Conflict, Social Capital and Managing Natural Resources A West African Case Study

Conflict, Social Capital and Managing Natural Resources A West African Case Study

Edited by Keith M. Moore

Sustainable Agriculture and Natural Resource Management (SANREM)

Collaborative Research Support Program (CRSP) West Africa Site

Office of International Research, Education, and Development Virginia Polytechnic Institute and State University

CABI Publishing CABI Publishing is a division of CAB International

CABI Publishing CABI Publishing CAB International 875 Massachusetts Avenue Wallingford 7th Floor Oxfordshire OX10 8DE Cambridge, MA 02139 UK USA

Tel: +44 (0)1491 832111 Tel: +1 617 395 4056 Fax: +44 (0)1491 833508 Fax: +1 617 354 6875 E-mail: [email protected] E-mail: [email protected] Website: www.cabi-publishing.org

© CAB International 2005. All rights reserved. No part of this publication may be reproduced in any form or by any means, electronically, mechanically, by photocopying, recording or otherwise, without the prior permission of the copyright owners.

A catalogue record for this book is available from the British Library, London, UK.

Library of Congress Cataloguing-in-Publication Data Conflict, social capital, and managing natural resources : a West African case study / edited by Keith M. Moore. p. cm. “Sustainable Agriculture and Natural Resource Management (SANREM), Collaborative Research Support Program (CRSP).” Includes bibliographical references and index. ISBN 0-85199-948-4 (alk. paper) 1. Natural resources–Africa, West–Management. 2. Land tenure– Environmental aspects–Africa, West. 3. Grazing–Environmental aspects–Africa, West. 4. Soil conservation–Africa, West. 5. Sustainable agriculture–Africa, West. 6. Infrastructure (Economics)–Africa, West. 7. Conflict management–Africa, West. 8. Sustainable development–Africa, West. I. Moore, Keith M. II. SANREM.

HC1000.Z65C66 2004 333.7Ј0966–dc22 2004012409

ISBN 0 85199 948 4

Typeset in Melior by Servis Filmsetting Ltd, Manchester Printed and bound in the UK by Cromwell Press, Trowbridge Contents

Contributors vii Preface ix Acknowledgements xi

Part I: Background

1. Conflict and Agropastoral Development in the Sahel 1 Keith M. Moore, Michael Bertelsen, Salmana Cissé and Amadou Kodio 2. The Lifescape: Production Systems and Social Institutions 23 Salmana Cissé, Keith M. Moore and Charlene Brewster 3. Landscape and Soil Characterization of Madiama Commune 35 Oumarou Badini and Lassana Dioni 4. Investigation into the Movements of Cattle, Sheep and Goat Herds Through the Commune of Madiama 53 Adama Ballo and Abdoul Karim Ouattara 5. Land Use Changes in Madiama Commune 71 Randolph H. Wynne, Mohamed Sidi Mohamed Touré, Nina Sengupta and Adama Ballo

Part II: Interventions

6. Participatory Landscape/Lifescape Appraisal 77 Julia Earl and Amadou Kodio 7. Building Social Infrastructure for Sustainable Development 89 Keith M. Moore, Salmana Cissé and Abdoulaye Touré 8. Conflict Management Training 101 Jeff Goebel, Catherine Ampagoomian and Abdoulaye Touré

v vi Contents

9. Holistic Management Applied to SANREM Research and Development in Madiama 117 Sam Bingham

Part III: Findings

10. Farmers’ Knowledge and Perceptions of Soil Fertility 129 Todd Crane and Boureima Traoré 11. Biophysical Assessment of Alternative Soil Fertility Technologies Through Modelling 143 Oumarou Badini, Boureima Traoré and Mohamed Sidi Mohamed Touré 12. An Economic Evaluation of Alternative Soil Management Technologies 159 Peter Wyeth, Boureima Traoré, Oumarou Badini, Mohamed Sidi Mohamed Touré and Charlene Brewster 13. Economic Linkages Among Occupational Groups in the Madiama Commune 181 Charlene Brewster, Daniel Kaboré, Michael Bertelsen and Peter Wyeth 14. Dry Season Feed Supplements: the Potential Role of Cassia tora 195 Meriem El Hadj, Ozzie Abaye, Amadou Kodio and Moussa Keïta 15. Controlled Grazing: Botanical Response and Animal Performance 211 Ozzie Abaye, Meriem El Hadj, Amadou Kodio and Moussa Keïta 16. Social Capital and Improved NRM 229 Keith M. Moore and Salmana Cissé 17. What We Have Learned 247 Keith M. Moore, Charlene Brewster and Michael Bertelsen

Index 257 Contributors

Abaye, O., Department of Crop, Soil and Environmental Science, Virginia Tech, Blacksburg, Virginia 24061-0404, USA. E-mail: [email protected] Ampagoomian, C., Tulane School of Public Health and Tropical Medicine Tulane University, New Orleans, Louisiana, USA. E-mail: [email protected] Badini, O., International Programs/Development Cooperation, Washington State University, Pullman, Washington 99164-5121, USA. E-mail: [email protected] Ballo, A., ESPGRN Team Leader, Centre Régional de la Recherche Agronomique, Institut d’Economie Rurale, BP 205, Mopti, Mali, Africa. E-mail: [email protected] Bertelsen, M., College of Agricultural and Life Sciences, Virginia Tech, Blacksburg, Virginia 24061-0334, USA. E-mail: [email protected] Bingham, S., Allan Savory Center for Holistic Management, 394 Vanderbilt Road, Asheville, NC 28803, USA. E-mail: [email protected] Brewster, C., Office of International Research, Education and Development, Virginia Tech, 840 University City Boulevard, Blacksburg, Virginia 24061-0378, USA. E-mail: [email protected] Cissé, S., Centre Régional de la Recherche Agronomique, Institut d’Economie Rurale, B.P. 205, Mopti, Mali, Africa. E-mail: [email protected] Crane, T.A., Department of Anthropology, University of Georgia, Athens, Georgia 30602, USA. E-mail: [email protected] Dioni, L., Centre Régional de la Recherche Agronomique-Sotuba, Institut d’Economie Rurale, BP 258, Bamako, Mali, Africa. E-mail: [email protected] Earl, J., Preventing Harm Minnesota, Minneapolis, Minnesota 55407, USA. E-mail: [email protected] El Hadj, M., Department of Crop and Soil Environmental Science, Virginia Tech, Blacksburg, Virginia 24061-0404, USA. E-mail: [email protected] Goebel, J., Goebel & Associates, PO Box 2503, Redmond, Oregon 97756, USA. E-mail: [email protected] Kaboré, D., Institut de l’environnement et des recherches agricoles, Burkina Faso, 04 BP 7192, Ouagadougou, Africa. E-mail: [email protected] Keïta, M., Centre Régional de la Recherche Agronomique, Institut d’Economie Rurale, B.P. 205, Mopti, Mali, Africa. E-mail: [email protected]

vii viii Contributors

Kodio, A., Centre Régional de la Recherche Agronomique, Institut d’Economie Rurale, B.P. 205, Mopti, Mali, Africa. E-mail: [email protected] Moore, K.M., Office of International Research, Education and Development, Virginia Tech, Blacksburg, Virginia 24061-0378, USA. E-mail: [email protected] Ouattara, A.K., Centre Régional de la Recherche Agronomique, Institut d’Economie Rural, BP 205, Mopti, Mali, Africa. E-mail: [email protected] Sengupta, N., Center for Environmental Applications of Remote Sensing, Virginia Tech, Blacksburg, Virginia 24061-0324, USA. Touré, A., Local Governance Advisor, CARE/Tombouctou, Mali, Africa. E-mail: [email protected] Touré, M.S.M., Centre Régional de la Recherche Agronomique, Institut d’Ecomonie Rurale, BP 205, Mopti, Mali, Africa. E-mail: [email protected] Traoré, B., Centre Régional de la Recherche Agronomique, Institut d’Economie Rurale, BP 205, Mopti, Mali, Africa. E-mail: [email protected] Wyeth, P., International Programs-Development Cooperation, Washington State University, Pullman, Washington 99164-5121, USA. E-mail: [email protected] Wynne, R.H., Center for Environmental Applications of Remote Sensing, Virginia Tech, Blacksburg, Virginia 24061-0324, USA. E-mail: [email protected] Preface

This book is the product of a 6-year pro- Georgia, and trainers from Tufts University gramme involving the population and terri- and the Center for Holistic Management. tory of one decentralized local government Despite SANREM’s academic roots, it is also unit in West Africa. On one hand, it highly participatory and seeks to ensure rel- recounts the efforts of research and devel- evance and encourage adoption of improved opment agents engaged in developing and NRM processes by developing the tools and transferring the appropriate social and bio- methodologies collaboratively with local physical technologies and decision support institutions and partners. This inclusive tools for sustainable development. On the approach provides the added benefit of other hand, it describes the experience and being able to build on all SANREM partners’ conditions of the local population seeking experience. In Mali, SANREM’s principal a pathway from poverty and food insecu- partners are led by the Institut d’Economie rity to a healthy economy and environ- Rurale (IER), CARE-Mali and Groupe de ment. Recherche Action pour la Développement The book’s contributors are part of a (GRAD) as well as the men and women, global programme, the Sustainable Agri- farmers, pastoralists and fisherfolk of the culture and Natural Resource Management Commune of Madiama. (SANREM) Collaborative Research Support SANREM began Phase II (1998–2004) of Program (CRSP) initiated in 1993 with sites its activities in West Africa with the objec- in Ecuador, the Philippines and Mali. The tive of advancing NRM in the Sahel by SANREM CRSP is financed jointly by the addressing what was increasingly being per- United States Agency for International ceived as the major stumbling block to Development (USAID) and participating progress in community-based NRM: violent institutions. Through SANREM, US univer- conflict over access to and use of natural sities partner with host country institutions resources. We immediately realized that as for the purpose of building and testing inter- partners in a research programme we had no disciplinary tools and methodologies to comparative advantage in conflict resolu- assist agricultural and natural resource tion per se. However, we did feel we had management (NRM) decision makers. something to contribute through the devel- In the USA, SANREM West Africa is led opment and testing of a model institutional by Virginia Tech in partnership with Wash- framework by which communities could ington State University, the University of learn to work together to improve manage-

ix x Preface

ment of both their natural resources and the This community-based NRM would be both conflicts embedded in their exploitation. good science and good development.

x Acknowledgements

A collaborative effort such as this involves a Madiama has been a source of inspiration large supporting cast who have built the and renewal for our work. We especially SANREM CRSP West Africa site programme would like to thank Commune Mayor from its base in Madiama through IER and Konaté and his staff who have always been their partners on the local and national scene available and encouraging in our activities. in Mali, to Virginia Tech, Washington State, Our gratitude is also extended to the chiefs Tufts, and the University of Georgia, and of the ten villages composing the commune, finally, to the Land Resources Management their councillors and those who shared in Team in the Office of Natural Resource our research activities whether in the house- Management at USAID/EGAT/NRM/LRM in hold or in the field throughout the seasons. Washington, DC. The SANREM CRSP West In addition, we deeply appreciate the sup- Africa team would like to thank USAID for port and advice of Macké Cissé of the supporting the research and development Sub-Prefecture of Sofara, as well as that of activities described in this book. We are Amadou Hadj, President of the Nérékoro especially grateful to Bob Hedlund, SAN- Herders Association and M. Askofaré of the REM CRSP Project Manager, and Christine Mopti Office Riz. We also want to thank the Bergmark, who preceded him. We would Institut d’Economie Rurale (IER) technicians also like to thank the SANREM Management Ibrahima Diallo, Issa Kané and Albedja Entity at the University of Georgia: Carlos Touré, who are based in Madiama. Of course, Perez and the late Bob Hart, Program we owe a deep debt of gratitude to the new Directors, Carla Roncoli and Constance leaders of Madiama on the Natural Resource Neely, Associate Program Directors, and Management Advisory Committee (NRMAC) Tammy Cotton and Gina Thomas, Adminis- who have been our guides and assistants. trative Assistants for their support and assis- Among those we would particularly like to tance. This is the team that assured the thank are President Issa Sao, Secretaries effective implementation of this effort over Moussa Sao and Abdoulaye Cissé, Herder the past 6 years. Representative Amadou Cissé, Hunters Rep- Applied researchers cannot function resentative Malicki Camara, and Women’s without the assistance of cooperative part- Association Representative Kadidia Konaté. ners who may or may not benefit from the This multi-institutional programme null hypothesis. The faith and cooperative could not have functioned without the coor- support of the residents of the Commune of dination and support of our Malian partners. xii Acknowledgements

We wish to express our gratitude to Director laborative research endeavour. Generals Alpha Maïga and Bino Témé and I would personally like to express my Scientific Director Bouréma Dembélé of IER deep appreciation for the colleagueship of for facilitating our work. We deeply appreci- the authors, practitioners, researchers, tech- ate the leadership provided by SANREM nicians and friends who have brought this CRSP-Mali National Coordinators, Lassine work to its present state. Thank you all. In Diarra and Salmana Cissé (in particular for particular, I would like to extend my grati- his wisdom and humour) and the hospital- tude to my predecessor as Program Manager, ity of Centre Régionale de la Recherche Mike Bertelsen, a constant source of inspira- Agronomique (CRRA)/Mopti Center Direc- tion, Oumarou Badini for late night discus- tor, Amadou Kodio. Thanks are owed to sions, Charlene Brewster for calming advice Christie Collins and Garth Van’t Hul at and council, Catherine Ampagoomian for CARE/Mali and Idrissa Maïga and Gouro research assistance, John Lipovsky for Diallo of Group de Recherche Action pour la clerical support, and copy editors Sara Développement (GRAD) for backstopping Thorne-Thomsen and Kirk Neal for making the field team in Djenné and Madiama. All of this all readable. us owe a deep debt of gratitude to Abdoulaye Finally, this book was made possible Touré (CARE/Djenné) who as NRMAC through support provided by the G/EGAD/ Facilitator kept the wheels turning through AFS Global, United States Agency for all seasons. Thanks are also due to Chéry International Development under terms Traoré (GRAD) for picking up the reins in the of Grant No. PCE-A-00-98-000-19-00 to last year of the project. Virginia Polytechnic Institute and State We would also like to thank S.K. De University (Virginia Tech). The opinions Datta, SANREM CRSP Principal Investigator expressed herein are those of the authors and Director, Office of International Re- and do not necessarily reflect the views of search, Education, and Development, Vir- the United States Agency for International ginia Tech, and the late Jan Noel, Director of Development. International Programs, Washington State University, for lightening the load of this col- Keith M. Moore 1 Conflict and Agropastoral Development in the Sahel

Keith M. Moore1, Michael Bertelsen2, Salmana Cissé3 and Amadou Kodio3 1Office of International Research, Education and Development, Virginia Tech, Blacksburg, VA 24061-0378, USA; 2College of Agricultural and Life Sciences, Virginia Tech, Blacksburg, VA 24061-0334, USA; 3Centre Régional de la Recherche Agronomique, Institut d’Economie Rurale, B.P. 205, Mopti, Mali

The Inland Delta and hinterlands of the weak involvement of the local population Niger River, like many natural resource sys- in the conception, design, implementation tems throughout the world, are transitioning and evaluation of these interventions. to more intensified agriculture and animal This, in turn, led to new community-based husbandry production. These dominant sec- approaches such as gestion de terroir1 and tors serve as the engines of sustainable the development of pastoral organizations. economic development that provide food These sectoral approaches despite their security and alleviate poverty. Although more participatory nature have suffered open range opportunistic grazing manage- from three interrelated weaknesses: the ment by transhumant herders has been a limited scope of intervention with respect way of sustaining life for centuries in this to the natural resource base to be managed; region, increasing population pressure, inherent biases towards sedentary popula- changing political structures, declining and tions; and the lack of mechanisms to cope erratic rainfall, and degrading natural with natural resource-based conflicts resources have forced both agricultural and at the local level. Indeed, until recently, pastoral communities to transform their non-governmental organizations (NGOs) production systems and the social relations implementing local development pro- on which they are based. Unfortunately, grammes have systematically avoided this transformation has brought about all situations where conflict was mani- increasingly violent conflicts over natural fested. resources. Historically, the State has been either After the introduction of animal trac- unwilling or unable to effectively reverse tion effectively increased cultivated land, the trend towards increased conflict and subsequent technological fixes have largely related natural resource degradation. In fact, failed to increase agricultural or livestock the State has often incited conflict by intro- productivity and have even resulted in ducing projects, programmes, and policies higher rates of natural resource degrada- favouring one group of resource users tion. The reason for such poor results is over another. Throughout the Sahelian generally perceived as stemming from the region of West Africa there is a growing © CAB International 2005. Conflict, Social Capital and Managing Natural Resources (ed. K.M. Moore) 1 2 K.M. Moore et al.

consensus that central governments are where appropriate, created. Although par- poorly placed to make many of the decisions ticipation is both necessary and appropri- critical to their citizens’ welfare. The ate, what matters most is the quality of Interstate Committee for the Fight against participatory relationships and how those Desertification in the Sahel (CILSS)-spon- relations are interconnected (Pretty and sored Nouakchott Conference (1984) was Buck, 2002). the first time West African governments had NRM is one of the powers being formally recognized the need for local decentralized to the local level across West involvement in development projects. This Africa, shifting governance and consensus conference was followed by two other building to those best equipped to address CILSS-sponsored conferences, one in Ségou these issues. However, this has placed a sig- (1989) that further stressed the need for nificant burden on emerging local NRM decentralized natural resource management decision-making institutions. Local govern- (NRM) governance, and the other in Praia ments are poorly equipped and trained (1994) that highlighted the relationships to deal with NRM issues. In addition, between decentralization and land tenure. members of the Rural Councils are fully Subsequently, West African states embarked occupied in becoming acquainted with the on a policy of decentralization, devolving large number of additional duties and res- responsibility for governance to local ad- ponsibilities they have assumed under ministrative structures with far-reaching decentralization. A consistent and effective implications for governance in the region. approach to building and reinforcing the At the same time, multi-national and bi- social and human capital necessary to lateral donor agencies (The World Bank, facilitate local NRM decision making is United States Agency for International needed. Development (USAID), etc.) and NGOs This book details an approach to have promoted local civil society organiza- social and human capital development for tions (CSOs) as a core element in their improved NRM decision-making through new strategies for rural development and the Sustainable Agriculture and Natural NRM. Resource Management (SANREM) Colla- The advent of decentralization has pro- borative Research Support Program (CRSP)- vided a renewed impetus mobilizing rural West Africa programme. Within these civil society in the transition to more inten- chapters we describe the development of sified production systems and transforming networks among the local population, the vicious cycle of poverty into a virtu- NGOs and local and foreign researchers, ous circle of poverty alleviation. Recent and we explore the biophysical technolo- research on social capital has shown that gies and decision support tools for facilitat- participation in development activities is a ing improved and locally driven, natural more nuanced phenomenon than once resource decision making. In this chapter imagined (Dasgupta and Serageldin, 2000; we review efforts that have been made to Isham et al., 2002; and Grootaert and van accelerate agricultural and pastoral devel- Bastelaer, 2002). In order to confront the opment in the Sahel since independence. over-determined conflict between produc- We further analyse the origins and impacts tion system actors as they transition to more of natural resource-based conflicts. We dis- intensified and sustainable livelihood sys- cuss the four pillars on which SANREM tems, the interconnections between groups work is based (participation, interdis- sharing common resources need to be ciplinarity, multi-stakeholder involvement, addressed. Bonding, bridging and linking landscape/lifescape scales) and explore the relationships that build trust and common SANREM approach to improving natural interests and the networks that support resource and conflict management at the them within, between and beyond commu- commune level. The chapter concludes nities (collectively referred to as social cap- with an overview of the chapters in this ital) need to be examined, strengthened and, volume. Conflict and Agropastoral Development in the Sahel 3

Development Assistance Strategies and cultivars has been a major thrust of the Agropastoralism IARCs (following the Green Revolution suc- cesses) and National Agriculture Research Development assistance actors Services (NARSs) have also followed this path (Sanders et al., 1996). Extension activ- Leadership and support for development ities have focused on the introduction of assistance to agropastoral populations have inorganic fertilizer and new cultivars. traditionally included multinational banks, Despite some successes, the adoption of bi-lateral donor agencies, international agri- new varieties has not caught on with cultural research centres (IARCs), national African farmers and consequently popula- ministries, research institutes, and locally tion growth has outdistanced agricultural based NGOs. The most significant of these growth (Lynam and Blackie, 1994). While have been the national ministries and American range science began to influence research institutes whose legacy of control livestock research during this period (Pratt over long-term programme implementation et al., 1997), increased livestock productiv- often dates from the colonial period. These ity has not occurred either. Livestock devel- organizations have provided sectoral guid- opment and extension activities have ance and set national priorities. The role of stressed little more than providing veteri- multinational and bi-lateral agencies has nary services and finding new sources of been to set the trends in development pro- water. ject design, building on the lessons learned The Sahelian drought of 1968–74 from successive waves of pastoral and agri- brought about many changes in the way cultural development projects. agropastoral development issues were add- Various government ministries have ressed. In 1973, the Permanent Inter-State promoted agricultural and pastoral policies, Committee for Drought Control in the Sahel programmes, and projects by sector. These (CILSS) was established with the objective of ministries have rarely coordinated their finding permanent solutions to food self-suf- work. Often the ministry for agriculture is ficiency and socioeconomic well-being in separate from that of livestock (or subsumed the Sahel through the coordinated efforts of as a separate component). Recently evoked nine countries (Burkina Faso, Cape Verde, environmental issues have been the domain Chad, The Gambia, Guinea Bissau, Mali, of a third ministry. If, as Sanders et al. Mauritania, Niger and Senegal). Their first (1996) suggest, government policies have actions involved emergency aid by coordi- neglected agriculture, they have even further nating the international relief efforts of neglected livestock. Overall government the time. CILSS has moved on to mobiliz- development policies have favoured the ing development resources, conducting intensification of agricultural production at research through the Institute of the Sahel the expense of transhumance. This bias (INSAH) and national research institutes, extends to the almost complete omission of disseminating information, and promoting pastoralist concerns in the development awareness throughout the region and inter- of famine/drought early warning systems nationally on drought and desertification (Sommer, 1998). (Djalbord Diard, 1992). After independence in the 1960s, the World Bank, IARCs, and other donor agen- cies supported this sectoral approach. Projects and programmes Agricultural research and development focused on the development and introduc- To alleviate the effects of periodic drought tion of new crop varieties. Productivity and promote development, projects and gains were expected to result from genetic programmes using different types of strate- enhancement of export crops, rather than gies have been implemented to assist from improved agronomic practices (Lynam agropastoral groups. These include drou- and Blackie, 1994). The introduction of new ght relief actions, targeted interventions, 4 K.M. Moore et al.

systematic or programmatic interventions, land areas where herders traditionally bring and participatory interventions. their herds during the dry season. Both pas- tures and sources of drinking water are diminishing rapidly due to this trend. Drought relief Governments and NGOs have supported the Drought relief actions consist of provid- intensification of agriculture and livestock ing food to people under current drought production. Both require fixed investments conditions. Pastoralists who lost their live- in land and, therefore, the permanent settle- stock or farmers who lost their crops during ment of the producing population(s). Lack the drought benefit from such actions. Food of success has occurred for many reasons, is distributed to needy households. Such including an increased concentration of ani- interventions are frequent in all Sahelian mals contributing to environmental degra- countries. However, recurrent drought relief dation and increased conflicts between has done little more than further debilitate herders and farmers. existing production systems. Commercial ranching, group ranching and permanent settlements designed to increase meat and milk production were Targeted interventions the primary emphasis of pastoral develop- Apart from the above punctual drought relief ment projects during this period. Although actions, specific interventions have often some World Bank projects addressed land been implemented for pastoral develop- tenure and policy issues, priority was given ment. The objectives of such interventions to capital investments, centralized plan- have been: (i) to increase herd size (or re- ning, and reduced stocking rates (Pratt et establish/restock herd), (ii) to increase milk al., 1997). National and bi-laterally funded yield, (iii) to maintain appropriate herd projects followed this general model. For structure, and (iv) to develop disease resis- the most part, these efforts have been char- tance by selective breeding (Niamir, 1991). acterized as either remarkable for their uni- Vaccines, the eradication of tsetse flies, dams versal lack of success (Horowitz and Little, and boreholes have contributed to in- 1987) or simply a failure of farmers and creased livestock production. Unfortunately herders to adopt the proposed technologies drought relief actions, livestock salvage (McIntire et al., 1992). Preston and Leng operations, and targeted interventions have (1994) suggest that animal science made been conducted on an ad hoc basis, and not little or no contribution to increasing live- as a part of a coordinated, coherent long- stock productivity. Pratt et al. note the term strategy. As a result, they are relatively primary causes of failure as a lack of under- costly and have not been sustainable (Okai, standing of the traditional pastoral systems; 1987). an assumption that these systems were market-driven; an imposition of rigid orga- nizational forms, tenure rights systems, and Systematic and programmatic interventions inappropriate incentive frameworks; and Systematic and programmatic interventions institutional weaknesses of the implement- are intended to develop the whole system of ing agencies. production (agriculture or livestock) instead Livestock market development has been of targeting actions on only restricted and generally coupled with permanent settle- isolated system components. Irrigation pro- ments. The aim has been to facilitate live- jects originally determined the model for stock exports and taxation of pastoralists. this type of development intervention. Since livestock products represent 20–30% While major new irrigation systems are not of Malian exports, there is good reason to currently being developed, older systems believe that developing and formalizing have been improved or expanded, the such markets is worthwhile. Agreements number of micro-systems increased, and between Communaute Economique du agricultural production expanded into wet- Betail et de la Viande (CEBV) countries Conflict and Agropastoral Development in the Sahel 5

theoretically delineate corridors for live- production circumstances. Intensification is stock inter-state movement, mostly from already beginning to occur, as Sanders et al. Mali, Burkina Faso and Niger to coastal (1996) note, with respect to the use of zaï countries (Côte d’Ivoire, Benin and Togo) in and bunds combined with manure. They order to ensure international transhumance recommend going beyond simple labour- patterns that minimize conflicts between intensive technological innovations to farmers and herders. Unfortunately, these increased investment by farmers in combi- agreements are less and less respected by nation with capital intensive technologies farmers who expand food crop production to including inorganic fertilizer. meet the demand of the growing population. NGO contributions to agropastoral Furthermore, the efficiency of corridors is development typically began with emer- questioned even at the national level. gency relief activities. While this led to a In Burkina Faso, the government, in col- top-down approach, with increasing regu- laboration with the World Bank, established larity NGOs have become involved in more four ranches (a pastoral management zone) routine development activities (Brown, on land for which full title was not estab- 1993). These activities have focused on lished. In doing so, use rights were left small scale interventions targeting villagers undefined for project beneficiaries. The and local problems. The NGOs have begun majority of colonists in these ranches were to recognize the importance of natural farmer–herders (80%) and the rest, Peul resource tenure issues and associated con- herders. The agricultural communities grew flicts, but they have tended to avoid involve- cereals, cotton, citrus, bananas and vegeta- ment in these issues. Their lack of bles. The bananas and vegetables were cul- partnerships and contact with government tivated along the streams and in wetland services, the small scale of activities, the areas. In fact, the cultivators left little or no absence of baseline and monitoring data, room in their land use plans for animal cor- and low level of technical expertise have all ridors, access to water points, or dry season diminished NGOs’ impact (Toulmin and pasture (Sanon, 1996). In a similar project in Moorehead, 1993). Nevertheless, NGOs are Sideradougou (Bary, 1996), conflict result- beginning to support pastoral systems that ing from the grazing of a herd in a cultivated attempt to enable pastoralists to exploit spa- field left a Peul herder dead, and later led to tial and temporal availability of rangeland the death of the Karboro father of the resources. assailant. Both Peul and Karboro communi- ties accused government forces (which Participatory interventions defused the immediate confrontation) of siding with their opponents. Although most interventions are systematic Consequently, sector-based strategies or implemented in a programmatic frame- have come into disrepute throughout the work, they fail to increase agricultural or Sahel. Current strategies apply integrated livestock production and result in even approaches focusing on natural resource greater natural resource degradation. The management, a prioritization of food pro- reason for such bad results is the weak duction, flexibility and mobility in herd involvement of the local population, as they management, and an emphasis on local were not involved in project conception, institution building (particularly among design, implementation or evaluation. Such pastoralists). These approaches shift agri- failures called for new community-based cultural research priorities from breeding to approaches such as gestion de terroir and improved soil management technologies the development of pastoral organizations (Lynam and Blackie, 1994) in order to in the Sahelian countries. increase agricultural and livestock produc- Gestion de terroir is a multi-sector and tivity. Farmers and herders should be pro- global strategy aiming to establish a new vided with a menu of technological options socioeconomic and ecological equilibrium from which they may select or adapt to their in order to achieve food self-sufficiency 6 K.M. Moore et al.

