion and the ultimateadoption of the recommended options.

Table 1. Consortium partners and potential research activities in the MSEC collaborating countries.

Country National partners International institutions Proposed research activities

India CRIDA, BAIF, IISS, IRD, ICRISAT Agronomy; farming systemsJNKKV

Indonesia CSAR, BAPEDA, CIRAD, IRD Agronomy; hydrological studies;BPTP, CSES institutional arrangements

Laos SSLCC IRD, ICRAF, NORAGRIC Hydrological studies; nutrientdynamics

Nepal NARC ICIMOD, University of Farming systems; nutrientBayreuth, IRD, IFPRI dynamics; hydrological studies;

institutional aspects

Philippines PCARRD, CMU, BSWM ICRAF, IRD, SEARCA, Hydrological studies;DA, DENR, SANREM, ACIAR institutional arrangements;UPLB, local government policy studies; off-site impact;

farmers’ adoption; modeling

Thailand RFD, DLD ICRISAT, IRD, AIT, Farming systems; off-siteUniversity of Bayreuth impact; nutrient dynamics and

pollutants; hydrological studies;remote sensing

Vietnam NISF, NEU, VASI ICRISAT, CIRAD, IRD, Farming systems; hydrologicalIFPRI, IRRI studies; institutional

arrangements

214

Project implementation

In the case of MSEC, the consultations and meetingswith various stakeholders have taken much time, andthe start of full-scale implementation of the projectwas greatly delayed. A memorandum ofunderstanding (MOU) with the NARES was stillneeded to formalize the implementation of the projectin the participating countries. Attracting donors toprovide funds for the project was likewise a slowprocess. For the ASIALAND network, theparticipation of major stakeholders proceeded withthe development of the proposals by phases. In theearlier phases, mainly researchers were involved. Inthe later stages, the farmers and extension workersplayed an active role.

Both projects have steering committees to providedirection and guidance for the operation of thenetwork or consortium. However, there is stillconcern on the effectiveness of such committees inproviding the expected inputs. The committees meetonly once or twice a year. The members of thecommittee are usually the national coordinators whothemselves are busy with their regular responsibilityin their respective institutions. Several suggestionshave been forwarded to strengthen these committees.For MSEC, a smaller cluster or task force has beencreated to look into the more specific issues ofresearch, capacity building, and informationdissemination. Another suggestion was to have justone committee for the two projects, but there were anumber of concerns that still have to be clarified.

Monitoring and evaluation is usually done throughfield visits, annual reviews, and reports. BothASIALAND and MSEC conduct annual review andplanning meetings to review the progress and plan forthe following year. These meetings also serve as thevenue to share experiences and discuss problems ofimplementation.

Communication between IWMI and the NARESand among the NARES themselves has much more tobe desired. Exchange of information and monitoringis very critical in this kind of research, whichimplements new methodologies and involves anumber of partners. Thus, communication betweenand among partners needs to be further strengthened.Transaction cost was initially very significant. Theexistence of the network and the consortium as a

long-term collaborative arrangement can beascertained heavily by the full commitment of allpartners, particularly the NARES.

The leadership role of the network or consortiumcoordinator is quite crucial. Even without a network,researchers can work together and help one another.However, such bilateral and direct cooperation rarelytakes place. By working in a network mode, thenational coordinators and other researchers relatedwith the project were benefited through exchange ofideas and support from the network coordinator. Jointauthorship and cross-country analysis of results canbe facilitated by the network coordinator.

Project outputs

Alternative technology and its adoption

Results of technology validation in the ASIALANDnetwork showed that some of the technologies testedwere effective in reducing runoff and soil erosion andin maintaining good growth and yields of crops. Thealley cropping system, which was a major feature inthe improved conservation farming, has beenconsidered an effective conservation measure and animportant option for sustainable land management inall member countries. However, there were varyingdegrees of acceptance of the system. Understandably,farmers would select certain technologies that theyperceive to be most appropriate to their socioeconomicconditions. Suthipradit and Boonchee (1998) indicatedthat reluctance to adopt improved technologies couldbe due to high cost, high labor requirement, and noimmediate return or benefits. From the experience ofthe network, Maglinao and Phommassack (1998) andSajjapongse and Maglinao (1998) suggested toconsider a number of issues to achieve sustainableadoption by farmers. These include more involvementand participation of the farmers in the project,provision of short- and long-term benefits, properlyimplemented provision of subsidy, and long-term andconstant follow-up by extension.

Monitoring of soil erosion in MSEC catchmentsshowed that it is affected by land use, farm activities,and catchment size (Maglinao et al. 2002, Maglinaoand Penning de Vries 2003). More intensivelycultivated areas produced more erosion than thoseplanted to perennials or left idle with grasses. Erosion

215

was also higher when high amount of rainfall occurswhen the soil had little cover (at planting time).Runoff and erosion start when the soil becomessaturated and infiltration is low. On a per-hectarebasis, calculated soil loss is higher in smaller sub-catchments probably because of the short distancetraveled by the moving sediments and less depositionoccurring along the way. In general, the landmanagement options identified with the farmers arevariants of the hedgerow cropping technology. Whilethe farmers expressed interest to apply thesetechnologies, they mentioned some constraints likeinadequate labor for establishment and land tenureconcerns. Recommendation of land managementoptions for a particular area also depends oncatchment size. The role of the farmers in decidingthe options that will be introduced is crucial.

These observations will be also useful inidentifying erosion “hot spots” where application ofsoil conservation measures should be prioritized.Also, these will be important in the development andapplication of the methodology for extrapolation orscaling up of potential interventions for sustainableupland development. As the interventions introducedin the catchment will surely affect other sectorsdownstream, recognition of their concerns becomesnecessary.

Capacity building

During the long tenure of the ASIALAND network, ithas accumulated voluminous data, information, andexperiences. The network has contributed to variouspersonal and professional changes in the differentpartners involved both at the national and theinternational levels. For example, several staff of theNARES have been promoted to lead importantdivisions or departments. Directly or indirectly, thishas benefited and strengthened the network. Many ofthe promoted staff were able to provide greatersupport to the network (Santoso et al. 2000).

Information materials and publication

For the annual meeting, each country prepares anannual report. This has been a good practice to reportthe network’s achievements. They are published inproceedings, which can be used as reference byresearchers, extension workers, or other agricultural

officials from member and non-member countries.The report is also for the project and the donor. Theproceedings, however, generally contain only countryreports and no network analysis of the data is done. Asa research network with research sites spread widelyacross different agroecosystems in Asia, the networkhas exceptional opportunities to do and produceanalyses across sites. It is only recently that thisactivity is being pushed through. Analyses acrosscountries can actually strengthen the researchnetwork.

The ASIALAND network has tremendousopportunities to also produce technical papers orscientific publications in national and internationalscientific journals. This remains a great challenge forall members involved in the network. Similaractivities are also done by the MSEC project. Thetechnical sessions during the annual meetings haveconsidered the presentation of cross-country analysis,but annual reports are still submitted. MSEC alsomaintains a web page, which is now redesigned to fitinto the IWMI site. In addition, it has started tooperate a list server called SLMNet.

Tools, guidelines, and methodologies

The ASIALAND network produced the decisionsupport system for sustainable land management(DSS-SLM) for evaluating sustainable landmanagement for sloping lands. A second version wasdeveloped by MSEC to understand the interactionamong erosion, nutrient depletion, and conservationpractices. However, a prognosis module is still neededto make it more user friendly (Santoso et al. 2000).

The soil and land database (SALAD) was alsodeveloped to store and retrieve the data from theASIALAND project. Training of partners was done toorient them on the use of the program. MSEC’semphasis on both research and research methodologyproduced tools and guidelines to support decisionmaking and improved implementation of its researchactivities. One such output is an earlier publicationwhich provides the guidelines for model catchmentselection for MSEC. The site selection was based oncriteria agreed upon by the consortium partners.

The minimum data sets for biophysical andsocioeconomic site characterization was preparedand employed. Protocols on the biophysical datacollection, analysis, storage, and retrieval have been

216

discussed during the country visits of IWMI staff. Theexisting methodologies for the economic assessmentof soil erosion and nutrient depletion and on-farmtrials have been adapted and applied in the MSECsites. A soil erosion and hydrology model wasdeveloped to predict the consequences of land usechanges, technology interventions, and other factorson soil erosion. This will be helpful in decidingdevelopment initiatives that have to be introduced in aparticular area.

The consortium approach

The network and consortium arrangement has been agood approach to organize many activities covering alarge geographical area. It provides avenues forexchange of experiences between and amongpartners. Through this mechanism, a broadercollaboration base is established and inputs to projectplanning, implementation, and evaluation are viewedat a wider perspective. Researchers and extensionworkers under the ASIALAND network have speededup the transfer of research results throughmethodologies developed by combining experiencesfrom different countries. The network also broadensthe view of researchers and enhances researchers’experiences by visiting other centers in the network(Santoso et al. 2000).

The experience of MSEC may not still be sufficientto make any conclusion on the approach that itemploys. However, it has in itself added a newdimension to soil erosion management, with thepotential to enhance the adoption and sustainability ofintroduced interventions. In a self-appraisal conductedby the project, benefits from the approach are alreadyevident. These include increased awareness of decisionmakers and research leaders, increased cooperationacross disciplines and institutions, choice of morerelevant research topics, and applications of findingsacross sites (Maglinao and Penning de Vries 1999).

While there are some positive aspects that can bederived from the consortium arrangement, it is alsoworth noting that there are other concerns that mustbe looked at. For instance, consultations amongpartners to come up to an agreement has proved to bea slow and time consuming process. With a number ofpeople and sectors involved, communicationbecomes critical. Likewise, different partners have

different levels of resources and capacity. Thus,strategies to address these concerns should bestrongly considered. Commitment of the partners tofulfill their obligations to the project is crucial.Leadership appears to play a major role. In essence,the approach should be combined with otherapproaches and this will depend on the conditionswhere and when the project is implemented.

Conclusion and RecommendationThe consortium approach to the implementation ofsoil erosion management research has resulted insome benefits. These include increased awareness ofdecision makers and research leaders on the impactsof resource degradation, increased cooperation acrossdisciplines, choice of more relevant topics, andapplication of findings across sties.

IWMI’s experiences in employing the approach insoil erosion management research on sloping lands inAsia have provided some challenging tasks stillahead. Looking at a wider perspective, the approachcan be modified to suit the requirements for anintegrated land and water management researchplanning and implementation.

With stronger and continuing partnerships amongstakeholders, particularly the farmers, it is believedthat the network and consortium arrangement willbear its fruits in the longer term. IWMI will continueto employ this approach and the promising outputswill further be validated at different scales ofapplication and expanded to a much wider area forgreater impact.

Acknowledgment

The authors appreciate the contributions of thedifferent NARES partners of both the ASIALANDNetwork and MSEC in terms of feedback on theirown experiences in the implementation of theprojects. They also acknowledge the funds providedby SDC and ADB for these projects.

References

IBSRAM (International Board for Soil Research andManagement). 1985. IBSRAM highlights 1985. Bangkok,Thailand: IBSRAM.

217

IBSRAM (International Board for Soil Research andManagement). 1997. Model catchment selection for theManagement of Soil Erosion Consortium (MSEC) ofIBSRAM. Report on the Mission to Thailand, Indonesiaand the Philippines. Bangkok, Thailand: IBSRAM.

Maglinao, A.R., and Penning de Vries, F. 1999. MSEC:Implementing a new paradigm in research on soil erosionmanagement in Asian catchments. Presented at the ThirdInternational Symposium on Systems Approaches forAgricultural Development (SAAD III), 8–11 November1999, Lima, Peru.

Maglinao, A.R., and Penning de Vries, F. 2003. TheManagement of Soil Erosion Consortium (MSEC): Linkingland and water management for sustainable uplanddevelopment in Asia. Pages 31–41 in Integrated watershedmanagement for land and water conservation and sustainableagricultural production in Asia: proceedings of the ADB-ICRISAT-IWMI Project Review and Planning Meeting,10–14 December 2001, Hanoi, Vietnam (Wani, S.P.,Maglinao, A.R., Ramakrishna, A., and Rego, T.J., eds.).Patancheru 502 324, Andhra Pradesh, India: InternationalCrops Research Institute for the Semi-Arid Tropics.

Maglinao, A.R., Penning de Vries, F., Agus, F., Ilao,R.O., and Toan, T.D. 2002. Land management options forreducing soil erosion in sloping uplands. Presented at theInternational Symposium on Sustaining Food Security andManaging Natural Resources in Southeast Asia –Challenges for the 21st Century, 8–11 January 2002,Chiang Mai, Thailand.

Maglinao, A.R., and Phommassack, T. 1998. Enhancingand sustaining technology adoption through appropriateincentives. Pages 71–82 in Farmers’ adoption of soil-conservation technologies. Proceedings of the 9th AnnualMeeting of the ASIALAND Management of the Sloping

Lands Network, Bogor, Indonesia, 15–21 September 1997(Sajjapongse, A., ed.). IBSRAM Proceedings no. 17.Bangkok, Thailand: International Board for Soil Researchand Management.

Maglinao, A.R., Wannitikul, G., and Penning de Vries, F.2001. Soil erosion research in catchments: initial MSECresults in Asia. Pages 51–64 in Soil erosion managementresearch in Asian catchments: Methodological approachesand initial results. Proceedings of the 5th Management ofSoil Erosion Consortium (MSEC) Assembly (Maglinao,A.R., and Leslie, R.N., eds.). Bangkok, Thailand:International Water Management Institute, Southeast AsiaRegional Office.

Sajjapongse, A., and Maglinao, A. 1998. Technologytransfer: The ASIALAND Management of Sloping LandsNetwork approach. Pages 51–57 in Farmers’ adoption ofsoil-conservation technologies. Proceedings of the 9th

Annual Meeting of the ASIALAND Management of theSloping Lands Network, Bogor, Indonesia, 15–21September 1997 (Sajjapongse, A., ed.). IBSRAMProceedings no. 17. Bangkok, Thailand: InternationalBoard for Soil Research and Management.

Santoso, D., Penning de Vries, F., and Boonchee, S. 2000.The ASIALAND management of sloping lands forsustainable agriculture in Asia network: Lessons learned.Presented at the 5th MSEC assembly, Semarang, Indonesia.

Suthipradit, S., and Boonchee, S. 1998. Constraints totechnology adoption of farmers on sloping land. Pages 83–91 in Farmers’ adoption of soil-conservation technologies.Proceedings of the 9th Annual Meeting of the ASIALANDManagement of the Sloping Lands Network, Bogor,Indonesia, 15–21 September 1997 (Sajjapongse, A., ed.).IBSRAM Proceedings no. 17. Bangkok, Thailand:International Board for Soil Research and Management.