and to preserve/regenerate the productive types of institutional arrangements that reg- potential of natural resources (Rochette, ulate individual and collective actions by 1989). Gestion de terroir involves the local pastoralists to safeguard and promote their people in natural resources planning and economic, social, cultural and political gives them full responsibility for NRM interests. They are legally organized by the within the limits of their terroir. Land tenure government and dependent on the state for must be secure enough in this unit in order technical assistance, supplies and financial to ensure a sustainable use of resources. All resources (Vedeld, 1992). PO functions resource users in the terroir should be include acquiring secure land tenure, included in the management unit (e.g. resource management, provision of services, ethnic, age and gender groupings). communication of information, external This approach was developed in the relations and the building and maintenance early 1980s and first initiated in some of community coherence and morale. Key Sahelian countries like Burkina Faso and factors that determine the viability of POs Senegal. In Burkina Faso, the Programme include food security, water security, land National de Gestion des Terroirs (PNGT), security, herd ownership, credit, veteri- financed by the World Bank, has been nary services, marketing, economic self- conducted for a decade in regions of sufficiency and literacy (Shanmugaratnam the country where NRM is critical. The et al., 1992). Implementation problems have subsequent World Bank Projet de Gestion lessened PO impacts. In particular, drought des Ressources Naturelles (PGRN) has dis- has led pastoralists to disperse as migrants in seminated the methodology throughout search of food. In addition, the identification the Sahelian countries in order to implement of beneficiaries has not been clear in com- local level NRM. In Senegal, the USAID- munities and there has been a noted inca- sponsored Community-Based Natural Re- pacity of POs to implement project plans. source Management (CBNRM) Project also Shanmugaratnam et al. provide a complete worked with a similar approach but applied list of the problems. it at the multi-village level. Despite these Increasingly, international donor agen- attempts, many problems have been noted, cies and national NGOs are recognizing the including: (i) lack of fundamental research- importance of conflict resolution as a major based data that could help projects or NGOs component of natural resource management in implementation; (ii) village terroirs impre- research and extension efforts. This has cisely defined and limited in scope; (iii) the resulted from their increased understanding agricultural area has nearly always been the of pastoral management systems and their unit of analysis instead of livestock; and (iv) problems shifting project focus to institu- avoidance of conflict situations (Lowenberg- tion building, and decentralized and flexi- DeBoer et al., 1994; Painter et al., 1994; ble decision making processes (Pratt et al., Benjaminsen, 1997). 1997). For their part, the NARSs only indi- rectly address issues of land tenure and con- flict in agropastoral systems. NARS major Pastoral organizations contributions continue to be the develop- An emerging general consensus on the ment of resource-intensifying technologies importance of livestock to the Sahelian designed to increase and diversify incomes, economies led decision-makers to consider thereby increasing food security and reduc- pastoral development in their policies. ing the likelihood of resource conflict. In particular, pastoral organizations sup- ported by The World Bank, the African Development Bank, IFAD, France, and NRM-related Conflicts Norway were developed in Mali, Senegal, Mauritania and Niger as a new form of com- The connection between food insecurity, munity-based NRM (Shanmugaratnam et al., NRM and conflict is well documented in the 1992). Pastoral organizations (POs) cover all literature. Messer et al. (1998) point out that Conflict and Agropastoral Development in the Sahel 7

food insecurity and natural resource scarci- intensive systems of the modern era. ties are major causes and consequences of Traditional systems have used the natural conflict. The authors cite data indicating resource base of the Sahel in multiple and that conflict in countries of sub-Saharan complementary ways. With increasing pres- Africa has been associated with per capita sure on the resource base, however, live- food production declines of more than stock and farming systems are coming into 12% per year. In the Sahel, the most severe conflict over the schedule of land use by the conflict has generally involved NRM prac- different systems (Soumaré, 1996). For tices associated with agricultural and pas- example, at one time near Nioro du Sahel in toral production systems. Mali, the farming Soninké and the herding The heightened competition bet- Toucouleur benefited from an exchange of ween livestock and agricultural production manure for grain and established a tradition through the conversion of pastureland into of shared natural resource tenure. The cropland and the expansion of cultivation Toucouleur eventually established a village into irrigated agriculture along water points in the region and continued their transhu- has restricted access to water and to dry- mance. Today, with decreased rainfall, land season pasture of bourgou (Echinochloa degradation and adoption of livestock rais- stagnina). Additionally, by cultivating ing by the Soninké, the exchange has lost its livestock corridors (enriched by animal value. Although the Soninké rely on the manure), sedentary farmers have disturbed Toucouleur for tending to their herds, the transhumance patterns and herders have shared use of resources has become prob- been obliged to change their strategies in lematic. The herders have more difficulty order to adapt to the new landscape. accessing dry season crop residues for their Maintaining control of local resources in the herds, while the sedentary farmers feel their face of non-local users or migrant ‘strangers’ agricultural calendar is restricted by the can be a powerful force mobilizing a seden- incursion of the herds. Exchanges have tary agricultural community. For herders, become monetized, but there is no unified herd mobility is crucial to survival and the authority to regulate local land use. While trampling and grazing of cultivated fields farmers are governed by the village council, before harvest consequently leads to the herders are governed by their clan, which most common flash point initiating overt has no fixed location. conflicts. Pairs of actors involved in natural These conflicts have been particularly resource conflicts are often used to develop acute where competition for resources a conflict typology. This usually involves between managers of ethnically differenti- pairing farmers and/or herders with like or ated production systems is highest. Conflicts other actors (fisherfolk, non-local private are found not only in conditions where com- sector, migrants, the State or its agents). It plementary systems once co-existed, but may be more useful, however, to categorize they are also becoming more frequent among conflicts in terms of whether they are within members of the same community. Inter- or between production systems. On the one and intra-community conflicts increase as hand, conflict within production systems the natural resource base shrinks due usually involves attempts to maintain or to higher population pressure and diminish- increase the size of the actors’ production ing annual rainfall. The disputed resources units and is based on issues of inheritance, revolve around access to land, but are land boundaries and proprietorship. To the particularly focused on specific rights to extent that unambiguous rules and proce- cultivate or to graze, water and move dures exist, these conflicts can normally be livestock. resolved within the village, but have These conflicts are endemic to the tran- become a source of socio-economic differ- sition between the extensive livestock and entiation within rural communities. cropping systems characteristic of the pre- On the other hand, conflicts between colonial past and the expanding, more production systems are more serious, 8 K.M. Moore et al.

because they threaten not only the immedi- while the Peul continued their transhumant ate livelihoods of the disputants, but also practices (Maïga and Diallo, 1998). The the way of life of distinct ethnic communi- advent of colonialism legally ended slavery. ties. When disputants are from different However, the Dina, a socio-religious land ethnic groups, a conflict may be further tenure organization established during aggravated and larger constituencies drawn Peul dominance, remained in effective oper- into the fray. ation. This allowed the Peul’s former slaves Increasing population, expansion of continued access to the land they farmed, but cultivated land (with the consequent disap- also assured perceptions (at least in the pearance of fallow), the introduction of cash medium term) of Peul proprietorship (Cissé, cropping, and the accompanying degrada- 1996). tion of the vegetative cover have conspired Inter-state relations are also involved in across the Sahel to create situations of natural resource conflicts. Nomadic or tran- confrontation between rural stakeholders. shumant herders have little regard for These types of conflicts, particularly bet- national frontiers. Key informants indicate ween farmers and herders, are increasing that this is often the case when herders (Souley, 1996; Maïga and Diallo, 1998; arrive in northern Ivory Coast or Ghana from Blench, 1996). Burkina Faso or Mali, along transhumant On a larger scale, the government of routes between western Mali, Mauritania Niger set a limit in the northern part of the and Senegal, or between Niger, Nigeria and country (called the Sorghum Line), north of Benin. This type of conflict may provoke an which agricultural producers were not international confrontation, as in the case of allowed. This region was reserved for rainy the Mauritanian military action killing two season grazing only. Today, migrant herders Senegalese (Soninké) cultivators following a have begun to cultivate land around water confrontation between Mauritanian Peul points north of the sorghum line on herders and Soninké cultivators (Park et al., lands once managed for livestock grazing 1993). This killing in turn led to a serious (Sidikou, 1995). The official deadlines of 30 international incident involving the death May (on which livestock should be moved of hundreds. out of the cultivation zone) and 30 Lo et al. (1996) note the importance of November (the date for their return) are no distinguishing between conflicts which are longer respected. Furthermore, the changing horizontal and those which are vertical in rainfall regime has obliged herders to orientation. Throughout the region conflicts move north later, and consequently, they are not only between cultivators and herders are tempted to take advantage of the new- and among themselves and their communi- growing grasses in the southern agricultural ties (horizontal conflicts), but also between zone. these communities and the state or specific These conflicts can become extremely development projects (vertical conflicts). violent when disputants contesting the right Often the state, or a development project, to use a resource have a history of ethnic con- is a background partner setting the stage flict. This was the case of the conflict between for and/or reawakening latent conflicts. the Sosobe and the Salsalbe in the Niger Delta Privatization of the economy has led to of Mali. Contested claims to a grazing terri- access to forest reserves by the private sector tory dating back to 1936 exploded into a aggravating this trend. In Senegal, despite violent confrontation in December 1993, encouraging the decentralized management leaving 29 dead and 42 wounded (Maïga, of natural resources, the government has 1996; Vedeld, 1994). Often, contested claims issued permits to harvest wood for charcoal in the Niger Delta are linked to traditional production to urban-based charcoal mer- ethnic relationships between the Peul chants. For example, the Wolof and and the Rimaibés (their ancient slaves). Mandinke communities of Maka Coulibanta According to tradition, the Peul allowed the attempted to assert their right to regulate the Rimaibés to farm Peul-dominated territory, wood cutting trade only to have the Forestry Conflict and Agropastoral Development in the Sahel 9

Service negotiate use rights to urban mer- At the heart of the perennial nature of chants (Kane and Winter, 1997). In Niger, most of these conflicts is the lack of a single the government denied access of local accepted authority for the resolution of nat- Toureg and Peul populations to the Gadabeji ural resource rights and associated tenure Reserve. In so doing, conflicts were created security in the management of natural not only between the government and resources (Touré, 1996; Maïga and Diallo, herders grazing their animals on Reserve 1998; Lo et al., 1996; Ngaido, 1996). For lands, but also between herding communi- example, when local authorities and exten- ties as they concentrated around the only sion agents were asked to describe natural remaining watering holes and private wells resource tenure in Senegal (Moore, 1996), (Sidikou, 1995). they would begin by describing how the While the most frequently recounted National Domain Law was supposed to (and often more violent) conflicts are operate. They explained that when the law between herders and farmers, there are was not explicitly followed it was due to nearly as many less notorious affairs between lack of understanding of the texts and poor farmers and farmers, or herders and herders. interpretation by illiterate local officials. Cissé (1996) reports that 42% of the conflicts However, when pressed, they admitted that in the Niger Delta of Mali are between traditional land allocation practices domi- farmers and herders, and 40% are between nate: a household or compound head that farmers and farmers. Conflicts were equally requires additional land will turn to a neigh- divided between inter- and intra-village. An bouring household head to borrow it. If, example of farmer–farmer conflict can be after some years or in the next generation, seen in the Burkina Faso village of Nébourou conflicting claims arise, village elders or the where conflict erupted between the original village head may arbitrate. Only when vil- Nuni inhabitants and the in-migrant com- lage authorities cannot resolve a dispute munity of Mossi (Laurent and Mathieu, does litigation proceed to the Rural Council, 1994). Nébourou is located in the land-rich, and perhaps even to the Sous-Prefet. Even in and once lowly populated, western region of these cases, resolution of the problem is Burkina Faso. With increasing soil degrada- dependent on the political power of the liti- tion in the Mossi Plateau and worsening gants as much as on the technical require- drought impacts, Mossi populations began ments of any law (see Blundo, 1996; or immigrating to Nuni villages, particularly Laurent and Mathieu, 1994). after the establishment of the Agrarian Conflicts over access to natural re- Reform of 1985, which opened all Burkina sources and their benefits between genera- lands to those who farmed them. The Nuni tions and between men and women in the chief distributed lands to the incoming same lineage have also been documented. Mossi through their Chief (the first Mossi set- Women’s tenure security is often very weak. tler). By 1986, the population of Nébourou For example, in the village of Sikore (Mali) was 74% Mossi. When an NGO presented a daughters receive only one-third of their proposal to establish a village woodlot, the father’s inheritance while sons receive two- community council agreed. When it came thirds. Even with a third share, it is unlikely time to plant the trees, however, the Mossi that land will be passed on to women. One wanted to plant theirs separately. Because village Imam even attempted to seize the tree planting is considered sufficient to vali- land inheritance of a deceased man who was date rightful ‘ownership’ of land, this was survived only by a wife and daughter, which not acceptable to the Nuni, who did not want led to a conflict in which the whole village the Mossi to establish a claim to their land. became involved (Maïga, 1996). In addition, Although the confrontation was diffused the natural resource rights of pastoral women through informal mechanisms, tenure secu- are rarely considered (Pointing, 1995). rity was never established for either Nuni or Young men, eager to earn money desire Mossi and the latent conflict over land rights access to land for cash crops and resist still exists. attempts by outsiders taking what they feel 10 K.M. Moore et al.

is, by right, their land. Within the Nuni com- multidisciplinary approach integrates the munity of Nébourou (Laurent and Mathieu, fields of sociology, economics, agriculture, 1994), the young men were the most con- natural resources and animal science to pro- cerned about the continued distribution of vide decision-making tools for community- lands to incoming Mossi, because they felt based management. that they were being dispossessed on behalf Local participation is crucial to the suc- of people outside of the family. cess of NRM projects of USAID, the World Bank, and numerous NGOs throughout Sahelian West Africa. Not simply participa- The SANREM CRSP Response tion, but the type and form of local partici- pation are important for project success. The SANREM CRSP West Africa Project This participation implies more than a designed and implemented a programme to simple presence in a situation or relation- develop and test an approach to address ship. An active commitment by the ‘partici- issues surrounding decentralization, con- pant’ is necessary for the benefits of flict, and NRM. The driving force behind the participation to be assured. This commit- approach is the need to find long-term solu- ment is responsible for all successes using tions to complex natural resource manage- the participatory approach. Commitment ment problems. Although in dealing with can be seen when the local population spec- conflict over natural resources, it is impor- ifies the needs to be addressed, defines tant to implement short term conflict reso- the questions to ask, and sets the agenda lution/management strategies, the approach to follow. These needs, questions, and also focuses on long-term consensus build- actions must be locally meaningful; that is, ing and provision of social infrastructure as they must correspond to circumstances as a platform for change and improved agricul- they are perceived within villagers’ daily tural and natural resource decision-making. lives. The same tools being used to resolve con- Interdisciplinarity draws the research flict and to develop a set of organizational process into a holistic dialogue across disci- skills (human capital investment) for con- plinary boundaries. Without taking into sensus building and local governance are account the sum of all that influences and being used simultaneously to establish a shapes humankind’s relationship with its new form of social infrastructure (social environment, improved management is capital investment) to bridge relationships, unlikely to occur without setting off a chain build trust between local communities and of unrecognized and unintended conse- link these communities with scientific quences. Working in disciplinary isolation, research services to identify and introduce focusing on a single production system, technologies that increase the productive emphasizing one resource over another, or capacity of the natural resource base. The assisting one ethnic group or village intro- working hypothesis of SANREM in Mali has duces a bias that, if left unchecked, leads to been: When a local population is provided overall system imbalance. with: (i) methods for natural resource and However, the social and biophysical conflict management, and (ii) an institu- environment contains an infinite number of tional vehicle for inter-village, inter-ethnic aspects, some perhaps more critical than dialogue, the population can then become others, depending on timeframe and situa- proactive in addressing major agricultural tional interdependencies. Scientists or productivity and natural resource manage- research managers are forced to choose, to ment issues. isolate a single aspect in order to fully grasp Globally, the SANREM CRSP approach it and manipulate it for the purpose of is based on a holistic view of NRM involv- improving livelihoods. Disaggregating ele- ing four pillars: participation; interdiscipli- ments from the whole to isolate them for rig- narity; multi-stakeholder involvement; and orous testing and investigation is the landscape/lifescape scales. This holistic, hallmark of the advances provided by Conflict and Agropastoral Development in the Sahel 11

modern science. Unfortunately, it also leads aggregate local (terroir/village) level partic- to monoculture production and the ipation into commune-wide participation. attendant weaknesses of an unbalanced In addition to trial and error learning, landscape. substantial effort has been necessary to This tension between holism and par- determine the most feasible mechanisms for ticularity has pervaded implementation of aggregating village priorities and concerns. the SANREM approach. It is characteristic Local knowledge and practice, in fact, pro- of multi-institutional, multi-disciplinary vide one of the most reliable ways to iden- teams; and it is also reflected in community tify mechanisms for this aggregation. relations. Any process that involves multiple stakeholders must address how benefits can The SANREM-West Africa Approach be optimized for the group as a whole. Stakeholder involvement implies not only The SANREM approach to improving participation, but also the recognition that natural resource and conflict manage- not all participants have the same goals, or ment at the commune level is composed of the same power to achieve them. Successful five steps: negotiation of a shared vision involves pool- 1. Ascertain local perceptions and priority ing of prior knowledge and reasoning needs; processes and is predicated on a collective 2. Build commune consensus and estab- need or desire for results by and for the lish local management capacity through group. Confronting the reality of opposing NRMACs; stakeholder interests is required if the sense 3. Build institutional capacity for impact: of ownership in a shared vision is to be cre- •Training in literacy and governance; ated and the plan embedded in the social •Training in natural resource management consciousness. and conflict management; SANREM CRSP captures this holistic 4. Complement local knowledge with bio- element by building on landscape concepts physical and socioeconomic research and drawn from watershed management and development that lead to technologies and completing the metaphor with a new term, decision-making tools; lifescape. Landscapes are constructed reali- 5. Monitor and evaluate. ties, the mirror image of which is people con- structing lifescapes for their livelihoods. In These steps are summarized below. the global lexicon of SANREM, working at the landscape/lifescape scale means going beyond what has come to be known as ges- Step 1: Ascertain local perceptions and tion de terroir in the West African Sahel. As priority needs the landscape/lifescape scale increases, a wider range of stakeholders, often having no Implementing the SANREM process is pred- direct interest in specific local ecosystems icated on the interests of local partners and become involved. SANREM research focuses local government. Further, local perceptions on understanding both the complex bio- of the constraints and potentials of the com- physical and social processes within and munity’s landscape and lifescape must be across ecosystems. understood. Generally, perceptions vary For small groups, it is easy to gather in within communities and among stakehold- one location and quickly take these steps at ers – farmers, pastoralists, village associa- one or two meetings. However, this is not tions, NGOs and technical service providers. possible in the context of decentralized The best way to reveal and document these administrative units, because the popula- perceptions is with a rapid participatory tion of these communities can range from survey at a multi-village level, such 2000 to 15,000 individuals. In this case, a as the Participatory Landscape/Lifescape standardized methodology is necessary to Appraisal (PLLA). 12 K.M. Moore et al. SANREM West Africa process SANREM West flowchart. ig. 1.1. F Conflict and Agropastoral Development in the Sahel 13

The PLLA is based on the proposition • disseminating awareness of these activi- that any successful NRM project must be ties and any recommendations resulting grounded in a balanced and thorough appre- from them; ciation for the biophysical and socioeco- 2. To work with farmers, pastoralists, nomic milieu of the target community. The local chiefs, and other stakeholders to PLLA forms multi-disciplinary teams of avoid and manage disputes over natural researchers and local stakeholders to exam- resources; ine the natural resource base and the socioe- 3. To act as a liaison between the local pop- conomic realities of the community. Rapid ulation and the commune’s mayor and other Rural Appraisal techniques identify key government authorities regarding natural biophysical and socioeconomic constraints resource issues; and and opportunities. Groups and individuals 4. To provide training and act as mentors from all strata of the local population for the development of holistic management become involved and a representative pic- and conflict management strategies in the ture of the Commune becomes apparent commune’s villages. during the course of the week-long exercise. Great care was taken to ensure that all Particular attention is paid to institutions social strata in the commune were equitably and traditions that provide an interface represented on the advisory committee, par- between the landscape and the lifescape, ticularly in terms of socio-professional because management decisions affecting groups, ethnicity and gender. Because per- both landscape and lifescape are made ceived constraints and solutions vary con- through them. The PLLA also provides the siderably, any excluded group will not opportunity to inform the local population attach much legitimacy to decisions or about the nature of the SANREM interven- advice coming from the committees. In tion and helps form realistic expectations Madiama, 18 representatives were elected to within the community about what collabo- the commune level NRMAC from among ration may mean in the future. representatives of each village-level user The PLLA in Madiama Commune took group. Each village committee has at least place in February of 1999 and identified one representative. One-third of the com- three major NRM constraints: the poor and mune level committee members are women. degrading condition of soil fertility, pasture, Although SANREM encourages diversity, and water points for livestock. These con- the village committees are less diverse straints formed the basis for prioritizing the because the villages themselves are less het- activities that became the work plans and erogeneous in terms of ethnicity, socio- research programmes prepared by the part- professional groups, and in their views of ners, including representatives from the gender roles. local population. The NRMAC can only function well if village chiefs, the mayor and other authori- ties are fully informed, provide support, Step 2: Build commune consensus and and understand that their own role in establish local management capacity through the local hierarchy will remain secure NRMACs and undiminished by collaboration. In Madiama, the NRMAC has established good The roles of the NRMACs are: relations with the mayor who regularly 1. To collaborate in: participates in meetings, often officially • drawing up each year’s work plan; opening and closing them. The chef d’ar- • choosing farmers and pastoralists to par- rondissement has also attended. Both have ticipate in field tests; been outspoken in encouraging committee • monitoring and evaluating the results of activities. field tests and progress towards fulfilling NRMAC sustainability is an important work plan objectives; issue. Sustainability in this context is 14 K.M. Moore et al.

dependent on the participants acquiring the planning and lobbying. These workshops training and skills necessary to manage were designed to create a framework of everyday problems and on the capacity to exchange and dialogue between the access and interact with deconcentrated NRMAC and the other actors implicated technical service providers. The ever-chang- directly or indirectly in NRM. The NRMAC ing landscape and lifescape of a commune also benefited from the support of CARE means that new and difficult challenges will in conducting inter-village negotiations for arise in the future. The NRMAC has been the establishment of agreements for wet- designed to build social capital within lands management. Additional institutional Madiama to meet these challenges, whether coaching included assistance in developing through the solidarity of horizontal bonds statutes and rules of procedure, formal between villages and ethnic groups, or registration of the organization, financial through bridges connecting the community management, the mobilization of the exter- to service providers. nal resources, and the development of link- ages with other local service providers and NGOs. Step 3: Build institutional capacity for impact Training in natural resource management and conflict management Training in literacy, governance, and financial Holistic management (HM) concerns itself management specifically with the degradation of nat- To be effective, NRMAC members require ural resources, including how degradation significant training in organization, democ- occurs, and how to reverse it. As such, it ratic organizational procedures, literacy and is very appropriate to the West African numeracy skills. The value of literacy and Sahel. numeracy training is that it gives the HM as an approach to natural resource NRMAC the capacity and confidence to act management was first conceived in south- autonomously – gathering information, ern Africa and has been applied throughout writing plans and contacting government the world. It helps to foster skills necessary authorities. Training is an essential element to maintain purposeful dialogue between in the processes of decentralization and stakeholders. HM introduces visioning tech- democratization. niques to help channel dialogue towards Institutional strengthening was pro- consensus, and conflict resolution skills to vided to the committee by the Djenné office help negotiate and manage the inevitable of CARE/Mali through formal training conflicts. HM training and implementation workshops, informal exchanges, monitor- in Madiama concentrated initially on the ing and technical assistance and study contentious issue of bourgou2 management tours. The institutional reinforcement train- and is moving on with the help of the ing programme was initiated with an NRMAC to more common dryland pasture institutional diagnosis that identified the issues (see Chapter 13). strengths and weaknesses of the organiza- Conflict management training provides tion. Focusing on the identified weak- techniques for managing conflict within the nesses, a plan for institutional development contexts of scarcity, diversity and unequal was elaborated. Thus, the formal training power distributions. The training is well programme applied functional literacy adapted to conflict avoidance and manage- methodologies and focused on the princi- ment where natural resources are con- ples of democratic governance and assured cerned. NRMAC members have been very comprehension of national texts, laws and pleased with conflict management training codes for NRM and decentralization. and assert that it has had the most impact of Subsequent training addressed financial all early SANREM interventions, because management and accounting, strategic there are already fewer conflicts. Conflicts Conflict and Agropastoral Development in the Sahel 15

have been avoided or managed using the tative cover and agricultural land uses techniques learned during the training (Chapter 5). (Chapter 16). Conflict management tools The dialogue with farmers builds on the used successfully help legitimize the farmers and herders own considerable NRMAC and the SANREM programme, par- knowledge of the land and on prescriptions ticularly when combined with agricultural regarding crops and NRM techniques and NRM interventions planned in collabo- suggested by the modelling. Researchers cor- ration with the NRMAC and jointly imple- relate their scientific soil classes training mented with local partners. with traditional soil concepts farmers have devised, and enter into discussion with farmers using the local terminology. Thro- Step 4: Complementing local knowledge: ugh this dialogue, researchers confront the biophysical and socioeconomic modelling local reality and farmers gain access to the researchers’ knowledge of alternatives and the benefits they can bring. Chapter 10 Science-based information and tools for provides a complete discussion. decision makers at the commune level SANREM’s science-based tools include bio- Socioeconomic modelling physical and socioeconomic models that can be used to generalize the experience of SANREM researchers have developed a one commune to other, broader geographi- social accounting matrix (SAM) for the com- cal areas. These models are described in mune of Madiama. The SAM is a simple Chapter 11. In Madiama, SANREM has model that traces the sources of income and applied CROPSYST, a cropping systems demonstrates how that income is distrib- model that predicts the influence of cli- uted among households. Consequently, it is matic, soil and management practices on a good tool for showing how economic productivity and sustainability. CROPSYST growth affects different groups within a serves three purposes: commune. A SAM model can be developed at any •To simulate results with variables such level, from a village to a nation with the as yield, soil fertility and soil erosion number of sectors it contains depending on over a period of 20 years, allowing for the complexity of the economy being studied long-term evaluation and comparison and the available data. Four different types of of current and alternative farming and households were identified according to NRM technologies without long-term socio-professional status by community field tests; leaders as being most important for under- • to identify crops and technologies best standing the commune: farmers, whose suited for ecosystems in the region; and income is mainly from crops; agropastoral- • to provide a basis for dialogue with the ists, who have a significant amount of live- NRMAC, local farmers, and herders. stock as well as crops, sedentary pastoralists; The biophysical models require rainfall whoraisecropsbutregardlivestockraisingas data, obtained from official meteorological theirprimaryoccupation;andtranshumants, statistics and supplemented with rain who move their livestock seasonally to grass- gauges placed by the project and monitored lands in or outside of the commune along tra- by local stakeholders. The models also use ditional routes. Major findings illustrated soil data collected by a cost-effective alter- how benefits from increased production in native to classic intensive soil surveys (see any activity are passed on, through repeated Chapter 3) that combines computer-based cycles of spending and income, to benefit remote sensing analysis and Geographic other production activities and the incomes Information System (GIS) technologies with of the other household types. In Madiama, it limited field assessment and ground tru- was found that an increase in production thing to classify and map soil types, vege- in any sector, including livestock, benefits 16 K.M. Moore et al.

transhumants far less than the other three Evidence from the commune indicates that socio-professional groups. Thus interven- since SANREM began its activities, conflicts tions must take particular care to target trans- have been reduced as the community has humants or they will tend to fall further begun to work together on priority problems. behind economically, a situation that could While still a work-in-progress, SANREM has contribute to greater conflict. The Madiama developed a replicable process that will lead SAM results are relevant to neighbouring to successful locally controlled participatory communes with similar socioeconomic decision-making and the decentralization of groupsandnaturalresourcebases.TheSAMs development services and community initia- are discussed in Chapter 12. tives to improve resource management throughout the region. The contributors hope that this book will stimulate further Step 5: Monitor and evaluate work and development towards this impor- tant end. The broad-based SANREM strategy requires The first part provides an overview of an equally broad-based monitoring and the landscape and lifescape in which the evaluation (M&E) system to assess progress, SANREM project intervened. and to identify and correct problems. Much In Chapter 2, Cissé et al. introduce of the M&E system is participatory, with the social and historical context shaping the indicators and measurements identified lifescape of the Inland Delta of the Niger, the and followed by the local participants. To core of Mali’s 5th Region. Various ethnic take advantage of local learning and apply groups have co-existed in this area for cen- it horizontally across communes and verti- turies, generating livelihoods with comple- cally to the regional level, the universal mentary systems of production. The chapter language of science must be employed. discusses how recent changes in agricul- While local definitions of resources, indica- tural and pastoral production systems have tors and results are indispensable for local unbalanced this symbiosis and increased applications, on their own they may not be competition for scarce resources, thereby widely applicable. The fundamental strat- leading to land tenure confrontations that egy SANREM uses to apply results to con- are not as easily resolved as they were in texts beyond the commune is to ensure that the past. In addition, decentralization and biophysical and socioeconomic researchers democratization have complicated the situ- work with those variables and categories ation in which the modern state and civil that local populations understand and are society have been superimposed on the familiar with. This choice of variables and modified, but not displaced, customary gov- categories is the responsibility of the ernance systems. The discussion concludes, researchers so as not to over-burden local with some qualifications, that NGO-driven participants with data gathering they do not opportunities for local dialogue and prob- understand or for which they have no prac- lem solving hold considerable promise for tical use. rural Mali. In Chapter 3, Badini and Dioni present a detailed description of the landscape types Overview of the Book and soils of the Commune of Madiama. Combining knowledge gained from informal This book describes the process and results surveys, field observations, biophysical of the 5-year SANREM Phase II Program in monitoring, transects, remote sensing and West Africa, detailing its approach, compo- pit holes, the chapter characterizes soil nents and outcomes. Working with our part- types and distributions, climate and hydrol- ners in Madiama, SANREM has succeeded in ogy, cropping patterns, land use systems developing an approach that leads to sus- and potentials. The database on these land- tainable social and economic development scape types, their location, potentials and while minimizing NRM related conflicts. constraints is at an appropriate scale for use Conflict and Agropastoral Development in the Sahel 17

by village, commune and regional level the desire of the communities for improved planners, as well as for providing input to soil fertility and pasturelands in order to biophysical models to evaluate technologies increase productivity and incomes, and the (as in Chapter 11). underlying concern about conflict. In Chapter 4, Ballo and Ouattara In Chapter 7, Moore et al. describe describe the systems of animal husbandry the development of the commune-level within and around Madiama, thereby broad- NRMAC providing the social infrastructure ening the perspective of the landscape/ that is both adapted to the exigencies of lifescape scale. The chapter characterizes recent governmental decentralization and each of three livestock management types compatible with customary governance (sedentary, semi-transhumant and transhu- structures at the village level. They argue mant) involving cattle, oxen, milk cows, that it is not sufficient to simply assemble a sheep and goats. Pastoralists are either tran- group of men and women representing vari- shumant or resident, and often tend to the ous ethnic groups and occupational cate- herds of local farmers much of the year. The gories, but it is also necessary to develop pastoral resources available to the commune each member’s individual capacities (func- are limited and though traditional grazing of tional literacy, leadership skills, association crop residues and fertilization of soil exists, management, knowledge of codes and laws, there is increasing loss of organic matter. The etc.) in order to stimulate mutual trust and chapter concludes that under Mali’s new network building (social capital formation) Pastoral Code, conditions could improve, between villages and clans and to help them but improvement will require the concerted to define their mission. Of particular impor- efforts of agriculturalists and pastoralists. tance is training in conflict management to In Chapter 5, Wynne et al. combine build individual self-confidence and to pro- remote sensing and ground truthing in the vide a credible and valued service in the analysis of land use change during a 50 year eyes of villagers. period (1952–2002) in the Commune of In Chapter 8, Goebel et al. provide an Madiama. The chapter documents the dra- overview of an alternative conflict manage- matic shift in land use from pastoral to crop- ment approach and discuss how it differs based production systems. from common approaches to conflict resolu- The second part of the book describes tion. The approach is based on building con- the elements of SANREM West Africa’s flict management and consensus building multi-pronged approach to development skills rather than simple conflict manage- intervention. The intention is not to provide ment, per se. The training programme con- a history per se, but to highlight the essen- sists of a series of workshops focused on tial and potentially reproducible elements building skills and empowering local lead- of the interaction between researcher and ers by their learning a sustainable process community. Each chapter in Part II presents for facilitation and management of diverse components modelling the conduct of a suc- conflict situations. This process has been cessful community-mobilizing develop- central to the evolution of the NRMAC from ment intervention. a group of village representatives to a com- In Chapter 6, Earl and Kodio describe mittee with a Commune-level mission to how the SANREM team introduced them- improve NRM in the face of resource selves to members of the Madiama commu- competition. nity through the Participatory Landscape/ In Chapter 9, Bingham introduces the Lifescape Appraisal (PLLA), an informal set Holistic Management™ (HM) Model, an of participatory survey methodologies that approach developed specifically for open bring out not only important information range animal husbandry, but well adapted to about the natural resources of Madiama and provide intuitive analyses, insights and the production systems shaping the popula- decision-making information for commu- tion’s livelihood, but also its primary con- nity-based management of natural resources. cerns and priorities. In particular, they note Consistent with the SANREM approach, this 18 K.M. Moore et al.

holistic methodology was introduced to sci- suggest that corralling livestock in the fields, entists and community members in order to spreading manure and micro-dosing with facilitate diagnostics of and behavioural chemical fertilizer are adoptable within the change in the management of the natural range of farmers’ risk preferences. resource base in the Commune of Madiama. In Chapter 13, Brewster et al. model the He describes how HM has been a source of economic linkages between different groups tension between scientists and HM promot- of natural resource users and analyse the ers, but has also led to innovative attempts at effect of the potential growth strategies. changing resource management within the Using data from the PLLA (Chapter 6) and community. an extensive household and enterprise level Following on these capacity building data set, a Social Accounting Matrix (SAM) interventions, the third part presents the model for the Commune of Madiama was results of various research activities involv- developed. Application of this model ing the development and testing of technolo- demonstrated that the largest impacts from gies and decision-making tools appropriate exogenous changes in demand are associ- for the community and environment of ated with rice and livestock sectors. Madiama. Furthermore, impacts are shared differen- In Chapter 10, Crane and Traoré com- tially among socioeconomic groups with pare indigenous perceptions and models of agropastoralists benefiting most and the soil fertility management with those of transhumant group benefiting least. modern soil science. The differences In Chapter 14, El Hadj et al. describe between these perspectives are discussed research in response to a request by Peul vil- along with the cultivators’ perceptions of lage women to deal with Cassia tora, an the scientific research conducted in invasive plant that is unpalatable for grazing Madiama. The authors argue for implicating animals. As pasture lands have degraded, end users in technology development from Cassia tora has spread throughout their the beginning so that folk knowledge can fields out-competing other plants. However, guide scientific research and research find- this noxious plant has potential as dry ings can be better communicated to farmers season forage when ensiled. The authors to increase productivity. describe their analysis of the forage poten- In Chapter 11, Badini et al. demonstrate tial and their work with village women to the use of a combination of results of data develop ensilage techniques adapted to collected from 3 years of on-farm field trials local conditions. and long term simulation with CROPSYST In Chapter 15, Abaye et al. investigated modelling to evaluate crop rotations, and the regenerative potential of pastureland in organic and inorganic fertilizer practices for two villages through a controlled experi- yield efficiency, stability and soil sustain- ment with tethered grazing of small rumi- ability over a simulated 30-year period. nants. This work builds on the Holistic Organic fertilization produced the best Management™ (Chapter 9) insight that results in the analysis and the chapter con- animal impact is not simply a function of cludes by recommending increased efforts numbers of livestock or duration of grazing in the development of strategies to produce time in order to provide management indi- and apply more manure and other organic cators that optimize the potential of forage fertilizers. regeneration/biomass production rates, In Chapter 12, Wyeth et al. pursue the plant biodiversity, and animal performance. issue of sustainability of technologies for The chapter concludes that grazing vegeta- enhancing soil fertility from economic and tion down to a 3 cm height on any particular financial perspectives. This analysis takes parcel is likely to limit forage regeneration. into account the results of 3 years of on-farm In Chapter 16, Moore et al. address trials and combines those findings with the the issue of social capital as a factor of output of the computer modelling analyses development within Malian civil society. of Badini et al. (Chapter 11). Their results Social capital is investigated with a focus on Conflict and Agropastoral Development in the Sahel 19

embedded and autonomous social relations Projet Régional d’Appui au Secteur at the commune and village levels. Using de l’Elevage Transhumant (PRASET), household survey data from the Commune Ouagadougou, Burkina Faso, pp. 1–42. of Madiama, they investigate the extent Benjaminsen, T. (1997) Natural resource manage- to which the NRMAC provides a social ment, paradigm shifts, and the decentraliza- tion reform in Mali. Human Ecology 25(1), infrastructure on which to build commu- 121–141. nity-level social capital. The analysis demo- Blench, R. (1996) Aspects of resource conflict nstrates the importance of building on in semi-arid Africa. Natural Resource Pers- traditionally valued social relationships and pectives 16. Overseas Development In- combining them with linkages across groups stitute. http://www.odi.org.uk/nrp/16.html for the management of conflict situations. (accessed 1 May 2004) They conclude that the NRMAC does indeed Blundo, G. (1996) Gérer les conflits au Sénégal: le provide a platform for building inter-ethnic rôle de l’administration locale dans le sud- and multi-village social capital. est du bassin arachidier. In: Mathieu, P., In Chapter 17, Moore et al. review how Laurent, P. and Willame, J. (eds) Démocratie, enjeux fonciers et pratiques the four pillars of SANREM (participation, locales en Afrique. Cahiers Africain landscape scale, multiple stakeholders and 23–24, Institut africain/L’Harmattan, Paris, interdisciplinarity) were applied in the pp. 101–119. Commune of Madiama. In the list of lessons Brown, M. (1993) Non-governmental organiza- learned, they highlight the difficulties tions and natural resources management involved in establishing and maintaining in Africa’s pastoral sector: Where to go full participation of and communication from here? United States Agency of between all stakeholders in the context of International Development, The PVO- power relations and traditionally excluded NGO/NRMS Project, Washington, DC, groups. Building social capital and co-man- pp. 1–76. Cissé, S. (1996) Problématique foncière et litiges agement agreements is a long term and iter- fonciers en 5e région. In: Maïga, I. and Diallo, ative process requiring that project and G. Atelier de restitution et de reflexion sur les government development agents be well- litiges fonciers dans la région de Mopti. trained and integrated within the commu- Groupe de Recherche-Actions pour la nity in order to empower the population to Développement/Institut International pour act on its own behalf. l’Environnement et le Développement, Bamako, Mali, pp. 24–26. Dasgupta, P. and Serageldin, I. (2000) Social Notes Capital: a Multifaceted Perspective. World Bank, Washington, DC. Djalebord Diard, A. (1992) Le stratégies et le rôle 1 The terroir is a biophysically and socio-cul- du CILSS dans la sous-région. In: La zone turally defined space for production and living Agro-Sylvo-Pastoral au Sahel : Limites et generally associated with a village. Gestion Potentialités de Développement contre la der terroir is the development approach Désertification. Actes de la Conference de emphasizing community participation associ- Assissi. Comité Inter-Etats de Lutte contre la ated with this space. Sechéresse au Sahel, pp. 25–30. 2 Echinoloa stagnina or bourgou in French is a Grootaert, C. and van Bastelaer, T. (2002) The wetland grass that grows with the rising river Role of Social Capital in Development: waters during the rainy season. It is prized as an Empirical Assessment. Cambridge Uni- a very nutritious animal feed. versity Press, Cambridge, UK. Horowitz, M.M. and Little, P.D. (1987) African Pastoralism and Poverty: Some Implications References for Drought and Famine. Cambridge University Press, Cambridge, UK. Bary, H. (1996) Etude sur la gestion des conflits Isham, J., Kelly, T. and Ramaswamy, S. (2002) liés aux aménagements pastoraux: cas de la Social Capital and Economic Development: zone pastorale de Sidéradougou. Ministère Well-Being in Developing Countries. Edward de l’Agriculture et des Ressources Animales, Elgar Publishing, Limited, Cheltenham, UK. 20 K.M. Moore et al.

Kane, O. and Winter, M. (1997) La gestion décen- KAP study (Summary Assistance Report tralisée des ressources naturelles au Sénégal. No. 9). Ministry of Environment and Pro- In: La Gestion Décentralisée des Ressources tection of Nature, Community-Based Natural Naturelles dans Trois Pays du Sahel: Resources Management Project, Dakar, Sénégal, Mali et Burkina Faso. Le Projet Sénégal, pp. 1–6. d’Appui au Développement Local au Sahel Ngaido, T. (1996) Redefining the boundaries of (PADLOS). Ouagadougou, Burkina Faso, pp. control: post-colonial tenure policies and 13–131. dynamics of social and tenure change in Laurent, P., and Mathieu, P. (1994) Authority and Western Niger. Ph.D. dissertation, University conflict in management of natural resources of Wisconsin, Madison, Wisconsin. – a story about trees and immigrants in Niamir, M. (1991) Traditional African Range southern Burkina Faso. Forests, Trees and Management Techniques: Implications People 25, 37–44. for Rangeland Development.(Pastoral Lo, H., Sylla, C., Ndione, P. and Agne, A. (1996) Development Network Paper No. 31d). Ressources forestières et conflits au Sénégal: Overseas Development Institute. London, quelles approches pour une typologie? UK. http://www.odi.org.uk/pdn/papers/ Bulletin Arbres, Forêts et Communautés paper31d.html (accessed 1 May 2004) Rurales 8, 26–36. Okai, M. (1987) Agricultural production, food Lowenberg-DeBoer, J., Kaboré, D. and Abdoulaye, security and poverty in West Africa. In: T. (1994) The opportunity cost of capital for Asenso-Okyere, W.K., Benneh, G. and Tims, agriculture in the Sahel: Case study evidence W. (eds) Sustainable Food Security in West from Niger and Burkina Faso (Staff Paper Africa. Kluwer Academic Publishers, 94–2). Purdue University, Department of Boston, Massachusetts. Agricultural Economics, West Lafayette, Painter, T., Sumberg, J. and Price, T. (1994) Your Indiana. terroir and my ‘action space’: Implications of Lynam, J.K. and Blackie, M.J. (1994) Building differentiation, mobility and diversification effective agricultural research capacity: The for the approche terroir in Sahelian West African challenge. In: Anderson, J. (ed.) Africa. Africa 64(4), 447–463. Agricultural Technology: Policy Issues for Park, T., Tabor, J., Frankenberger, T., Lynham, M., the International Community. CAB Baro, M., Ngaido, T. and Rogers, G. (1993) International, Wallingford, UK, pp. 106–134. Risk and Tenure in Arid Lands: the Political Maïga, I. (1996) Etude sur la prévention et la ges- Ecology of Development in the Senegal River tion des conflits. Centre Canadien d’Etude et Basin. University of Arizona Press, London, de Coopération Internationale (CECI), Prog- UK. ramme de Promotion de la Démocratie et des Pointing, J. (1995) Impact of social and economic Droits de la Personne (PDDP), Bamako, Mali, change on pastoral women in East and West pp. 1–55. Africa. In: Stiles, D. (ed.) Social Aspects of Maïga, I. and Diallo, G. (1998) Les conflits Sustainable Dryland Management (United fanciers et leurs gestion dans la 5ème Région Nations Environment Program). John Wiley du Mali (Dossier No. 76). Groupe de & Sons, Chichester, UK, pp. 253–265. Recherche Actions pour le Développement. Pratt, D.J., Le Gall, F. and de Haan, C. (1997) International Institute for Environment and Investing in Pastoralism: Sustainable Development, Programme Zones Arides, Natural Resource Use in Arid Africa and London, UK. the Middle East (World Bank Technical McIntire, J., Bourzat, D. and Pingali, P. (1992) Paper No. 365). The World Bank, Crop–Livestock Interaction in Sub-Saharan Washington, DC. Africa. World Bank Regional and Sectoral Preston, T.R. and Leng, R.A. (1994) Agricultural Studies, Washington, DC. technology transfer: Perspectives and case Messer, E., Cohen, M.J. and D’Costa, J.D. (1998) studies involving livestock. In: Anderson, J. Food from Peace: Breaking the Links (ed.) Agricultural Technology: Policy Issues between Conflict and Hunger (Food, for the International Community. CAB Agriculture and the Environment Discussion International, Wallingford, UK, pp. 267–286. Paper 24). International Food Policy Pretty, J. and Buck, L. (2002) Social capital and Research Institute, Washington, DC, pp. social learning in the process of natural 1–44. resource management. In: Barrett, C.B., Moore, K.M. (1996) Status report on CBNRM Place, F. and Aboud, A.A. (eds) Natural Project Monitoring System and NRM Resources Management in African Conflict and Agropastoral Development in the Sahel 21

Agriculture. CAB International, Wallingford, Sommer, F. (1998) Pastoralism, drought, early UK, pp. 23–33. warning and response. MA thesis. Rochette, R.M. (1989) Le Sahel en lutte contre University of Sussex, Brighton, UK. la désertification: leçons d’expérience. Souley, A. (1996) Rapport d’Etude: Prévention et Gesellschaft für TechnischeZusammenarbeit Gestion des Conflits au Niger. Centre (GTZ), Weikersheim, Germany. pp. 1–59. Canadien d’Etude et de Coopération Sanders, J.H., Shapiro, B.I. and Ramaswamy, S. Internationale, Niamey, Niger, pp. 1–52. (1996) The Economics of Agricultural Soumaré, S. (1996) Pastoral land tenure: the Technology in Semiarid Sub-Saharan exploitation of natural resources against a Africa. Johns Hopkins University Press, backdrop of actors from different ethnic Baltimore, Maryland. groups, diverse production systems and Sanon, Y. (1996) Etudes sur la gestion des conflits frontier cohabitation. In: Managing Land liés aux aménagements pastoraux: cas de Tenure and Resource Access in West Africa. la zone pastorale Samorogouan. Ministère Groupe de recherche et d’échange tech- de l’Agriculture et des Ressources Animales, nologiques (GRET)/IIEE, Goré, Sénégal, pp. Projet Régional d’Appui au Secteur de 173–178. l’Elevage Transhumant (PRASET), Ouaga- Toulmin, C. and Moorhead, R. (1993). Local dougou, Burkina Faso, pp. 1– 36. strength and global weakness: NGO experi- Shanmugaratnam, N., Vedeld, T., Mossige, A. and ence with pastoralists in Africa. Paper pre- Bovin, M. (1992) Resource Management and pared for the PVO.NGO/NRMS-IIED-CCA. Pastoral Institution Building in the West ONG Workshop on NGOs and Natural African Sahel (World Bank Discussion Resource Management in the Pastoral Sector Papers Africa Technical Department Series of Africa – Strategies for Enhancing Per- No. 175). World Bank, Washington, DC. formance and Impact. 22–28 February 1993, Sidikou, H.A. (1995) La gestion alternative des Mopti, Mali. conflits liés à la gestion des ressources Touré, D. (1996) Prévention et Gestion des naturelles en zone pastorale: brève réflexion Conflits. Centre Canadien d’Etude et de inspirée pas le cas de la reserve de Gadabeji Coopération Internationale, Ouagadougou, (Nord Dakoro). Paper prepared for Atelier Burkina Faso, pp. 1–50. national sur la gestion alternative des con- Vedeld, T. (1992) Local institution building and flits liés à la gestion des ressources resource management in West African Sahel. naturelles, Niamey, Niger, pp. 1–8. Forum for Development Studies 1, 23–50.

2 The Lifescape: Production Systems and Social Institutions

Salmana Cissé,1 Keith M. Moore2 and Charlene Brewster2 1Centre Régional de la Recherche Agronomique, Institut d’Economie Rurale, BP 205, Mopti, Mali and 2Office of Intenational Research, Education and Development, Virginia Tech, Blacksburg, VA 24061–0378, USA

In this chapter, we outline the social and Historical Development of Productive institutional context shaping the lifescape and Ethnic Relations of the Inland Delta of the Niger, the core of Mali’s 5th Region. This historical analysis Three general production systems have of the region reveals the roots of many of characterized the Sahel: agro-sylvo, agro- today’s conflicts across the Sahel. The sylvo-pastoral, and sylvo-pastoral (Sunzini, lifescape of the Inland Delta can be viewed 1992). The Inland Delta of the Niger has as an ethnic melting pot, whether in the a fourth, aquaculture. Historically, the inundated or rainfed zones, nested within numerous ethnic groups in the region complementary and conflictual production (Bambara, Peul, Bozo, Bobo, Malinké, systems. The rules that regulate these pro- Marka, Dogon, etc.) are principally associ- duction systems are organized into compet- ated with different principle production ing customary local and modern national systems, based on millet, rice, livestock, governance systems. This arrangement fish, etc. These systems, however, are in complicates the livelihoods of the popula- transition as herders, farmers and fisherfolk tion. The social institutions governing seek out new means of survival in a chang- ethnic and production relations have ing ecology. Herders and fisherfolk are evolved over time with regard to the natural becoming sedentary; agro-pastoralists are resource base (see Chapters 3, 4 and 5 for a increasingly using trees for animal feed and discussion of the landscape), and have soil fertilization; and farmers are raising become ineffective for managing contempo- more livestock. Today, ethnic flexibility rary technological and population exigen- with respect to production practices can be cies. Consequently there is an impasse in found in all production systems of the two natural resource management (NRM) rule zones of the 5th Region. An individual can administration. This chapter describes the be a farmer, herder or other regardless of eth- historical development of social and eco- nicity. In addition, significant and increas- nomic systems and relations between the ing numbers of each of these groups seek State and civil society (an associational urban employment, thereby reducing the sphere intermediary between the family available labour pool and shifting manage- and the state). ment responsibilities to already overbur- dened and under-prepared women. © CAB International 2005. Conflict, Social Capital and Managing Natural Resources (ed. K.M. Moore) 23 24 S. Cissé et al.

Traditional production systems in the minimized due to the abundance of land, as Sahel have existed for centuries in various well as the spatial and temporal distribution proportions dependent on available re- of different production activities. The Dina sources and ethnic groups. Although Blench divided the delta region into 37 pastoral ter- (1996) disagrees, Watts (1987), Park et al. ritories (leydi) each under the administra- (1993), and Pratt et al. (1997) all agree that tion of a local pasture and herd master sustainable NRM was a conscious, if not (jowro), in order to maximize mobility and always explicit, practice. The production coordinate opportunistic grazing in the pas- systems were primarily extensive with a low toral management system. The leydi ratio of output to input. New fields were cul- included not only the pasture lands, but also tivated nearly every year, while large areas farmer and fisherfolk villages and their pro- were left fallow or used for animal grazing. duction spaces. Although village chiefs Herds were moved across vast expanses in a (jom) and land chiefs had control over local seasonal pattern of migration opportunisti- uses of the village terroir (the physical cally seeking fresh pastures. Fishing and and social territory attributed to a village gathering were also extensive, each utilizing or group of villages), jowros administered a relatively large natural resource base (water pasture use by regulating access rights for or forest). Describing similar conditions in different groups, scheduling use and negoti- the Senegal River Basin, Park (1993) has ating annual exchanges of pasture with pointed out that the key adaptive strategy other jowro (Crowley, 1991). across production systems was to maximize The Dina provided the minority Peul returns to labour on a selective temporal and (20% of Inland Delta population) with a spatial basis. mechanism to protect their transhumant This intersection of production systems herds, pastures and routes, maintain local in time and space led to a wide variety of dominion and regulate relations with com- social groups in the 5th Region and shaped plementary production systems. This form intra-community and inter-ethnic relations of social organization was able to persist, over the centuries (Turner, 1992). A complex because all the local inhabitants, whether and hierarchical social structure, composed Peul herders, Marka farmers, or Bozo fisher- of successive settlements of Sorongo fisher- folk, were considered part of the same patri- folk (Bozo), Nono (Marka) rice cultivators, lineal extended family or clan (suudu and Fulbe (Peul) pastoralists, was estab- baaba) in relation to other groups and in the lished in the delta region by the time of the face of various social or other adversities Mali Empire (14th century). Central author- (Crowley, 1991). This ‘family’ nature of rela- ities (from the Mali and Songhay Empires tionships between different groups of the through those of the Moroccans and same leydi was founded on political Bamanan (Bambara)) were paid tribute but accords, marriage alliances or simply on had limited authority over the mostly self- particular social relations where cohabita- governing populations of the Inland Delta. tion could produce feelings of sharing in the Local Peul families and clans tended to same social entity. dominate the region through their slave A unified Peul State dominated social cultivators (often Bozo and Marka). The and economic relationships until the successive historical waves introducing Toucouleur invasion of the mid-19th cen- each group (Bozo, Marka, Bobo, Peul, tury. The Peul loss of political power Bambara, Maures, Dogon) culminated in reduced the state’s formal administrative the Hamdallahi regime (1818–1862) that structure supporting the Dina. Successive formalized the political, economic and new states were designed with a territorially social structure and created a land tenure fixed administrative structure favouring the and pasture management code called the sedentary (farming) majority and their local ‘Dina’. leaders. Nevertheless, the Dina’s economic Under the Dina, competition for space and non-politicized social organization of between different productive activities was production relations has persisted despite Production Systems and Social Institutions 25 the abolition of slavery by the French at Crop and Livestock Production the turn of the 19th century and the attemp- Relationships ted institutionalization of a sharecropping system. Although farming and herding have diver- The scale of management for each ged as different specializations, there is a system was related to its need for natural long history of complementary interaction resources. Organizing their production on a between the two groups through livestock concentric basis within their terroirs, seden- sharing and bartering milk for grain (Sturm, tary farmers first cultivated fields around 1996). Under this system, the products and their dwellings, then in the village, and by-products of one system constituted finally in the bush. Fisherfolk migrated an important input for the other, and there along the river with the fish after the flood was coordinated action between the actors. wave each year (Ambach, 1996). Operating For example, cultivators have obtained on a larger scale, pastoralists went from animal manure and labour in return for pro- zone to zone seeking pastures for grazing viding cash, crop residues or water rights and water for their animals according to sea- (wells) to herders (Heasley and Delahanty, sonal availability (Pratt et al., 1997). 1996). Following the rainfall and searching for In the case of manure, a formally nego- fresh pasture, this transhumance shifted tiated contract has traditionally allowed between dry and humid regions. During the herders to exchange manure from their rainy season they moved out of the Delta herds as compensation in kind for cereals, into the rainfed grasslands and returned animal drinking water, clothing, etc. during the dry season as the pastures and Although the initiative for such a contract water holes dried out. It was only in the often came from the farmer who simply pro- 1980s that the Peul became a predominantly vided a well in his field for the animals, sedentary population, expanding their pas- herders in need of millet stocks or rice straw toral production system into agricultural left in the fields after harvest could also ini- production. Transhumance still occurs but tiate contracts. In this case, milk, or more now all extended families provide members rarely an animal, would be offered in with a permanent residence. exchange. In all cases, however, the size of The current land management and pas- the herd and/or the value of the pasture was toral movement is embedded in the Dina- an important element in the contract negoti- imposed social structure despite the form of ations. These relationships have decreased the governing state shifting from Toucouleur as crop farmers have increasingly come to to French colonial rule and later to national manage their own cattle herds and herders independence. Some ethnic groups, includ- become sedentary (Toulmin, 1983). ing the Marka, Bambara and Dogon, occupy In recent decades, the advent of har- and control a clearly defined territory for nessed traction for cultivation set in motion cropping and pursuing other livelihoods. a dramatic transformation in the relation- Other groups, including the Peul and Bozo, ships between rural production systems in occupy space in a more transitory fashion, the 5th Region. The passage from hand cul- with shifting campsites scattered through tivation with the daba (a short handled hoe) space and time according to the seasonal to the animal drawn plough has resolved a availability of natural resources. This space number of problems (while often creating is open and, according to the season, may others) related to bottlenecks in the labour engulf or demarcate a village space where supply. It set in motion significant modifica- relations between social groups are either tions in the organization and management of complementary or conflictual, depending field space, including encroachment culti- on the relations of power: demographic rela- vation on agricultural trails, designation of tions (age and gender), and social relations pastures reserved for working cattle and (classes), and political relations (dominant recruitment of permanent herders to clan groupings). manage this newly created herd category 26 S. Cissé et al.

(see Chapter 4). Harnessed traction has also However, it is very difficult to encounter a had an impact on the transport of agricul- pure production system that has evolved tural inputs to the fields and field products through space and time on its own. All the to the family granary or the market. systems have evolved in the same space, Although herders and farmers have tra- using the same human tools, with the same ditionally used what appeared to be two sep- general objectives to satisfy fundamental arate landscapes, livestock now occupy a human needs. Consequently, there is a his- significant place in sedentary farming sys- tory of multiple confrontations over access to tems (McIntire et al., 1992). In addition to space and labour. the need for traction animals, farmers are diversifying into livestock production to reduce the risk of crop failure. While diver- Competition for space sification in sedentary agriculture is taking place, livestock mobility – the comparable Both between and within production units, risk-averting behaviour in pastoral systems – pastoral spaces are being lost to crop pro- has been restricted as potential grazing lands duction. Even within the same production are converted to crop lands and herding fam- system, competition can be found between ilies have established permanent villages two coincidental practices. One, the con- for crop production and other livelihood frontation between production practices in needs. Such crop/livestock integration and a given space, can be partially explained by accompanying intensification of production the willingness of one social group to systems is generally viewed as a positive exclude another. For example, successive development for both productivity and envi- use of pastures by herds of different origins ronmental conservation (Mortimer and can create conflicting uses when the first Turner, 1994; Williams et al., 1995; Sturm, herd is late to leave the space and impedes 1996). However, Wolmer (1997) points out the entrance of the next. Competition, how- that such integration may not be environ- ever, involves not only spatial occupation of mentally sustainable and rejects the notion the same nature (agriculture/agriculture or of a single diversification development tra- livestock/livestock), but also of different jectory for farmers and herders. natures (agriculture/livestock). This latter In Mali, there is strong evidence that competition is more commonly perceived, users are concentrating in areas of higher because it is easily discernible in the quality resources (e.g. the Inland Delta of the destruction of crop fields by animals and the Niger), and it is apparent that competition is destruction of animal corridors by non- replacing the more complementary and herders. The immediate consequences of cooperative resource use of the past. Also this competition have typically been mani- increasing is competition between farmers fested in violent confrontation (Maïga, and sedentary or transhumant herders for 1996). crop residues and grazing areas. The tradi- tional complementary exchange of stubble grazing for improved soil fertility has disap- Competition for labour peared, while competition for woody vege- tation used as fuel or fodder has increased. A division of labour exists within house- holds, because members are typically assigned specific tasks by age and gender. As Competitive Relationships a general rule, planning and decision making are left to the senior members of the Schematically, three principal land-based household – most frequently the male production systems can be identified in the household head (Cissé, 1985). For the most zone – an agricultural system, a pastoral part, women are charged with household system and an agro-pastoral system which and reproductive tasks (food preparation, combines elements of the other two. collecting water and firewood, childbearing, Production Systems and Social Institutions 27

childcare and so on). They also tend the manent settlements and raised crops as a livestock kept around the homestead and, in survival mechanism. Traditional support particular, are charged with milking and mechanisms such as cow sharing no longer processing dairy products. In nomadic function (Bovin, 1990). Permanent settle- households, women are also often responsi- ment reduces herd size and restricts the ble for setting up and taking down their tem- economic returns from raising livestock porary household structures. Men’s tasks (Horowitz and Jowkar, 1995). As men mig- include clearing and preparing land for cul- rate in search of seasonal employment, tivation, herd management (herding, castra- women’s workload increases. tion, vaccination and slaughter), and Specialization, such as in the intensifi- digging wells (Pointing, 1995). However, as cation of milk, meat or horticultural pro- population increase puts pressure on the duction, has led to further individualization limited resource base – intensifying live- and inequities in wealth. However, another stock production, establishing agro-pastoral type of entrepreneurialism is growing in production systems, and increasing migra- West African countries: the absentee herder tion in search of employment – women or farmer, who owns livestock or a farm but appear to be taking on more of these tradi- lives in town. This category of entrepreneur tionally male tasks (Pointing, 1995). has been created by ongoing changes Family labour allocation decisions are brought on by drought and consequent eco- made between rainfed and irrigated crop nomic problems that have impoverished production activities, or between herding rural people relative to urban-based govern- and cropping activities, or between fishing ment employees, merchants, and so forth. and cropping activities. For the organization These new urban entrepreneurs hire poor of production, a double consequence can be herders, who lost their livestock during pre- noted: on one hand, the periodic appearance vious droughts and have been unable to re- of bottlenecks assures that optimal produc- stock their herds, to tend their livestock tion strategies are not followed, and, on the (often the same animals bought from the other, the use of unqualified labour and inap- herders during the crisis) in return for a propriate production implements results in salary or other in-kind arrangements. inefficient use of resources. Competition for Village solidarities have traditionally labour is not limited to intra-family alloca- mitigated against lost access to resources. tion, but can also involve an entire village or Increased competition for resources, how- group of villages. ever, is expanding the gap between rich and poor. This differentiation is growing both between, and within, villages. Resource Social differentiation tenure for village livelihoods has become highly insecure and conflicts over resources The level of household wealth determines have increased in recent years. In the next vulnerability to such disasters as drought, section, we will see how the social institu- disease epidemics and subsequent famine. tions have or have not evolved to manage Wealth may be indicated by housing materi- resource access and conflict. als, size of holding or size and composition of the herd. Poor farmers may be forced to sell off productive assets, while richer farm- Customary Authority, State and Civil ers may even gain because of such a disaster Society (Frankenberger and Lynham, 1990). The poorer the pastoralist, the more likely he is Contested terrain: modern vs. customary law to disappear as an independent herdsman after such a disaster, thereby widening intra- In recent years, Sahelian governments have ethnic disparities (see Brewster et al. in tried to reconcile the realities of customary Chapter 13). Some marginalized and impov- law at the village level with the need for erished populations have established per- standardized national laws (Bohrer and 28 S. Cissé et al.