218

A Consortium Approach for Sustainable Managementof Natural Resources in Watershed

S P Wani1, T K Sreedevi2, H P Singh3, T J Rego1, P Pathak1, and Piara Singh1

Abstract

The natural resource management in the dry regions imposes challenging pressure on the fragileecosystems, as they are the source of livelihoods of people whose key occupation is agriculture. ICRISAT’searlier experiences indicated that in the past, watershed management emphasized soil and waterconservation measures. Lack of holistic approach to natural resource management in conventionalwatersheds has led to the emergence of a new integrated watershed management model. Importantcomponents of this new model are farmer-participatory approach, use of new science tools, knowledge-flow from on-station to on-farm watersheds, holistic systems approach with integrated genetic and naturalresource management (IGNRM) strategy providing site specific solutions, a consortium of institutions fortechnical backstopping, continuous monitoring and evaluation by the stakeholders, community andwomen empowerment, and environmental protection. The main features of the consortium approach aretechnical backstopping by the consortium of multi-institutions, linking strategic and developmentalresearch on farmers’ fields, reducing the lag for transferring results from research fields to farmers’ fields,empowering the development workers and farmers to manage natural resources sustainably, andharnessing the strengths of the partners to make a win-win situation for all the partners. The consortiumstrategy has facilitated the exchange of knowledge and technologies amongst the consortium partners,reduced land degradation, and improved rural livelihoods through increased incomes.

The natural resources in the semi-arid tropics (SAT)are the “life line” of rural livelihoods, the keyoccupation being agriculture. These dry eco-regionsare predominantly rainfed, marginal, and fragile, andprone to severe land degradation. Unpredictableweather, limited and erratic rainfall with longintervals of dry spells, and intense rainfall causingrunoff and severe soil erosion characterize these dryregions. The overexploitation and reduced rechargeof groundwater, along with low rainwater useefficiency is another serious threat to scarce waterresources in the dry regions. Low levels of soilorganic matter, accompanied by high rates of organicmatter degradation aggravated by low literacy andpoverty are the major causes of low productivity and

depleting natural resource base in the dry regions. Thechallenge, therefore, is to develop sustainable andenvironment-friendly options to manage naturalresources in this fragile ecosystem to increase theproductivity and incomes of millions of poor farmerswho are dependent on the natural resources for theirsurvival. The way forward to address this gigantic taskis by sustainable management of natural resources in amanageable land unit, which is a watershed.

A watershed is a logical, natural planning unitfor sustainable resource management. Integratedwatershed management is the rational utilization of allthe natural resources for optimum production to fulfillthe present need with minimal degradation of naturalresources such as land, water, and environment.

1. International Crops Research Institute for the Semi-Arid Tropics (ICRISAT), Patancheru 502 324, Andhra Pradesh, India.2. Andhra Pradesh Rural Livelihoods Programme (APRLP), Rajendranagar, Hyderabad 500 030, Andhra Pradesh, India.3. Central Research Institute for Dryland Agriculture (CRIDA), Santoshnagar, Hyderabad 500 059, Andhra Pradesh, India.

219

ICRISAT’s ExperiencesScientists of the International Crops ResearchInstitute for the Semi-Arid Tropics (ICRISAT) havelearned the following lessons over years working withwatershed technologies in partnership with nationalagricultural research systems (NARS) (Wani et al.2002a) for possible reasons of low adoption ofwatershed technology package.• Efficient technical options to manage natural

resources for sustaining systems are needed.• Key processes and institutional lessons such as

mere on-farm demonstration of technologies bythe scientists do not guarantee adoption by thefarmers.

• Contractual mode of farmers’ participationadopted during Vertisol technology evaluation didnot present the expected results. There is a need tohave higher degree of farmers’ participationthrough consultative to cooperative mode fromplanning stage up to evaluation stage.

• Appropriate technology application domains forregion specific constraints need to be identified,e.g., broad-bed and furrow (BBF) for all Vertisols.

• Developmental watershed projects implementedby non-governmental organizations (NGOs)lacked technical support; thus technicalbackstopping is essential. Any single organizationcannot provide answers to all the problems in awatershed. Thus, a consortium of organizations isneeded for technical backstopping.

• Process of partnership selection for eachwatershed has to be undertaken carefully. Ageneralized formula-based selection does notguarantee success. For example, not all NGO-implemented watersheds adopted participatoryapproach and were successful.

• Technical change is intimately bound with broaderinstitutional context of the watershed and the roleof institutions and different players varies fromlocation to location.

• Individual farmers should realize tangibleeconomic profits from the watersheds; only thenthey would come forward to participate incommunity-based activities in the watershed.

• Most farmers considered watershed programs assource of employment in the project for soil andwater conservation measures and not as programs

which could generate long-term employment orincrease incomes of most of the small farmersindividually.

• Holistic systems approach through convergence ofdifferent activities is needed and it should result inimproved livelihood options and not merely soiland water conservation in the watershed.

• Technological packages as such are not adoptedand farmers adopted specific components that theyfound beneficial.

• Capacity building for all the stakeholders iscritical.

• Women and youth groups play an important role indecision making in the families.

• Sustainability although desired is rarely visibleafter project duration is over. Exit strategies arenot planned in almost all the projects.The watershed programs were undertaken for

managing natural resources and improvingagricultural productivity thereby improving the rurallivelihoods. However, the expected benefits fromthese investments were not realized mainly due tolack of people’s participation, lack of scientificinputs, compartmentalized approach with maximumemphasis on construction of rainwater harvestingstructures (many of which are of poor quality), lack oftangible economic benefits to individuals,involvement of contractors for executing works, andnon-involvement of landless families and marginallandholders (Farrington and Lobo 1997, Kerr et al.2000, Wani et al. 2002a, 2002b).

New Integrated WatershedManagement Model for EfficientManagement of Natural ResourcesA new model for efficient management of naturalresources in the SAT has emerged from the lessonslearned from long-term watershed-based research byICRISAT and NARS partners (Wani et al. 2002a).The important components of the farmer participatoryintegrated watershed management model are:• Farmer participatory approach through

cooperation model and not through contractualmodel with stakeholders’ involvement at all thelevels right from inception (planning andimplementation) to managing the process andsharing the benefits in the watersheds.

220

• Use of new science tools such as remote sensing,geographic information system (GIS), digitalterrain modeling, and crop simulation modelingfor monitoring and management of watersheds.

• Link on-station and on-farm watersheds andfacilitate the ‘knowledge flow’ of the successes ofon-station watersheds at ICRISAT to on-farmwatersheds, and to use feedback to guide furtherresearch in on-station watersheds.

• A holistic systems’ approach with integratedgenetic and natural resource management(IGNRM) strategy as a new paradigm.

• A consortium comprising several institutions fortechnical backstopping of the on-farm watersheds.


• A holistic approach to improve livelihoods of peopleand not merely conservation of soil and water.

• Cost-effective technology approach such as low-cost soil and water conservation structures.

• Amalgamation of traditional knowledge and newlydeveloped technologies.

• Minimize free supply of inputs for undertakingevaluation of technologies. Farmers are encouragedto evaluate new technologies themselves withoutfinancial subsidies.

• Emphasis on individual farmer-based conservationmeasures for increasing productivity of individualfarms along with community-based soil and waterconservation measures.


• Empowerment of community individuals andstrengthening of village institutions for managingwatersheds with emphasis on women empowerment.

• Environmental protection.

Consortium model for developing andmanaging watersheds

The concept of consortium is an integral part of thenew integrated watershed management model. Theconsortium model is a participatory watershed systemwith a multi-disciplinary and multi-institutionalapproach to technically support a process involving

people who aim to create a self-supporting system forsustainability (Fig. 1.). The approach is built on theprinciple of harnessing the strengths of the consortiumpartners for the benefit of all the stakeholders includingthe farmers. The main features of the approach are:• Technical backstopping by the consortium of

multi-institutions.• Links strategic and developmental research on

farmers’ fields.• Cuts down the lag for transferring the results from

research fields to farmers’ fields.• Empowers the development workers and farmers

to manage natural resources sustainably.• Harnesses the synergies of the strengths of the

partners to make a win-win situation for all thepartners.The model is a holistic systems approach and it

demands collective efforts of all the stakeholders toaddress the complex problems in watersheds. Theapproach is also a knowledge-driven managementsystem, which operates in a watershed as a unit forefficient management of natural resources.

Process components and execution strategy

• Participatory and bottom-up approach to identifythe problem, possible solutions, and approaches.

• Holistic systems approach and use of new sciencetools for managing and monitoring the watershedsto sustain/increase productivity, and improvelivelihoods.

• Site-specific solutions through refinement of existingoptions.

• Consortium approach to address complex problemsthrough technical backstopping. The consortiumpartners include farmers, NGOs, governmentorganizations, advanced research institutions,extension agencies, private entrepreneurs, andfarmers’ training centers (FTCs), who share thebroad goal of improving rural livelihoods in thewatershed and not restricted sectoral goals such aswater and soil conservation.

• Emphasis on empowerment of stakeholders toenable them to take decisions, implement theprograms, and manage the processes.

• Continuous monitoring and mid-course refinementof technologies to meet the local needs.

221

Scaling-up of the watershed approach

To scale-up the benefits of integrated watershedmanagement observed in operational-scale watershedsat research station to the real world on-farmwatersheds, the approach followed is in a participatorymode in Asia under ICRISAT’s project “ImprovingManagement of Natural Resources for SustainableRainfed Agriculture”, funded by the AsianDevelopment Bank (ADB). All the on-farm technologyevaluation trials are conducted on benchmarkwatershed sites in partnership with farmers. The on-farm watersheds vary from 30 ha to 10,000 ha withvarying agro-ecological potential. Currently, we areevaluating the model of technical backstopping the on-farm watersheds, which are planned, developed, andmonitored in partnership with NARS, NGOs, andfarmers, using new science tools. Five on-farm

watersheds in India, Thailand, and Vietnam are inoperation. The model followed adopts amultidisciplinary and multi-institutional consortiumapproach for technical backstopping the developmentprojects. “Islanding approach” is the strategy forlinking strategic research done in micro-watershedswithin a community watershed with applied on-farmresearch for development to provide effectivemechanisms to more effectively transfer technologiesfor managing natural resources to farmers. Holisticfarming systems approach to sustain productivity andto improve land and environment quality is adopted. Atthe village and community level women have beenempowered through group training. Women are usuallythe critical group involved in decision-makingregarding natural resources management. Continuousmonitoring and impact assessment is considered anintegral part of the program right from the initial stage.

Figure 1. Consortium model for watershed management.(Note: AP Govt. Depts. = Andhra Pradesh Government Departments; FTCs = farmers’ training centers; KVKs =Krishi Vigyan Kendras; CRIDA = Central Reserch Institute for Dryland Agriculture; NGOs = non-governmentalorganizations; ANGRAU = Acharya NG Ranga Agricultural University; ARIs = advanced research institutions.)

Adarsha Watershed Consortium

222

Community Participation – AnEssential Element for SuccessfulWatershed ManagementPrograms of development and management of naturalresources have suffered due to inadequateparticipation of local people. For success of anystrategy of natural resource management involvementof local people is important. People and livestock arean integral part of the watershed community andshould be given utmost importance. They depend onthe watershed for their needs and in turn influence thegood or bad events in the watershed. Thusparticipation of the people is essential for the successof the watershed programs. The detailed analysis ofsuccessful watersheds revealed that communityparticipation played a significant role in making thewatersheds successful.

In the past, watershed management wassynonymous to soil and water conservation. In thenew approach, it is more synonymous with people’slivelihoods and is used as a vehicle for overalldevelopment of rural people through povertyalleviation and sustainable development for thewelfare of the people. With the new focus on povertyalleviation and food security through appropriatenatural resources management, the people rather thanthe natural resources become the first focus forwatershed management. The degree of peoples’participation in watershed management varies fromlocation to location and is described below:• Contractual – Contract farmers to provide land

and/or services for experiments to be conducted byscientists.

• Consultative – Farmers consult scientists abouttheir problems and solutions but decision is madeby the researchers.

• Collaborative – Farmers and scientists collaborateas partners in the research process.

• Collegiate – Farmers conduct the research andresearchers provide technical advisory support.Participatory watershed management aims at

farmers’ and community involvement in planningand management of natural resources in a watershedfor sustainable use. Since farmers and other landusers are the main stakeholders in watershedmanagement, they themselves are to take charge ofthe processes for development of watershed

resources. Participation means the act of partakingby farmers in all the stages of watershed programsright from planning, designing various structures,execution, monitoring and evaluation of theirperformance. Such participation requires that thetarget farmers voluntarily spend their time andenergy for the program and adopt the recommendedmeasures and practices, repair and maintain them ingood condition on a sustained basis.

The traditional systems of use of natural resourcesin the village communities have evolved over a periodof centuries. However, the traditional systems thatonce met the test of sustainability have not been ableto respond adequately to modern rates of growth indemand as demanded by current population pressuresand rapidly declining quality of land and waterresources. To achieve sustainable use of naturalresources there is a need to increase farmers’participation in efficient management of naturalresources.

Basic principles for effective communityparticipation

Some basic principles which facilitate effectivecommunity participation are: compelling vision;strong and shared leadership; shared problemdefinition and approach; power equity;interdependency and complementarity; mutualaccountability; attention to process; communicationlinkages; explicit decision-making process; trust andcommitment; and credit and recognition. Theparticipation process also includes a combination ofindigenous and traditional approaches, which maypave the way for long lasting participation. This iscritical in case of integrated watershed management.

Promoting community participation

The previous projects had sufficient expertise inimplementing soil and water conservation measuresand were largely based on technical perspective andinvolved only land and water management activities.The activities in those projects did not involve peoplewho are actually the important players within thewatershed and whose activities have a significantimpact. To make the watershed program successful,the primary goal should be the participation of the

223

local community. Project implementation can only besuccessful if the people participate and contributeadequately to the development program. In an effortto achieve community participation for managingnatural resources in the watersheds, ICRISAT isworking to build stronger partnerships with state andlocal agencies, community leaders, and people. Theseefforts are based on a strong commitment to involvethose affected by or responsible for environmentalregulation in finding the most effective workablesolutions possible. Successful partnerships arecritical for understanding participatory watershedmanagement, as several players with varying interestsare involved.

Model Application in ProjectThe project “Improving Management of NaturalResources for Sustainable Rainfed Agriculture” wasfunded by ADB in 1999 in an effort to improve thenatural resource base and to have sustained increasein food production by SAT farmers. The projectinvolves watershed research in three countries (India,Thailand, and Vietnam) at both on-station and on-farm watersheds. The on-farm benchmark watershedsin India, Thailand, and Vietnam are in operation since1999. This project demonstrated the consortiumapproach model application in a number of ways byencouraging community participation in watershedmanagement, by empowering the farmers, and also bybuilding stronger working relationships with state andlocal governments, and NGOs and encouragingvoluntary initiatives for improving sustainable use ofresources. Five on-farm and three on-stationwatersheds covering varying agroecological,socioeconomic, and technological situations wereselected. A case study of one on-farm watershed, i.e.,Adarsha watershed in Kothapally village, RangaReddy district in Andhra Pradesh, India is described.