Hobbs, 1996; Delville, 1998). Those in- to resource. Access to land for crop produc- volved in conflict resolution relating to tion is determined at the village or clan level, these laws cite either inappropriate laws or first through the lineage, then through the failure to apply them correctly (Maïga and household. Access to pastureland has Diallo, 1996; Ngaido, 1996; Touré, 1996; depended on the jowro. Leaders of the origi- Moore and Thiongane, 1998). In fact, what is nal or conquering lineage or ethnic group are typically considered the region’s land often the only people deemed capable of tenure framework is no more than a spatial ‘owning’ property (Cissé, 1985; Ngaido, mosaic of productive activities, a physical 1993). For the most part, male household manifestation of the social organization, or heads inherit the use right to the majority of ‘logiques sociales du territoire’ (Mathieu, land they farm (Cissé). Some may borrow 1996). This mosaic is rarely linked to from others in the same lineage or from out- modern land tenure status. In order to side the lineage (Matlon, 1993). Temporary secure their livelihoods, the majority of use rights may be arranged through share- actors completely ignore all legal documen- cropping or rental. Slaves may access land tation, thereby causing frequent renewed through their masters (Cissé, 1985; Ngaido, legal challenges, despite the administra- 1993). Women and other subordinate family tion’s efforts to settle these disputes. This members have traditionally had access to explains why, until recently, administration lands through the male household head. officials have had little respect for peasant Inheritance of land by women, although traditions and practices. occasionally practiced, is rare as it tends to There once might have been coherence fragment household and lineage land hold- between production practices and politico- ings (Cissé; Ngaido). Within pastoral sys- administrative systems. Resource tenure in tems, women are not authorized to own West Africa has traditionally balanced the cattle; however, childbearing may establish land needs of various households within rights in, and responsibility for, milk pro- low population production systems (Cissé, duction (Pointing, 1995). 1985; Sanders et al., 1996). Use rights to a Initially, the modern state did little to resource varied seasonally, providing the change customary land tenure arrangements flexibility needed for multiple production in rural areas. Docking (1999) notes that systems. Resource conflicts were mostly despite French and Malian government resolved amicably. At the core of this cus- attempts to manage local government in tomary law1 is a process of reconciliation Mali, village-based decision making pro- between disputants, rather than a winner- cesses changed little. In 1934, the contempo- take-all consequence, as in modern western rary context was set by the Supreme Court of law. Procedures in traditional courts were French West Africa which decreed that the simple and informal, although not easily African village was ‘a legal entity with cus- accessible to women or younger men. In tomary rights and the village chief was the modern courts, while access is universal in defender of those rights’ (Mamdani, 1996, principle, limits still exist due to trans- p. 114). An intermediary infrastructure of portation costs and the need for a lawyer to cantons was adapted as the mechanism for advocate one’s case. Currently, the simulta- organizing relations between the central neous existence of customary and modern state and the villages. Moore et al. (2003) legal systems offers multiple opportunities contend that this preservation of customary for contestation if resource users are dissat- rights defined a subordinate, but partially isfied with the existing determination of autonomous state apparatus. The colonial their rights. This legal pluralism often con- administration in Mali sought to govern indi- tributes to, rather than resolves land tenure rectly by relying on local authorities to conflicts. implement colonial policies while limiting In the Inland Delta region, customary the organizational integrity of Malian civil rights to the use of natural resources tend to society through the singular administrative vary from season to season and from resource binding of villages to the state. Production Systems and Social Institutions 29

As the demand for scarce resources struggles have been unfolding, because how increased, conflict could only be avoided if the state ultimately resolves the contradic- access was structured according to mutually tions inherent between legal texts and cus- agreed on criteria of legitimacy. Asserting its tomary practices has important implications authority, the state saw this as a matter of for the livelihood of the rural population. defining a single system of legal ownership This is a profoundly political issue, involv- and usufruct. Such a universal system does ing relations between the modern state, the not account for the multiple uses and sus- customary local state it has nurtured and a tained sharing of the resources that have nascent rural civil society (Ribot, 1999). customarily governed resources at the local During the socialist era after indepen- level. Customary law, on the other hand, is dence, the party and government were based on local practice and the local legiti- essentially fused in the state. The totalizing, macy of traditional leaders (herd masters, and thus centralizing, approach of the Keïta land chiefs, village chiefs, etc.). These lead- regime (1960–1968) tried to bypass the local ers, who are perceived to represent the best customary state, because it severely limited interests of the parties involved in any con- the associational sphere of both urban and flict over natural resources, have been able rural society. Canton chiefs were eliminated to resolve many conflicts through adapta- and villages came into direct relationship tion and interpretation of customary law. In with state power. However, as Rawson the eyes of the state, however, customary (2000) notes, ‘the party misjudged the power law and traditional leaders have often been of village elders’ (p. 271), who helped to perceived as an impediment to progress, move conditions at the national level even when customary laws and authorities towards increasing anarchy. Consequently, are officially recognized. the modernizing mission of the socialist In the Niger Delta, overlapping spatial regime was thwarted at the rural level. In claims are seen among livestock product- response, the military regime of Traoré ion systems at the intersection of pas- (1968–1991), which toppled the Keïta toral territories (leydi) managed exclusively regime, instituted an extensive reform of ter- by herders and the village pasturelands ritorial administration. This administration (harima) managed by villagers. This compe- sought to co-opt the customary system of tition between the leydi and the harima has local governance by incorporating it into the occurred as the size of village terroirs has state. ‘The chef [de village] was now ‘the increased at the expense of pastoral terroirs. representative of the chef d’arrondissement’ It is a manifestation of the power of village and hence of the territorial administration’ chiefs, legitimized by the support of village (Rawson, 2000, p. 273). Thus, local auton- elders and legalized by the laws of the repub- omy was severely restricted, much as it was lic. This process has reduced the power and under French colonial rule. Customary customary legitimacy of the herd masters authorities were treated as an appendage of (jowro) since their legality has not been rec- the state. Civil society under the Traoré ognized by the Malian administration. With regime began to open up, but participation this de-legitimization of their roles, local without the benefit of democratization was herd masters and other resource masters restricted primarily to urban populations. (land and water) are finding it more justifi- Both national regimes attempted to end able to privatize the resources for which they the phenomenon of a state within a state by have historically held a public responsibility changing national law, either by denying (Keïta, 2003). customary authorities effective recognition (Keïta) or by incorporation (Traoré). Within this modern legal framework, the village National policy and institutional context chief was to function as an appendage of the central state. Village elders would propose a It is worth considering the macro-institu- candidate for village chief to the chef d’ar- tional context in which these local legal rondissement who would either accept or 30 S. Cissé et al.

reject him. The traditional counter-power of these territorial collectives involved fre- secret societies and elder councils played a quent consultations at the local (or regional) moderating role in customary decision- level and was often controversial. Con- making. But, in all matters in which the testation between villages and within the State was concerned, the village chief acted national assembly was common. Despite as the local representative of the chef d’ar- serious attempts to organize ‘communes rondissement, executing decisions handed based on criteria of social solidarity, eco- down by him. Autonomous voices of rural nomic viability, and logic of geography and civil society were not to be heard beyond the space’ (Rawson, 2000, p. 278), regional gov- village terroir. ernment officials imposed many commune boundaries, often re-establishing the territo- rial limits of the cantons during the colonial Decentralization/deconcentration period (Cissé, 1997). Nevertheless, the current Malian gov- It is argued that Mali’s rich history of com- ernment has accepted the principle that the plexity and differentiation has provided a central government is poorly placed to make long and healthy tradition of civil society appropriate decisions for the local level and (Fox et al., 2001). Rawson (2000) claims that has moved to implement decentralization ‘Associational activity has existed in vary- policies (i.e. relinquishing decision-making ing degrees of autonomy from political authority over local matters, particularly authorities for centuries; civil society is not land use and natural resource allocation; a new idea in Mali’ (p. 269). Docking (1999) Mission de Décentralisation). Ostensibly, suggests that ‘rural associational groups are this increases the local populations’ role in key actors within the village level political deciding appropriate land uses and owner- structure and form the backbone of rural ship/usufruct issues. Such local control is communal life’ (p. 24). However, he notes expected to optimize the use of natural that ‘the concept of civil society as articu- resources and encourage investment. How- lated in the West fails to account for the ever, the reality is that local decision-makers dynamics unfolding at the grassroots level have very little legal discretion. The local in Mali’ (p. 395). Social relations in rural institutions established to govern natural civil society, such as they are, have been resources rarely have budgetary power, largely coincidental with the nearly local legitimacy or a mandate for more than autonomous local state at the village level awareness building. Furthermore, the cen- and external relations have been structured tral government, while delegating the through the network of the national state responsibility for decision making to local apparatus. authorities, has required that they enforce Opportunities for the growth of a national laws and provided few resources vibrant civil society were given new life for implementing their decisions (Johnson, with the downfall of the Communist regimes 2001; Ribot, 2002). In effect, the current of Eastern Europe and the Soviet Union. policy of decentralization has become Indeed, the fall of the Traoré regime can be largely a matter of deconcentration of state dated from this historic world event. The power (i.e. strengthening the national gov- new regime in Mali took past lessons to heart ernment at the local level). and instituted, during the 1990s, a policy of administrative decentralization and democ- ratization throughout the country (Rawson, Local Institutions: Gestion de Terroir and 2000). Substantial powers, including land Beyond tenure decision-making, were devolved to newly created Rural Communes. However, Recognizing the Malian State’s positive step regulations specifying how this power trans- towards democratization, the donor commu- fer will be implemented have yet to be nity (United States Agency for International drafted. The delineation (découpage)of Development, the World Bank, etc.) has Production Systems and Social Institutions 31

promoted the role of non-governmental (and oppositional) attributes of Malian organizations (NGOs) in order to encourage political life. However, democratic voting the growth of civil society. This effort builds behaviours are essentially unknown in on NGO experiences working within urban Malian village traditions (Kassibo, 1997). civil society under the Traoré regime. Failure to recognize the significance of this However, since 1991, the customary local fact has led to a blurring of the distinction state, presided over by village chiefs, has not between the organization and practices of been affected by these transformations the customary local state, on the one hand, (Davis, 2000), including the election of rural and rural civil society, on the other. Despite commune councils in 1999.2 concern about fostering an intermediary Many NGOs, donors and national gov- associational sphere between individuals ernments have implemented the gestion de and the national state, ‘rural civil society’ in terroir approach using socio-professional donor lexicon has come to be equated with groups to represent civil society at the local NGO-supported extensions of the custom- level. Local here is defined, by default, most ary state (Ribot, 2002). often as a single village. Indeed, what passes Nevertheless, decentralization has cre- for a local community has remained largely ated opportunities for local dialogue unexamined (for example, see Fox et al., and social adaptation. Recently, the devel- 2001). Common practice when a ‘stranger’ opment of local agreements for collec- (state or NGO development agent) desires to tive, multi-village management of natural work with villagers is for them to address resources has begun (Bocoum et al., 2003; the village chief requesting entry into the Hilhorst and Aarnink, 1999). Confronted by village and explaining their mission. The real and shared natural resource-based con- chief then calls together the village associa- flicts affecting the reproduction of their tion or a collection of representative elders livelihoods, local populations are learning to discuss the matter. After this consulta- to take advantage of the organizing space tion, the chief either indicates the village’s created by NGO-led efforts to develop and acceptance of the mission by designating apply innovative, locally controlled, prob- villagers to participate with the ‘strangers’ lem-solving approaches. Local open-forum or declines the offer. dialogues may concern one or several vil- The local management committees lages. Dialogue at the village level tends to developed in the context of these projects do reduce divisions within the community afford a level of participation and local con- which draw on rights of first settlement, etc. trol over decision making, but are still too by an explicit recognition of rights afforded few to have a major impact on NRM or rural to all engaged in productive activities. civil society. Although these forms of local When this dialogue concerns a group of vil- management show promise for the future of lages, the moment of communal manage- NRM, the approach has been criticized for its ment of space is introduced. These limited scope, avoidance of conflict, and bias discussions advance towards accords which against pastoral management (Benjaminsen, are the fruit of vigorous discussions and 1997; Painter, 1994). As a mechanism for negotiations. improving local NRM governance, local Although civil society in rural Mali is management committees still need consider- weak, decentralization has created condi- able support in the form of literacy and bud- tions ripe for growth. Civil society, how- getary training. Diversity of committee ever, is confronted with a subordinate membership is encouraged, but the tendency customary state’s clan-based and ethnic sol- is for local elites to dominate and for women, idarities that can serve as both a constraint youth and pastoralists to be excluded and a building block for the social capital (Bohrer and Hobbs, 1996). needed to strengthen rural civil society (see This recent promotion of rural civil Moore et al., Chapter 16). Functional hori- society has been based on donors’ normative zontal linkages between socio-professional expectations in evaluating the democratic groups across villages are beginning to be 32 S. Cissé et al.

established, and inter-group interaction Bohrer, K. and Hobbs, M. (1996) Post-Praia will be required in order to strengthen those Progress Towards Tenure Security and ties. Given the objective of trying to build Decentralization: Review of CILSS-member civil society in rural Mali, effectiveness country legislative reforms. University of requires that social relationships be built Wisconsin, Land Tenure Center, Madison, Wisconsin, pp. 1–41. that extend beyond the local customary Bovin, M. (1990) Nomads of the drought: Fulbe state, because by monopolizing the trust and Wodabee nomads between power and and reciprocal expectations of rural social marginalization in the Sahel of Burkina Faso capital, it inhibits the development of and Niger Republic. In: Bovin, M. and viable alternative social networks and rela- Manger, L. (eds) Adaptive Strategies in tionships that can promote a viable consen- African Arid Lands. Uppsala, Sweden, pp. sus concerning access rights and equitable 29–58. management of natural resources. Cissé, S. (1985) Land tenure and development problems in Mali: the case of the Niger Delta. In: Hill, A.G. (ed.) Population, Health and Nutrition in the Sahel. Kegan Paul, London, Notes UK, pp. 140–151. Crowley, E.L. (1991) Resource Tenure in Mali: 1 While one cannot speak of African customary an Anthropological Analysis of Local law as a single body of legal customs applica- Institutions. Club du Sahel/OECD, The ble to all of the different societies found in Hague, Netherlands, pp. 1–65. Africa, it can be used ‘as a generic term for Davis, J.U. (2000) Classrooms of democracy? The more or less similar traditional systems of law, educational prospects of Malian civil soci- or standards of behavior’ (Rugege, 1995, p. 1). ety. In: Bingen, R.J., Robinson, D. and Staatz, This is how we use it here. However, we rec- J.M. (eds) Democracy and Development in ognize that customary law in West Africa Mali. Michigan State University Press, East includes Islamic influences and is a living law, Lansing, Michigan, pp. 289–320. not dictated by rigid tradition. Delville, P.L. (1998) Harmonizing formal law and 2 In contrast, urban unions and syndicates have customary land rights in French-speaking experienced considerable separation from the West Africa. In: Politique des Structures et organs of the state and their networks of social Action Foncière au Service du Déve- relations are expanding. loppement Agricole et Rurale. Groupe de Recherche et d’Echanges Technologiques (GRET), Paris, France, pp. 1–29. Docking, T.W. (1999) International influence References on civil society in Mali: the case of the cotton farmers’ union, SYCOV. Ph.D. Ambach, B. (1996) Nomads on the river: Mali’s dissertation. Boston University, Boston, fishermen threatened by change. Deve- Massachusetts. lopment and Cooperation 3, 20–23. Fox, L., Ouattara, R., Boly, A., Ba, M.B., Konaté, Y. Benjaminsen, T. (1997) Natural resource manage- and Charlick, R. (2001) Francophone Civil ment, paradigm shifts, and the decentraliza- Society Organization Assessment: Mali and tion reform in Mali. Human Ecology 25(1), Côte d’Ivoire Synthesis Report (West Africa: 121–141. Civil Society Strengthening for Conflict Blench, R. (1996) Aspects of resource conflict Preventions Study). Associates in Rural in semi-arid Africa. Natural Resource Pers- Development (ARD). United States Agency pectives 16. Overseas Development In- for International Development Broadening stitute. http://www.odi.org.uk/nrp/16.html Access and Strengthening Input Market (accessed 1 May 2004) Systems (BASIS), Burlington, Vermont. Bocoum, A., Cochrane, K., Diakité, M. and Kane, Frankenberger, T.R. and Lynham, M.B. (1990) O. (2003) L’inclusion sociale pour une ges- Household food security and coping strate- tion équitable et durable des ressources gies along the Sénégal River Valley. In: Park, naturelles: deux expériences au Mali T.K. (ed.) Risk and Tenure in Arid Lands: the (Securing the commons No. 7). SOS-Sahel Political Ecology of Development in the International Institute for Environment and Senegal River Basin. University of Arizona Development, London, UK. Press, Tucson, Arizona, pp. 51–86. Production Systems and Social Institutions 33

Heasley, L. and Delahanty, J. (1996) Politics of mentation (Summary Assistance Report manure: resource tenure and the agropas- No. 22). Ministry of Environment and toral economy in Southwestern Niger. Nature Protection, Community-Based Natu- Society and Natural Resources 9, 31–46. ral Resource Management Project, Dakar, Hilhorst, T. and Aarnink, N. (1999) Co-managing Sénégal. the Commons: Setting the Stage in Mali and Moore, K.M., Cissé, S. and Touré, A. (2003) Zambia (Bulletin 346). Royal Tropical Building social capital in the Sahel. Rural Institute, Amsterdam, Netherlands. Sociological Society Annual Meeting. Horowitz, M.M. and Jowkar, F. (1995) Gender and Montreal, Quebec, Canada, pp. 1–36. participation in environment and develop- Mortimer, M. and Turner, B. (1994) Crop–Livestock ment projects in the drylands. In: Stiles, D. Farming Systems in Semi-arid Zone of Sub- (ed.) Social Aspects of Sustainable Dryland Saharan Africa: Ordering Diversity and Management (United Nations Environment Understanding Change (Agricultural Re- Program). John Wiley & Sons, Chichester, UK. search and Extension Network Paper No. 46). Johnson, C. (2001) Local democracy, democratic Overseas Development Administration, decentralization and rural development: the- London, UK. ories, challenges and options for policy. Ngaido, T. (1993) Land Use conflicts in West- Development Policy Review 19(4), 521–532. ern Rural Niger: Kollo and Tillabery Kassibo, B. (ed.) (1997) La decentralization au Arrondissements (Land Tenure Center Dis- Mali: état des lieux (Bulletin 14). LitVerlag, cussion Paper No. 1). Madison, Wisconsin, Hamburg, Germany. pp. 1–30. Keïta, N. (2003) Décentralisation et responsabilité Ngaido, T. (1996) Redefining the boundaries of dans les modes de gestion des ressources control: post-colonial tenure policies and naturelles et des redevances y afférentes: les dynamics of social and tenure change in bourgoutières Yallarbé de Youwarou (Work- Western Niger. Ph.D. dissertation, University ing Paper). Centre Universitaire Mande of Wisconsin, Madison, Wisconsin. Bukari/World Resources Institute, Bamako, Painter, T., Sumberg, J. and Price, T. (1994) Your Mali. terroir and my ‘action space’: implications of Maïga, I. (1996) Etude sur la prévention et la ges- differentiation, mobility and diversification tion des conflits. Centre Canadien d’Etude et for the approche terroir in Sahelian West de Coopération Internationale (CECI) Africa. Africa 64(4), 447–463. Programme de Promotion de la Démocratie Park, T., Tabor, J., Frankenberger, T., Lynham, M., et des Droits de la Personne (PDDP), Bamako, Baro, M., Ngaido, T. and Rogers, G. (1993) Mali. Risk and Tenure in Arid Lands: the Political Maïga, I. and Diallo, G. (1998) Les conflits fonciers Ecology of Development in the Senegal River et leurs gestion dans la 5ème Région du Mali Basin. University of Arizona Press, London, (Dossier No. 76). Groupe de Recherche UK. Actions pour le Développment (GRAD), Pointing, J. (1995) Impact of social and economic International Institute for Environment and change on pastoral women in East and West Development, Programme Zones Arides, Africa. In: Stiles, D. (ed.) Social Aspects of London, UK. Sustainable Dryland Management (United Mamdani, M. (1996) Citizen and Subject: Nations Environment Program). John Wiley Contemporary Africa and the Legacy of the & Sons, Chichester, UK. Late Colonialism. Princeton University Pratt, D.J., Le Gall, F. and de Haan, C. (1997) Press, Princeton, New Jersey. Investing in Pastoralism: Sustainable Mathieu, P. (1996) La sécurisation foncière entre Natural Resource Use in Arid Africa and the compromise et conflits: un processus Middle East (World Bank Technical Paper politique? In: Mathieu, P., Laurent, P. No. 365). World Bank, Washington, DC. and Willame, J. (eds) Démocratie, Enjeux Rawson, D. (2000) Dimensions of decentraliza- Fonciers et Pratiques Locales en Afrique. tion. In: Bingen, J.R., Robinson, D. and Statz, Cahiers Africains, No. 23–24. Institut J.M. (eds) Democracy and Development in africain/L’Harmattan, Paris, pp. 26–44. Mali. Michigan State University Press, East McIntire, J., Bourzat, D. and Pingali, P. (1992) Lansing, Michigan. Crop-Livestock Interaction in Sub-Saharan Ribot, J.C. (1999) Decentralization, participation Africa. World Bank, Washington, DC. and accountability in Sahelian forestry: Moore, K.M. and Thiongane, S. (1998) Review Legal instruments of political-administrative and Analysis of CBNRM Activity Imple- control. Africa 69(1), 23–65. 34 S. Cissé et al.

Ribot, J.C. (2002) African Decentralization: Local Network Paper 15d). Overseas Development Actors, Powers and Accountability (Gover- Institute, London, UK. nance and Human Rights, Paper 8). United Touré, D. (1996) Prévention et gestion des conflits. Nations Research Institute for Social Centre Canadien d’Etude et de Coopération Development, Programme on Democracy, Internationale, Ouagadougou, Burkina Faso. Geneva, Switzerland. Turner, M.D. (1992) Living on the edge: Fulbe Rugege, S. (1995) Conflict resolution in African herding practices and the relationship customary law. Africa Notes, 1–3. between economy and ecology in the Inland Sanders, J.H., Shapiro, B.I. and Ramaswamy, S. Niger Delta of Mali. Ph.D. dissertation. (1996) Economics of Agricultural Tech- University of California, Berkeley, California. nology in Semiarid Sub-Saharan Africa. Watts, M.J. (1987) Drought, environment, and Johns Hopkins University Press, Baltimore, food security: some reflections on peasants, Maryland. pastoralists and commoditization in dryland Sturm, H. (1996) Traditional crop-livestock-inter- West Africa. In: Glantz, M.H. (ed.) Drought action systems as a basis for sustainable farm- and Hunger in Africa. Cambridge University ing systems in the subhumid savanna. In: Press, Cambridge, UK. West, N.E. (ed.) Rangelands in a Sustainable Williams, T.O., Powell, J.M. and Fernández- Biosphere: Proceedings of the Fifth Rivera, S. (1995) Manure utilization, drought International Rangeland Congress,Vol. 1. cycles, and herd dynamics in the Sahel: Society for Rangeland Management, Denver, implications for cropland productivity. Colorado, pp. 536–537. In: Powell, J.M., Fernández-Rivera, S., Sunzini, P. (1992) Limites et potentialités Williams, T.O. and Renard, C. (eds) Livestock pour un développement soutenable dans la and Sustainable Nutrient Cycling in Mixed zone agro-sylvo-pastoral au Sahel. In: La Farming Systems of Sub-Saharan Africa, zone Agro-Sylvo-Pastoral au Sahel: Vol. II, Technical Papers. Proceedings of an Limites et Potentialités de développement International Conference, Addis Ababa, contre la Désertification. Actes de la Ethiopia. Conference de Assissi. Comité Inter-Etats de Wolmer, W. (1997) Crop–Livestock Integration: Lutte contre la Sechéresse au Sahel, the Dynamics of Integration in Contrasting pp. 51–62. Agroecological Zones, a Review (Sustainable Toulmin, C. (1983) Herders and Farmers Livelihood Program Working Paper 63). or Farmer–Herders and Herder–Farmers Institute of Development Studies, University (Agricultural Adminstration Unit, Pastoral of Sussex, Brighton, UK. 3 Landscape and Soil Characterization of Madiama Commune

Oumarou Badini1 and Lassana Dioni2 1International Programs/Development Cooperation, Washington State University, Pullman, WA 99164–5121, USA; and 2Centre Régional de la Recherche Agronomique-Sotuba, Institut d’Economie Rurale, BP 258, Bamako, Mali

Knowledge of the relationships between chapter ends with a characterization of each landscape, soil types and land use systems landscape type and associated soils accord- is essential to improving sustainable natural ing to their importance in the commune, resource management. This chapter uses as well as their locations, potentials and computer-based remote sensing and geo- constraints. graphic information systems (GIS), com- bined with field observations, to describe, classify and map the landscape of the General Setting and Description of the Commune of Madiama. This supervised Commune classification is conducted at a scale of 1/40,000 and is suitable for commune and Site localization village, but not farm level decision making. Most existing work of this nature is only Madiama commune is located in the ext- applicable for national level planning, but reme southern portion of the south-Sahelian this verified characterization of spatial vari- Zone (less than 650 mm isohyets). The com- ability and distribution of landscapes and mune is about 25 kilometres from Djenné soil types provides a basis for optimizing (capital of the administrative Cercle)and land uses down to the village level. It 120 kilometres south of Mopti (capital of the becomes possible to recommend the best 5th Region of Mali) (Fig. 3.1). It lies between way to plan, manage and use different latitude 13° 45 N to 13° 52 N, and longitude landscape and soil types, and to identify 4° 22 W to 4° 30 W in the north-central part appropriate agricultural practices and tech- of Mali. Madiama commune covers an area nologies. In addition, the output and infor- of some 16,970 ha (169.7 square kilometres) mation generated can serve as input for gradually rising from the Bani River flood biophysical models for extrapolation to plain in the west to plateaus and occasional other similar sites or through time. rock outcrops in the east. In 2001, there were The chapter begins with a general about 7973 inhabitants (with a density aver- description of the commune, followed by an age of about 47 inhabitants per km2) com- account of the historical formation of the prising people of numerous ethnic groups. physical and biological landscape (geology, The dominant ethnic groups are Bambara, agroecology, climate and hydrology). The Marka and Peul. © CAB International 2005. Conflict, Social Capital and Managing Natural Resources (ed. K.M. Moore) 35 36 O. Badini and L. Dioni

Fig. 3.1. Physical location of Madiama Commune in the 5th Region of Mali (Landsat 7 Image from November 1999).

Agricultural systems fertility are considered the primary con- straints to crop production across the Sahel Reduced vegetative cover, shorter fallows, (Van Keulen and Breman, 1990). declining soil fertility, and increased erosion Agriculture is the primary economic are consequences of the climate changes and activity. Millet (Pennisetum spp.) and demographic pressures affecting the Sahel sorghum (Sorghum vulgare), generally inter- (Taylor-Powell, 1991). Expansion of farming cropped with cowpea (Vigna unguiculata), on marginal lands and conflicts over limited dominate the cropping pattern. The lowland resources are common in the study area. Like areas near the Bani River are used for rice other areas of the West African semi-arid cultivation. More favourable microenviron- Sahel region, the study zone has suffered ments of the sandy and loamy plains with diminished food production on a per capita favourable moisture and nutrients condi- basis since the early 1970s. Long-term tions such as the lower areas (micro basins) rainfall throughout the region declined dra- are cultivated in sorghum, watermelon matically in the early 1970s and has not (Citrullus lanatus), okra (Hibiscus esculen- returned to earlier levels. Although exacer- tus) and dah (Hibiscus sabdarifa). Based on bated by population growth, the fundamental availability, organic fertilizer (manure) is problem is physical. Unfavourable rainfall used in dryland farming to a great extent. patterns, soil degradation, and declining soil Chemical fertilizers are used mostly in irri- Landscape and Soil Characterization of Madiama Commune 37

gated rice plots and watermelon cultivation. some villages or substantially reduced to 2 In general, all crops respond well to fertiliz- or 3 years in others. ers, but fertilizer use is insignificant (applied Factors that influence agricultural pro- by less than 2% of farmers) and the rates duction decisions and natural resource man- applied are very low (e.g. 20 kg/ha nitrogen agement strategies include: anticipated diammonium phosphate (DAP) for water- rainfall (based on duration and intensity of melon and 50 kg/ha of urea on rice). Farmers the cold season), soil types, availability and in the commune dry seed or seed with the access to land, family size, farming equip- first sufficient rains after ploughing in mid- ment, market prices for cash crops, and to late June. Reseeding is often necessary due social factors (e.g. bride price, funerals, etc.). to early season drought. Planting can con- The major natural resource conflicts in the tinue through July in most years. Two weed- commune are between farmers and herders ings (sarclage) are customary, although these and occur because increased areas under cul- are dependent on labour availability and the tivation are causing reduced pasture areas, prospect of potential crop yield. Harvesting reduced animal passage routes, and compe- is usually done in late October to mid- tition for seasonal wetland and bourgou. November for millet and sorghum and late December for rice. Overall, the crop calendar varies depending on the rainfall pattern. Physical and Biological Environment Final crop yields are strongly dependent on (Landscape) time of planting, with higher yields obtained in years of early planting dates. The study zone, which is part of the Inland Animal husbandry is the second major Niger Delta region, is characterized by a economic activity in the commune. Nearly diverse potential dependent on the site spe- all households own livestock, principally cific combination of water, soils and plants. mixes of cattle, sheep and goats. Animals The following section describes the land- are an important source of income and scape of the commune of Madiama in regard provide energy for fieldwork (e.g. tillage), to its geology, geomorphology, physiogra- transport, manure and food. Animal traction phy, agroecology, climate and hydrology. is used in about 80% of dryland farming, and the remaining 20% is cultivated using hand-held hoes. Pastures are not under spe- Geomorphology and geology cific management regimes, except that herders move their animals from place to Like much of Mali, the commune of place depending on the seasons in order to Madiama has a monotonous relief, consist- maximize the grazing of the animals. Most ing of mainly flat with low plateaux and available land within village limits is used basins, and a mean altitude of 279 m. It is for agricultural production regardless of soil part of the Inland Niger Delta flood plain fertility, and the size of grazing lands is and influenced by the Bani River (an afflu- diminishing due to increased field sizes. ent of the Niger), which is on its western Principal production problems are side towards the city of Djenné. related to the geoecological changes affect- The Commune is located in the sedi- ing the Sahel region due to inadequate and mentary basin of the Bani River that was irregular rainfall, low soil fertility and soil formed by alluvial deposits during the degradation, lack or insufficient animal Quaternary period (1.8 million years ago to manure, and poverty. Land degradation and present – the most modern geological low soil fertility are common features of the period). Infra-Cambrian sandstone called the landscape. The traditional practice of long Bandiagara sandstone formed the area, bush fallow as a means to restore soil pro- which is relatively elevated with rock out- ductivity is breaking down. Fallow is either crops on its eastern border. The inner valley non-existent (the same parcel is cultivated of the Bani and Niger rivers started its forma- year after year without crop rotation) in tion during the beginning of the Quaternary 38 O. Badini and L. Dioni

with the Bandiagara sandstone covered with its ligneous component is dominated by ferruginous hardpan (cuirass). Remnants of small shrubs such as Combretum gluti- this cuirass are still found at the borders of nosum, large shrubs such as Pterocarpus eri- the commune. Sea level oscillations as well naceus and Terminalia spp. and other as climatic variations during the Quaternary woody species such as Parkia biglobosa produced diverse alluvial deposits that still (African locust bean or néré in French), a tree determine the soil types and landscape of the with pods whose seed and inner gelatin are study area today. eaten. The herbaceous stratum is mainly The soils of the Inland Delta region composed of annual grasses such as Loudetia were developed over a time sequence of togoensis. The forage potential is average. alternating subhumid and hyper-arid cli- The main crop is millet. However, sorghum matic periods during the Quaternary. The can be planted in low-lying areas. time sequences of the alluvial sedimenta- tion and the impact of sea oscillations and The Moyen Bani-Niger (the Bani-Niger climatic variations have shaped the final Midlands) makeup of the landscape. These variations explain the general characteristics of the This transition zone to the floodplain is pedogenesis and the potentials of soils characterized by soils with slightly sandy found in the region. Details concerning this texture and average natural fertility. The geomorphology can be found in many pre- main ligneous plants are small shrubs such vious works (Blanck, 1967; Gallais, 1967; as Piliostigma reticulatum and other woody Bertrand, 1973). species such as the nitrogen fixing Acacia albida (apple-ring acacia, ana tree, winter thorn), a widely used tree well documented Agroecological characteristics for increasing the yields of crops grown under it. The herbaceous strata are diverse The agroecological zone is defined here as and dominated by Loudetia togoensis a homogeneous geographic entity from in non-inundated areas and vetiver grass the standpoint of its climate, geology, geo- (Vetiveria nigritana) in the often flooded morphology, soil and vegetation. In Mali, zone. The forage potential is very high. Rice the Project d’Inventaire des Ressources cultivation is possible on low flooding lands Terrestres (PIRT) project (1986) identified 49 and sorghum is cultivated on the higher agroecological zones and 14 natural regions. elevations. As part of the south Sahelian Bio-climatic zone, most of the Madiama commune is The Delta Vif (the Active Delta) located in the Inland Niger Delta region, but its extreme east stretches to the Bandiagara- In the commune, the Delta Vif is composed Hombori Plateau (cf. Carte du Zonage Agro- of the floodplains of the Bani River (an afflu- écologique du Mali, PIRT, 1986). Therefore, ent of the Niger). These lands are part of the from east to west of the commune, the agroe- lacustrian zone seasonally flooded by the cological zones are: the Bas-Plateau Bobo, rising waters from the Bani. The type of the Moyen Bani-Niger, and the Delta Vif. vegetation found here is a function of the length and patterns of flooding. On the soils where floodwaters remain longer are found The Bas-Plateau Bobo (the Low Bobo Plateau) some perennial grasses such as bourgou1 Higher elevation with indurated laterite and (Echinocloa stagnina). On the soils with soils on hardpan (cuirass) and rock outcrops shorter duration of floodwaters are found on non-arable lands characterizes the relief vetiver grass (Vetivera nigritana) and woody in this zone, in which intermittent streams species such as Mitragyna inermis. The and rivers are found. The arable lands have forage (pastures) potential is very high. The loamy sand, sandy clay loam textures or lat- receding floodwaters permit the cultivation erite clay. The vegetative cover is diverse and of rice. Landscape and Soil Characterization of Madiama Commune 39

Climate and Hydrology of years to develop. If these areas continue to be degraded, they will probably not Brief climatic history recover under today’s arid conditions.

Like all the earth’s regions, the Sahel has been subjected to repeated climatic fluctua- Climatic zone and rainfall patterns tions throughout its history. Cold, warm, dry and moist periods have altered the face of The study area is part of the South Sahelian the region many times. These fluctuations zone (below 650 mm isohyets) with a short have taken place over the last 18,000 years. rainy season, considerable variability in the Approximately 20,000 years ago the Sahara rainfall amount and distribution, severe expanded southward into the current Sahel drought due to high evapotranspiration, and zone (including the study area) during a pro- high daily and annual temperature. There nounced dry period. Rivers dried up, vege- are no noticeable seasons that can be tation died out, and sand dunes formed. defined by temperature as in temperate About 15,000 years ago, climatic conditions zones. Differences in temperature are began to improve. Moist tropical air masses greater between day and night than between advanced far to the north during the summer and winter. summer months, bringing a considerable The 3–4 month rainy season occurs increase in rainfall (Leisinger and Schmidt, during the summer from June to Septem- 1995). ber, with rainfall ranging from 400 to 600 This favourable climate lasted for about mm. However, because of high tempera- 5000 years (from about 8000–3000 BC). The tures and intense solar radiation, much of Sahara desert receded to small, isolated this precipitation evaporates before it can areas in northern Africa. An enormous be used for agriculture. There is sufficient drainage network came into existence in the humidity for agriculture for only 2–4 regions south of the Sahara. Rivers months during the year, with dry months deposited sediments over broad areas such predominating the rest of the time. The as in Inland Delta region of the Niger, where annual rainfalls, the long-term mean and land use today is based on extensive irriga- the general rainfall trend in the area of tion. Previous arid, desolate areas came to Madiama in the period from 1950 to 2000 life again, and the tropical rain forest are plotted in Fig. 3.2. advanced far to the north. Sand dunes were The annual rainfall in the commune gradually covered with vegetation and sand and surrounding area varies considerably masses no longer migrated southward. Red, from year to year, a very common character- fertile tropical soils were formed in areas istic in the semi-arid tropics. From 1950 to where it is virtually impossible for soils to 2000, as shown in Fig. 3.2, the mean annual be formed today due to aridity. Humankind rainfall is 544 mm (Badini, 2001). The is still profiting today from the favourable lowest annual rainfall of 274 mm was climatic conditions of that time, because recorded in 1987, while the highest annual soils that were formed then continue to ben- rainfall in the past 51 years was 914 mm, efit present-day agriculture. received in 1957. After this relatively ‘wet’ period, the cli- The average rainfall for the period mate once again worsened. Precipitation 1950–1969 was 636 mm compared to an became highly variable and the decreased average of 482 mm for the period 1970–2000. amount of rainfall caused lakes, river sys- The average rainfall loss of around 154 mm tems and rain forests to recede, which led to between the two periods highlights the a renewed advance of the desert. Today cli- declining rainfall over the last three decades matic history still determines to a consider- (Fig. 3.2). The consequences for this drop able extent the potential for agriculture in rainfall amount are increased aridity in the Sahel. In many areas, people live in and more risks for crop failure due to an environment that took many thousands limited water availability for rainfed crops 40 O. Badini and L. Dioni

1000

900

800

700

600

500

400

Annual precipitation (mm) 300

200 1950 1952 1954 1956 1958 1960 1962 1964 1966 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 Year

Fig. 3.2. Precipitation distribution, mean and trend from 1950 to 2000 in Djenné-Madiama (Service National de Meteorologie, Bamako, Mali).

and diminishing pastures. These same con- August (155Ϯ68 mm) shows a pronounced straints caused by the drop in rainfall are peak in the Madiama area (Fig. 3.3). Short common for the many parts of the Sahel but heavy rainstorms and strong gusts of region since the droughts of the 1970s. wind distinguish the beginning and the end Due to the seasonal movement of the of the rainy season, whereas more gentle Inter Tropical Convergence Zone (ITCZ) that rainfalls that tend to endure for longer peri- regulates the rainfall pattern, precipitation ods of time characterize the peak period in in Madiama and in the Sahel region in gen- August. The wide annual variability in rain- eral is monomodal (Sivakumar et al., 1984). fall is characteristic of the whole Sahel Compared to the wetter zones in West region and represents a main agroclimatic Africa, rainfall in July (125Ϯ53 mm) and constraint to crop production.

35

30

25

20

15

% annual total rain 10

5

0 Jan Feb March April May June July Aug Sept Oct Nov Dec Months

Fig. 3.3. Monthly rainfall distribution in Madiama. Landscape and Soil Characterization of Madiama Commune 41

Hydrology for biophysical modelling (see Chapter 11) to further the assessment of land qualities In the study zone, flooding follows the start and the impact of agricultural practices on of the rainy season, but the onset date for the productivity and the environment. The data rise in the water levels varies from year to consisted of geo-referenced physical and year (Gallais, 1967). The water rise and chemical characteristics of soils that were flooding pattern in the study area are nor- recorded in the field, determined in the lab- mally slow and gradual. The waves of the oratory and/or extracted from remotely rising water from the Niger at Ké-Macina sensed imagery. They were used in interac- (upstream from Madiama) take up to 2 or 3 tions with the geomorphology (e.g. land- weeks before reaching Belentieny and then forms) and the hydrology of the area to Sofara (downstream from Madiama). resolve landscapes into mappable areas in Due to its location in relation to the which the soil is less variable than the over- Bani River, the Commune of Madiama is all landscape (Beckett and Webster, 1971). mainly an intermediary zone lying between These areas, are presented as landscape the flood plain in the west and the ironstone types, and comprise one or two dominant plateau on the east. Although many rivers soil types. and streams (called yamé in Fulfulde and The area covered by each examination local Bambara) run through the commune, point (mapping unit or pixel) is 16 ha. In the the most important are the seasonal streams commune, a farmer’s parcel size is far less flowing through Promani-Tatia to the Bani than the 4 ha found in the Sahel region in River at Baratou village. Another less impor- general. An examination point of 4 ha is tant stream runs through Kessédougou and only achievable with a detailed soil survey Toumadiama. At the extreme south of the and large-scale map of 1/10,000 or less. commune, a third, more or less steep-sided Although these large-scale maps could stream is found. better assist in farm level planning and even detailed fertilizer recommendations depen- dent on field variability, their implementa- Landscape and Soil Types tion is labour intensive and very expensive. Characterization To compensate for these limitations, the approach has been to combine semi- Methodology detailed soil survey (1/40,000) and biophys- ical modelling with field studies to evaluate A morpho-pedological approach combining the impact of agricultural practices on pro- computer-based remote sensing and GIS ductivity and the environment (see Chapter analysis with field observations of land- 11). The existing wealth of knowledge scape and soil types made it possible to derived from the landscape and soil charac- describe, classify, and map land and soil terization study of the commune that allows resources of the Commune of Madiama for a description of the landscape and a (Badini and Dioni, 2001). This approach determination of the soil types, distribution, takes into account the interactions between potentials, constraints and suitability the geomorphological characteristics (i.e., should contribute to sustainable natural landform, relief, nature of soil parental resource development and better land use material), the hydrology, the soil profile and planning and decision-making. the pedogenesis. The analysis of these inter- The present chapter describes how soils actions determined the specific landscape with different potentials and constraints are and soil types of the area. distributed within the commune bound- The objective of the soil survey investi- aries. It also provides a semi-detailed map gations (Badini and Dioni, 2001) was to pro- (Fig. 3.4) and analyses of agricultural suit- vide data for the rational planning and abilities and ecological vulnerabilities of the management of land use. Also, the results classified landscape types (see Tables 3.1 from this work met the data requirements and 3.2). Because this work identifies the 42 O. Badini and L. Dioni

precise nature and distribution of the land perennial grasses such as bourgou and soils in the commune, it can serve as a (Echinocloa stagnina). On the soils with basis for the planning and management of shorter duration of floodwaters are found agricultural projects. Furthermore, it also vetiver grass (Vetivera nigritana) and woody provides a framework for specifying appro- species such as Mitragyna inermis and priate land management techniques related Acacia (Fig. 3.5). to each land type. Two types of soils are found in this land- scape: the pseudo gley hydromorphic soils (or in French sols hydromorphes à gley Description of Landscape and Soil Types réduit)inthe low-lying areas, and the amphigley hydromorphic soils (sols hydro- For the total land area of 16,970 ha of the morphes à amphigley)inhigher areas. They commune, eight landscape types are identi- can be classified as United States Depart- fied and characterized according to specific ment of Agriculture (USDA) Aqualfs (Fig. landform, pedogenesis, hydrology, topogra- 3.4 and Table 3.1) (Survey Soil Staff, 1999). phy, land use and soil cover. They range These are deep (95 cm) silty clay loam from hydromorphic flood plains on the west hydromorphic soils with gleying at the side near the Bani River to plateaux and bottom of the profile. Because of their loamy some rock outcrops on the east side outside texture, they have a high water holding the inundated area of the commune. Based capacity. Organic matter (OM) levels vary on the age of each formation, and starting from 0.4% (very low) to 1.4% (very high) from the youngest to the oldest (in geologic with an average of 0.9% (high). The terms) landforms, the following landscape carbon/nitrogen (C/N) ratio of around 10 types are shown in Fig. 3.4 and Table 3.1. In shows that high nitrogen mineralization is each landscape type, one or two main soil taking place. The soils are acidic in surface types are identified and described. The soil (pH 5.0 on average), with low assimilable texture, fertility level, production con- phosphorus content of 4.3 ppm, and low straints, and overall agricultural or pastoral nitrogen content between 0.02 and 0.04%. suitability are described for each landscape The major constraints to production type unit (Tables 3.1 and 3.2). (Table 3.2) are due to the hydrology of these lands. The underground water is at shallow depth, with very high risks of inundation Recent (young) formations and waterlogging excluding any possibility of rainfed agriculture. The receding flood- waters permit the cultivation of rice. The Hydromorphic floodplains (Unit Ci) soils can also be used for gardening and hor- This landscape, called ‘lateral basin’ or ‘inun- ticulture. The forage (pasture) potential is dation basin’ (or in French Cuvette very high. d’inondation – Ci), is the most recent alluvial formation. These hydromorphic floodplains Hydromorphic alluvial levees or ‘sand banks’ located in the upper west side of the com- (unit Tr) mune are the youngest geological landform. Their formation is still taking place today. This landscape called ‘recent alluvial ter- They are part of the lacustrian zone season- race’ (or in French terrasse récente – Tr) is a ally flooded by the rising waters from the Bani relatively recent alluvial formation. It is River. They cover a total land area of 746.8 ha characterized by alluvial levees or ‘sand or 4.4% of the commune and are called banks’ forming a convex strip around the ‘goubé’ by the people in the commune. edges of the inundation basin (unit ci) The type of vegetation found in this above. The landform is slightly elevated and landscape is a function of the length and made of coarse sandy soils often covering patterns of flooding. On the soils where some pebbles of quartz. It covers a total land floodwaters remain longer are found some area of 1391.5 ha or 8.2% of the commune. Landscape and Soil Characterization of Madiama Commune 43

Fig. 3.4. Land and soil units of the Commune of Madiama. 44 O. Badini and L. Dioni

Fig. 3.5. Hydromorphic flood plains.

The main ligneous plants are small Table 3.1). The soils of this landscape are shrubs such as Piliostigma reticulatum. The very deep (120 cm) with sandy clay loam herbaceous layer (Fig. 3.6) is mainly consti- texture covered by alluvial sand deposits. tuted of Ipomea repens and vetivers These soils are acidic and have a very low (Vetiveria nigritana). chemical fertility. The dominant soils found in this land- Besides their low fertility, the major scape are the soils of limited development constraints of the soils of this sand bank from alluvial deposits (or in French sols peu landscape are their low water holding cap- évolués d’apport alluvial). They can be clas- acity and risks for temporary inundation sified as USDA Psammaquents (Fig. 3.4 and (Table 3.2). Their sandy texture permits a

Fig. 3.6. Sand banks. Landscape and Soil Characterization of Madiama Commune 45 squelettique très peu profond ou squelettique très peu profond engorgement de surface engorgement et sols argileux hydromorphes à hydromorphes et sols argileux gley de profondeur tropicaux peu lessivés à tâches tropicaux et concrétions Soil classification Soil classification Area Total Lithosol Names banks) Landscape types Landscape Legend and landscape/soil types classes. landscape/soil types Legend and latéralerécente (banc de sable) alluvial flood plainssubactuelle levees (sand dougoukolo terraces alluvial Bahtientien réduit et à amphigley grossière sableux à texture sableux faiblement hydromorphes limoneuse/ ou dépression ancien bras mort ancienne loamy plains materials Tientienblen à réoxydés et sols hydromorphes peu lessivés et Sols ferrugineux de la terrasse t2induré (Regosols) loamy materials cuirassée croupe pediment ou Djè dougoukolo vertique et nodules calcaires cuirasse ou régosols (Lithosols) able 3.1. Code Français English Bambãna CPCS) (French US (tentative) (ha) (%) Area citr Cuvette d’inondation Hydromorphic Terrasset1 Farah-lèh Terrasse Hydromorphic à gley Sols hydromorphes ou Tientien Old levees and Aqualfs Sols peu évolués d’apport alluvial Tientien-fing Psammaquents 746.8 Sols peu evolués d’apport alluvial 1391.5 Tropaqualfs Typic 4.4 8.2 5243.7 30.9 cl Cuvette t2 Terrasse Hydromorphic Bona dougoukolo Plains of sandy à gley Sols hydromorphe Nanganblen ou Haplustalfs non à Sols ferrugineux tropicaux Ultic Haplustalfs 2104.3 4327.3 12.4 25.5 T cavi Cuvette ancienne Plains of silty andbk Glacis- versant Boua dougoukolo Bas plateau ou à tendance Sols hydromorphes Laterized indurated Fara-dèbè Plinthic Haplustalfs plateau Ironstone 1612.1 Fuga dougoukolo sur évolués d’erosion Sols peu 9.5 Pratiquement pas de sol sinon sol Cambic Cuirorthids Aridic Cuirustalfs 526.1 882.4 3.1 5.2 46 O. Badini and L. Dioni

high permeability and does not allow good surface), which corresponds to USDA retention of water and nutrients. These soils Haplustalfs. The soils in this landscape are have a very low agricultural value and their deep (100 cm). Their texture is sandy loam improvement would necessitate costly to clay loam in the surface horizon with peb- investments. The use of fertilizers on these bles of quartz at a depth around 60 cm where soils must be accompanied by soil amend- the clay content is between 35% to 40%. ment in order to improve nutrient retention Due to their texture, they have a high water and decrease soil acidity. However, these holding capacity (15–20%), and OM content soils exhibit some potentialities for the averaging 0.5–1%. The C/N ratio is around development of fire wood trees and pastures. 10 due to a very active mineralization process. Major constraints to production (Table Hydromorphic loamy plains (unit Cl) 3.2) in these soils include possibilities of This landscape, called ‘loamy basins’ (or in flooding, soil compaction, difficulties for French Cuvette Limoneuse – unit Cl), is tillage due to high soil adhesivity and weed located below the old levees and alluvial (adventices) invasion. However, these soils terraces (unit t1). The relief is uniform and are deep with a silty clay loam texture and a regular. Mounts of sandy soil trapped under- good capacity to hold water. Besides graz- neath small shrubs called ‘nebkass’ charac- ing, rainfed rice cultivation is possible in terize the present morpho-dynamic. This years of average rainfall. landscape type, the third largest in the com- mune, has a total land area of 2104.3 ha or Levees and alluvial terraces (unit t1) 12.4% of the commune (Table 3.1). The nat- ural vegetation is dominated by Acacia This landscape, called ‘young alluvial for- albida and grasses (Fig. 3.7). mation’ (or in French formation alluviale Two types of soils are found in this subactuelle – t1) was formed during the landscape. They are called hydromorphic moist period that succeeded the dry period soils with reoxydated gley and hydromor- 15,000 years ago (see climatic history in pre- phic soils with surface saturation (or in vious section). This land type, which covers French sols hydromorphes à gley réoxydés 5243.7 ha or 30.9% of the land area in the and sols hydromorphes à engorgement de commune dominates the commune (Fig. 3.4

Fig. 3.7. Loamy plains. Landscape and Soil Characterization of Madiama Commune 47

loam soils with gleying at lower depth. OM content is between 0.5% and 1%, and a C/N ratio of 10. The soils are only slighly acid with an average pH of 6.2. Phosphorus content and nitrogen content in the sur- face are 9 ppm and 0.02% (very low), respectively. Major constraints to production in these soils include soil crusting, limited soil fertility due to acidity and low soil nutrients reserve. The landscape type covers much of the commune and the soils are suitable to agriculture with a good capacity to hold water. These soils allow the rainfed crop- ping of an array of crops, including millet, sorghum, groundnuts and cowpea and rain- fed rice cultivation.

Old formations

Old alluvial plains and terraces (unit Ca) This landscape, called ‘old basin’ (or in French cuvette ancienne – Ca) is character- Fig. 3.8. Recent alluvial terraces. ized by slightly obliterated depressions with a regular relief. The present morphody- namic is shaped by mounds of sand and silt and Table 3.1). The landscape is made of trapped underneath the small shrubs and low-lying peneplains (terraces) with depres- forming what is called ‘nebkass’. This land sions (micro-basins) of loamy clayey texture type covers a total land area of 1612 ha or alternating with oval-shaped sandy levees. 9.5% of the commune (Fig. 3.4 and Table It has an undulating relief with beige- 3.1). The natural vegetation is dominated by coloured materials located between the thorny shrubs and trees such as Acacia seyal sand banks (unit tr) and the old alluvial and Balanites aegyptiaca and parkland trees terraces (unit t2 below). The vegetation such as Tamarindus indica. cover is composed of Acacia albida and The dominant soils in this landscape Piliostigma reticulata on the higher eleva- are slightly ‘vertique’ hydromorphic soils tion (sandy levees) and Mitragyna inermis (or French sols hydromorphes à tendance and Diospyros mespiliformis in low lands vertique avec nodule calcaire), which are (micro-basins) (Fig. 3.8). classified as USDA Plinthic Haplustalfs. Two dominant soil types are found in They are deep (105 cm) with a texture vary- this landscape: alluvial sandy soils with ing from sandy loam to clay with depth. As limited development and hydromorphy (or a result, the water storage capacity is high, French sols peu evolués d’apport alluvial with soil water availability of about 20% on faiblement hydromorphes sableux) on the average. The soil OM content is low (0.5%) higher elevations (sandy levees), and the and decreases with depth. Temporary flood- clayey hydromorphic soils with gley (or ing and waterlogging due to runoff water, as French sols argileux hydromorphes à gley well as high soil compaction and glaciation de profondeur) in micro-basins areas. These are the main constraints. These soils are, soils, which can be classified as USDA Typic however, suited to sorghum and rice, and Tropaqualfs, are deep (105 cm) sandy clay have high pastoral value (Table 3.2). 48 O. Badini and L. Dioni

Table 3.2. Agropastoral constraints and potentialities.

Code Constraints Agricultural suitability Grazing suitability Forest suitability ci Very high risk of inundation Restricted only to rice Good Marginal and water logging cropping tr High grounds, soils with Marginal, sandy soils Good Good coarse sandy texture, high with low fertility permeability, low fertility t1 Seasonal water logging Good to sorghum Good High and risk of inundation and rainfed rice cl Risk of inundation and Moderate; suitable for Good Good water logging, costly sorghum; more crops can conservation needed for be cultivated if soil water improvement conservation and management techniques are implemented t2 Very fragile soils very highly Moderate overall; suitable Good Good prone to erosion, deep water for millet; more crops can table and low fertility be cultivated if soil water conservation and management techniques are implemented ca Possible slight inundation Suitable for rice and Good Good due to runoff water, weak sorghum permeability due to high soil compaction vi Shallow soil on hardpan Marginal for millet and Marginal Good (cuirass), very low water sorghum holding capacity, problems of roots penetration, high risk of erosion bk Very shallow soil with rock Unsuitable Unsuitable Unsuitable outcrops

Plains of sandy to loamy materials (unit t2) Two types of soils are found in this This landscape called ‘old alluvial terrace’ landscape: the tropical ferruginous soils (or in French terrasse alluviale ancienne – t2) with limited development (or French sols is made of sandy to sandy loam peneplains ferrugineux tropicaux non à peu évolués) (terraces) found on the eastern and western associated with tropical ferruginous soils edges of the Delta. It covers a land area of with concretions and ‘tints’ (or French sols 4327.6 ha or 25.5% of the commune (Fig. 3.4 ferrugineux tropicaux peu lessivés a taches and Table 3.1). It is almost never flooded et concrétions). The soils are very deep (115 (inundated) due to its higher profile. The cm) with sandy loam texture. The water parent material is made from wind blown storage capacity is low in surface to average sandy and silty materials from the in deeper layers. Soil OM content is low Quaternary period. The land can be distin- (0.5%) on average, and the pH varies from guished by its reddish (due to rubefication) 4.6 (acid) to 6.0. Average phosphorus con- overall colour. Natural vegetation is domi- tent is 3.3 ppm. Nitrogen content is low with nated by protected parkland trees such an average of 0.02%. as Acacia albida (apple-ring acacia) and These soils have no major physical con- Vitellaria paradoxa (shea-nut tree) (Fig. 3.9). straints (Table 3.2) except their low water Landscape and Soil Characterization of Madiama Commune 49

Fig. 3.9. Plains of sandy and loamy materials. holding capacity. Constraints are mainly cuirasse), classified as USDA Aridic Cuiru- chemical with low pH (high acidity) and stalfs. These are shallow (60 cm) gravelly and deficiency in phosphorus and nitrogen. Soil laterized soils. Their hydrology is character- amendments as well as organic and mineral ized by a diffuse and intense runoff that fertilization are needed in order to improve causes sheet erosion. These soils are prone to productivity. These soils constitute the erosion with a slope of about 2–3%. Soil second largest rainfed cultivated unit in the crusting and surface glaciation in addition to commune. Cultivated crops include millet, the shallowness of the soils are the main con- groundnuts and cowpeas. Sorghum is only straints. Cultivation of sorghum and millet is planted in the low-lying areas of the field. To possible but yields are generally low. These ensure a sustainable management of this fragile soils must be protected and soil con- unit, trees and productive pastures could be servation measures as well as appropriate planted. cultural practices should be implemented.

Land underlain by laterite (unit Vi) Ironstone plateau (unit Bk) This landscape type, called ‘laterized This landscape called ‘low hardpan plateau’ indurated pediment’ (or in French versant (or in French bas plateaux cuirassés – Bk) glacis induré – Vi), is made of peneplains covers a total surface area of 526.07 ha or (terraces) with skeletal soil of loamy clayey 3.1% of the commune. The relief in this texture on a substratum of rocks or cuirass zone is characterized by higher elevation, found at a shallow depth of 60–70 cm. It with indurated laterite and soils on hardpan covers an area of 882.4 ha or 5.2% of the (cuirass) and rock outcrops on non-arable commune. On the map in Fig. 3.4, it can be lands. The vegetation is stunted and domi- seen as ring-like shape surrounding the nated by diverse small shrubs. ironstone plateau (unit Bk) with a slight There is almost no soil in this landscape slope towards the landscape type t2. Natural except skeletal and thin gravels covering a vegetation is dominated by diverse small cuirass found at a very shallow depth of shrubs (Fig. 3.10). about 10–20 cm. These very shallow soils The main soils are soils of limited devel- are called Lithosols or USDA Cambic opment on cuirass (hardpan) or Regosols Cuirorthids soils. It has the highest eleva- (or French sols peu évolués d’érosion sur tion (300 m) in the commune. The hardpan 50 O. Badini and L. Dioni

Fig. 3.10. Laterized indurated pediment and ironstone plateau.