Consortium partners and process

The consortium partners involved in integratedwatershed management in Adarsha watershed were:• ICRISAT – international agricultural research

center• Central Research Institute for Dryland Agriculture

(CRIDA) – NARS

• Drought Prone Area Programme (DPAP) –government organization

• M Venkatarangaiya Foundation (MVF) – NGO• National Remote Sensing Agency (NRSA) –

national institute• Farmers – Kothapally village

In Adarsha watershed, the total irrigable area wasvery less and no single water harvesting structure forhuman and animal use was seen in 1998, i.e., at thestart of the project. A large area is under rainfedfarming in the village. ICRISAT, DPAP, and MVFjointly selected this watershed to evaluate integratedwatershed management options for improving rainfedagricultural production through integrated watersheddevelopment and thus reduce poverty throughincreased system productivity. A micro-watershed of30 ha was selected in partnership with the farmers.The watershed is equipped with hydro-meteorological equipment and is also monitored forinputs, outputs, productivity, incomes, etc., forpreparing detailed budgets for water and nutrients atcatchment level and also to assess the impact oftechnical interventions. All the activities in thewatershed are planned, executed, and evaluated bythe farmers through the watershed committee andwatershed association with technical support fromICRISAT. These prime committees form further sub-committees for specific activities such as siteidentification for check-dams and farm ponds, and foridentifying farmers to evaluate the improved options.User groups were formed for development of waterharvesting structures. Self-help groups (SHGs) wereformed to undertake watershed developmentactivities. A system of social auditing is also anintegral part of the integrated watershed developmentactivity. New tools such as remote sensing and cropsimulation models were used for planning andmonitoring the development activities. Humanresource development was considered an importantcomponent of the model. Farmers were encouraged toundertake income-generation activities.

Monitoring and impact assessment

Continuous monitoring of several parameters wasdone in Adarsha watershed as described below:• Weather: An automatic weather station was installed

to continuously monitor the weather parameters.

224

• Runoff and soil loss: Runoff, soil, and nutrientlosses were monitored using automatic water levelrecorders and sediment samplers.

• Groundwater: To monitor the groundwater levels,open wells in the watershed were geo-referencedand regular monitoring of water levels was done.

• Crop productivities: Productivities were recordedfor every crop in each year.

• Nutrient budgeting: Studies on optimum doses offertilizers were conducted to have balancednutrient budgets.

• Biological nitrogen fixation (BNF): Quantificationof BNF in farmers’ fields was done using Ndifference method and 15N isotope dilutionmethod.

• Satellite monitoring: Changes in croppingintensity, greenery, water bodies, and groundwaterlevels were monitored. Also, GIS maps indicatingsoil types, soil depths, and crops grown duringrainy and postrainy seasons were prepared.Community-based soil and water conservation

measures such as grassed waterways and gabionstructures were constructed in Kothapally. Ninety-seven gully control structures, 60 mini-percolationtanks, 4 water storage structures, and 1 gabionstructure for increasing groundwater recharge werecompleted. Wasteland development was undertakenby contour trenching, planting horticultural andagroforestry plants, and developing grasslands. Alongwith water harvesting for enhancing water useefficiency, several improved land, crop, pest, andnutrient management options and soil conservationmeasures were taken up and all of these togethermade farmers reap rich rewards. Ten SHGs wereformed to undertake vermicomposting as a micro-enterprise in the village. Improved cropping systemswith high-yielding stress tolerant crop cultivars wereintroduced in the watershed. Bullock-drawn tropicultorwas used for sowing and fertilizer application.

The normalized difference vegetation index(NDVI) images showed that the spatial extent ofmoderately dense vegetation cover in Kothapallyincreased from 129 ha in 1996 to 152 ha in 2000. Thegroundwater level increased. Crop productivities,farmers’ incomes, and profits also increased. Farmerswere exposed to new methods and knowledge for

managing natural resources through training, videoshows, and field visits to on-station and on-farmwatersheds. Educated youth were trained in skilledactivities such as HNPV (Helicoverpa nuclearpolyhedrosis virus) production. Adarsha watershed isa model watershed with significant achievements(Wani et al. 2003).

Emerging Issues• Scaling-up from benchmark watersheds.• How to institutionalize consortium.• Harmonization of existing village institutions and

watershed-based institutions.• Ensuring effective functioning of user groups.• Common property resources – How to ensure

sharing of benefits between user groups andpanchayats.

• Equity and gender issues need special attention.• Efficient and sustainable use of water resources

(water use policies, water markets).• Financial operations and resources for functional

viability of associations.• Linking with markets and enterprises.• Identification of quantitative indicators for build-

up of social capital and processes.

ConclusionsA holistic consortium approach in watersheds enablesto have “win-win” situations for sustainingproductivity and reducing land degradation which arethe main causes of poverty in the rainfed areas ofAsia. The current model of watershed researchfollowed at ICRISAT links on-station research to on-farm situation and adopts the consortium approach bytechnical backstopping. This model seems to havevery high potential for bringing favorable changes indrylands of the SAT. On-farm watersheds managedthrough community participation could sustainproductivity of drylands and preserve the quality ofthe land resources and environment in the SAT.Holistic systems approach through integratedwatershed management can result in sustainablemanagement of land resources and in achieving foodsecurity in the SAT.

225

ReferencesFarrington, J., and Lobo, C. 1997. Scaling-upparticipatory watershed development in India: Lessonsfrom the Indo-German Watershed Development Program,Natural Resource Perspective No. 17. London, UK:Overseas Development Institute.

Kerr, J., Pangare, G., Pangare, Vasudha L., and George,P.J. 2000.An evaluation of dryland watershed developmentin India. EPTD Discussion Paper 68. Washington, DC,USA: International Food Policy Research Institute.

Wani, S.P., Pathak, P., Tam, H.M., Ramakrishna, A.,Singh, P., and Sreedevi, T.K. 2002a. Integrated watershedmanagement for minimizing land degradation andsustaining productivity in Asia. Pages 207–230 inIntegrated land management in dry areas: proceedings of aJoint UNU-CAS International Workshop, 8–13 September2001, Beijing, China (Zafar Adeel, ed.). Tokyo, Japan:United Nations University.

Wani, S.P., Sreedevi, T.K., Pathak, P., Rego, T.J., RangaRao, G.V., Jangawad, L.S., Pardhasaradhi, G., andShailaja R Iyer. 2003. Minimizing land degradation andsustaining productivity by integrated watershedmanagement: Adarsha watershed, Kothapally, India. Pages79–98 in Integrated watershed management for land andwater conservation and sustainable agricultural productionin Asia: proceedings of the ADB-ICRISAT-IWMI ProjectReview and Planning Meeting, 10–14 December 2001,Hanoi, Vietnam (Wani, S.P., Maglinao, A.R., Ramakrishna,A., and Rego, T.J., eds.). Patancheru 502 324, AndhraPradesh, India: International Crops Research Institute forthe Semi-Arid Tropics.

Wani, S.P., Sreedevi, T.K., Singh, H.P., Pathak, P., andRego, T.J. 2002b. Innovative farmer participatoryintegrated watershed management model: Adarshawatershed, Kothapally, India – A success story! Patancheru502 324, Andhra Pradesh, India: International CropsResearch Institute for the Semi-Arid Tropics. 24 pp.

226

Improving Rural Livelihoods through Convergencein Watersheds of APRLP

T K Sreedevi1

Abstract

The rural livelihoods in the dry agro-regions are dependent on the natural resources. Unpredictableweather, erratic rainfall, poor soil fertility, land degradation, lack of improved varieties, poor knowledgebase on improved farm technology, resource-poor farmers, low farm productivity and income levels, andburgeoning problem of rural poverty characterize these dry regions. The challenge therefore is to improvethe rural livelihoods and alleviate poverty by increasing the farm productivity and income of rural people.The Andhra Pradesh Rural Livelihoods Programme (APRLP) has adopted watershed as an entry point forefficient management of natural resources for improving rural livelihoods. The APRLP has taken Adarshawatershed as a model, which is a farmer participatory watershed with a multi-disciplinary and multi-institutional approach featured by consortium and convergence mode. It is a system involving people whoaim to create a self-supporting system with a vision towards sustainability with increased productivity andprofitability of complex farming systems at the farmer level. Convergence in the watersheds has evolvedwith integrated watershed management model, which apart from integrated genetic and natural resourcemanagement strategy has watersheds as an entry point for converging the entire livelihood relatedactivities based on natural resource use. The convergence takes place at different levels facilitating theprocesses that bring about synergy in all the watershed related activities. Micro-enterprises, equity issues,income generating options for landless and women groups, and micro-finance which in turn bringincreased incomes and improve the rural livelihoods in a sustainable way are all features of this holisticapproach. The model provides scope for issues related to suitable processes for change in micro-practices,macro-policies, convergence of information and management systems, and socioeconomic, institutional,and policy needs to increase adoption of improved options by the rural people. For sustaining the benefitsfrom watershed management through convergence approach, monitoring and evaluation mechanisms formid-course corrections and impact assessment are adopted to identify the mechanisms that build thecapacity of the community for self-regulation and management.

Natural resource base forms the lifeline for rurallivelihoods. Agriculture is the key occupation in ruralareas and farming is mainly dependent on rainfall.Amongst the natural resources, land is the mostvaluable natural resource and fundamental to life,especially to the farming community as it is theprimary basis for production (Wani et al. 2002). Aclose picture of land in the state of Andhra Pradesh,India reveals that 42% of the total land area isdegraded. The problem of land degradation isparticularly serious where local food productioncannot adequately provide survival options to the

rural poor. Low agricultural yields and highpopulation pressure have forced small farmers tocultivate fragile marginal lands and clear forestscausing soil erosion and thus increasing landdegradation.

The rainfed areas with dry agro-ecosystems arecharacterized by unpredictable weather, and limitedand erratic intense rainfall with long intervals of dryspells. Erratic rainfall additionally burdens the ruralpoor by causing soil erosion, as the torrentialdownpours are lost as runoff often carryingsignificant quantities of soil. These rainfed areas have

1. Andhra Pradesh Rural Livelihoods Programme, Rajendranagar, Hyderabad 500 030, Andhra Pradesh, India.

227

higher proportion of poverty as the majorsocioeconomic and environmental constraints forsustaining increased productivity revolve aroundwater (water is the most critical resource and posesgreat risk to water productivity in the dry regions).Factors such as low soil fertility, inappropriate soiland water management practices causing landdegradation, lack of improved varieties, pest anddisease attack, resource-poor farmers, decliningland:man ratio, and poor rural communities, who areunable to meet even minimum standards of health andnutrition, add to the burgeoning problem of ruralpoverty (Wani et al. 2002). The situation calls forreducing poverty through proper management ofthese limited resources in the dry regions byincreasing systems productivity without causingfurther degradation of natural resource base. Hence,under the existing situation the mission of the AndhraPradesh Rural Livelihoods Programme (APRLP),Hyderabad, Andhra Pradesh is to help reduce povertyby protecting the fragile environments, promotinginclusiveness through participatory and convergenceapproach and creating diversified opportunities to therural poor.

APRLP’s ApproachFor realizing the goal of sustaining rural livelihoodsand effective utilization of existing resources,convergence of activities mode was chosen. The termconvergence means a tendency to meet at a point.Adoption of convergence in APRLP is to improverural livelihoods, which implies that all activitiesunder APRLP should bring in betterment in the rurallivelihoods. For maximizing the efforts so as to meetstrategic and practical livelihood concerns of thepoor, small and marginal farmers, landless people,and women, the convergence system forms thestrategy of APRLP. An appropriate unit needs to beworked out for rural livelihood enhancement in thefive target districts, through convergence of activitiesand proper utilization of resources. For efficientmanagement of the existing resources, one has to lookfor a suitable unit of management so that theseresources are managed effectively, collectively, andsimultaneously. The APRLP has chosen watershed asa logical unit for efficient management of naturalresources thereby sustaining rural livelihoods where

focus is on the scope and priorities for developmentof rural people.

Watershed as an Entry PointFor improving the rural livelihoods, watershed formsa logical unit for efficient management of naturalresources thereby sustaining rural livelihoods. Ahydrological watershed is a delineated area fromwhich the runoff drains through a particular point inthe drainage system. Since soil and vegetation canalso be conveniently and efficiently managed in thisunit, the watershed is considered the ideal unit formanaging the vital resources of soil, water, andvegetation. Watershed management is the integrationof technologies within the natural boundaries of adrainage area for optimum development of land,water, and plant resources to meet the basic needs ofpeople and livestock in a sustainable manner.

Integrated Watershed ManagementApproachThe conventional watershed approach attempts tooptimize the use of precipitation through improvedsoil, water, nutrient, and crop management. In anagricultural watershed approach management ofwater and land is most important. People andlivestock being an integral part of the watershed,traditional watershed programs alone, which arestructure driven, cannot offer solutions to improverural livelihoods. Though watershed serves as anentry point, a paradigm shift is needed from thesetraditionally structure driven watershed programs. Aholistic system’s approach is needed to alleviatepoverty through increased agricultural productivityby environment-friendly resource managementpractices.

APRLP believes that watershed, as an entry point,should lead to exploring multiple livelihoodinterventions. The overall objective of the wholeapproach being poverty elimination, the newintegrated watershed management model fits into theframework as a tool to assist in sustainable rurallivelihoods. For the development of rainfedagriculture-based livelihoods, the integratedwatershed model conceptually provides an envelopethrough which many of the steps for sustaining

228

agriculture and agriculture-related activities can beimplemented. The task is to intensify complexagricultural production systems while preventingdamage to natural resources and biodiversity and toimprove the welfare of the farmers.

The new integrated watershed model developed bya consortium led by the International Crops ResearchInstitute for the Semi-Arid Tropics (ICRISAT)provides technological options for management ofrunoff water, in situ conservation of rainwater,appropriate nutrient and soil management practices,waterway system, crop production technology, andappropriate farming systems. The current model ofwatershed management as adopted by ICRISATwatershed team, involves environment-friendlyoptions and use of new science tools along with theconcept of consortium approach and emphasis onempowering farmers through capacity building. Thenew science tools emphasized in the model includegeographic information system (GIS), satelliteimageries, and crop and climate simulation modelsemployed in understanding the constraints forincreasing productivity in the dry agro-ecoregions.The model includes the consortium approach andadopts the concept of convergence in every activity inthe watershed (Wani et al. 2002).

APRLP’s working mode to improve the rurallivelihoods through watershed approach has adoptedthe Adarsha watershed (in Kothapally, Ranga Reddydistrict in Andhra Pradesh), as an example which is amore holistic vision that brings the concept ofsustainability and eco-regionality and focuses onincreased productivity and profitability of complexfarming systems at the smallholder level. Theintegrated watershed approach adopted by theconsortium at Adarsha watershed encompasses thenew modeling tools and technologies for harvestingand managing natural resources on a watershed scalewithout undermining the natural resources. Adarshawatershed team led by ICRISAT has clearlydemonstrated increased productivity from rainfedsystems through integrated watershed approach,which further helped in improving the soil quality andreducing the land degradation. Farmers adoptedimproved management practices such as sowing onbroad-bed and furrows (BBF) landform, Gliricidiaplanting along bunds, integrated nutrient managementtreatment including inoculation with Rhizobium or

Azospirillum sp, environment-friendly integrated pestmanagement, using improved bullock-drawntropicultor for sowing and interculture operations, insitu conservation, and harvesting of excess rainwaterand storage for use as supplemental irrigation and forincreased groundwater recharge.