(cuirass) is covered by gravelly materials officials as well as researchers and develop- from broken indurated laterized rocks. ment agents can use the information to plan Major constraints for production are the land use and identify special practices shallowness of the soil, the presence of needed to ensure proper performance. cuirass, the very low water holding capacity, From a total of eight landscape types, and the high susceptibility to water erosion. four types, representing a total 13,197 ha These soils are unsuitable to agriculture. If (67.0% of the commune), are suited to rain- agriculture has to be attempted on these fed agricultural production with varying soils there is a need for improved and labour degrees of constraints as described in this intensive cultural practices, including the chapter. These landscapes types are the old removal of gravel, tillage techniques allow- levees and alluvial terraces (unit t1), the old ing an increase of soil thickness and organic alluvial plains and terraces (unit Ca), the and mineral fertilization. hydromorphic loamy plains (unit Cl), and the plains of sandy to loamy materials (unit t2). The hydromorphic floodplains (unit Ci) Conclusion and alluvial levees, representing a total of 2141 ha (12.6% of the commune), are partly This chapter has shown how land and soils used as grazing lands, and in some cases for with different potentials and constraints irrigated rice cropping due to their proxim- are distributed within the boundaries of ity to the Bani river. The two last landscape Madiama Commune. Soil properties for types, found mostly on the east side of the making agricultural and pastoral land use commune, are lands underlain by laterite decisions are identified. Information and and hardpan (unit Vi), and Ironstone Plateau soil parameters generated from this study (unit Bk). They both represent a total of have also served as input for the biophysical 1408 ha or 8.3% of the commune. They models (see Chapter 11) and helped in the are degraded (about 40% bare soil) with assessment of agricultural practices and stunted vegetation, constituted mainly of technologies best suited to the various soils annual pasture grasses such as Loudetia and landscapes of the commune (see and Schoenefeldia, and used as wet season Chapter 12). Community members and their pastures. Landscape and Soil Characterization of Madiama Commune 51

Note (under the direction of Pr Tricart),Strasbourg, France. 1 The term bourgou, derived from the Peul word Gallais, J. (1967) Le Delta Interieur du Niger, No. bourgou meaning forage, refers to a grass grow- 70 Tome I. Institut Fondamental d’Afrique ing in deep basins of the Active Delta that pro- Noire (IFAN) Dakar, Sénégal. duce a lush (rich) pasture as the flood waters Leisinger, K.M. and Schmidt, K. (1995) Survival recede. These flood retreat pastures (pâturages in the Sahel: an Ecological and Deve- de décrue) become the focal point for transhu- lopmental Challenge. International Service mant herds during the dry season. for National Agricultural Research (ISNAR), The Hague, Netherlands. PIRT Project (1986) Projet d’Inventaire des References Ressources Terrestres. Zonage Agroecologi- que du Mali. Ministere de l’Elevage et de Badini, O. (2001) Agro-climatic Assessment l’Environnement, Bamako, Mali. (Description and Analysis) of Madiama Sivakumar, M.V.K., Konaté, M. and Virmani, S.M. Commune in Mopti Region, Mali (West (1984) Agroclimatology of West Africa: Mali Africa). Washington State University, Inter- (Information Bulletin, 19). International national Programs, Pullman, Washington. Crops Research Institute for the Semi-Arid Badini, O. and Dioni, L. (2001) Etude Tropics, Bamako, Mali. Morphopédologique de la Commune de Taylor-Powell, E. (1991) Integrated management Madiama, Cercle de Djenné – Technical of agricultural watersheds: Land tenure and report SANREM West Africa. Washington indigenous knowledge of soil and crop man- State University, International Programs, agement. TropSoils Bulletin No. 91-04. Pullman, Washington. United States Department of Agriculture Soil Beckett, P.H.T. and Webster, R. (1971) Soil vari- Survey Staff (1999) Soil Taxonomy. A Basic ability: a review. Soils and Fertilizers 4, 1–15. System of Soil Classification for Making Bertrand, R. (1973) Systeme de Paysage des and Interpreting Soil Surveys, 4th edn. Plaines du Delta Vif et Moyen du Niger: (USDA-NRCS Agricultural Handbook 436). Etude Morphopédologique de la Quelques Washington, DC. Plaines du Delta (Tome I et II). Institut de Van Keulen, H. and Breman, H.B. (1990) Agri- l’Economie Rurale and Office de la Vallée du cultural development in the West African Niger, Bamako, Mali. Sahelian Region: a cure against land hunger? Blanck, J.P. (1967) Etude Morphologique de la Agriculture, Ecosystems and Environment vallée du Niger. Université de Strasbourg 32, 177–197.

4 Investigation into the Movements of Cattle, Sheep and Goat Herds through the Commune of Madiama*

Adama Ballo and Abdoul Karim Ouattara Centre Régional de la Recherche Agronomique, Institut d’Economie Rurale, BP 205, Mopti, Mali

The Commune of Madiama suffers from a fields during the dry season. Those farmers lack of pasture land, particularly rainy mainly invested in livestock production season pastures (ESPGRN/SANREM-CRSP, own both cattle and small ruminants, and 1999). This lack of rainy season pastures is labour is distributed within the family so extreme in certain villages that some between cultivating crops and herding live- locally owned sedentary and semi-transhu- stock. Finally, there are a small number of mant herds must leave the commune for residents that rely only on livestock, and pastures in neighbouring communes or go pastoralists that visit the commune with on transhumance. However, other terroirs of their herds during the crossing to and from the commune are sites of livestock concen- the rich grazing available in the bour- tration during the return from rainy season goutières of the Niger delta during the dry pastures before crossing into the bour- season. Within these broad generalizations goutières.1 These livestock concentrations are a multitude of specialized husbandry vary in duration from year to year and often techniques, relationships between resident cause multiple conflicts between farmers and non-resident herders and farmers, tradi- and herders. tional pastures within and around the com- This chapter focuses on the system of mune for both resident livestock and animal husbandry within and around visiting pastoralists, and rights of access to Madiama. Within the commune reside resources such as water and grazing. approximately 5100 cattle, 2550 sheep and To document the circumstances sur- 2100 goats. Although nearly all of the com- rounding these pastoral issues in the mune’s families own livestock, the type of Commune of Madiama, a survey was made livestock ownership, and thus husbandry, of all livestock production systems and can be broken into three broad categories. accompanying strategies. This survey helps Those farmers mainly invested in crop pro- to identify and propose sustainable tech- duction own a predominance of small rumi- niques for managing pasture resources and nants and oxen that hired pastoralists herd raising livestock. The survey was carried out during the growing season, and that are left in two stages. The first stage involved focus to graze on crop residues in the owners’ groups, assembled in each of the ten villages

* Translated and adapted by Keith M. Moore. © CAB International 2005. Conflict, Social Capital and Managing Natural Resources (ed. K.M. Moore) 53 54 A. Ballo and A.K. Ouattara

of the commune. In the second stage, a vil- Promani are primarily of the Peul ethnic lage level inventory of livestock resources group with raising livestock as their princi- and management practices was conducted. pal activity. A few villagers are employed in This chapter reports the results of these sur- ancillary activities related to agriculture or veys in two parts. The first part provides a small trade, and others work in small craft general description of the area’s livestock industries in these villages. In contrast, in and pasture management practices. The Bangassi a significant share of the cattle pop- second part of the chapter focuses on tran- ulation consists of oxen owned by farmers shumant herd management during their for ploughing and cultivation of crops movement through the various terroirs of during the rainy season and for transport in the commune. the dry season. In the villages of Téguégné, Siragourou and Toumadiama, the livestock are nearly all oxen. Milking cows are found There are three systems of livestock man- primarily in Nérékoro, Promani and Nouna, agement in the commune of Madiama: where milk production is a major activity. the sedentary, semi-transhumance and transhumance. Cattle and Pasture Management

Sedentary Herd Management Dry season Feed Cattle herds of Madiama commune From the end of the millet and sorghum har- Table 4.1 lists the cattle herds by villages of vests in November to the first rains in June, the commune. local herds of sedentary cattle graze on crop The village of Nérékoro has approxi- residues (millet, sorghum and rice straw) mately 33% of the cattle population in the remaining after the passage of the transhu- commune, followed by Madiama (15%) and mant cattle. Bush and shrub fodder growing Bangassi (12%). Torokoro and Toumadiama on rare pockets of uncultivated land and on have the fewest cattle (2%). Nérékoro has fallow land are also grazed. The grazing of the highest average number of cattle per crop residues is done collectively within household (31.48), followed by Nouna2 each village terroir. All animals have access (with 14.29), Promani (3.47) and Bangassi to these resources once harvests are (3.25). The villages of Nérékoro, Nouna and completed. Towards the end of the dry

Table 4.1. Number of cattle by village and average per household.

Principal Number of Number of Average number of Villages activity cattle* households cattle per household

Torokoro Agriculture 100 74 1.35 Nérékoro Livestock 1700 54 31.48 Bangassi Agriculture 630 194 3.25 Promani Livestock 520 150 3.47 Nouna Livestock 500 35 14.29 Téguégné Agriculture 165 58 2.84 Siragourou Agriculture 410 82 5.00 Tombonkan Agriculture 200 76 2.63 Toumadiama Agriculture 100 150 0.67 Madiama Agriculture 750 365 2.05

Estimated total 5075 1238

* The number of heads indicated in the table is an estimate made in village assembly. Movement of Cattle, Sheep and Goats 55

season (May–April), the availability of these from two (Nérékoro) to 20 (Siragourou). In resources becomes limited and the animals the villages of Bangassi and Promani, in suffer from malnutrition. To minimize the addition to the wells, the animals water consequences of under-feeding during this themselves at the river (Table 4.2). In the vil- period, some livestock owners harvest and lage of Nouna, livestock are watered only at store stocks of millet and sorghum straw for the river. The village of Nouna is located feed before the animals graze the harvested closest to the river. fields. In addition to storing these crop residues, straw from shrubs and bushes are Herding of the livestock stored, as are bourgou, cowpea and ground- nut hay. The agroindustrial feed Bétail In the dry season, herdsmen under contract Huicoma (ABH) is also used as a feed sup- or young boys of the village lead some of the plement. In general, the feed supplementa- livestock. Other animals are left to them- tion strategy observed in all the village is to selves. In many cases, herdsmen are con- distribute ABH to production animals tracted for the rainy period until the animals (dairy, pasture-fattened animals, oxen, asses enter the bourgoutières. Generally Peuls, the and horses), and to weakened and sick ani- herdsmen, are remunerated in kind and/or mals. However, in the villages of Téguégné in cash, and according to the season. The and Siragourou, ABH is not used (Table 4.5). methods differ from village to village (Table This could be explained by the weak pur- 4.3). For instance, in the case of milk cows, chasing power of these agropastoralists. It milk production is used to compensate the should be noted that these types of dry herdsman. season feeding practices apply to sedentary, semi-transhumant and even certain trans- Stabling humant herds that do not enter the bour- goutières. In the dry season, the animals are kept in enclosures at night in order to collect their manure. In Madiama, a villager will often Watering take on the cost of the herdsman in order to The animals are watered at the river and at benefit from the manure produced while the wells. The number of wells per village varies animals are stabled. Some livestock owners

Fig. 4.1. Cattle grazing on the Bani flood plain. 56 A. Ballo and A.K. Ouattara

Table 4.2. Number of watering points per village.

Well with large Villages Traditional wells diameter Bore hole wells Observations

Torokoro 11 2 1 All functional Nérékoro 2 Ϫ 1 All functional Bangassi 16 Ϫ 1 All functional Promani 11 2 1 All functional Nouna ϪϪϪThe animals water themselves at the river Téguégné 10 ϪϪAll dry out during the dry season Siragourou 20 ϪϪAll functional Tombonka 2 ϪϪFunctional Toumadiama 7 Ϫ 2 Functional wells; 1 bore hole in good condition Madiama 11 Ϫ 6Well with low flow and 3 drillings in good state

Total 90 4 12

Table 4.3. Methods of remuneration for herdsmen by village.

Dry season Rainy season

Cash In kind Cash In kind Villages (fcfa) (kg millet) (fcfa) (kg millet) Other remuneration

Torokoro 125/head/mo 0.5/head 1000/head Development of 0.5 ha cereal Nérékoro 9000/herd/mo Ϫ 15,000/herd Bangassi 100/head/mo 1.0/head 500/head Purchase of boots, torch and raincoat for the shepherd Promani 125/head/mo 0.5/head 5000/herd Nouna 7500/herd/mo Ϫ 15,000/herd Téguégné 5000/herd/mo 45/herd 7500/herd Purchase of boots, torch and raincoat for the shepherd; Siragourou 125/head/mo 0.5/head 1250/head Production of 1 ha of cereal Tombonka 125/head/mo 1.0/head 1250/head Toumadiama 600/head/mo Ϫ 12,500/herd Madiama 150/head/mo Ϫ 1000/head Two working days in the field of the shepherd

stake their oxen in their compounds and The rainy season provide them with feed supplements. Feed During the rainy season, cattle graze fodder Health resources on the occasional patches of Both cattle and small ruminants are vacci- grass, fallow land and the borders of fields. nated annually against principal diseases. These animals are primarily oxen that are Other treatments such as the de-parasiting kept in these areas until the end of field cul- are done sporadically. tivation (August). At this time, they are led out of the terroirs and join the other ani- Movement of Cattle, Sheep and Goats 57

mals on neighbouring or distant rainy to avoid damage to the growing crops. The season pastures. Only the animals of cost of employing a herder varies from 5000 Téguégné, Toumadiama and Bangassi to 7500 FCFA/herd/month in the dry season remain in their respective terroirs during and from 7500 to 12,500 FCFA in the rainy this period. In Promani, Nouna and season (Table 4.3).3 Nérékoro, some milking cows are kept close to the village to insure family milk consumption throughout the rainy season. The Number of Small Ruminants in the The milking cows graze on the field borders Commune and in a few pockets of undeveloped land. All categories of livestock in Bangassi and The number of small ruminants (sheep part of the animals of Promani are taken to and goats) in the commune is provided in a 257 ha wooded pasture, called wala-wala, Table 4.4. in the terroir of Torokoro and remain there for the rest of the rainy season. However, to avoid a concentration of animals in the Small Ruminant Herd and Pasture wala-wala,Torokoro livestock do not graze Management these pastures. The use by others of Torokoro’s terroir constitutes a regular Dry season source of conflict. Feed Natural pasture resources and crop residues Watering found in the fields provide dry season feed The animals primarily water themselves in for small ruminants. Feed supplements are temporary ponds existing on the grazed pas- rarely given to small ruminants, except for tures. the animals, particularly sheep, being fat- tened. These sheep receive a wide range of feed resources, including groundnut, Shepherding cowpea and millet hay, millet grain, ABH, During the rainy season the animals are bourgou straw, and kitchen wastes. under the supervision of herdsmen in order

Table 4.4. Number of sheep and goats by village and average per household.

Number of small Average number ruminants* per household Principal Number of Villages activity Sheep Goats households Sheep Goats

Torokoro Agriculture 250 200 74 3.38 2.70 Nérékoro Livestock 300 400 54 5.56 7.41 Bangassi Agriculture 200 250 194 1.03 1.29 Promani Livestock 250 150 150 1.67 1.00 Nouna Livestock 200 300 35 5.71 8.57 Téguégné Agriculture 100 160 58 1.72 2.76 Siragourou Agriculture 250 100 82 3.05 1.22 Tombonka Agriculture 400 200 76 5.26 2.63 Toumadiama Agriculture 200 100 150 1.33 0.67 Madiama Agriculture 400 200 365 1.09 0.55

Estimated total 2550 2060 1238

* Estimates made in village assembly. 58 A. Ballo and A.K. Ouattara

Toumadiama, Nouna and Téguégné spend Watering the rainy season in their respective terroirs Like cattle, small ruminants are watered at grazing on field borders, fallow lands and wells in all the villages except at Nouna undeveloped spaces (Table 4.6). In the where watering is done at the river. Small evening they are tied to the stakes or put in ruminants are watered twice a day. the enclosures. In the villages of Nérékoro and Torokoro, some are maintained by the women and graze at field borders tied to Herding stakes. Others are semi-transhumant in the The small ruminants are left on their own to neighbouring terroir of Timissa for most of browse and graze where they will. However, the season. In Madiama, two-thirds of the in the evening, they are brought back to the herds graze during the day in the terroirs of household compound, typically by the Timissa and return to the village in the small children (boys and girls) of the village. evening.

Stabling Watering In both the dry and rainy seasons, small During the rainy season, animals are ruminants are stabled in the compounds of watered in temporary ponds along their the owners in enclosures made of thorn- pathways to and from grazing areas. bushes or they are tied to stakes. Shepherding The rainy season Unremunerated villagers, generally young boys and/or girls, often herd small rumi- Feed nants. Herding duties are rotated between Small ruminants in the villages of Ban- various family members in Toumadiama. gassi, Tombonkan, Promani, Siragourou, The herder might also be a Peul (local or

Fig. 4.2. Small ruminants staked in a village. Movement of Cattle, Sheep and Goats 59

Table 4.5. Cattle feed supplements, rainy season pastures, watering points and markets frequented by village.

Feed Supplements Watering points distributed in dry Rainy season (frequency/day dry Trade (markets Villages season pastures season) attended)

Torokoro Crop residues, ABH * Timissa Well Madiama- (April–June) 2 times Sofara Nérékoro Crop residues, ABH Tominian, Timissa, Well Madiama (March–June) Fangasso, Ouan 2 times Bangassi Crop residues, ABH Wala-Wala ** Well and river Madiama (April–June) 1 to 2 times Promani Crop residues, ABH Wala-Wala, Timissa, Well and river Madiama- (April–June) Fangasso 1 to 2 times Djenné Nouna Crop residues, ABH Timissa, Fangasso, River Madiama- (April–June) Ouan, Toumidian 2 times Djenné-Sofara Teguégné Crop residues Fallow land Well Madiama- (April–June) 2 times Konio Siragourou Crop residues Laramba Well Madiama (April–June) Timissa 2 times Madiama Crop residues, ABH Timissa Well Madiama- (March–June) 2 times Sofara-Djenné Tombonkan Crop residues, ABH Timissa, Fangasso Well Madiama- (April–June) 2 times Sofara Toumadiama Crop residues, ABH Fallow land Well Madiama (April–June) 2 times

* ABHϭfood Huicoma cattle; ** Wala-Walaϭforest installation pertaining to the terroir of Torokoro (257 ha). non-local), recruited according to a contract, Semi-transhumant Herd Management with varied remuneration. In Promani, in addition to villagers planting a millet field Semi-transhumance is defined in this con- for him, the local Peul herder also receives 4 text as a low frequency of movement by kg of millet per animal. Non-local Peul cattle and/or small ruminants into and out herders receive 8 kg of millet per animal, of the terroir. The animals leave their terroir without profiting from labour and the millet for neighbouring terroirs only when in field. Generally, payment for herding is only search of rainy season pasture. calculated for the ewes and does having had their first labour. The herding of males (rams and bucks) is not remunerated. The payment Cattle is made in kind and in cash. For example, in Madiama, the herder receives 350 FCFA/ Semi-transhumance is practised in the head of ewe or doe plus 500 g millet and villages of Madiama, Nérékoro, Torokoro, 25 FCFA (for the cost of condiments). In Promani, Nouna, Tombonkan and Siragou- Téguégné, the females that have had their rou. With the first rains (June, July), most of first labour cost 200 FCFA per animal per the cattle except oxen are moved to the month plus seven bowls of millet (equiva- neighbouring Commune of Timissa in the lent to 3.5 kg) per month during the rainy search of better pastures (Table 4.5). The season. cattle are the first to leave their respective terroirs for these rainy season pastures. Once the field cultivation is completed, the 60 A. Ballo and A.K. Ouattara

oxen and the few dried-up milk cows join mant management. During the rainy season, the other animals. The duration of the stay these animals are led up on to the plateau in varies from 3 to 4 months for oxen and from search of fresh grazing and watering points 4to5months for the rest of the herd. During in the pastures of Seno (in the cercles of this period, the cattle of seven villages of Bankass, Koro and Douentza) and of Konari Madiama are found on the pastures of (cercle of Mopti). Some travel as far as the Timissa where they spend the whole rainy frontier with Burkina-Faso. These transhu- season exploiting the fodder and waterways. mant herds are generally mixed and made up This zone experiences a strong concentra- of dry cows, heifers, bulls and calves from tion of the animals during this period. the villages of Nérékoro and Nouna. This Timissa is not the only space exploited management practice, characteristic of these by the animals. Animals of the villages of two villages, is partly related to the pastoral Nérékoro, Promani, Nouna and Tombankan vocation of the dominant ethnic group. exploit pasture resources of other neigh- The animals leave for these rainy season bours in the Communes of Fangasso, Ouan pastures in June–July (according to the early and Tominian. They remain there through- or late onset of the rainy season) and return out the rainy season until the harvest or the in November–December (period of har- water of the temporary ponds dries up. The vests). These herds return at the same time as return to their respective terroirs starts in the herds of non-local transhumants. During November with the beginning of harvesting. this passage, all herds graze the same spaces In addition to oxen kept for crop culti- (fallow fields and the borders of fields and vation until August, some milking cows are roads) and use the same watering points. maintained in the villages to ensure family This period lasts for approximately a month milk supply as well as provide milk for mar- on average, until the rice fields are opened keting. Maintaining a core of milk cows is a for crossing into the bourgoutières. Among common practice among the Peuls in the vil- the locally based transhumant herds, 75% lages of Nouna, Promani and Nérékoro. continue in the bourgou, another 5% are milking cows and calves that remain in the village terroir to ensure the family milk Sheep and Goats supply, and 20% remain all year in the dry zone. Ten per cent of the animals that go into Herds of small ruminants are also found on the bourgou are dairy cattle, oxen, or weak or the same rainy season pastures of neigh- sick animals that graze the young growth of bouring communes as the cattle. They are bourgou for a month. Once the harvests are primarily animals from Madiama and finished, these animals go up to the millet Torokoro (Table 4.6). In these villages, the fields to graze on the crop residues until livestock are stabled for the night in enclo- the next season of rains. Those remaining in sures of thorn-bush to avoid crop damage. the bourgoutières leave with the non-local Nevertheless, sometimes crop damage does transhumant herds at the beginning of the occur and friendly solutions are always rainy season. found. The return of small ruminants to their respective terroirs occurs at the same Sheep and goats time as the cattle. The local small ruminants do not practice transhumance. Transhumant Management

Local herds Non-local herds Cattle During October and November, the terroirs Some cattle herds of the commune of of certain villages of the commune become Madiama are continually under transhu- zones of livestock concentration for local Movement of Cattle, Sheep and Goats 61 and girls stake to Timissa them out a semi-transhumant in remains herd Timissa Sheep fattening for out 0 Women stabled the house stabled Rainy season Control around villagearound the house Promani stabled Promani Siragourou roaming Laramba Staked roaming stabled Dry Sedentary small ruminant management practices by villages of Madiama. practices by villages ruminant management Sedentary small okoro okoro roaming Free Staked out Staked out near once/year Women A part goes illages season Day Night Vaccination cash income Observations able 4.6. ombonkanoumadiama roaming Free roaming Free Fallows Fallows Staked out Staked out once/year once/year Women Women Siragourou Free Free Siragourou Téguégné Dougoukourani Free T T Madiama roaming Free Fallow land roaming Free Staked out or Timissa 0 Staked out Women once/year Men and women During rainy season Nérékoro roaming Free Plain Staked out near once/year Men and women Bangassi Promani roaming Free roaming Free Wala-wala Space Staked out or Staked out or once/year Women once/year Women Nouna roaming Free Plain Staked out or once/year Men and women The evening boys T V Tor 62 A. Ballo and A.K. Ouattara

and non-local herds returning from rainy RESIDENCE TIME OF THE ANIMALS IN THE COM- season pastures in the direction of the bour- MUNE Each year the date of crossing is goutières. fixed by the administration and technicians, in collaboration with the local authorities of the various zones concerned. This date is Cattle communicated and diffused by all forms of The feed and water points of the terroirs of media. Prior to the date of river crossing to all the villages are used by the non-local the bourgoutières, the duration of stay varies herds. according to village (Table 4.7), with the ani- mals spending an average of 1 month while ORIGIN OF THE ANIMALS The herds come waiting for the crossing. Before the end of from neighbouring communes (Konio, the harvests, the animals graze on the bor- Sofara Timissa), certain Cercles of the area ders of the fields of millet and rice, unde- (Bankass, Bandiagara, Koro) and even from veloped lands, fallow lands and residues Burkina-Faso. The majority of herdsmen are from harvested fields. The watering takes young Peul men. Dogon, Sore and Mossi place in the plains. During this period, there ethnic groups are also involved (Table 4.7). are numerous conflicts, causing the villages of Torokoro, Tombonkan, Bangassi and PERIOD OF ARRIVAL OF THE ANIMALS IN THE Promani, to organize their populations into COMMUNE OF MADIAMA The herds gener- monitoring groups to prevent livestock ally arrive at the moment of harvests during damage to the crops. November. This date is not fixed since the To reach the bourgoutières, the only descent of the animals is a function of the point of crossing in the commune is Nouna. unavailability of fodder and water resources Three groups of herds can be distinguished on the plateau. During the years of bad rain- in this moment in particular: fall, the animals go down towards the delta Group 1 (95%) herds cross to reach the bour- earlier than normal because water becomes goutières and remain until the first the limiting factor on the rainy season pas- rains, a period lasting from 7 to 8 ture. months (November to June–July). Group 2 (3%) herds, after having crossed, FLUX OF THE ANIMALS IN THE COMMUNE return at the end of 1 or 2 weeks to graze The pressure of the animals on the fodder the fields of millet of the terroirs of resources of the various terroirs is related to Nouna, Torokoro and Bangassi, and the proximity of these terroirs to the remain in the villages during the rest livestock corridors and river crossing points. of the dry season (November to The principal passage of the animals through June–July). the commune follows this path: Sira- Group 3 (2%) herds do not cross and remain gourou-Nérékoro-Promani-Nouna. There is in the area only to graze crop residues also a passage through Konio-Kansara- in the millet and rice fields of the Kessédougou-Toumadiama-Téguégné. The commune. number of herds that pass daily through these terroirs varies. In the villages of Once harvesting is completed, local animals Nérékoro, Promani and Nouna, an average of as well as transhumants graze the crop up to 300 herds per day may pass with the residues (millet and rice) of all terroirs. number of heads per herd varying between Cattle are watered initially at the river and 40 and 200. In the villages of Bangassi, ponds, and later at the wells. Siragourou, Toumadiama and Téguégné, an The livestock return from the bour- average of 20 herds pass by daily in the direc- goutières during June and May with the tion of the bourgoutières. The duration of coming of the rains and the rising of the passage in the terroirs varies from 1 day to 15 river. However, the stay in the bourgoutières days, with an average of 7 days for the can be prolonged until August if the herba- passage. ceous biomass and watering points are Movement of Cattle, Sheep and Goats 63 to rainy Stay (15 days) Stay (2–3 days) Stay (1–2 days) Stay (10–15 days) Stay (30 days) Period of return from bourgoutières bourgoutières February–April June–August Duration of residence Duration of able to pass perable to pass In the Commune before In the Number of herds herds Number of day at the time ofday at the time size Average the date of returnto graze on before Séno, 15 40–60 30 – May–June Bankass, 1000 50–100 30 – June onikoro onikoro ogo, Tiocki Bankass Bankass Faso, Kessédougou W Koro Sofara, Koro, Konguéna Sofara, Ouenkoro, T Bankass, Bandiagara, Timissa, Sofara Bankass, Ouenkoro, Sofara, Fangasso, Konio Gonkoro, Burkina- Gonkoro, ranshumant cattle herds movement by village in the Commune of Madiama. by village in movement cattle herds ranshumant T okoro okoro Burkina-Faso, Séno, 5 100 1–7 3 December–February June–July mbonkan Burkina-Faso, 30 30–200 15–20 – June illages Origin the crossing of herd (days) the crossing (months) residues crop season pastures able 4.7. oumadiama Bankass 1–5 20–100 30 – June Siragourou Burkina-Faso, 25To 400 60 3 T Bangassi Bandiagara, Bankass, Promani Burkina-Faso, Bankass, 20 400 50–200 50–300 7–15 10–15 6 January–June 2 May–June July–August June–August V Tor T Burkina-Faso, Nérékoro 500 100–150 30–60 – May–June Nouna Téguégné Burkina-Faso, Burkina-Faso, Séno, 100–200 50–200 2–7 7 December–June June–July Madiama Séno, 64 A. Ballo and A.K. Ouattara

Fig. 4.3. Cattle grazing stubble in a farmer’s field.

late in occurring in the rainy season pas- the terroirs in the direction of the bour- tures. Under these conditions, damage is goutières. noted on the crops traversed by the few late These small ruminants arrive from the herds (1–2%). In contrast, when the rainy Cercles of Koro, Bankass, Bandiagara and season pastures are established early, the Korientzé, the Cercle of Niafunké in the area animals leave the bourgoutières before of Tombouctou, and from Burkina Faso. The planting. herders are either Peul or Dogon. Normally, The livestock return from the bour- the small ruminants precede the cattle and goutières is often disorderly with each generally arrive during September, before herder taking his own path. The duration of the harvest begins. These herds of small this passage of the animals varies according ruminants are not held to await the date of to terroir.Inthe majority of the villages, the crossing. passage is short, not exceeding two to three The livestock corridor is the same as the days. However, the duration of the passage one followed by the cattle leading to the through the villages of Téguégné, Nérékoro bourgoutières. All the categories of herds are and Toumadiama can reach 15–30 days if found: mixed herds (sheep and goats), herds the rainfall comes early to these villages. The of sheep and herds of goats. The passages of village of Siragourou constitutes a special the goat herds are announced in Madiama, case. Indeed, some transhumant herds can Siragourou and Téguégné. Three to 20 herds remain in the terroir on the pastoral zone, of sheep can pass each day in some villages Darilalan, some 2–3 months. Some herds with the number of head per herd varying can remain during the entire rainy season in from 50 to 400. For goats, the number of the terroir on the pastoral zone, Laramba. herds per day varies from 5 to 30, with The longer stay of these animals in the vil- 30–400 head per herd. lage terroirs generally leads to conflicts While the duration of passage for a related to crop damage. single herd is quite brief, normally lasting only as long as the time it takes to walk through, which rarely exceeds 3 days, Small ruminants the complete passage of all of these herds Like cattle, small ruminants also exploit the can last from 1 to 3 months. During this resources of the commune as they traverse time, the borders of fields are grazed. Movement of Cattle, Sheep and Goats 65 (staying 3 days) bourgoutières Commune before return from crossing size herd Average Days in Date of the moment of the moment Number of herds herds Number of passing per day at passing per Koro Koro 5 3 200 100 – June Saraféré, 40 30 60–150 100–300 – June Borated 40 20 800 800 – May–June Koro, 3–10 – 40–200 – – May–June Koro, 20 – 50–100 – – – Gouma, 6 8 100–150 20–100 – – Douentza, 2–5 5–10 30–100 100–300 15 – Niafunké, Douentza, Gouma, Me Binga, Sarè-Yamou Me Binga, Sarè-Yamou Bandiagara, Sofara Séno-Bankass, Bandiagara Borated Douentza, Borated, Korientzé, Konna, Fatoma Borated, Douentza Bankass, Koro Bankass, Koro or October and of Guimbala October and of Guimbala Month Small ruminant transhumant movements by village in the Commune of Madiama. by village in the transhumant movements Small ruminant oko September oko September mbonkan September Burkina-Faso, illages of arrival Origin Goats Sheep Goats Sheep crossing able 4.8. oumadiama November Séno-Bankass, Bangassi September Promani September of Kounari Zone 30 20 15–80 50–200 15 – Nérékoro October Nérékoro Burkina-Faso, Siragourou October Siragourou Bandiagara, Téguégné September Zone of Kounari, 10 3 60–100 60 1–30 May Nouna Mid-October To T Tor Madiama September Séno-Bankass Madiama September 1–10 – 50–100 – – – V T 66 A. Ballo and A.K. Ouattara between Existence of herdsmen and herdsmen Number of livestockNumber of social bond erroir T verage stay in verage stay A Rainy season pastures (months)pastures (days) Cattle Sheep Goats the village (%) Most common Seno (36) Koula (9) Soufroulaye(9) Seno (38) Konari (23) Saréyamu (13) Seno (13) Burkina (5) Burkina (33) Seno (16) Kougué (9) Burkina (6) 18Commune 16 27 377 38,692 150 13,032 177 12,077 43 43 59 (average/total) rainy season Seno (24) Koula (14) Who (14) Fangasso (53) Ouan (20) Frequently used rainy season pastures, average stay, number and destination of the transhumant herds for eight villages in the Commune of Madiama. okoro Timissa (38) 4 4 4.5 34 14 21 44 4326 34 3943 36 3585 56 38 44 illage (%) pastures Cattle Sheep Goats Cattle Sheep Goats Herds Head Herds Head Herds Head Cattle Sheep Goats able 4.9. ombonkan Timissa (62) 4.5 4.5 5 39 45 50 7 532 6 329 7 422 71 50 71 oumadiama Timissa (27) 5 5 5.5 3 3 3 14 1140 8 640 6 465 29 25 50 T Bankassi Timissa (26)Nérékoro 3 Timissa (78)Téguégné 3 Timissa (47) 5.5Nouna 3 5 5.6 Timissa (53) 18 5.5 4.5 5 23 3 5.5 19 – 10 3 23 – 1 4 2175 25 10 83 16 1130 9770 56 – 12 5550 69 – 3490 810 2 126 64 39 12,519 700 2170 42 41 100 – 3600 67 100 – 100 0 – 14 23 – 48 – – V Madiama (28) Timissa 5T 5 5 15 21 81 24 2680 15 2800 11 1025 4 33 55 T Tor Movement of Cattle, Sheep and Goats 67

Because the passage of the small ruminants species and the maximum stay 180 days for causes less crop damage than cattle and the amixed herd of sheep and goats. pressure of the small ruminants on the nat- During the days of the survey, 24 herds ural resources of the commune is much less of cattle, 15 of sheep and 11 of goats were compared to that of the cattle, the conflicts listed. The total number of livestock was related to their passage are limited. Indeed, approximately 2680 head of cattle, 2800 of the small ruminants pass through all the vil- sheep and 1025 of goats, amounting to 2527 lage terroirs in the commune. After the UBT (tropical cattle unit: 1 cattleϭ0.8 UBT; crossing, the herds remain in the bour- 10 sheep/goatsϭ1 UBT). All transhumant goutières until the beginning of the rainy cattle herds listed moved to the bour- season. As with the descent into the bour- goutières after their stay in the terroir. As for goutière, the small ruminants return quickly the small ruminants, 32% of the sheep and to their rainy season pastures, although each 59% of goats remained to use the resources shepherd follows the path of his choice. of the terroir for the following 6 months. The transhumant livestock pressure of cattle on the resources of the terroir is less than that A detailed study on the movement of tran- of the small ruminants. In addition to the shumant livestock was conducted in eight of transhumant herds, it is necessary to add the the ten villages of the Commune of Madiama. local herds that amount to 577 head of cattle The study involved collaborating members and 60 of goats. These cattle and goats of the Natural Resource Management remain in the terroir for a full 8 months of Advisory Committee (NRMAC). A collabora- the year. The majority of the cattle (96%) tor villager was chosen in each village and and sheep (67%) herdsmen do not have a provided with a book for data collection. The local landlord in Madiama, while less than data collected included the number of herds half of goat herders (45%) do not have a (cattle, sheep or goats), the approximate local landlord. This could partly explain the number of animals in each herd, duration in prolonged stay of the goat herds. the terroir,location during the rainy season, destination during the descent and finally the existence of a social bond between the Toumadiama herdsmen and villagers. This investigation was designed to complement the previous Transhumant livestock listed as passing data collection discussed in the first part of through this village use the rainy season pas- this chapter. The survey data cover the tures of Fangasso (53%), Timissa (27%) and period from harvest until the return of the Ouan (20%). The average time of stay on livestock from the bourgoutières. The results these pastures is 5 months for cattle and are summarized below by village. sheep and 5.5 months for goats. Contrary to the terroir of Madiama, Toumadiama constitutes a crossing point of the trans- Madiama humant livestock towards the bourgoutières. Indeed, the time spent in the terroir does not Among the rainy season pastures most fre- exceed 3 days. During the investigation, 14 quented by the livestock encountered in the herds with a total of 1140 head of cattle, eight terroir of Madiama are Timissa (28%), herds totalling 640 sheep and six herds Douentza (24%), Koula (14%) and Woh totalling 465 goats were listed. This corre- (14%). All species stay on these pastures for sponds to 1023 UBT. The locally owned ani- an average duration of 5 months. The dura- mals included a cattle herd with 70 head, one tion of the stay in the terroir of Madiama sheep herd with 45 head and one goat varies according to the animal species, with herd with 30 head. All the herds listed as sheep remaining nearly 3 months, cattle only transhumant pass through in the direction 52 days, and goats just 2 weeks. The mini- of the bourgoutières. Therefore, the fodder mum duration in the terroir was 2 days for all resources of the village terroir of 68 A. Ballo and A.K. Ouattara

Toumadiama do not experience high animal Tombonkan is an average of 39 days for grazing pressure. cattle, 45 days for sheep and 50 days for goats. The minimum time of stay is 5 days for cattle and sheep and less than a day for goats. Torokoro The maximum duration of stay is 180 days. After this stay on the terroir of Tombonkan, The livestock passing through the terroir of the majority (82%) of the cattle and sheep Torokoro are most likely to have come from moves down to the bourgoutière. The rainy season pastures in Timissa (29%), remainder of the cattle and sheep stay in the Bankass (24%), Douentza (12%), Koula terroir. The goats are dispersed between the (9%), Tominian (9%), Soufroulaye (9%), bourgoutière (47%), the terroir of Madiama Bandiagara (5%) and Dialassagou (3%). The (47%) and Tombonkan (6%). In addition to cattle and sheep herds remain in Torokoro those transhumants that make a prolonged for an average of 4 months and the goat herds stay on the terroir of Tombonkan, there are for 4.5 months. During 43 days of investiga- also local herds of 125 cattle, 55 sheep and 55 tion, there were 44 cattle herds with a total goats. Nearly three-quarters of goat and cattle of 4326 head, 34 sheep herds totalling 3943 herdsmen and half of the sheep herders head, and 36 goat herds with 3585 head. maintain social bonds with the villagers. This totals some 4214 UBT. The fodder resources of this terroir are grazed on aver- age 34 days for cattle, 14 for sheep and 21 for Bankassi goats. The minimum time of stay is 1 day, with a maximum stay of 270 days for cattle The herds passing through Bankassi come and 60 days for small ruminants. After their from the rainy season pastures of Timissa stay on the terroir, all of the cattle, and two- (17%), Fatoma (15%), Saréyamou (13%), thirds of the sheep and goats head into the Diallassagou (13%), Tominian (9%), Jama- bourgoutière. The remainder of sheep and kan (9%), Konari (8%), Bankass (6%), Niako- goats migrate towards the prefecture of San. ungo (4%), Farimaké (4%) and Burkina Faso However, some animals remain in the terroir (2%). The residence time on these pastures is for up to 8 or 9 months. These include 2% of on average 3 months for all livestock. During transhumant goats in addition to local live- the 34 days of the survey, transhumant herds stock consisting of 156 cattle, 57 sheep and identified included 23 cattle herds of 2175 96 goats. Many herdsmen have social bonds head, 16 sheep herds of 1130 head and 12 with the village. Indeed, 56% of cattle herds- goat herds of 810 head. These animals, men, 38% of sheep herdsmen and 44% of which correspond to 1934 UBT, graze the goat herders have a landlord in Torokoro. forage resources of the terroir for an average of 18 days for cattle, 23 days for sheep and 19 days for goats. The minimum time of stay is Tombonka 7 days for cattle and sheep and less than 1 day for goats. The maximum stay is 28 days The results of the investigation show that the for the cattle and sheep and 30 days for goats. pastures most frequented by transhumant All the animals listed in the terroir during animals on the terroir of Tombonka are those the period of investigation enter the bour- of Timissa (31%), Tominian (31%), Bandi- goutières. The cattle (39%), sheep (42%), agara (23%) and Koro (15%). These rainy and goat (67%) herdsmen maintain relations season pastures are used on average for 4.5 with villagers. months by the cattle and sheep and 5 months by goats. There are seven cattle herds estimated at 532 head, six sheep herds of 329 Nérékoro head, and seven herds of 422 goats. These 20 herds are estimated at 500 UBT. The time of Cattle, sheep and goat herds counted on the stay of the animals on the terroir of terroir of Nérékoro spend the rainy season Movement of Cattle, Sheep and Goats 69

on pastures primarily in Timissa (78%) and from the rainy season pastures of Timissa Burkina (5%). These pastures are grazed for (53%), of Seno (16%), Kougué (9%) and 5.5 months by all livestock species. During Burkina Faso (6%) where they have been the period of the survey (37 days), 83 cattle during the preceding 5 months. On the herds totalling 9770 head, 69 sheep herds whole, 126 herds amounting to 12,519 head, totalling 3490 head, and 64 goat herds or about 10,015 UBT, were listed. The dura- totalling 2170 head were listed. These num- tion of the stay on the terroir averages 25 bers yield a total of 8382 UBT. The herds days with a minimum duration of 1 day and remained on the terroir for a brief time, 3–4 maximum of 120 days. All of these cattle days, with a minimum stay of 1 day and a descend into the bourgoutière. Half of the maximum stay of 7 days for sheep, 12 days herdsmen do not have any social ties with for goats and 15 days for cattle. All the tran- villagers of this terroir. shumant animals descend into the bour- goutière. All the herdsmen (100%) have a social bond with the village. In addition to Conclusion the transhumant herds, the local livestock includes seven herds of cattle totalling 138 The rainy season pastures of the Commune head, three herds of sheep totalling 28 head of Madiama are insufficient. Only the live- and two herds of goats totalling 90 head, stock in three villages out of the ten remain accounting for a total of only 141 UBT. in their terroir (sedentary livestock manage- These dairy animals remain on the terroir ment) for the rainy season. In the other vil- during the entire year. This is for both auto- lages, the few ruminants and oxen that consumption and sale of milk and dairy remain survive on small pockets of undevel- products. oped land, the rare fallow, and the borders of the fields. The vast majority of commune- owned livestock are outside of their own ter- Téguégné roirs, either in semi-transhumance or in full transhumance. The rainy season pastures of Timissa (47%) Many non-local transhumant herds of and Seno (33%) are the source of transhu- cattle, sheep and goat herds collectively mant livestock identified in the terroir of graze the crop residues being used as feed Téguégné. The sheep stay here 4.5 months, supplements for livestock during the dry the cattle 5 months and goats 5.5 months. On season. Their herders generally camp in the terroir, the average stay of transhumant proximity to the croplands, often before the animal is 10 days for cattle and goats and less end of harvesting, thereby creating many than 1 day for sheep. The maximum dura- conflicts between their herders and local tion of stay is 30 days. There are 56 cattle farmers. During the dry season, some tran- herds (5550 head), seven sheep herds (700 shumant herds continue their movement head), and 41 goat herds (3600 head). These through to the bourgoutières, while others correspond to a total of 4870 UBT. All the remain in the terroirs of the commune to cattle descend into the bourgoutière, while exploit the crop residues. The majority of the small ruminants stay on the non-flooded the transhumant animals inventoried in the pastures. The vast majority of herdsmen do Commune of Madiama from November to not have any social ties with the villagers. December 2001 return from the rainy season pastures of Timissa and Seno where they had been for an average of 5 months. Nouna The length of time that these herds remain in the commune varies from 3 to 50 The data collected in the hamlet of Nouna days, depending on the terroir. The average (part of the officially twinned village of Tatia stay of the cattle and sheep herds in the Nouna) refer only to cattle. The transhumant commune is 17 days, which is lower than cattle counted on the terroir of Nouna arrive that of goats (27 days). The number of UBT 70 A. Ballo and A.K. Ouattara

and the time of stay allow for a crude esti- between some agriculturalists and pastoral- mate of livestock pressure on the feed ists in which the herder spends time with resources of the commune. The terroir of his herd on the farmer’s fields to graze Nouna experiences the greatest pressure, residues and fertilize the soil, the amount of followed by the terroirs of Torokoro, time spent by herds on Madiama’s fields is Téguégné and Nérékoro. decreasing, resulting in a loss of organic The Commune of Madiama is a live- matter. With the adoption of Mali’s Pastoral stock corridor for passage from rainy season Code in 2001, rules that resemble the his- pastures to the zones of bourgou. Indeed, toric consultative process of the Dina have nearly all of the inventoried livestock pass been set in place to determine the dates for through the commune to the bourgoutières. entrance and exit from the Niger Delta. This position as a zone of transit exposes the These dates are based on harvest and plant- commune’s feed resources to intense ing. This pastoral code begins to address exploitation. Indeed, more than 30,000 UBT this conflict between agriculturalists and remain for an average of 20 days within the pastoralists in the crossing of Madiama into terroirs of the commune. A rate of con- the delta. However, to improve soil fertility, sumption of 2 kg dry matter/UBT corre- there remains a need to encourage the tradi- sponds to the removal of 1200 tonnes of feed tional relationships between the two groups resources (dry matter weight). Only a small for fertilizing agricultural field. part of these resources is restored to the soils through manure deposited mainly on the croplands. During their stay in the com- Notes mune, the transhumant herdsmen may work 1 under contracts that allow them to stable The bourgoutière is where bourgou, Echino- chloa stagnina, is grown. Bourgou is a highly their own livestock in the fields. However, productive plant that grows underwater with this is most often dependent on the existing the seasonal rise of the Niger River and the relationship between herdsmen and the associated system of lakes and marshes. As the landlords of the various terroirs. In this water falls the bourgou becomes the major regard, goat herders have more and better source of fodder for a wide region. relationships with the villagers in these ter- 2 Nouna is the Peul dominated hamlet of the roirs than cattle or sheep herdsmen. administratively twinned village of Tatia The results showed that during the Nouna. 3 rainy season the natural pastures of the rural FCFA refers to the currency in Mali, franc de district of Timissa are of cardinal impor- la Communauté Financière d’Afrique (franc of the African Financial Community). The tance to the management of the transhumant exchange rate varies, but usually about 500 herds in the region as well as the vast major- FCFAϭ$1.00. ity of the locally owned livestock from the Commune of Madiama. This region deserves further investigation in order to Reference better understand the management of these natural pastures and the characteristics of SANREM CRSP – West Africa (1999) Parti- the transhumant herds that impact the cipatory Landscape/Lifescape Appraisal: commune. Synthese du diagnostic participatif réalisé à The natural resources available for live- Madiama, Nérékoro, et Tombonkan du 1er au 14 février 1999 (Sustainable Agriculture stock production within the Madiama and Natural Resource Management Colla- Commune are limited. Water and grazing borative Research Support Program – West resources are scarce, and there is significant Africa Site, Working Paper No. 99–02). conflict over crop damage by herds as they Office of International Research, Education, cross the commune for the delta. Although and Development, Virginia Tech, Blacks- the traditional relationship still exists burg, Virginia, pp. 1–94. 5 Land Use Changes in Madiama Commune

Randolph H. Wynne,1 Mohamed Sidi Mohamed Touré,2 Nina Sengupta1 and Adama Ballo2 1Center for Environmental Applications of Remote Sensing, Virginia Tech, Blacksburg, VA 24061–0324, USA; 2Centre Régional de la Recherche Agronomique, Institut d’Ecomonie Rurale, BP 205, Mopti, Mali

A thorough understanding of the transfor- current land use (see Wardell et al., 2003) mation of opportunistic grazing manage- and (ii) unpredictable random elements ment by transhumant herders is critical resulting in sudden changes (see Reenberg, to understanding and avoiding natural re- 2001). source-based conflicts as well as to ensuring Having a sound picture of land use economic and food security. Unfortunately, pathways in this region is, therefore, impor- however, there is very little quantitative tant for a variety of reasons. The remainder information on the magnitude and type of of this chapter elucidates the methods used changes in land use that have occurred in to determine land use change over an recent decades. This is particularly problem- approximately 50 year period, reports the atic given that recent transformations in gov- results, and discusses the implications of ernance in the region have created a cadre the findings. of newly empowered local-level decision makers who are tasked with natural resource management, yet are often only partially Methods aware of important changes in the natural resource base. High quality information on A field campaign was carried out in August land use trends is, of course, equally import- 2002. In situ data collected at that time ant to scientists and managers focused on focused on documenting the location and a range of scales. type of land use at a minimum of 100 loca- Further exacerbating the potential prob- tions in and around each of the project study lems facing managers of a natural resource areas. Two QuickBird images were acquired base that is changing in ways not fully for Madiama, one in 2002 (Fig. 5.1) and a understood is the often untested assump- second in 2003. Fortunately, there was also tion of unidirectional expansion of fields on an historical high quality panchromatic to marginal land once reserved for other stereo photograph of the Madiama area uses. As noted by Reenberg et al. (1998) the acquired on 30 January 1952 (Fig. 5.2). dynamics of cultivation patterns are quite Such imagery provided a truly unique complex, and are affected by both biophysi- opportunity to infer land use and carbon cal and socio-economic factors at a range of change over 50 years. scales. In addition, there are often: (i) sub- A team from the United States Depart- stantial lags between historical policies and ment of Agriculture–Agricultural Research © CAB International 2005. Conflict, Social Capital and Managing Natural Resources (ed. K.M. Moore) 71 72 R.H. Wynne et al.

Fig. 5.1. QuickBird image of Madiama and vicinity acquired 11 November 2002.

Service (USDA–ARS) performed a super- sizes, the availability of only a single date of vised classification of Landsat 7 ETM+ ima- QuickBird imagery, and 2002 rainfall that ges acquired over Madiama (Wynne et al., led to poor yields or outright crop failure, 2003). Unfortunately, due to the small field crop separation was very difficult. The team Land Use Changes in Madiama Commune 73

Fig. 5.2. Panchromatic aerial photograph of Madiama and vicinity acquired 30 January 1952. (Note: The small black circles are the random points at which land use was interpreted on this photograph, obtained from the Institut Geographique National of France.) used a radiative transfer modelling appro- fit of classification schemes assuming the ach to develop leaf area index maps for all separation of these categories in space. crops (aggregated), which are useful for gain- The panchromatic photography was ing a rough understanding of the spatial dis- not suitable for any sort of spectrally based tribution of biomass. None the less, it was discriminant analysis of land use or cover clear from the 2002 results that the land since it only contained one spectral band cover in the region was not very amenable to (black and white). Furthermore, because even quite advanced semi-automated classi- there are so few fixed cultural features fication techniques. Possible issues con- (building footprints, roads, etc.) suitable as tributing to this include the following: horizontal ground control points in either (i) average field area, whether farmed or the photos or the 2002 and 2003 QuickBird fallow, is at least an order of magnitude images, accurate image co-registration was smaller than comparable agricultural fea- not feasible. The aerial photos could be tures in most of the developed world, (ii) viewed in stereo, which (for the trained ana- frequent interspersion of farmed land, trees lyst familiar with local conditions) slightly and cultural features is the rule rather than offset the disadvantage of not having infor- the exception, thereby leading to a very poor mation in colour. 74 R.H. Wynne et al.

The classification scheme was based ing system (GPS) receivers for navigation on the work of Badini et al.(2001; see improved the quality of the interpretations Chapter 3), which provided detailed infor- by allowing the analysts to learn how par- mation concerning the soils and land use ticular land uses in situ appeared on the practices in the commune. The United States imagery. Geological Survey (USGS) Level I classifica- tion scheme (Anderson et al., 1976) was used as a point of departure, but the USGS num- Results bering scheme and class descriptions were amended to better reflect local conditions. The results are shown in Table 5.1. The classes used are as follows: bare soil/soil Rounded to the nearest per cent, the with litter, brush/brush land, cultivation, principal trends are as follows: bare (uncul- park agroforestry, riparian vegetation, road/ tivated) soil has decreased by 23%, woody settlement, pasture, denuded area, river/ vegetation has decreased by 27%, cultiva- water and stone quarry. tion has increased by 40%, park agroforestry The protocol developed required care- has increased by 10%, the area of degraded/ ful implementation by the Malian analysts denuded soil has tripled (2% to 6%) and who visually interpreted the land use in pasture has decreased by 29% (from 32% to these broad categories at 300 random points 3%). In this case the general observation on the 1952 aerial photos (Fig. 5.2) and the that increasing population has increased 2002 QuickBird image, with different the cultivation of marginal lands once reser- random samples used for each date. The ved for other uses has been verified and number of points in the first data set was quantified. later circumscribed to the 205 points that fell on the photograph that covered nearly all of the study area due to some difficulties Discussion in utilizing the point grid on adjacent pho- tographs. While this reduction to 205 points With the reduction of pasture and brushland had a random element, the fact that the data from 63% of the landscape to less than sets were slightly different in size, even 8% of the landscape, opportunistic grazing given the same study area, is a clear limita- management by transhumant or resident tion. Field visits to the random points used herders is no longer sustainable in the for the 2002 imagery using global position- commune. Furthermore, careful use of the

Table 5.1. Land use change in Madiama: 1952 and 2002–2003.

1952 2002–2003

Land use Frequency Per cent Frequency Per cent Per cent difference

Bare soil 21 10.3 37 12.3 2.0 Brush/brush land 65 31.7 15 5.0 Ϫ26.7 Cultivation 35 17.1 170 56.7 39.6 Park agroforestry 2 1.0 34 11.3 10.3 Riparian vegetation 2 1.0 4 1.3 0.3 Road/settlement 5 2.4 5 1.7 Ϫ0.07 Pasture 65 31.7 8 2.7 Ϫ29.0 Denuded area 4 2.0 18 6.0 4.0 River/water 6 2.9 6 2.0 Ϫ0.9 Stone quarry 0 0.0 3 1.0 1.0 Total 205 100.0 300 100.0 Land Use Changes in Madiama Commune 75 remaining grazing resources will be req- References uired to avoid increasing the denuded/ degraded area even further. The qualitative Anderson, J.R., Hardy, E.E., Roach, J.T. and observations indicate that many of the Witmer, R.E. (1976) A Land Use and Land remaining wooded areas are so marginal in Cover Classification System for Use with soil quality that they are unlikely to be cul- Remote Sensor Data (United States Geolo- gical Survey Professional Paper 964). US tivated even with additional population Government Printing Office, Washington, pressure. However, they are likely to face DC. increasing grazing pressure that may impact Badini, O. and Dioni, L. (2001) Etude morphopé- their ability to provide a wide variety of dologique de la Commune de Madiama, forest products, including firewood, con- Cercle de Djenné – Technical report struction materials, and traditional foods SANREM West Africa. Washington State and medicines. Reducing grazing press- University, International Programs, Pull- ure through community-managed rotational man, Washington. grazing is therefore indicated from the Reenberg, A. (2001) Agricultural land use pattern perspective of both long term food security dynamics in the Sudan-Sahel: towards an event driven framework. Land Use Policy 18, and natural resource management. In terms 309–319. of cultivation, one bright spot is that approx- Reenberg, A., Nielsen, T.L. and Rasmussen, K. imately 8% of the current cultivated area is (1998) Field expansion and reallocation in currently fallow. However, this is minimal the Sahel: land use pattern dynamics in a given past practices of 15 year fallows. It fluctuating biophysical and socio-economic was not possible to determine whether this environment. Global Environmental Change percentage of land left out of cultivation has 8(4), 307–327. changed in the last 50 years, but the fact that Wardell, D.A., Reenberg, A. and Tøttrup, C. (2003) traditional practices ensuring long term soil Historical footprints in contemporary land viability for agriculture have not been aban- use systems: Forest cover changes in savan- nah woodlands in the Sudano-Sahelian doned even in the face of population pres- zone. Global Environmental Change 13, sure is encouraging. 235–254. In summary, with excellent cooperation Wynne, R.H., Moore, K.M., Doraiswamy, P.C., between US and Malian scientists, we have Badini, O., Touré, M.S.M. and Ballo, A. quantified trends in land use in the (2003) Remote Sensing of Land Use and Soil Madiama Commune, Mopti Region, Mali. Carbon Changes in a Semi-arid Agro- The information provided by this study has pastoral System: Case Study in the Madiama implications for sustainability, food secu- Commune, Mopti Region, Mali (Technical rity, and (potentially) income from evolving Papers, Africa GIS, 3–7 November 2003). carbon/climate exchanges. Dakar, Sénégal.

6 Participatory Landscape/Lifescape Appraisal

Julia Earl1 and Amadou Kodio2 1Preventing Harm Minnesota, Minneapolis, Minnesota 55407, USA; 2Centre Régional de la Recherche Agronomique, Institut d’Economie Rurale, BP 205, Mopti, Mali

In keeping with SANREM CRSP’s participa- ticipatory research, interdisciplinary teams, tory research methodology developed inter-institutional collaboration and a land- during its initial years of field work and scape scale of analysis. This platform investigation, the SANREM West Africa pro- was designed to bring scientific and local ject employed the Participatory Landscape/ knowledge systems together with ecologi- Lifescape Appraisal (PLLA) in Mali. The cal, agricultural and social scientists and PLLA is a method for participatory rural development and research practitioners to appraisal that goes beyond the village level address natural resource and agricultural to the more inclusive landscape scale. It problems. SANREM selected the landscape was used to gain an initial understanding of focus, because it encompasses a broader the issues facing the farmers and herders analytical framework that incorporates in Madiama Commune, Mopti Region, Re- cross-ecosystem linkages addressing upstr- public of Mali, an area of increasing resource eam–downstream interactions. scarcity and growing conflict. The chapter SANREM learned many lessons from begins with an overview of the PLLA meth- the five preceding decades of development odology, followed by a description of how it work conducted by national and interna- was applied in Mali. A synthesis of the find- tional institutions in developing countries. ings then leads into a brief introduction to Local participation in the design of develop- the activities that have been simulated and ment efforts is necessary for effective project provide the basis for the rest of the book. implementation. Therefore, in SANREM’s ‘research for development’ approach, local farmers and other project participants, in- The Methodology cluding government, research and non- government organization (NGO) partners, SANREM CRSP Global Programme devel- participated in the full spectrum of the oped the PLLA methodology during its ini- research process that encompassed problem tial years of operation in the Philippines, identification and diagnosis, identification Burkina Faso and Ecuador in the early and testing of potential solutions, and evalu- 1990s, and SANREM investigators subse- ation and adoption of practices or policies quently used it in a variety of research sites. based on the research findings. The approach was based on four philosoph- A second key tenet of the PLLA app- ical and methodological cornerstones: par- roach is interdisciplinary research teams. © CAB International 2005. Conflict, Social Capital and Managing Natural Resources (ed. K.M. Moore) 77 78 J. Earl and A. Kodio