The Adarsha model is a participatory watershedsystem with a multi-disciplinary and multi-institutional approach, a process involving peoplewho aim to create a self-supporting system essentialfor sustainability. The process begins with themanagement of soil and water, which eventually leadsto the development of other resources. Humanresource development and large-scale communityparticipation is essential since finally it is the peoplewho have to manage their resources. Access toproductive resources, empowering women, buildingon local knowledge and traditions, and involvementof local farmers or villagers in the local communitiesin watershed activities contributed to the successstory at Adarsha watershed.

APRLP has adopted this path with technicalbackstopping from research organizations likeICRISAT, Central Research Institute for DrylandAgriculture (CRIDA), and Acharya NG RangaAgricultural University (ANGRAU) for improvingthe rural livelihoods in the state. The lessons learnedfrom Adarsha watershed envisages that farmers’participation and involvement is critical in integratedwatershed management. Organizing farmers andcommunities for effective management of resourcesis complex and needs careful consideration. There is aneed to harmonize working between existinginstitutions such as panchayats and watershedmanagement and users’ associations.

Consortium model for developing andmanaging watersheds

The concept of consortium is an integral part of thenew integrated watershed management model. Aconsortium approach of institutions is adopted fortechnical backstopping of the watersheds. Expertisefrom different international, national, andgovernment organizations as well as non-governmentorganizations (NGOs) is utilized to help the farmerson the system under operation. Establishment of aneffective and credible consortium mechanism to

229

support the watersheds technically and empoweringthe stakeholders (mainly the farmers and NGOs) willhelp to expand the effectiveness of the variouswatershed initiatives. The activities are:• Establishing a process for technical backstopping

for efficient use of natural resources in thewatersheds.

• Addressing the activities of empowering farmers,NGOs, and other stakeholders through trainingand dissemination of information.

• Establishing an information and communicationtechnology (ICT)-enabled farmer-centered learningsystem for sharing the knowledge (traditional andimproved) and mapping the information flows in therainfed areas for facilitating the exchange oftechnologies and information amongst thestakeholders, including scientists and policy makers.

Convergence approach in integratedwatershed management

Convergence in the watersheds has evolved withintegrated watershed management model, which apart

from integrated genetic and natural resourcemanagement strategy encompasses several otherentities. By adopting a holistic watershed managementprogram, the watershed is used as an entry point forconverging all the livelihood related activities based onnatural resource use. The approach mode inconvergence is to explicitly link watershed developmentwith rural livelihoods and effective poverty eradicationand in the process identify policy interventions atmicro-, meso-, and macro-levels.

In the new adopted mode, emphasis should be onencouraging the convergence of rural developmentprogram at the watershed level. Any project designshould encourage a more holistic understanding of theneeds and priorities of the poor in integration withpolicy and institutional structures. An example ofconvergence for agriculture-related activities in thewatershed and its link with other micro-enterprises isshown in Figure 1.

Convergence can take place at different levels.Convergence at the village level requires facilitationof processes that bring about synergy in all thewatershed-related activities. Scope for issues related

Figure 1. An example of convergence for various activities based on use of natural resources.

230

to suitable processes for change in micro-practices,macro-policies, convergence, and information andmanagement systems also form part of the APRLPmandate. Socioeconomic, institutional, and policyneeds to increase adoption of improved options by therural people are adapted in the convergence approach.

Adarsha watershed is an excellent example ofconvergence in the watershed. This model ofwatershed with technical backstopping is beingevaluated by ICRISAT, Drought Prone Area Programme(DPAP) [now District Water Management Agency(DWMA)], CRIDA, and ANGRAU, with theparticipation and involvement of farmers. Theactivities in integrated watershed managementapproach where convergence mode works include:• Establishing village seed banks through self-help

groups.• Availability of quality seeds to farmers at

reasonable rates.• Processing for value addition (seed material,

poultry feed, animal feed, grading andmarketability, quality compost preparation).

• Poultry rearing for egg and meat production andlocal hatching to provide chicks.

• Vermicomposting with cow dung, fodder waste,and weeds provides quality compost locally.The above activities are income-generating options

for landless and women groups, which in turn bringincreased incomes and improve the rural livelihoods ina sustainable way through a participatory approach.

Equity issues

The benefits in watershed development generally go tothose who own or control the land and water resourcesin the watershed. However, the landless and marginalfarmers’ group benefit from the watershed needs to beensured. The issues of equity for all in the watershedcall for innovative approaches; institution and policyguidelines for equitable use of water resource areneeded. Along with water use, equity issues concerningsustainable use of common property resources in thewatershed also needs to be addressed.

Micro-enterprises

The provision of training and development programsto farming communities in micro-enterprises forms abetter way to disable movement to urban areas for

seeking employment during off-farm season.Building up on micro-enterprises can be decided onthe locally available resources and technicalbackstopping for training the farmers. Some suchtechnologies include:• Vermicomposting: Providing training to women

farmers can empower women.• Preparation of bio-fertilizers.• Livestock-based activities: Improved fodder

production to improve livestock productivity,improved genotypes, and animal health.

• Fisheries and related activities: Fish or prawn culturein the water channels where excess rainwater isavailable. This option can be made available to thelandless people in the rural communities.

• Poultry-based activities: Agro-wastes (e.g., frommaize cultivation) can be diverted for poultry feedalong with other supplemental food. Rearing ofimproved breed like broilers can increase thereturns and improve the livelihood options.

• Horticulture- and forestry-based activities: Teakplanting, pomegranate cultivation, and custard applecultivation along the bunds and marginal lands.

Micro-finance

The rural poor need a market place in which they receivefair prices for their crops and livestock. The APRLPgroup in partnership with self-help groups and localfinancial institutions like Grameen Bank aims to developand link micro-credit and revolving loan programs toresource-poor farmers so that they can purchase basicmaterials and make farm improvements that increaseharvests, improve product quality, and reduce losses.

Monitoring and EvaluationMechanisms for Mid-courseCorrections and Impact AssessmentThe major concern of watershed development effortshas been attaining sustainable impacts on poverty andthe environment after the end of interventions. Forsustaining the benefits of convergence throughwatershed management approach beyond the projectperiod, it is essential to identify the mechanisms thatbuild the capacity of the community for self-regulation and management. Moreover, special exitstrategies, which provide minimal technical and

231

organizational support to the community, are neededto ensure smooth transition.

The need for project monitoring and impactassessment becomes clear as it helps in mid-coursecorrections. The project strategy being evolutionaryand based on the lessons learned duringimplementation, experiences from the current projectwould benefit other parallel projects working in thesame mode. Also cross-site lessons can be learnedand experiences would benefit to improve projectimplementation. Project monitoring and impactassessment will assist decision makers and policymakers to evaluate the project objectively.

The strategic approach for project monitoring andimpact assessment begins with the preparation ofinventory baseline data covering socioeconomic, naturalresource base, and inputs and outputs for eachwatershed. This is followed by continuous monitoringand documentation by the project implementingagencies (PIAs) and periodic monitoring by APRLP.The approach also encompasses impact assessmentbefore project completion and process documentation.The key instruments are participatory rapid appraisals,stratified sampling, detailed surveys, objectiveverifiable indicators (qualitative and quantitative), GIS-based analysis, feedback from the PIAs through regularreports, tour notes by project staff, feedback evaluationfrom experts/visitors, and impact assessment reports.

For each program activity details of specificstrategy, methods, and instruments for monitoringand assessment will be worked out with all thestakeholders. The approach adopted would be aparticipatory approach involving the stakeholders.There would be greater reliance on quantitativeindicators for objective assessment along withqualitative aspects. Further, heavy reliance oninformation flow and information technology usingscience-based methods such as GIS, remote sensing,and ICT would be adopted in the process.

Conclusion and Looking ForwardRural development through sustainable management ofland and water resources gives plausible solution foralleviating rural poverty and improving the livelihoodsof rural poor. In an effective convergence mode forimproving the rural livelihoods in the target districts,with watersheds as the operational units, a holisticintegrated systems approach by drawing attention on thepast experiences, existing opportunities and skills, andsupported partnerships can enable change and improvethe livelihoods of rural poor. The well-being of the ruralpoor depends on fostering their fair and equitable accessto productive resources. The rationale behindconvergence through watersheds has been that thesewatersheds help in “cross learning” and drawing widerange of experiences from different sectors. A significantconclusion is that there should be a balance betweenattending to needs and priorities of rural livelihoods andenhancing positive directions of change by buildingeffective and sustainable partnerships. Based on theexperience and performance of the existing integratedwatersheds in different socioeconomic environments,appropriate exit strategies, which include propersequencing of interventions, building up of financial,technical, and organizational capacity of localcommunities to internalize and sustain interventions,and the requirement for any minimal external technicaland organizational support need to be identified.

Reference

Wani, S.P., Pathak, P., Tam, H.M., Ramakrishna, A.,Singh, P., and Sreedevi, T.K. 2002. Integrated watershedmanagement for minimizing land degradation andsustaining productivity in Asia. Pages 207–230 inIntegrated land management in dry areas: proceedings of aJoint UNU-CAS International Workshop, 8–13 September2001, Beijing, China (Zafar Adeel, ed.). Tokyo, Japan:United Nations University.

232

Integrated Water Resources Assessment at the Meso-scale:An Introduction to the Water and Erosion Studies of PARDYP

R White and J Merz1

Abstract

Water resources in mountain regions are under increasing pressure. With rapidly growing population,demand on food increases. This increasing food demand is presently met by intensifying agriculture andexpansion into marginal lands. Both can result in degradation and depletion of land and water resources.However, the impact of the intensification is not yet fully understood. Often, water resources are unequallydistributed with people upstream having no access to water and people downstream being affected bywater diversion from mid-stream. This leads to conflicts within communities and watersheds. However,using available data to assess resource availability and subsequent improvement of planning andmanagement decisions is difficult.

The People and Resource Dynamics in Mountain Watersheds of the Hindu Kush-Himalayas Project(PARDYP) was launched to provide an integrated and interdisciplinary approach to these issues with along-term perspective. PARDYP provides an impetus for continuing a long-term monitoring program thatis essential for understanding the environmental dynamics and rates of change in selected watersheds ofthe Hindu Kush-Himalayas. It is based on an interdisciplinary approach, and involves disciplines such asforestry, agriculture, sociology, economics, hydrology, meteorology, and soil science. Hydrometeo-rological studies with a dense measurement network in five watersheds across the Hindu Kush-Himalayasfocus on generation of relevant and representative information about water balance and sedimenttransport related to degradation at a watershed level. Proven, and for the region, new technologies areapplied for the monitoring of a variety of parameters. PARDYP’s research includes studies on wateravailability, possibilities to retain water within the watershed for agricultural and domestic use, andrecharge of heavily used groundwater bodies. In terms of high flows, PARDYP is looking at processes ofrunoff generation under different land uses, technical measures, and management options to preventrunoff generation and surface erosion. In general all studies focus on the assessment of the possible impactby changing conditions such as climate change and population pressure.

The paper presents an introduction to the PARDYP project followed by an insight into its water anderosion studies. The methods and techniques employed and the concepts behind the measurement networkare discussed and experiences of the last 6 years highlighted.

1. International Centre for Integrated Mountain Development (ICIMOD), Kathmandu, Nepal.

Land and water resources in the Middle Mountains ofthe Hindu Kush-Himalayas are under constant threatof degradation. Rapid population growth andmismanagement of the natural resources are alreadycausing concern and could lead to immense problemsin the near future. For example:• Top soil erosion is threatening the soil fertility of

many rainfed agricultural fields in the MiddleMountains of Nepal. Nakarmi and Shah (2000)showed that losses of nutrients through erosion are

significant. About 10% of the nitrogen loss andabout 20% of the calcium loss are due to erosion.According to Scherr and Yadav (1996), this regionis one of the hot spots in South Asia in terms ofland degradation.

• Increasing irrigation demand for cash cropproduction with up to four crops per year may leadto decreasing water tables and drying up of riverbeds. A number of farmers have mentioneddecreasing irrigation water availability in the

233

Yarsha Khola watershed (Merz et al. 2000). Thesituation in the Jhikhu Khola watershed isexpected to be even worse due to the closeproximity of markets for vegetables and cashcrops.

• Access to adequate drinking water supply isincreasingly a concern to local residents. Peoplealong the watershed divide do not have access torunning water at a convenient location and have tospend up to 2 hours per trip to collect water (Merzet al. 2002).

• Water quality has become a major issue in manyparts of the region. Sharma et al. (2000) compiledwater quality information from Nepal. The majorarea of focus is the Kathmandu Valley whereimmediate improvements have to take place if thelives of the people living in the valley are not to beendangered.However, the reasons and the impacts of these

processes are not yet fully understood partially due toinappropriate or missing data sets. The People andResource Dynamics in Mountain Watersheds of theHindu Kush-Himalayas Project (PARDYP) waslaunched to provide an integrated and inter-disciplinary approach to these problems with a long-term perspective.

The PARDYP ApproachPARDYP is a regional research for developmentproject active in many fields of natural resource andwatershed management. It succeeded two successfulprojects funded by the International DevelopmentResearch Centre (IDRC), Canada, the 7-year“Mountain Resource Management Project”, whichundertook resource dynamic studies in the JhikhuKhola watershed of Nepal (1989–96), and the“Rehabilitation of Degraded Lands in MountainEcosystems Project” (1992–96) in China, India,Nepal, and Pakistan involving research on therehabilitation and “re-greening” of small patches ofdegraded land. PARDYP combines the regional andthe integrated approach from its two predecessors.Phase I was from 1996 to 1999 while the currentsecond phase continued till the end of 2002.

The Jhikhu Khola watershed work that started in1989 has continued without a break and now 13 yearsof data on soil and water dynamics is providing new

insights into both intensification and degradationprocesses. The lessons learned in Jhikhu Khola havebeen adopted and guide the current phase of PARDYPin watershed management research across theHimalayas. During the first phase, PARDYP aimed atfurther improving the understanding of environmentaland socioeconomic processes associated withdegradation and rehabilitation of mountain ecosystemsand generating wider adoption and adaptation ofproposed solutions by stakeholders in the Hindu Kush-Himalayas (ICIMOD 1996).

PARDYP Phase II aims at contributing tobalanced, sustainable, and equitable development ofmountain communities and families in the HinduKush-Himalayan region (ICIMOD 1999). Theproject activities are concentrated in six majorcomponents:1. Understanding community institutions and their

dynamics2. Social and gender inequity, marginalization3. Water resources for irrigation and domestic use4. On-farm resources5. Common property management6. Livelihood potentials of mountain communities

All PARDYP project components are carried outin each of the five watersheds in the MiddleMountains of the Hindu Kush-Himalayas in China,India, Nepal, and Pakistan (Fig. 1). The activitiesinclude agronomic and horticultural initiatives,socioeconomic and market studies, rehabilitation ofdegraded lands and forestry, soil fertility studies,participatory conservation activities, and water anderosion studies. PARDYP encourages regional dataexchange, and generation and dissemination ofknowledge.

The overall coordination, guidance, andadministration support is by the International Centrefor Integrated Mountain Development (ICIMOD).The national partners are: the Pakistan ForestryInstitute, Peshawar, Pakistan; the Kunming Instituteof Botany in China; the GB Pant Institute forHimalayan Environment and Development, India;and ICIMOD itself in Nepal together with theDepartment of Forest and the Department of SoilConservation and Watershed Management. Thesenational focal research institutions implement,manage, and supervise the activities with theassistance of national and international partners and

234

collaborators. The two main international partnersare the Institute for Resources and Environment at theUniversity of British Columbia (UBC), Canada, andthe Hydrology Group at the Department ofGeography, University of Berne (UoB), Switzerland.Financial support comes from the Swiss Agency forDevelopment and Cooperation (SDC) and IDRC.

The PARDYP network focuses on watersheds ofsimilar size (50–100 km2), at similar elevations (800 to3,000 m), and those that carry out similar activities andsurveys, and use similar instrumentation and the samesoftware so that results are directly comparable. Thebroad cropping systems are the same; rice and wheat inthe irrigated valley areas and maize in the rainfedareas. The teams work closely with farmers and areable to observe what works and what does not.

The approach is that farmers employed as erosionplot or hydro-meteorology readers are also theproject’s point of contact with farmers for researchand demonstration trials and dissemination offindings. This is a cost-effective approach. Runningcosts for each watershed average US$ 50,000 peryear; this includes salaries for the technical staff,travel costs, equipment maintenance, site runningcosts, and all survey costs.

Water and Erosion Studies inPARDYP

One of the main project components is research fordevelopment and demonstration focusing on water,erosion, and related matters. This componentspecifically aims at the generation and exchange ofinformation on water as a resource and its role in landdegradation, and at identifying and testing of optionsto enhance water management decisions (ICIMOD1999). The main activities in this context aremonitoring and collection of baseline information,water quality investigations, soil conservation, andwater management. As shown in Figure 2, theseactivities are investigated in the biophysical and thesocioeconomic environment. The resourcemonitoring is mainly done on water resources.

Resource monitoring and mapping

During the first three years between 1997 and 1999,major emphasis was on data collection. Long-termdata collection was initiated with the setup of ameasurement network in five watersheds in the HinduKush-Himalayas. A total of 89 measurement sites

Figure 1. The five PARDYP watersheds.

235

were operational in June 2001 (Table 1), monitoredby local residents who get annual training and newinstructions if needed.

The setup of the station network in all watershedsfollowed the same principle of the nested approach.This approach helps in investigating the processes onmicro- to meso-scale, i.e., from the plot to thewatershed level, and subsequently determining scaledependency of these processes. Erosion plots and,more recently, surface flow collectors were used forthe plot level investigations at 100 m2 and 2–5 m2

respectively. Sub-catchments and catchments rangingfrom a few hectares to several km2 were monitoredwith hydrological stations equipped with differentinstruments. The main watersheds ranged from about20 to 110 km2. The nested approach principle isgraphically presented in Figure 3.

During an inception workshop the methods of datacollection were as far as possible standardized or atleast kept similar. Hofer (1998) gives background forthe recommended data collection in all watersheds.Different manuals were prepared and distributed tothe respective people in the watersheds: Dischargemeasurement by salt dilution (Merz 1998) and bycurrent meter technique (Dongol et al. 1998), andmeasurement of runoff and soil loss by erosion plots(Nakarmi 1999). In three of the watersheds the sameequipment is used for the collection of rainfallintensity and temperature data. The monitoredparameters include water level, discharge, sedimentconcentration, rainfall, air temperature, water quality,and others.

The collected data is being thoroughly checkedand then stored in a watershed database running on

Figure 2. Water and erosion studies in PARDYP.

Table 1. Measurement sites in the five PARDYP watersheds, June 2001.

Hydrological MeteorologicalWatershed stations stations Erosion plots

Xi Zhuang (China) 4 10 6Bhetagad (India) 6 5 4Jhikhu Khola (Nepal) 5 10 7Yarsha Khola (Nepal) 4 11 4Hilkot (Pakistan) 4 6 3

236

the hydrological software HYMOS 4. The use of thesame software in all watersheds ensures theexchangeability of data between the different countryteams. Annually the final data is published in the formof a yearbook; the latest of PARDYP-Nepal is on aCD-ROM. In addition to the regular monitoringvarious surveys are undertaken. This is mostly doneto obtain the necessary baseline information forfurther hydrological studies. Some examples from thetwo Nepal watersheds, Yarsha Khola and JhikhuKhola watersheds, are given below:• Land use was mapped with the help of aerial

photographs of the scale 1:20,000. Major changesoccurred in the last 15 years in Yarsha Khola; theforest cover increased and the rainfed agriculturalareas decreased. In the Jhikhu Khola, forest coverand rainfed agricultural area increased while shruband grassland areas decreased between 1972 and1990. Shrestha (2000) has discussed the results indetail.

• In both watersheds of Nepal geological baselinemaps give information on the geological bed rockof the watersheds (Nakarmi 2000).

• In Jhikhu Khola a geomorphological map wasproduced giving information on predominantprocesses. The same map for the Yarsha Kholawatershed is in its final stages.

• A land systems map of the Jhikhu Khola watershedgives information on soil types and landforms.Recently a sediment source map including soildepth was completed.

• During December 1998 and September 1999 awater demand and supply survey was done in theYarsha Khola and the Jhikhu Khola respectively,to give an indication of water surplus and waterdemand areas. The same survey highlightsagricultural production and agrochemical inputs asbaseline information for water quality surveys.The results of these surveys are presented in Merzand Nakarmi (2001) and Merz et al. (2002).

• For the allocation of water resources a detailedpublic water sources survey was done in bothwatersheds. A total of 319 springs were mappedand basic physical parameters measured in JhikhuKhola watershed (Shrestha et al. 2000). The same

Figure 3. Principle of the nested approach (schematic).

Erosion plot

Hydrological stationMeteorological station

237

survey in the Yarsha Khola yielded 215 springs(Shrestha et al. 2001).In addition to the topical studies mentioned above

the water and erosion study team of PARDYP isconducting comparative hydrometeorological andsediment transport studies between the PARDYPwatersheds and other watersheds across the region.This includes:• Integration of the results from each watershed

using geographic information system (GIS) andother technologies to construct a picture of thebehavior of water and sediment in terms of time,season, land cover, and extremes.

• Comparison of these results and key findingsbetween the watersheds to formulate and explainthe main similarities and differences across theregion.

• Modeling of scenarios under given and changingconditions to predict changes in the flow regimesand sediment transport.

Water quality

In all watersheds basic chemical and physical waterquality parameters are being tested on a regular basis.The main aim of these investigations is the study ofnutrient dynamics from agricultural land into thestreams. However, a survey of all health posts in theJhikhu Khola showed that water-borne diseasesaccount for about 25% of the patients visiting thesehealth organizations. PARDYP-Nepal therefore hasrecently undertaken a water quality assessment studyin the Jhikhu Khola watershed in collaboration withthe Department of Environmental and BiologicalSciences of the Kathmandu University. This study,funded by the Australian Agency for InternationalDevelopment (AusAid), aimed at assessing thepresent water quality of Jhikhu Khola and itstributaries, examining the water quality of the publicwater supply, raising public awareness of waterquality issues, and recommending simple waterquality assessment methods and treatmenttechnologies.

The main findings of this study were the highmicrobiological contamination of all sources and theelevated phosphate and nitrate levels in the surfaceand groundwater. This is attributed to the intensive

agricultural practices in the area. For water qualityimprovement, both water treatment and long-termsolutions such as spring and catchment protection areproposed and will be tested in detail in the near future.A number of springs have been renovated and will bemicrobiologically tested. To standardize amethodology for field-based microbiological testing,different methods were examined amongst otherpresence/absence tests.

For dissemination of the findings, workshops andtraining programs with different stakeholders wereorganized. Stakeholders included local residents,female health volunteers, science teachers, localauthorities, and scientists.

Soil conservation

Sediment loads and their sources are of major interestfor irrigation schemes and hydropower developmentand design. For the farmers and local users ofcommon land resources, the loss of topsoil is of directconcern. Furthermore, sediment in drinking watersupplies is also commonly high on the list ofhousehold concerns (Merz et al. 2002). The studies ofsediment balances at different spatial levels and therouting of sediment from the plot level to thewatershed were followed by the testing andrecommendation of possible measures against surfaceerosion.

Initial results showed that degraded landsproduced the highest surface erosion rates followedby rainfed agricultural land. To combat this soil lossPARDYP has initiated trials on the rehabilitation ofdegraded lands. A number of nitrogen-fixing tree andshrub species were tested (Shah et al. 2000). Onagricultural land, PARDYP-Nepal is conducting aterrace improvement program under supervision ofthe Department of Soil Conservation and WatershedManagement of His Majesty’s Government of Nepal.The first trial on cover crops failed due to dryconditions. New approaches will be studied in thenear future.

Water management

Too much water during the monsoon season and toolittle water during the dry season are major issues in

238

the Hindu Kush-Himalayas. Therefore, flooding,drought, and water availability are the priority issueson the list of many stakeholders. The data collectedon the routine monitoring network is therefore beinganalyzed.

Study of water dynamics

Water dynamics research included the study ofrainfall distribution in time and space, high flows andtheir generation, low flows and their occurrence, andrainfall-runoff relationships from areas of differentland use and cover. The results are of particularinterest to planners and engineers involved in water-related projects (irrigation, hydropower, drinkingwater).

Study of water availability

The activity on water availability is being carried outat different spatial levels, where water availability,water losses, and hence water balances are estimated.These calculations are leading to an indication ofwhere and when water is available, where and whenwater supply is critical, where water harvesting willlead to considerable benefit to the residents andfarmers, and how storage can be improved. Thisrefers as much to drinking water supplies as toirrigation.

To apply the results from the above studies,PARDYP has a strong interest in improving theefficiency of water use. This starts with the retentionof rainfall and storm water. In the Jhikhu Kholawatershed a trial site for the retention of storm waterwas constructed in collaboration with a Chineseexpert (Nakarmi and Neupane 2000). The knowledgeof rainfall patterns and runoff coefficients helped inthe design of the system. Another system of the sametype was constructed in the Yarsha Khola watershed.

A trial on alternative irrigation methods wasconducted using the water from the constructedunderground tank. Bucket and drip irrigation werecompared. In mid-2000, trials on rainwater harvestingwere initiated with the help of the jar technique asimplemented by the Finnida-supported Rural WaterSupply and Sanitation Support Programme

(RWSSSP) in Nepal. About 20 local masons fromBhutan, the Jhikhu Khola and the Yarsha Kholawatersheds were trained in collaboration with theWater Harvesting Project of ICIMOD and RWSSSP.Subsequently these trainees built 13 demonstrationunits in the Jhikhu Khola watershed and 9 in theYarsha Khola watershed.

Regional activities

The regional activities of the water and erosionstudies have been exchange of information onmethods and support in data collection andmanagement. While the first three years of the projectwere mainly taken up by the setup of the measurementnetwork, emphasis was put on data analysis in PhaseII. The project staff involved in water and erosionstudies were trained in hydrological data analysisduring a training workshop held in Kathmandu withthe main topics “basic analysis”, “low flow”, “wateravailability”, “high flow”, and “sediment analysistools”. On the occasion of the training a manual onwater and erosion analysis methods was produced,which could also be of value to other watershedmanagement projects. Currently all PARDYP countryteams work towards the synthesis of the water anderosion activities at a regional scale, which isenvisaged to be complete by the end of the currentphase in December 2002.

Experiences and RecommendationsThe experience of the last 6 to 10 years has shown thatbest results are achieved with a small, but dedicatedteam of 4 to 6 core staff of different background. Thefield teams should be based in the watershed and havea suitable field office. Their commitment has to befull time and coordination on the country level shouldalso be on a full time basis.

For water and erosion studies, 1 to 2 staff membersare assigned. Most of the staff members are youngand at the beginning of their professional career. Thisfits with the guiding principles of the project ofcapacity building. In water and erosion studies, thetraining and education of young hydrologists in thefour participating countries and from outside the

239

region is important. The aim is to familiarize themwith new and appropriate technology in the field ofwater and erosion research.

For well-defined research activities students anduniversity departments can be involved. In most casesthese temporary arrangements have produced goodresults. For hydrological data collection the setting upof guidelines is very important. A common datamanagement software helps in data exchange andsupport in case of difficulties. However, theexperience with setting up guidelines for data formatsfailed in this case. Analysis guidelines have proven tobe very useful for the synthesis and technical support.

Currently PARDYP is managed by one regionalcoordinator working in close collaboration with thecountry coordinators. For improved regionalcollaboration and synthesis of data in the differenttopic areas of the project it is suggested to haveactivity coordinators at the regional level. This shouldensure frequent exchange of information, technicalsupport, and strategic planning. The exchange ofinformation with other projects of the similar kind is aprerequisite.

ConclusionThe integrated approach of watershed managementresearch including natural and human resources at themeso-scale is proving to be very interesting. Theproject has gone from a pure data collection phase toa data consolidation phase and has now reached thetime of data analysis. With proper synthesis of allresults from the different fields, interesting findingswill come up on the dynamics of natural resourcesand people in the single watersheds and across theregion. Further information on the project can beobtained from http://www.icimod.org under projectsor [email protected].

Acknowledgment

The authors would like to acknowledge the financialsupport of the donors SDC, IDRC, and ICIMOD. Thework of all country teams in PARDYP is highlyacknowledged. All the data collection would not bepossible without the efforts of the local people. A veryspecial thanks therefore goes to them.

ReferencesDongol, B.S., Dhakal, M.P., and Merz, J. 1998.Discharge measurement by means of current metertechnique. Internal Working Paper. Kathmandu, Nepal:People and Resource Dynamics of Mountain Watersheds inthe Hindu Kush-Himalayas Project.

Hofer, T. 1998. Hydrometeorological measurements andanalysis in interdisciplinary watershed projects.Discussion Paper No. MNR 98/3. Kathmandu, Nepal:International Centre for Integrated MountainDevelopment.

ICIMOD (International Centre for Integrated MountainDevelopment). 1996. People and Resource Dynamics inMountain Watersheds of the Hindu Kush-Hiamalayas.Project Document. Kathmandu, Nepal: ICIMOD.

ICIMOD (International Centre for Integrated MountainDevelopment). 1999. PARDYP – The project document.Kathmandu, Nepal: ICIMOD.

Merz, J. 1998. Discharge measurement by means of saltdilution. Internal Working Paper. Kathmandu, Nepal:People and Resource Dynamics of Mountain Watersheds inthe Hindu Kush-Himalayas Project.

Merz, J., and Nakarmi, G. 2001. Water availability inrural watersheds of the Middle Mountains in Nepal. InProceedings of the International Symposium on theHimalayan Environments held in Kathmandu, Nepal,November 2000.

Merz, J., Nakarmi, G., Shrestha, S., Shrestha, B., Shah,P.B., and Weingartner, R. 2002. Water demand and supplyin rural watershed of the Nepal Middle Mountains – Thelocal perception in the Yarsha and the Jhikhu Kholawatersheds. Kathmandu, Nepal: People and ResourceDynamics of Mountain Watersheds in the Hindu Kush-Himalayas Project. (CD-ROM)

Merz, J., Shrestha, B., Dhakal, M.P., and Dongol, B.S.2000. An assessment of the water need and supply situationin a rural watershed of the Middle Mountains in Nepal. InPARDYP – Research for development in the HKH – Thefirst three years (1996 to 1999) (Allen, R., Schreier, H.,Brown, S., and Shah, P.B., eds.). Kathmandu, Nepal:International Centre for Integrated MountainDevelopment.

Nakarmi, G. 1999. Measurement of runoff and soil loss bymeans of erosion plots. Internal Working Paper. Kathmandu,Nepal: People and Resource Dynamics of MountainWatersheds in the Hindu Kush-Himalayas Project.

Nakarmi, G. 2000. Geological mapping and its importancefor construction material, water chemistry and terrainstability in the Jhikhu Khola and Yarsha Khola watershedsof Nepal. In PARDYP – Research for development in the

240

HKH – The first three years (1996 to 1999) (Allen, R.,Schreier, H., Brown, S., and Shah, P.B., eds.). Kathmandu,Nepal: International Centre for Integrated MountainDevelopment.

Nakarmi, G., and Neupane, P. 2000. Construction of awater harvesting tank – Experience from Kubinde in theJhikhu Khola watershed, Nepal. In PARDYP – Researchfor development in the HKH – The first three years (1996to 1999) (Allen, R., Schreier, H., Brown, S., and Shah,P.B., eds.). Kathmandu, Nepal: International Centre forIntegrated Mountain Development.

Nakarmi, G., and Shah, P.B. 2000. Soil nutrient lossesthrough soil erosion in the Middle Hills of Nepal.Presented at the Working Group Meeting on Improving theSoil Fertility Status of Maize Based Systems in the Hills,August 16–18, 2000, Lumle.

Scherr, S.J., and Yadav, S. 1996. Land degradation in thedeveloping world: Implications for food, agriculture, andthe environment to 2020. Food, Agriculture and theEnvironment Discussion Paper 14. Washington, DC, USA:International Food Policy Research Institute.

Shah, P.B., Schreier, H., and Nakarmi, G. 2000.Rehabilitation of degraded lands. In PARDYP – Researchfor development in the HKH – The first three years (1996

to 1999) (Allen, R., Schreier, H., Brown, S., and Shah,P.B., eds.). Kathmandu, Nepal: International Centre forIntegrated Mountain Development.

Sharma, S., Bajracharya, D.R., Dahal, B.M., Merz, J.,Nakarmi, G., and Shakya, S. 2000. Present status ofsurface water quality monitoring, research and training inNepal. Report submitted to HKH Friend Water QualityGroup, Kathmandu, Nepal.

Shrestha, B. 2000. Population dynamics and land use inthe Yarsha Khola watershed. In PARDYP – Research fordevelopment in the HKH – The first three years (1996 to1999) (Allen, R., Schreier, H., Brown, S., Shah, P.B., eds.).Kathmandu, Nepal: International Centre for IntegratedMountain Development.

Shrestha, S.K., Nakarmi, G., Merz, J., and Lamichhane,R. 2001. Survey of the public water sources in the YarshaKhola watershed. Kathmandu, Nepal: People andResource Dynamics of Mountain Watersheds in the HinduKush-Himalayas Project.

Shrestha, S.K., Nakarmi, G., Merz, J., and Osti, R.C.2000. Survey of the public water sources in the JhikhuKhola. Report to Capacity Building for Community WaterQuality Assessment Project. Kathmandu, Nepal: Peopleand Resource Dynamics of Mountain Watersheds in theHindu Kush-Himalayas Project.

241

A Dynamic Soil Erosion Model (MSEC 1): An Integration ofMathematical Model and PCRaster-GIS

A Eiumnoh, S Pongsai, and A Sewana1

Abstract

A physically-based dynamic Soil Erosion Model (MSEC 1) was developed by the Management of SoilErosion Consortium (MSEC) project by integrating revised Griffith University Erosion SedimentationSystem (GUESS) (to predict the movement of sediment and surface flow) and PCRaster-GIS (toincorporate spatial distribution of parameters and catchment behavior). The model was applied to HuayPano catchment, Luang Prabang, Laos using parameter values from literature and experiments. Theresults of runoff and predictions were satisfactory. The model results should be compared with field datafor further evaluation, validation, and improvement.

Generally the amount of soil loss and water runoffunder various soil erosion factors is obtained fromsmall plot experiments on a part of slope in awatershed. The actual soil losses from the plots arecollected in a tank before they are weighed andanalyzed. The surface runoff is measured in the weirgate. The results when expanded to whole slope andwatershed levels are not reliable and do not provideoff-site effect.

Mathematical models for soil erosion estimationhave been developed to provide on-site and off-siteinformation. These physical models do not providespatial distribution. Geographic information system(GIS) is a tool that provides spatial information. TheRaster GIS format, representing various sizes ofground resolution depending on the users, is notpopular as compared to the vector format due mainlyto computer memory. Recently, PC is very powerfuland convenient for data analysis. Therefore, thelinkage with physical soil erosion model is possibleand will provide a dynamic dimension and on-siteand off-site effects.

The MSEC 1 dynamic soil erosion model wasdeveloped by interfacing PCRaster-GIS with therevised GUESS soil erosion equations to calculatewater runoff and sediment transport and deposition inmicro- and macro-watersheds. This model on PC wasapplied to Huay Pano catchment, Luang Prabang,

Laos, using parameters obtained from literature andexperiments. The results of sediment and runoffprediction were satisfactory. However, the values ofparameters for different ecosystems in the Asianregion need further field experiments and evaluation.

The Dynamic ModelSeveral soil erosion models are based on soil,topographic, vegetation cover, and annual rainfallfactors. Soil erosion occurs during a rainstorm.Therefore, the amount and intensity of rainfall in anevent is very important to provide time seriesdetermination of soil erosion. There are two groups ofsoil erosion models: empirical models and physically-based models. The empirical models are simple anduse relationships of soil erosion factors under plotexperiment. The models are extrapolated to largerareas by using regression equation. Some of thesemodels are Universal Soil Loss Equation developed byWischmeier and Smith (1978), ELEMSA (Soil LossEstimation for Southern Africa) by Elwell (1978), andequations by Morgan and Finney (1987) and Morgan(1994). Generally these models cannot extrapolateoutside the experimental plots. Since it is time andlabor consuming, it also requires large budget.

The physically-based models are derived mainlyfrom mass characteristics and movement as defined

1. Asian Institute of Technology (AIT), Patumthani, Thailand.

Pg241_253.pmd 09-Feb-2004, 2:43 PM241

242

by the Law of Conservation of Mass which is also thebasis of the Continuity and Momentum Equation. Itpredicts the movement of sediment and surface flowof water, both on-site and off-site. These modelsimprove the weakness of the empirical models.However, these models require many parameters thatneed research support. Examples of the physically-based models are CREAMS (Chemical, Runoff, andErosion from Agricultural Management System)developed by Foster and Meyer (1972), WEPP (WaterErosion Prediction Project) by Nearing et al. (1989),GUESS (Griffith University Erosion SedimentationSystem) by Ross et al. (1983), EUROSEM (EuropeanSoil Erosion Model) by Morgan (1994) and Morgan etal. (1994). To integrate the soil erosion model into GIS,it is essential to know the details of the selected modeland the nature of soil erosion. The MSEC 1 model wasdesigned by integrating GUESS model and PCRasterand can be applied for micro- and macro-watersheds.

GUESS ModelThe GUESS model for soil erosion developed by Rosset al. (1983) illustrates the processes of soil erosion,transport, and deposition. Therefore, the model canbe used to predict on-site and off-site effects. Theconcept of the model is to use the equilibrium ofsediment in the area. It calculates the movement ofsediment under actual condition and then thedeposition area by runoff flow for each rainfall event.Generally, two types of soil are considered: originalsoil and newly deposited soil. The original soil hasdifferent cohesion and aggregation, while the newlydeposited soil has no cohesion and aggregation.Therefore, in each rainfall event the soil cohesion inthe area will not be the same. The degree of leaching,erosion, and transport of sediments varies accordingto Equation 1:

ei = Rate of detachment of soil particles ofsediment class i in the original soil byraindrop impact;

edi = Rate at which recently detached soil ofsediment class i is re-detached by raindropimpact;

ri = Rate of detachment of soil particles ofsediment class i by flow;

rdi = Rate at which recently detached soil ofsediment class i is re-detached by flow; and

di = Rate of deposition.

The rate of detachment of soil particles of sedimentclass i by raindrop impact is calculated by Equation 2:

where:a = Detachability of the soil;

Ce = Fraction of the soil surface exposed toraindrops;

I = Rainfall intensity; and

N = Number of particle size classes.

The rate of re-detachment of soil particles ofsediment class i is calculated as shown in Equation 3:

where:ad = Re-detachability of the soil.

The detachment rate of soil particles by flow (ri)derived from the activity of stream power (Ω) abovecritical (Ωc) and shear stress is shown in Equation 4:

where:

Qsi = Sediment load of sediment class i;

Ci = Concentration of sediment class i in theflow;

(4)

(2)

(3)

(5)

where:τ = Shear stress; and

V = Flow velocity.The shear stress is calculated by Equation 5:

where:γf = Specific weight of fluid;

R = Hydraulic radius; and

S = Energy slope (surface slope for steady stateflow).

(1)iddi+rir+di+ eie =t∂

hiC∂ +

χ∂siQ∂ )(

I/NeaC=ie

I/NeCda=die

Ω τV=

RSfγ=τ

Pg241_253.pmd 09-Feb-2004, 2:43 PM242

243

(6)

The detachment rate of soil particles by flow iscalculated as in Equation 6:

ri = (1H)F(Ω – Ωc)/Ιℑwhere:

(1H) = Fraction of the original soil surfaceexposed to runoff;

F = Fraction of the excess stream power ofoverland flow used in erosion; and

ℑ = Amount of unit stream power necessary todetach a unit mass of soil.

The re-detachment rate of soil particles is calculatedby Equation 7:

ν = Kinematic viscosity of the liquid;

γs = Specific weight of sediment;

γf = Specific weight of liquid;

µ = Absolute viscosity; and

ρ f = Density of fluid.

The Equation uses a theoretically derived expressionfor the maximum possible sediment concentrationthat can be sustained due to flow-driven erosionprocesses. It is based on the upper limit of sedimentconcentration, Cmax, which is reached when the rateof removal of sediment is equal to its rate ofdeposition. The transportation capacity of the soilerosion or water flow is calculated by Equation 11:

where:αi = A dimensionless parameter with a value

dependent on the depth of flow;

H = Fraction of the surface soil covered byrecently deposited material;

σ = Submerged sediment density;

ρ = Density of water;

h = Depth of flow;

Mi = Mass fraction of sediment class i; and

M = Total mass of material being re-detached.The rate of deposition of sediment of class i iscalculated as in Equations 8 to10.

where:υsi = Fall velocity of particles of sediment class i;

g = Gravitational constant;

Di = Particle diameter class i;

(7)MiM

Ω – Ωch

σσ ρ

rdi =αi HF

g

(11)

Proving the equation:From Equation 1,

It is observed that in the surface runoff, when thesediment concentration is at the highest the firstgroup in the equation is equal to zero. When Ωc hasvalue of 0, the maximum mass from erosion is equal tototal mass. Therefore, the H value is close to 1. In thisequation, it is given as 1 so that the calculation will behighest.

SVN

FCsi

max

−

=

ρσσ

υΣρ/

iddirirdieiethiCsiQ

−+++=∂

∂+

∂

∂ )(

χ

(10)fρ

µ=ν

(9)fγ

fγsγ.18ν

igD=siυ

2

(8)isiii Cνα=d

ℑΩ−Ω−++ IFHNICaNIaC cede /)()1(//

isiiici C

MM

hgHF υα

ρσσα =

Ω−Ω

−

+

iCsiiMiM

gh

HiI

H

FcdadaN

IeC

υαρσ

σα=

−

+ℑ

−

Ω−Ω++

)1(

)()(

Pg241_253.pmd 09-Feb-2004, 2:43 PM243

244

When Ω = τν and τ = γfYS in the equation arereplaced by assigning γf = ρg, the product isΩ = ργYSV.

From the balance of the equation, then

Therefore,

as shown in Equation 11.

Equation 11 is applied either to the overland flow orthe flow to the rills and it is possible to rewrite theequation as shown in Equation 12:

where:

φ = Mean settling velocity of sediment.

φ =

From Equation 12, the velocity is calculated using theoriginal Manning equation as shown in Equations 13and 14.

where:n = Mannings roughness coefficient;

S = Slope; and

R = Hydraulic radius (cross section area/lengthof wet zone).

(14)

When the equation is replaced by runoff discharge, itis written as Equations 15 and 16:

Q = VA (15)

where:

Q = Runoff discharge per unit area.

Therefore, the equation for velocity is shown inEquations 17 to 20:

For simplicity, the flow cross section is assumed asrectangular; hence the new equation is:

(16)

SVFCisi ρσρσυ−

=

SVN

FCsi

−

=

ρσσ

υΣρ/max

−

=1

max

ρσφ

σSVFC (12)

Nsiυ∑

(13)

or

213

21

SPA

nV

=

(17)

(18)21

32

32

32

1 SPV

Qn

V

=

(19)21

32

32

35 1 S

P

Qn

V

=

(20)5

2

525

3

P

QnS

V

=

52

525

3

L

QnS

V

= (21)

52

425

3

QLnS

V−

= (22)

21

321 SR

n=ν

R = AP

QA =

V

Pg241_253.pmd 09-Feb-2004, 2:43 PM244

245

When V is replaced in Equation 12, the newEquations 23 to 25 are derived.

(23)

(24)

where:

(25)

From the equation, it is seen that the transportcapacity of the flow is dependent upon k and runoffvalues. In the equation, the transport capacity iscalculated for each rainfall event. To have runoff in asteady state, it is necessary to obtain the effectiverunoff discharge (Qeff) and ensure that k value isstable. The new equation can be rewritten as inEquations 26 to 30.

(26)

(27)

(28)

(29)

where:

(30)

maxC = The flow-weighted or the average value ofCmax during an erosion event.

From Equation 28, it is then possible to calculate thesoil erosion in each rainfall event. To obtain precisecalculation, the soil cohesiveness and vegetation orsurface cover could be added as shown in Equations31 and 32.

(31)

where:β = Erodibility parameter (or soil erodibility); andC = Actual flow-weight sediment concentration.

(32)

where:c = Real soil erosion:

cb = Bare soil erosion;

Cs = Fraction of surface contact cover; and

ks = Non-dimensional (obtained by experiment).

52

525

3

1QL

nSSF

maxC−

−

=

ρσφ

σ

0.4kQmaxC =

53

52

1

−=

−

nSSLFk

ρσφ

σ

[ ]

[ ]

∗=

∗

52

1

52

3

53

51

523

/

1/

/

m

smkg

smsm

mmkg

rate runoffflux sedtot

ΣΣ

QQ0.4kQ

Cmax×

=

Q

0.4QkCmax Σ

Σ=

52

25

4.1

ΣΣ

=

QQkCmax

0.4effkQCmax =

[ ] [ ]3

52

1

1

52

3 //5

2

52 mkg

s

m

m

smkg =

∗

∗

25

1.4

QQ

effQ

=

ΣΣ

βmaxCC =

( )sCskbcc −= exp/

52

25

4.1

ΣΣ

=

QQkCmax

Pg241_253.pmd 09-Feb-2004, 2:43 PM245

246

GIS in PCRasterCurrently, GIS has been employed intensively tomanage and analyze data including watershedmanagement and soil erosion and modeling. It hasbeen applied in empirical models such as the USLE.However, the physical soil erosion models wouldrequire specific software that provide similarcondition as natural processes where time sequencesand changes of parameters are involved.

PCRaster is a GIS raster software that has bothfunctional and operational packages for real dynamicsoil erosion modeling. The software, similar to otherGIS packages, is capable to store, manipulate,analyze, and retrieve geographical data. Thissoftware is also capable for cartographic anddynamic modeling to simulate soil erosion, on-siteand off-site effects through surface water flow andsediment transportation. There is no digitization orscanning for data input available in the software.Rather, all data are transferred to and from other GISpackage. The information on PCRaster is in thePCRaster Manual Version 2 by Faculty ofGeographical Sciences (2000).

Main components in PCRaster

There are two main components in the PCRaster:cartographic model; and dynamic modeling module.

Cartographic model

The cartographic model is employed for spatial data.The value of each cell (pixel) changes according tothe point operations, neighborhood operationsand area operations. The point operations includefunctions that operate only on the values of the maplayers relating to each cell. The simplest of the pointoperations are the arithmetic, trigonometric,exponential, and logarithmic functions of mathematicaloperation. The neighborhood operations relate the cellto its neighbors by mathematical functions. The main

operations are local drain direction (LDD)operations and the friction path operations, fortransportation of material over LDD. The areaoperations are used for analysis of map operations.The map operation uses non-spatial (attribute data)that links to the map.

Dynamic modeling module

The dynamic modeling module is the advancedmodule of GIS function in PCRaster. This model usessimple language and is designed for movement ofmaterials under the cartographic model. Therefore,its operation is similar to most mathematicalcalculations. Hence, the users may not necessarily beprogrammers, but can do modifications for specificapplication. The dynamic model is used for modelingof soil erosion processes over time. In this model,new attributes are computed as functions of attributechanges over time. It is built with language providedby PCRaster. Within this language, the model can beprogrammed with the PCRaster operations ofcartographic modeling. A script, i.e., a program writtenin the dynamic modeling language, consists ofseparate sections. Each section contains a certainfunctional part of the script. The division in sectionsis an essential concept of the dynamic modelinglanguage. It tells the computer how to execute aprogram and it helps the user to structure thecomponent of a model.

The basic sections needed for building a dynamicmodel are the binding section and the area mapsection used at the dynamic section for the iteration atthe first time step. The dynamic section defines theoperations for each time step i. that result in a map ofvalues for that time step. Each time step consists ofone or more PCRaster operations that are performedsequentially. The results of time step i. are the inputvalues for time step i. + 1 and so on. This sectionreads dynamic data from the databases or storedmodel results in the database for each time step.

Data types

The data types used in PCRaster are Boolean,Directional, LDD, Nominal, Ordinal, and Scalar. Thescalar fields are used to describe intensities andpotential of the physical field such as precipitation.

Finally, the total mass of soil lost during an erosionevent (M) is shown in Equation 33:

(33)( )sCskQeffQkM −= expΣ4.0 ββ

Pg241_253.pmd 09-Feb-2004, 2:43 PM246

247

Directional fields apply to attributes that have acircular and continuous scale such as aspect of theterrain. The vector fields have both magnitude anddirection and can be used to represent (horizontal)fluxes and forces such as infiltration or rainfall.

The LDD data type has been introduced to providefor the definition of the direction of potential flow. Itis a data type that supplies the raster database with thetopological linkages needed by the operators thatdescribe lateral fluxes of the fluids or materials. Ingeneral, this LDD map is derived from a DigitalElevation Model (DEM), and represents the directionof surface flow through this elevation map, but anyscalar field showing potential difference can beprocessed to determine a LDD map.

From the models and structure of the PCRaster-GIS, it is very useful for physical soil erosionmodeling. The model is well used to analyze all soilerosion data through commands or script in thePCRaster. The model runs from the beginning stageof soil erosion processes to the end continuously.

Integration of PCRaster-GIS andModified GUESS Model

Principle

The main principle of integration of PCRaster-GIS andGUESS model is to simulate soil erosion in a realsituation or as virtual simulation in each rainfall event.The process is to integrate GIS function into operatorsin the PCRaster. The model will calculate soil erosionaccording to the GUESS model by time steps or rainfallevents. The results of soil erosion calculated at everytime step will show the direction of flow according tothe topography or DEM or LDD. The calculation of soilerosion starts from the first time step or the first rainfallevent and sediments transported and deposited beforethe second time step or second rainfall event begins.This calculation process will continue until the last timestep or the last rainfall event.

The amount of sediments transported anddeposited may be obtained in the pit or outlet of themicro-catchment or watershed. This effect can beconsidered as off-site effect. The model alsoprovides changes of sediments during each rainfallevent according to the LDD. Therefore, it is possible

to plan measures to reduce soil loss from the field.The structure and function of database in PCRasterwill support the GUESS mathematical model. Theraster format helps in the movement or transportationof sediments through LDD map. Therefore, it ispossible to simulate soil erosion according to thenature of the model. Several steps should be followedto meet the simulation.

Evaluation processes

Synchronize GUESS and PCRaster function

This step includes parameter selection and the rangeof each parameter according to the nature of soilerosion and the preparation of data structureaccording to the PCRaster format. The data typessuch as Boolean, Nominal, Ordinal, Scalar, orDirectional must be identified. They must also beappropriated for each parameter. The operators areidentified according to the mathematical function ofthe GUESS model.

Parameter setting according to GUESS equation

According to Equation 31, it is possible to set groupsof parameters. The soil parameters that are mostlyrelated to soil characteristics, both chemical andphysical properties, are sediment density, soilerodibility, and mean settling velocity of sediment.The sediment density is the ratio of dry soil mass andthe volume of soil particle. It is calculated from thefollowing equation:

Ds = ms/Vswhere:

Ds = Solid phase density/sediment density (kg m-3);

ms = Dry soil mass (kg m-3); and

Vs = Volume of solid phase (m-3).The dry soil mass can be obtained from the bulkdensity and volume of soil mass as below:

Ms = Db × Vbwhere:

Db = Bulk density; andVb = Total soil volume.

Pg241_253.pmd 09-Feb-2004, 2:43 PM247

248

The equation can be rewritten as:

Db = ms / Vb

Ds = Db × Vb/Vs

Assuming that a cubic meter of soil (solids and pores)weighs 1.33 t, the bulk density (Db) can be calculatedas:

Db = ms / Vb= 1.33/1= 1330 kg m-3

If Vs = 0.5 m3, then Ds = 1330 × 1/0.5 = 2600 kg m-3

The soil erodibility (β) is the resistance to flow ofwater, which is related to soil cohesion oraggregation. The cohesion has negativerelationshipwith distance between particles. But ithas positive relationship with specific surface area ofsoil particles. Clayey soil has high cohesion. Thecohesion reduces as the particle size increases due towater molecules between the particles. As thedistance between soil particles increases, the soilmoisture also increases, but the soil cohesiondecreases. The critical shear velocity increases withdecreasing particle size; the finer particles are harderto erode because of cohesiveness of the clay mineral.

Since the cohesion varies according to soilmoisture at different rainfall events, there is no fieldexperiment to find out the actual values of soilerodibility. But, from the mathematical calculation inthe GUESS model and statistics, the range of soilerodibility has values between 0 and 1. Thiscalculation, based on static, not dynamic, considersthe amount of clay, varying from 0 to 100% clay.Therefore, it is the potential cohesive element.

The velocity of sediment I can be calculated fromthe Stokes law:

V = (Ds Di)gd2/18n (cm s-1)where:

V = Speed of deposition of sediment class i;

Ds = Particle density (g ml-1);

Di = Density of fluid (g ml-1), for water (1000 kg m-3);

g = Gravity (980 cm s-1);

d = Diameter of particle size class i. (cm); and

n = Viscosity coefficient of fluid (poise).

When k = (Ds Di)g/18n, Strokes law will bewritten as:

V = kd2

Velocity is a function of distance divided by time.

V = s/twhere:s = Distance of sediment transportation; and

t = Time for deposition.As V = kd2

kd2 = s/t ort = s/kd2

The mean settling velocity of sediment (φ) iscalculated as:

φ = Rate of deposition of sediment class 1. +...class i./total sediment classes or

υi = settling velocity of any arbitrary size class i

I = Total number of equal mass size classes..

The land use/land cover in the GUESS model isapplied for Mannings coefficient and contact cover.Mannings (n) value is friction between sediment inwater and the soil surface developed during therunoff. It has values between 0 and 1. Land coverprotects rainfall detachment and reduces surfacerunoff. In the GUESS model, the contact cover (Cs)has value of 01. This contact cover is exponentiallyrelated to ks (an empirical, dimensionless factor) ofsoil erosion on bare soil as shown in the equationbelow:

where:

C = Amount of soil loss with land cover; and

cb = Amount of soil loss under bare soil.

The topographic factor including slope (%) andaspect of slope or LDD is obtained from DEM. Theamount of rainfall is obtained from each rainfallevent that collected from actual measurement usingautomatic rain recorder. The surface runoff dependson surface soil characteristics, contact cover, andamount of rainfall of that event. The runoff values

∑ == Ii I

i1υ

φ

( )sCskbcC −= exp/

Pg241_253.pmd 09-Feb-2004, 2:43 PM248

249

may be obtained from Rational Method, C-N Method,and Water Balance and Empirical Method. Thepopular method is the C-N method by using theequation given below:

where:

P = Precipitation (mm h-1); and

S = Maximum storage = 0.2* initial infiltration(mm)

The maximum infiltration is estimated by theequation below:

However, the GUESS model requires runoffdischarge (Q) (m3 s-1) and effective runoff rate (Qeff). Qcan be estimated by Equation 15. Since there is onlyevent data and no data on a smaller timescale, forinstance one minute, this equals the total amount ofrunoff per event (∑Q). According to the model, ∑Q is15% of the amount of rainfall per event. To obtainactual runoff, it is necessary to collect moreinformation of each rainfall event.

A fraction F of the stream power is used in theprocess of erosion by flowing water (commonly in therange of 0.10.2, sometimes higher). F can be calledthe effective excess stream power, as it is assumed tobe a function of the stream power minus the thresholdstream power (the limit of water velocity by which theparticles are not moving yet). In the steady statesituation, deposition is as big as erosion. Thissituation occurs theoretically even if conditions arenot steady.

Data Preparation for PCRasterIt has been mentioned that setting parameters mustfollow the requirement of the model. It is alsonecessary to prepare the data according to thedemand of the model. The file name must follow thescript as has been developed as MSEC 1.

The spatial data must be already prepared usingArc/Info or other programs, before the topology isconstructed. Then the database input according toparameters such as soil, land use, and contour line iscommenced. The data will be converted into ArcViewas Shape File and Converse File in the raster (grid)format. This is done by command shape → grid andexport to ASC format. All files must have the samegrid size. Grid files are obtained and ready to beimported into PCRaster. When the shape files areconverted into grid files, it is necessary that thecolumn must be the same name of the map; e.g.,cohesive or sedden.

The spatial data in the ArcView (grid file) can beimported into PCRaster by using command Arc2 ase.However, the table file may be prepared from anotherprogram, but need to have the format that can be readby PCRaster.

Written script and data analysis

The next step after parameter preparation is thecommand that will make PCRaster run. The dataprocessing under PCRaster follows the GUESSmathematics model that is employed by MSEC 1.There are five steps in the command or script as showbelow:

1. Binding section: To identify parameters that will beused in the analysis of the next step; for example:

Ldd = Ldd.map

It means that when referred to Ldd in the model, thedata can be taken from Ldd.map or Ldd map.

Manning = manning.Tbl

It means that when referred to manning in the model,the data can be taken from manning.Tbl or manningtable.

2. Area map section: To construct base map or MapExtent as reference for every map.

3. Timer section: To set the time or time step that willbe used in the dynamic model; for example, rainfallevent.

4. Initial section: To set the starting point of databefore any analysis is conducted; for example, set soilerosion is equal to zero before start.

S =25400

CN

254

Q =

(P 0.2 S)2

P + 0.8 S

0

P > 0.2 S

P ≤ 0.2 S

Pg241_253.pmd 09-Feb-2004, 2:43 PM249

250

Figure 1. Monthly soil erosion (t ha-1) during 2001 in Huay Pano catchment, Luang Prabang, Laos.

0–250.5–22–55–1010–20

SeptemberAugust

JulyJune

Pg241_253.pmd 09-Feb-2004, 2:43 PM250

251

5. Dynamic section: To calculate according to themodel for each time step by using operators andfunction of PCRaster.

To run script, it is recommended to use commandPCRaster-F, followed by model name that has beendeveloped in MSEC 1.

Interpretation of results

There are four outputs derived from the data analysisas given below: Runoff file: Shows runoff in cubic meter per plot or

per pixel of each raifall event. Erosion file: Shows soil erosion as ton per plot or

per pixel of each rainfall event. Eroflux file: Shows soil erosion at off-site of each

plot or each pixel according to the LDD of eachrainfall event.

Erostore file: Shows soil erosion or sediment that istransported and deposited in each plot or eachpixel of each rainfall event.

Test of model

The MSEC 1 model was tested using Huay Panocatchment in Luang Prabang, Laos as shown in Figure1. The amount of sediments varied from 5 to 30 t ha-1

depending on combination of soil erosion factors,particularly the amount of rainfall during the event. Ithas been observed that the rainfall data were ratherhigh. This model should be compared with field datafor validation and improvement in the future.

ReferencesElwell, H.A. 1978. Modeling soil losses in South Africa.Journal of Agricultural Engineering Research 23:117127.

Faculty of Geographical Sciences. 2000. PCRasterManual Version 2. PCRaster Environmental Software. TheNetherlands: Utrecht University.

Foster, G.R., and Meyer, L.D. 1972. A closed-form soilerosion equation for upland areas. In Sedimentation (Shen,H.W., ed.). Fort Collins, Colorado, USA: Department ofCivil Engineering, Colorado State University.

Morgan, R.P.C. 1994. The European soil erosion model:An update on its structure and research base. Pages 286299 in Conserving soil resources: European perspectives(Rickson, R.J., ed.). Wallingford, UK: CAB International.

Morgan, R.P.C., and Finney, H.J. 1987. Drag coefficientsof single crop rows and their implications for wind erosioncontrol. Pages 449458 in International geomorphology1986, Part II (Gardiner, V., ed.). Chichester, UK: JohnWiley & Sons.

Morgan, R.P.C., Quinton, J.N., and Rickson, R.J. 1994.Modeling technology for soil erosion assessment and soilconservation design: The EUROSEM approach. Outlookin Agriculture 23:59.

Nearing, M.A., Foster, L.J., and Finkner, S.C. 1989. Aprocess-based soil erosion model for USDA-Water ErosionPrediction Project Technology. Transactions of theAmerican Society of Agricultural Engineering 32:15871593.

Ross, C.V., Williams, J.R., Sander, G.C., and Barry, D.A.1983. A mathematical model of soil erosion and depositionprocess. I. Theory for a plane element. Soil Science Societyof America Journal 47:991995.

Wischmeier, W.H., and Smith, D.D. 1978. Predictingrainfall erosion losses. USDA Agricultural Research ServiceHandbook 537. USA: USDA.

Pg241_253.pmd 09-Feb-2004, 2:43 PM251

252

Gro

up P

hoto

grap

h

253

Participants

China

Yin Dixin Phone : 86 (851) 3762031Center for Integrated Rural Development Research Fax : 86 (851) 3762031Guizhou Academy of Agricultural Sciences (GAAS) E-mail : [email protected] Zhen, Guiyang CityGuizhou 550006

India

Acharya, C L Phone : 91 (0755) 730946Director Fax : 91 (0755) 733310Indian Institute of Soil Science (IISS) E-mail : [email protected] Bagh, Berasia RoadBhopal 462 038Madhya Pradesh

Patil, B RVice President Phone : 91 (0265) 654897BAIF Development Research Foundation Fax : 91 (0265) 6518023rd Floor, Indra Complex E-mail : [email protected] 390 004Gujarat

Sharma, R A Phone : 91 (0731) 701254, 492607Senior Scientist Fax : 91 (0731) 496989AICRP on Dryland Agriculture E-mail : [email protected] of AgricultureIndore 452 001Madhya Pradesh

Singh, H P Phone : 91 (040) 24530177Director Fax : 91 (040) 24531802, 24535336Central Research Institute for Dryland Agriculture E-mail : [email protected] (CRIDA)SantoshnagarHyderabad 500 059Andhra Pradesh

254

Sreedevi, T K Phone : 91 (040) 24001953, 24001954Additional Coordinator, Project Support Unit Fax : 91 (040) 24018656Andhra Pradesh Rural Livelihoods Programme E-mail : [email protected]/o A P Academy of Rural DevelopmentRajendrangarHyderabad 500 030Andhra Pradesh

Indonesia

Agus F Phone : 62 (251) 638987Center for Soil and Agroclimate Research (CSAR) Fax : 62 (251) 638987Jalan, Ir. H. Juanda 98 E-mail : [email protected] 16123

Sukristiyonubowo Phone : 62 (251) 383745/323012Center for Soil and Agroclimate Research (CSAR) Fax : 62 (251) 311256Jalan Ir. H. Juanda 98 E-mail : [email protected] 16123

Laos

Chanthavongsa, A Phone : 856 (21) 732047Soil Survey and Land Classification Center (SSLCC) Fax : 856 (21) 32047/414373PO Box 811 E-mail : [email protected]

Phommassack, T Phone : 856 (21) 732047National Agriculture and Forestry Research Institute Mobile : 020-520421

(NAFRI) Fax : 856 (21) 732047/414373Ministry of Agriculture and Forestry E-mail : [email protected] Box 811Vientiane

Valentin, C Phone : 856 (21) 502680IRD – Ambassade de France Fax : 856 (21) 412993BP 06, Vientiane E-mail : [email protected]

Nepal

Merz, J Phone : 977 (1) 525313International Centre for Integrated Mountain Fax : 977 (1) 524509

Development (ICIMOD) E-mail : [email protected] Box 3226Kathmandu

255

Maskey, R B Phone : 977 (1) 521147Soil Science Division Fax : 977 (1) 521197Nepal Agricultural Research Council (NARC) E-mail : [email protected] Box 5459 [email protected]

Shrestha, A B Phone : 977 (1) 262374/ 262974Department of Hydrology and Meteorology Fax : 977 (1) 262348PO Box 406, Babar Mahal E-mail : [email protected]

White, R Phone : 977 (1) 525313International Centre for Integrated Mountain Fax : 977 (1) 524509

Development (ICIMOD) E-mail : [email protected] Box 3226Kathmandu

Philippines

Ilao, R O Phone : 63 (49) 5360014 to 20Philippine Council for Agriculture, Forestry and Fax : 63 (49) 5360016/0132

Natural Resources Research and Development E-mail : [email protected](PCARRD)

Los BanosLaguna 4031

Villano, M GUPLB-CEATCollege, LagunaPhilippines

Thailand

Bandaragoda, D J Phone : 66 (2) 5614433 ext. 116Director Fax : 66 (2) 5611230International Water Management Institute (IWMI) E-mail : [email protected] Asia Regional OfficePO Box 1025, Kasetsart UniversityPhaholyotin Road, JatujakBangkok 10903

Boonsaner, A Phone : 66 (2) 5614292/3Royal Forest Department (RFD) Fax : 66 (2) 5798775Phaholyotin Road, JatujakBangkok 10900

256

Chanvid LusanandanaAgricultural Official C-7 Fax : 66 (43) 241286Office of Agricultural Research and Development E-mail : [email protected] III180 Mitra Parb RoadMoeng DistrictKhon Kaen Province

Eiumnoh, A Phone : 66 (2) 524-5588Asian Institute of Technology Fax : 66 (2) 524-6431PO Box 4, Klong Luang E-mail : [email protected] 12120

Janeau, J L Phone : 66 (2) 5614433 ext. 116International Water Management Institute (IWMI) Fax : 66 (2) 5611230Southeast Asia Regional Office E-mail : [email protected] Box 1025, Kasetsart UniversityPhaholyotin Road, JatujakBangkok 10903

Maglinao, A R Phone : 66 (2) 5614433 ext. 116International Water Management Institute (IWMI) Fax : 66 (2) 5611230Southeast Asia Regional Office E-mail : [email protected] Box 1025, Kasetsart University,Phaholyotin Road, JatujakBangkok 10903

Pongsai, S Phone : 66 (2) 524-5588Asian Institute of Technology Fax : 66 (2) 524-6431PO Box 4, Klong Luang E-mail : [email protected] 12120

Sewana, A Phone : 66 (2) 524-5588Asian Institute of Technology Fax : 66 (2) 524-6431PO Box 4, Klong Luang E-mail : [email protected] 12120

Sutham Paladsongkarm Phone : 66 (2) 5791792/ 5791562Department of Land Development (DLD) E-mail : [email protected] Road, JatujakBangkok

257

Wannipa Soda Phone : 66 (2) 5614433 ext. 116International Water Management Institute (IWMI) Fax : 66 (2) 5611230South East Asia Regional Office E-mail : [email protected] Box 1025, Kasetsart UniversityPhaholyotin Road, JatujaBangkok 10903

Vietnam

Chinh, N TDeputy Director General Phone : 84 (4) 8615480/87Legume Research and Development Center Fax : 84 (4) 8613937Vietnam Agricultural Science Institute (VASI) E-mail : [email protected] TriHanoi

Dan, N TChairman, Scientific Council Phone : 84 (4) 7332087Ministry of Agriculture and Rural Development Fax : 84 (4) 8433637Hanoi

Doanh, L Q Phone : 84 (4) 8615480/87Deputy Director General Fax : 84 (4) 8613937Vietnam Agricultural Science Institute (VASI) E-mail : [email protected] TriHanoi

Hong, N X Phone : 84 (4) 8385578Deputy Director Fax : 84 (4) 8363563National Institute of Plant Protection (NIPP)Hanoi

Long, T D Phone : 84 (4) 8615480/87Deputy Director General Fax : 84 (4) 8613937Vietnam Agricultural Science Institute (VASI) E-mail : [email protected] TriHanoi

Nghia, N H Phone : 84 (4) 8615480/87Director General Fax : 84 (4) 8613937Vietnam Agricultural Science Institute (VASI) E-mail : [email protected] TriHanoi

258

Phai, D D Phone : 84 (4) 8388958/8362379National Institute for Soils and Fertilizers (NISF) Fax : 84 (4) 8389924Tu Liem E-mail : [email protected]

Tam, H M Phone : 84 (4) 8615480/87Deputy Director Fax : 84 (4) 8613937Legume Research and Development Center E-mail : [email protected] Agricultural Science Institute (VASI)Than TriHanoi

Thai Phien Phone : 84 (4) 8362379/ 8385035National Institute for Soils and Fertilizers (NISF) Fax : 84 (4) 8389924Tu Liem E-mail : [email protected]

Thang, N V Phone : 84 (4) 8615480/87Head, Groundnut Research Department Fax : 84 (4) 8613937Legume Research and Development Center E-mail : [email protected] Agricultural Science Institute (VASI)Than TriHanoi

Toan, P V Phone : 84 (4) 8615480/87Head, Soil Microbiology Department Fax : 84 (4) 8613937Vietnam Agricultural Science Institute (VASI) E-mail : [email protected] TriHanoi

Toan, T D Phone : 84 (4) 8388958 / 8362379National Institute for Soils and Fertilizers (NISF) Fax : 84 (4) 8389924Tu Liem E-mail : [email protected]

Tuan, H D Phone : 84 (4) 8615480/87Head, Information Cooperation Fax : 84 (4) 8613937Vietnam Agricultural Science Institute (VASI) E-mail : [email protected] TriHanoi

Viet, N V Phone : 84 (4) 8615480/87Head, Plant Pathology and Genetics Department Fax : 84 (4) 8613937Vietnam Agricultural Science Institute (VASI) E-mail : [email protected] TriHanoi

259

ICRISAT

Singh, P Phone : 91 (40) 23296161 Extn. 2334Senior Scientist Fax : 91 (40) 23241239/23296182Global Theme on Agroecosystems E-mail : [email protected] 502 324Andhra PradeshIndia

Ramakrishna, A Phone : 91 (40) 23296161 Extn. 2317Senior Scientist Fax : 91 (40) 23241239/3296182Global Theme on Agroecosystems E-mail : [email protected] 502 324Andhra PradeshIndia

Rego, T J Phone : 91 (40) 23296161 Extn. 2173Senior Scientist Fax : 91 (40) 23241239/3296182Global Theme on Agroecosystems E-mail : [email protected] 502 324Andhra PradeshIndia

Wani S P Phone : 91 (40) 23296161 Extn. 2466Principal Scientist Fax : 91 (40) 23241239/3296182Global Theme on Agroecosystems E-mail : [email protected] 502 324Andhra PradeshIndia

About ICRISATThe semi-arid tropics (SAT) encompass parts of 48 developing countries including most of India, parts ofsoutheast Asia, a swathe across sub-Saharan Africa, much of southern and eastern Africa, and parts of LatinAmerica. Many of these countries are among the poorest in the world. Approximately one-sixth of the world’spopulation lives in the SAT, which is typified by unpredictable weather, limited and erratic rainfall, and nutrient-poor soils.

ICRISAT’s mandate crops are sorghum, pearl millet, chickpea, pigeonpea and groundnut – five crops vital tolife for the ever-increasing populations of the SAT. ICRISAT’s mission is to conduct research that can lead toenhanced sustainable production of these crops and to improved management of the limited natural resourcesof the SAT. ICRISAT communicates information on technologies as they are developed through workshops,networks, training, library services and publishing.

ICRISAT was established in 1972. It is supported by the Consultative Group on International AgriculturalResearch (CGIAR), an informal association of approximately 50 public and private sector donors. It is co-sponsored by the Food and Agriculture Organization of the United Nations (FAO), the United NationsDevelopment Programme (UNDP), the United Nations Environment Programme (UNEP) and the World Bank.ICRISAT is one of 16 non-profit, CGIAR-supported Future Harvest Centers.

About IWMIThe International Water Management Institute (IWMI) is a non-profit scientific research organization focusingon the sustainable use of water and land resources in agriculture and on the water needs of developingcountries. With the mission of “Improving water and land resources management for food, livelihoods, andnature”, IWMI’s research is organized around five themes. These are: (1) integrated water resource managementfor agriculture, (2) sustainable smallholder land and water management, (3) sustainable groundwatermanagement, (4) water resource institutions and policies, and (5) water, health and environment.

IWMI works with partners in the South and North to develop tools and methods to help these countrieseradicate poverty through more effective management of their water and land resources. It has researchprojects running in 21 countries in Asia and Africa. Work is coordinated through regional offices located inIndia, Pakistan, South Africa, Sri Lanka, and Thailand. The Institute has subregional offices in China, Nepal,Ghana, Kenya, Senegal and Uzbekistan. The Institute is a member of the Future Harvest group of agriculturaland environmental research centers that is supported by 58 member governments, private foundations andinternational and regional organizations known as the Consultative Group on International Agricultural Research(CGIAR).

About ADBAsian Development Bank (ADB) is a non-profit, multilateral development finance institution dedicated toreducing poverty in Asia and the Pacific. Established in 1966, it is now owned by 60 members with headquartersin Manila, Philippines and has 22 other offices in the borrowing countries around the world. ADB institutionengages mostly in public sector lending for development purposes in its developing member countries.ADB’s clients are its member governments, who are also its shareholders.

The adoption of poverty reduction as a strategy gives primacy to ADB’s fight against poverty in Asia and thePacific. It helps improve the quality of people’s lives by providing loans and technical assistance for a broadrange of development activities. In doing so, the institution emphasizes on promotion of pro-poor, sustainableeconomic growth, social development and good governance. ADB carries out activities to promote economicgrowth, develop human resources, promotion of gender and development thereby improve the status ofwomen, and protect the environment, but these strategic development objectives now serve its povertyreduction agenda. Its other key development objectives, such as law and policy reform, regional cooperation,private-sector development, and social development, also contribute significantly to this main goal.

Asian Development Bank formulates operational strategies for individual countries, including economic andpolicy analyses, and undertakes country performance reviews, which provide a basis for policy dialogue withthe governments of developing member countries. ADB develops country assistance plans, which includeidentification of individual technical assistance and loan projects and programs. It also establishes andmaintains relationships with DMC governments for overall country economic reporting and for loan negotiations.

The operations strategic focus of ADB is on promoting growth to reduce poverty in poor inland provinces,improving economic efficiency and improving environmental protection and natural resource management.Over the years, ADB has played a significant role in economic and social transformation in Asia and thePacific, boosting economic growth, fostering social development, and helping improve the quality of life formillions of people.

International Crops Research Institute for the Semi-Arid TropicsPatancheru 502 324, Andhra Pradesh, India

Asian Development Bank0401 Metro Manila, 0980 Manila, The Philippines

International Water Management InstituteP O Box 2075, Colombo, Sri Lanka

460-2003Order Code CPE 150ISBN 92-9066-466-5

Integrated Watershed M

anagement for Land and W

ater Conservation and Sustainable A

gricultural Production in Asia

international water asian development bank …oar.icrisat.org/3683/1/460-2003.pdfthe opinions...

Documents