The constraints on sustainability are not development. They include: interviews, delimited by discipline. Addressing such transects, mapping (social or physical), constraints required the recognition of a timelines, problem trees, wealth ranking, complex array of processes within and etc. Every effort is made to be as inclusive as between ecosystems, including the physical possible by accounting for any differentia- environment (soil, water, vegetation, cli- tion in a given community, be it religious, mate), the human dimension (population gender-based, wealth, ethnicity, caste, etc. patterns, social structure, cultural norms) PLLAs include incorporating any other data and their interactions in the domain of agri- that may be available regarding the locale or cultural and natural resource management issue. A complete description of the PLLA (NRM) practices. Therefore, PLLAs require process is found in Looking Back at the an interdisciplinary approach that tran- Landscape: Impacts and Lessons Learned scends traditional boundaries among ecolog- from the SANREM CRSP, 1999. ical, agricultural and social sciences, and that also encompasses a variety of concep- tual and methodological tools to produce an Pre-field Work Phase integrated understanding of landscape ecol- ogy and the systemic relationships between Prior to undertaking the field component of its components. the PLLA, the SANREM team conducted a A third key element of the PLLA process West Africa wide literature review on the requires inter-institutional collaboration. By issues of conflict and NRM (Moore et al., virtue of the cross-cutting nature and 1999). This synthesis was based on personal breadth of environmental and agricultural interviews with researchers and develop- sustainability issues, PLLAs incorporate the ment experts across West Africa, grey litera- diverse institutions implicated in them. ture and formally published documents. In Intersectoral partnerships and networks, the process, local partners were identified as including US and host country institutions, key to understanding the rich and dynamic with a shared concern for sustainability social and natural resource complexities include organizations such as universities, found in Mali. The national Malian natural non-profit organizations, research centres, resource and social research organization, development agencies, farmers’ organiza- Institut de l’Economie Rurale (IER), was a tions and women’s organizations. natural choice as a research partner because The fourth key PLLA element is the of IER’s long-standing history of applied, landscape/lifescape approach. It is based on rural and participatory research in Mali. The the premise that in order to understand sus- SANREM project then invited Sahelian tainability and thereby develop strategies to regional, Malian national and international achieve it, the relationships and linkages NGOs as well as local consultants to partic- within a landscape must be understood. ipate in a conference (25–27 January 1999) Landscape refers to a complex mosaic of at the Institut du Sahel (INSAH) in Bamako, biological and physical processes. Lifescape Mali. The conference achieved its objectives refers to the social or human factors that of formally introducing the SANREM pro- influence that landscape. In crafting the gramme, identifying a research site, build- PLLA approach, SANREM determined that ing relationships with local partners, and the landscape/lifescape level of analysis establishing working agreements. was the most comprehensive for under- standing as fully as possible the complex interactions between biophysical and social Research Site Identification components for a given area. Many of the tools used in a PLLA The SANREM-West Africa team and confer- are familiar to development agents, field ence participants needed to address a set of researchers, government employees and established criteria in its site selection non-profit agents working in sustainable process. The criteria included: the degree to Participatory Landscape/Lifescape Appraisal 79

which the site was representative of Mali which tools to use and divided into two and the broader Sahelian region; the pres- smaller working groups. One team worked in ence of a variety of agricultural or livelihood Tombonkan and the other in Madiama and activities (e.g. cultivation, animal hus- Nérékoro. (Nérékoro, the smallest of the bandry, fishing, transhumance, etc.); the three communities, is closest to Madiama presence of conflict related to natural and has a higher degree of interaction with resources; year-round accessibility; recep- Madiama dwellers than with more distant tivity and willingness to participate from villages.) rural partners; and availability of secondary data related to the site. An additional over- arching element was the degree to which the Fieldwork in the Commune of Madiama site might be representative of other areas of the world, semi-arid or otherwise, charac- The PLLA Teams spent 2 weeks conducting terized by increasing natural resource field work, sharing results with community scarcity and competition or conflict for members, triangulating findings between the those resources, with respect to replication two community-based teams, posing follow of lessons learned. In order to address these up questions, conducting a village meeting, criteria, the assembled partners, through an and synthesizing written results. Malian intensive dialogue, selected the Madiama IER research staff (both local and national), Commune (the commune being the smallest SANREM scholars from Virginia Polytech- administrative unit in the then newly estab- nic Institute and State University (Virginia lished decentralization rubric). Madiama is Tech) and Washington State University, located in the Djenné cercle, Mopti Region. a SANREM Management Entity Afri- The conference participants then identified canist, and a representative of the United two primary communities, Madiama (town) States Agency for International Develop- and Tombonkan, for inclusion, pending ment (USAID) participated on the multidis- their willingness to participate. Following ciplinary research teams (soil scientists, its arrival at the site and preliminary inves- animal scientists, agronomists, economists, tigation, the research team decided to sociologists, ecologists and rangeland spe- include Nérékoro, a small village of Peul cialists). The majority of researchers were pastoralists with active social and economic Malians. Each team resided in the commu- interactions as well as shared NRM conflict nity for the duration of the research effort. issues in Madiama and Tombonkan. Upon arriving in the Commune of Madiama the research team was introduced to the village chiefs by the resident IER Research Protocol Outline Established technician. Team members explained the research objectives of the PLLA, highlight- Following the site selection process, the ing that the purpose of the work was to research team members worked together to better understand NRM issues in the area design the research protocol. All members of and possibly work together towards their the team were well-versed in the Parti- resolution. However, the team also made a cipatory Rural Appraisal (PRA) or Méth- deliberate point to clarify that SANREM was ode Accélèrée de Recherche Participative not the type of project that could provide a (MARP) methodology. Therefore, they were tangible physical benefit such as a school- well prepared to conduct a PLLA, an house, birthing centre, rice processor, fertil- approach that draws heavily on PRA meth- izer or the like. Given Madiama’s history as ods applied at the watershed level, compre- an IER research site, the village elders were hensively assessing human (or lifescape) familiar both with research teams and par- and biophysical (or landscape) interactions. ticipatory approaches. The team explained Once in Mopti (the regional capital of the that it wanted to meet with women as well Mopti Region), the research team fully devel- as men and older and younger members of oped the interview questions, identified the community. Since the research occurred 80 J. Earl and A. Kodio

during the cold season, a time of year when The key research objectives identified agricultural activities do not occur or are at for the Madiama PLLA included: a minimum, participation in research groups was not detrimental to completing 1. Seek to understand the landscape/ key agricultural or resource garnering activ- lifescape interactions in the study area by ities. The community representatives agreed focusing on: to work together with the SANREM team. • natural resource use, access and control The PLLA research team employed a • food security variety of PRA-type tools that included: nat- • poverty ural resource management maps, problem • conflict related to natural resource use trees, activity and resource calendars, time- 2. Assess NRM practices and incentives lines of major environmental and historical that motivate these practices under varied events, access and control charts (regarding circumstances. natural resource management use and con- 3. Identify conflicts related to natural flicts), walking transects, pie charts, Venn resource use. diagrams and interviews. 4. Investigate natural resource conflict Following a week of field work, the management strategies in historical, current PLLA teams regrouped in Mopti to share and future scenarios. preliminary results, triangulate findings and 5. Identify decisions made by natural identify holes in data requiring further resource decision makers at the communal follow up. The teams then returned to level in historical, current and future con- Tombonkan, Nérékoro and Madiama in texts. order to pose remaining research questions 6. Prioritize research questions and infor- and to share their results at village-wide mation needs. meetings. Information sharing is a key tenet of participatory research, because it demon- All three participating communities strates respect and equality between the out- (Madiama, Tombonkan and Nérékoro) sha- side researchers and local partners and red similar concerns and research priorities. provides an opportunity for local partici- The citizens of Madiama and Tombonkan pants to respond to any key findings, clarify (the two communities dominated by seden- misunderstandings and ask the research tary agriculturalists with some animal hus- team questions. Subsequent to these meet- bandry activities) expressed a desire: (i) to ings, the teams reassembled once again in investigate how to improve agricultural and Mopti and began the written synthesis of animal production; (ii) to examine how eco- their work. nomic activities might be diversified in Although in close physical proximity to order to alleviate poverty; and (iii) to each other, the ethnic composition of the enhance food security. Tombonkan and three research sites varies. Madiama and Nérékoro (the two smallest communities) Tombonkan are both composed primarily of indicated a desire to investigate methods for sedentary agriculturalists dominated by the mitigating inter-village conflict. Although Marka ethnic group. Animal husbandry is the village elders in Madiama did not iden- practised at all three sites. Nérékoro, in con- tify enhancing conflict resolution practices, trast, is a sedentary village, whose composi- the research team strongly concurred that tion is Peul (Fulani or Fulbe) herders and this topic merited further investigation Rimaibé farmers. (The Rimaibé live in asso- given the level of animosity between the ciation with Peul and are culturally very communities, long smouldering disputes, similar. Historically however, the Rimaibé severely constrained natural resources, and were Peul slaves.) Although Nérékoro is a perceived ownership of the same resources permanent village, typically in Peul com- by both parties. Additionally, the village munities large numbers of men make sea- chief at the time gave the impression of sonal migrations with their cattle, following relative intolerance regarding resource use the rains to verdant pastures. conflicts with his Peul neighbours. Participatory Landscape/Lifescape Appraisal 81

In addition to these topics, each com- 20% of the population is rich; 60% is inter- munity highlighted its need for more water mediate and 20% is poor. Eighty per cent of both for human consumption and animal the village, or the rich and middle classes, husbandry. During the research period possess animal traction equipment. The vil- many of the wells in the village were lage infrastructure is well developed and already dry several months prior to the beg- includes a new school, a village health inning of the rainy season. The research centre, a birthing centre, a pharmacy, water team recommended that SANREM possibly pumps and a grain mill. Revenue sources are investigate area hydrology and groundwater quite diversified. Watermelon is a signifi- availability. cant cash crop in the area.

NRM practices Key PLLA Findings The area is characterized by very poor soil The PLLA yielded several key findings. fertility. Fallow is not practised; the same A summary of these results is presented in parcel is cultivated year after year without this chapter. However, the comprehensive crop rotation. Intercropping is practised, report containing all of the data is presented most notably with cowpeas and cereals such in SANREM-CRSP West Africa Working as millet or sorghum. Organic fertilizer Paper No. 99-02, Participatory Landscape/ (manure) is used to as great an extent as pos- Lifescape Appraisal: Synthese du Diagnotic sible in dryland farming. Chemical fertiliz- Participatif Realise a Madiama, Nérékoro et ers are used in irrigated rice plots. Animal Tombonkan du 1er au 14 fevrier 1999. The traction is used in 80% of dryland farming. summarized PLLA results that follow are The remaining 20% is cultivated using presented by village and research objective hand-held hoes. Fields are borrowed from in the voice of the villagers. neighbouring villages, particularly Kesse- dougou, Doukourani and Teguegne, for cultivation. Madiama (sedentary farming village) Factors that influence agricultural pro- duction decisions and NRM strategies include: soil type, anticipated rainfall Understanding landscape/lifescape (based on the duration and intensity of the interactions cold season), family size, farming equip- Most of the available land within Madiama’s ment, market prices for cash crops and village limits is used for agricultural pro- social factors (e.g. bride price, funerals, etc.). duction. Fallow is not practised. Extensive agriculture is conducted using animal trac- Natural resource conflicts tion and there is insufficient arable land. All available land is under cultivation regard- Conflicts between farmers and herders are less of soil fertility. Grazing land has dimin- the most frequent and contentious in rela- ished as field sizes have increased. Madiama tion to natural resources. These involve farmers determine access and control issues fields, pasture, water (seasonal wetland/ regarding natural resource management in pond) and trees. Damage in Madiama farm their territory. fields caused by Nérékoro animal herds was During a poor rainfall year, 80% of the the primary source of conflict identified. village population is food insecure. Various Passage routes for animals, either from strategies are employed to meet their food neighbouring villages or further away, have needs, including selling small ruminants diminished due to increased extensive agri- and engaging in small commercial activities. culture that is the result of declining soil fer- Twenty per cent of the population is food tility and the availability of animal traction. secure regardless of rainfall and harvest. Competition for a seasonal wetland/pond is According to the villagers’ self-assessment, intense between Madiama and Nérékoro. 82 J. Earl and A. Kodio

There are some conflicts between farmers levels of poverty (80% of Peul and 95% of within Madiama due to land tenure, but Rimaibé). However, it is possible that the they are settled amicably. indices used to establish wealth ranking were based on ideal circumstances and, therefore, not relative to current realities. Conflict management strategies Infrastructure in Nérékoro is poor or non- Historically, the village chief and his coun- existent. The villagers rely on the school sellors managed NRM conflicts, and cur- and health centre in Madiama for education rently, they resolve most NRM conflicts at and health services. The primary revenue the village level. However, if the conflict is source in Nérékoro is animal sales. not settled at this level, it will continue to the Chef d’Arrondissement level (former NRM practices pre-commune administrative unit), then the Justice Ministry at the cercle level (Djenné) Soil fertility is low, but fallow is practised, or the regional authorities in Mopti. The generally for 3-year periods. Organic fertil- Madiama village leaders currently see no izer (manure) is used in fields, to the extent areas for improvement in the conflict man- it is available. Zai (anti-erosion, moisture agement strategies, despite a long-simmer- retaining rock structures) are employed in ing conflict with the residents of nearby some farm fields. Pastures are not managed, Nérékoro. and herders move their animals in order to maximize the health of the herd and not the state of the pasture. Factors that influence Decisions made by commune-level authorities production strategies and NRM practices In historical and current contexts, supra-vil- include maximum use of available space lage decision making has not influenced nat- and maximal herd size. ural resource management practices in Madiama. Madiama residents have heard of Natural resource conflicts decentralization, but claim to be unaware of any of the details of it. The primary natural resource-related con- flicts the Nérékoro inhabitants identified were between farmers outside of Nérékoro Nérékoro (sedentary herder/farmer village) proper and Nérékoro herders. The reasons for these conflicts included: increased areas under cultivation, reduced pasture areas, Understanding landscape/lifescape reduced animal passage routes, damage interactions caused in fields by cattle and competition The majority of land within Nérékoro’s for the seasonal wetland/pond. village limits is used for agricultural produc- tion (i.e. cultivated fields). The size of pasture Conflict management strategies areas has decreased both within and outside of Nérékoro. The Nérékoro residents deter- Historically, the village chief and his coun- mine access and control issues in their imme- sellors managed NRM conflicts. If a conflict diate village territory. However, outside of could not be resolved at that level, it was Nérékoro during nomadic pastoral periods, taken to the Chef de Canton. Following inde- sedentary farmers have greater access and pendence unresolved conflicts were taken to control than transhumant herders. the Chef d’Arrondissement. Currently, NRM Food insecurity is a common problem conflicts are managed by a variety of struc- in Nérékoro. Strategies employed to res- tures including the village chief and coun- pond to this problem include selling cattle, sellors, the Chef d’Arrondissement and the reducing the number of meals consumed per Justice Ministry at the Cercle level (Djenné) day, and increasing consumption of wild or regional authorities in Mopti. foods. Village self-analysis indicates high A variety of practices are employed to Participatory Landscape/Lifescape Appraisal 83

mitigate conflict. They include taxes for over natural resources. The Djoula ethnic grazing rights, herder cooperatives, estab- group is the most prosperous. Eighty per cent lishment of a pastoral calendar regarding of the population own agricultural equip- herd movement, and animal corrals during ment and there are more than 150 traction the rainy season. Scenarios regarding more animals in Tombonkan. ideal management of natural resource Cereals compose the base of food pro- conflicts include inter-village conflict man- duction and consumption. The main agement structures, adherence to the gov- sources of protein are fish and beans. The ernment-determined pastoral calendar, and human hunger season lasts from August to a more just tax system. the beginning of September, while that for livestock is from May to July. Agriculture is the principal source of revenue. Animal Decisions made by commune-level husbandry is a strategy for saving money. authorities Tombonkan has poor socio-economic infra- Decision-makers historically included the structure, and the nearest school is 5 km village chief within the village. Tax systems away in Madiama. for grazing rights in the bourgou1 were implemented, but access to forest products NRM practices and water was free. Supra-village level decisions now include establishment of the All the inhabitants of Tombonkan farm, but pastoral calendar by the Malian govern- they also fish, raise livestock, make crafts ment, wood cutting permits issued by the and engage in petty commerce. Women col- administration, and tax systems established lect and process local fruits and participate by other authorities (i.e. river crossing fees in the more labour intensive planting and by the government, bourgou grazing rights harvesting activities. Monoculture domi- by residential villages and field damage nates dryland farming. Millet and sorghum fees levied on implicated villages). Ideally are planted in fields based on soil type. Crop NRM commune level decision-making rotation and cereal–legume intercropping should include recognition of Nérékoro are infrequently practised. Extensive animal land tenure rights, Nérékoro-authorized traction is the mode of cultivation. Tombon- natural resource use in their territory and kan farmers use early-germinating varieties. the creation of a fair tax system relative to Famine foods are cultivated. Rice is pro- animal passage and potential damage duced in the irrigated sector outside of the caused by herds in fields. Commune, and in lowlands where water is retained longer. Fishing occurs in those same neighbouring areas. Tombonkan (sedentary agricultural Two types of livestock production, community) nomadic and sedentary, are practised. Sedentary livestock are kept in corrals due to the lack of local pasture. Crop residues Understanding landscape/lifescape and bourgou grass constitute the forage interactions stocks. There is localized application of Tombonkan is experiencing severe land manure and compost. Participation in NRM degradation. Nearly all land is under cultiva- is gender-dependent. The hunters’ associa- tion without the possibility of extension. tion oversees environmental protection. Pastureland has been reduced to a minimum. There are insufficient permanent wetlands Natural resource conflicts and only seven wells in the community. There is no forest and little wildlife. The Disputes between farmers occur over loca- founding ethnic group, the Marka (agropas- tion of territorial boundaries, borrowed toralists), have most access to and control of land, and use of wetland areas. Farmer– natural resources. Women have no control herder disputes (the most frequent form of 84 J. Earl and A. Kodio

Fig. 6.1. Drawing a map of village resources.

conflict) involve livestock damage to crops, Ideally, it would be best if spaces were use of bourgou grass by the farmers (har- delimited between different territories and vested and stored as fodder), and cultivation between different production systems. of former and current livestock passage Agricultural intensification could free up routes. Women do not have secure tenure on more space for livestock production. Salt the land they borrow. licks and manufactured animal feed could There are also territorial disputes about be used to intensify animal production. The land tenure between villages. Villagers from former wetland areas could be managed in Madiama and Promani have borrowed land order to reintroduce bourgou cultivation from Tombonkan households that have and fish. Alternatively, soybeans could be requested its return. The dispute with produced for bouillon cube production to Promani has gone all the way to the Ministry replace sumbala, a traditional bouillon of Justice. Disputes also include access to made from Néré tree pods (Parkia biglo- bourgou grass in the wetlands outside of the bosa). Reforestation is also required. village territory. Improvements could be made through col- laboration among different partners as well as education and information on land use Conflict management strategies management. Historically, conflicts have been managed through the application of customary law. Decisions made by commune-level authorities Colonial law regulated disputes as well, but much Islamic and traditional legisla- In the past, the heads of households, the vil- tion of the Dina (established under Sekou lage chief and his advisers, or the Chef de Amadou) governs resource access and con- Canton in Promani made decisions. Today, trol. Currently, government extension the village counsel has more authority over agencies and the government administra- heads of households and contested admin- tion apply modern legislation. The govern- istrative decisions need to be referred to the ment-determined calendar of livestock justice system for conciliation. Ideally, res- passage to the bourgou pasture that is sim- ponsible commune-level decision-makers ilar to the Dina legislation is a major con- should be chosen judiciously, because this trolling factor. would help avoid improper influence. Participatory Landscape/Lifescape Appraisal 85

Fig. 6.2. Discussing women’s nature resource management practices.

Primary Problems Identified by Villagers est to lowest): (i) food insecurity; (ii) natural resources (i.e. water shortages); and (iii) As part of the participatory research conflict. Nérékoro villagers identified a process, the relevant villages were asked to widespread lack of food self-sufficiency identify and prioritize the issues discussed within their community as well as a lack of during the previous week’s work. The main production methods adapted to local condi- categories of discussion included natural tions. In reference to natural resources, they resources, food security, conflict and identified water shortages as the primary poverty. Although the priorities varied concern in terms of agricultural production, slightly from village to village, the overall domestic use and animal husbandry. areas of concern overlapped and are com- Conflict, the third priority, stems from the plementary. Information regarding the spe- long-simmering disagreements between cific problems they identified and their Nérékoro and neighbouring Madiama as respective prioritization by village follows. well as from the presence of sedentary farm- ers in areas where the Nérékoro herders travel to graze their cattle. Madiama

The village participants in Madiama were Tombonkan quite clear and adamant that their primary problems are related to natural resources. The types of natural resource and agricul- The top two priority problem areas identi- tural conflicts identified in Tombonkan were fied included: (i) lack of water (rainfall for similar to those of Nérékoro and Madiama. In cereal crops and river floodwater levels for terms of natural resources, the following irrigated rice); and (ii) decreased soil fertil- problems were identified: shortage of land, ity in relation to dryland cropping. reduced soil fertility, water shortages, insuf- ficient pasture, insufficient trees, and a shortage of fish. In reference to food security, Nérékoro the villagers cited several of the same natural resource limitations as well as crop pests and The residents of Nérékoro prioritized their a lack of farming equipment. In reference to problems in the following order (from high- conflicts, they cited as problems a shortage of 86 J. Earl and A. Kodio

arable land and poor harvests. The problem International Development (USAID) mis- of poverty was due to insufficient water, sion staff, is how the establishment of com- inadequate production, and natural resource munes will actually be operationalized, degradation. how natural resource management deci- sions will be made, and who will participate in those decisions. This is of particular rele- Summary vance in the Madiama commune given the village of Madiama’s dominance due to pop- In general, although their top priorities ulation, existing infrastructures, and land varied, all three participating communities area under cultivation and its recalcitrance shared similar concerns. The research team in reference to relations with Peuls in the did not attempt to reconcile inter-village nearby village of Nérékoro. This particular priorities. Instead, it presented the common topic of farmer–herder conflict is relevant to information needs and research questions Mali as a whole and is applicable in other the villages identified. The citizens of Sahelian countries as well. Madiama, Nérékoro and Tombonkan would Finally, it should also be noted that a like to investigate improved agricultural wide variety of development players, and animal production and examine how including UNICEF, CARE, government economic activities might be diversified in agencies and research organizations, have order to alleviate poverty and enhance food intervened in the Madiama commune. In security. Nérékoro and Tombonkan, the two particular, given recent government restruc- smaller villages, would like to investigate turing of its environmental ministries, gov- methods for mitigating inter-village conflict. ernment agents have become involved Although the largest village in the com- actively soliciting village problems, needs mune, Madiama, did not identify enhancing and solution identification. Government conflict resolution practices as a pressing and NGO colleagues have suggested that concern, the research team strongly con- SANREM first investigate what research has curred that this topic merited further inves- already been conducted in the study area tigation given the current level of animosity, prior to undertaking research that may be long smouldering disputes, severely con- repetitive or complementary to previous or strained natural resources and perceived ongoing activities. ownership of the same resources by both Research themes identified as a result of parties. the PLLA included the following: In addition to these topics, it should be noted that at every site the participants highlighted water shortages as a concern Agricultural production both in reference to agricultural productiv- ity and human consumption. Although • Rice production in both irrigated SANREM is not likely to get involved in and rain-fed areas, particularly in the well digging activities, etc., it should be context of decreased availability of made aware of these concerns and partici- water. pant eagerness to resolve these problems • Soil fertility for enhanced harvests and through an investigation of area hydrology related issues of wind and water ero- and groundwater availability. sion reduction. Although the topic of conflict has been • Intensification of agricultural produc- highlighted at various points in this sum- tion. mary, it has not been discussed in any detail • Fertilizer management strategies (che- with respect to decentralization and the mical and organic) and composting. establishment of communes. A major point • Pest management strategies. of concern from a number of interested par- • Short season varieties of cereal grains. ties, including villagers, NGO representa- • Potential for soybean introduction and tives and local United States Agency for useful tree species as well as investiga- Participatory Landscape/Lifescape Appraisal 87

tion of appropriate techniques for intro- ing rules, investigating methods for set- ducing leguminous crops. ting territorial boundaries). • Disease resistant species. • Information and training in NRM.

Water Poverty

•Availability of ground water and meth- •Women’s market gardening options in ods for determining the water table light of unavailability of water (wells go level. dry and pumps do not provide adequate •Water conservation and management volume). methods, with reference to rice cultiva- • Improvement of agricultural credit. tion, dryland cropping, animal water- ing and human consumption. •Techniques for improving water quality Building on the Participatory (related to human and animal health Relationship issues). Subsequent research was built on the initial relationship developed between SANREM Economic activities and the participating Madiama Commune communities and has led to core long-term •Diversification of economic activities in learning relationships developed between order to enhance household incomes researchers and community members. As a and food purchasing power. follow-up to the participatory appraisal, • Support for existing or moribund and in conformance with the SANREM village credit systems and/or cereal Annual Work Plan, researchers planned and banks. implemented more detailed formal socio- economic surveys at the household level (see Chapters 13 and 16). Also, biophysical Animal husbandry baseline data were collected on weather, soils, crops and crop management practices •Animal husbandry for production of (see Chapters 3 and 11), and a local inter- milk and meat. locutor, the Natural Resource Management •Preservation of animal products (e.g. Advisory Committee (NRMAC), was estab- dried milk). lished to represent and communicate • Improvement of animal health. between the full ten villages of the com- • Forage varieties. mune and SANREM on a sustained basis • Hay conservation. (see Chapter 7). Several partners worked •Techniques for equitable distribution of together to establish the formal work plans natural resources. that the SANREM CRSP Management Entity at the University of Georgia would review, and USAID-Washington, DC, would ulti- Conflict mately approved. These research-for-sus- tainable-development partners included a • Conflict management techniques. locally based NGO (CARE/Djenné), the IER’s •Predictive tools to avoid conflicts. regional agricultural research centre in • Information regarding use and improve- nearby Mopti (CRRA/Mopti), Virginia Tech ment of the official government-deter- and Washington State University. mined pastoral calendar. The CRSP work plan process entails •Research efficient conflict management extensive review based on technical merit, systems (consensus building among resources available and USAID policy focus. natural resource users before establish- Theresearchprioritiesthevillagersidentified 88 J. Earl and A. Kodio were paired with the scientific expertise of Note faculty at Virginia Techand Washington State University. The research topics were in turn 1 Bourgou refers to the rich grass that grows in developed into annual work plans. The the Niger River and its flooded delta during SANREM CRSP Management Entity at the the rainy season. It is a highly valued and University of Georgia, including a panel of nutritious source of animal fodder. Access to the bourgou may be contentious. neutral scientists composing the Technical Committee, and USAID-Washington, DC, reviewed the work plans. USAID made the final decision regarding what level of support References to provide for which projects. During a year-long process, the work Moore, K., Kaboré, D., Gnoumou, B. and Bertelsen, M. (1999) Conflict and natural plans were approved and funding initiated. resource management in agricultural and In preparation for that work plan, the newly pastoral systems of arid and semi-arid regions formed NRMAC took it upon itself to of West Africa (Sustainable Agriculture and conduct meetings in each village and Natural Resource Management Collaborative develop a list of priorities to focus SANREM Research Support Program –West Africa research activities. These lists, and the com- Site, Working Paper No. 99-01–E). Office mittee’s consolidation of them, were pre- of International Research, Education, and sented to IER and US scientists at a meeting Development, Virginia Tech, Blacksburg, on 1 February 2000 in Madiama. All NRMAC Virginia. members, development agents, technicians, Neely, C., Buenavista, G. and Earl, J. (eds) (1999) Looking Back on the Landscape: Impacts and and local officials were present. An lessons from the SANREM CRSP. University extended discussion ensued and two prior- of Georgia, Sustainable Agriculture and ity areas of biophysical focus were agreed Natural Resource Management Collaborative upon: (i) increasing soil fertility in crop- Research Support Program, Watkinsville, lands (see Chapter 12); and (ii) increasing the Georgia. quality of pastures (see Chapters 14 and 15). SANREM CRSP – West Africa (1999) Participatory The NRMAC also signalled the importance Landscape/Lifescape Appraisal: Synthese du of developing the institutional capacity of diagnostic participatif réalisé à Madiama, the committee to provide a commune-wide Nérékoro, et Tombonkan du 1er au 14 decision-making infrastructure for NRM in février 1999 (Sustainable Agriculture and Natural Resource Management Collaborative the commune (see Chapters 7, 8 and 9). In Research Support Program – West Africa addition, a work plan was proposed to Site, Working Paper No. 99-02). Office of develop agronomic and socio-economic International Research, Education, and modelling based on the priorities the com- Development, Virginia Tech, Blacksburg, mittee specified (see Chapter 11). Virginia. 7 Building Social Infrastructure for Sustainable Development

Keith M. Moore,1 Salmana Cissé2 and Abdoulaye Touré3 1Office of International Research, Education and Development, Virginia Tech, Blacksburg, VA 24061–0378, USA; 2Centre Régionale de la Recherche Agronomique, Institut d’Economie Rurale, BP 205, Mopti, Mali; and 3CARE/Tombouctou, Mali

Decentralization has changed the dynamics This chapter describes the establish- of natural resource management (NRM), but ment and operation of such an institutional it has not yet yielded a methodology for innovation in the socio-political organiza- effective local governance in the Sahel. In tion of a rural Sahelian population. This response to this insufficiency, the SANREM institutional innovation, the Natural Re- CRSP initiated a programme of action source Management Advisory Committee research to assist a local population in deter- (NRMAC), is an inter-village and pluri- mining a new modality for management, in ethnic civil society organization (CSO) in the accordance with its environmental and 5th Region of Mali. The primary hypothesis socio-cultural context. of this research was that sustainable, decen- We began with the idea that a new plat- tralized management of natural resources form for inclusive community action was can be achieved if a local population is required (Pretty and Buck, 2002; Vedeld, equipped with a methodology for managing 1994) based on the premise that it was insuf- their natural resources and the conflicts ficient simply to revive old institutions and associated with the exploitation of these traditions or to work with one group of resources, and a local organization to insti- stakeholders at the expense of another. gate inter-ethnic and multi-village dialogue. Recognizing that group decision makers Beginning with a description of the needed to be targeted, we set out to build a institutional context in which the popula- social infrastructure (Flora, 1998) that could tion of Madiama is situated, we then present shape social capital for rural civil society in the constituent components of the NRMAC an image of modern Africa. The objective and the process of its establishment. The was to increase the horizontal linkages discussion highlights the significance of the between stakeholders in the belief that such organization’s legitimacy, based as much in a common action platform can increase the customary practice as in modern legal for- chances that diverse ideas and people can: malities. It shows that associational life is (i) increase open debate; (ii) increase the flow not only a matter of gathering people of information; and (iii) lead to more care- together – men and women, regardless of fully considered management decisions and, ethnic or socio-professional status, but also as a consequence, increase trust between requires the development of individual communities. capacities (training in functional literacy, © CAB International 2005. Conflict, Social Capital and Managing Natural Resources (ed. K.M. Moore) 89 90 K.M. Moore et al.

association management, financial manage- power has centred on the Commandant of ment, knowledge of NRM texts, laws and the Cercle and the Chef de Arrondissement codes, management and reconciliation of (Sous-Prefet), more closely associated with the conflicts, etc.) in order for them to the local population. The village chiefs and assume active roles in the development of other assorted resource chiefs, who control their commune and to support the develop- the immediate allocation of resources for ment of extra-communal relations and, in household livelihoods, have held custom- particular, the role played by training in ary power. Overall, villagers have had little conflict management, given its importance say in resource allocation or development for the development of self-confidence and initiatives. building of credibility for the committee in The resources of the state have been the development and the safeguarding of the mobilized by an array of development ser- commune’s natural resources. vices that have attempted to assure environ- mental protection through threat of law and development initiatives designed to extract The Institutional Environment surpluses for the nation as a whole. For instance, the Office Riz has been a major This social infrastructure is designed to state initiative (reformulated from time-to- build linkages within the local institutional time) to enhance the productive power of environment and bridges to external organi- the peasantry in the production of rice. This zations and stakeholders (see Fig. 7.1). Two effort has been implemented by organizing bases of institutional power in rural areas farmers into village rice growing associa- have vied for control of resource mobiliza- tions based on traditional village associa- tion for development and environmental tions. These associations work state lands, preservation. The infrastructure of state flooded annually by the Office-controlled

Region of Mopti Prefect-Djenné Cercle of Djenné Livestock Service Forestry Service

Rural Development Service Rice Office Djenné Mayor’s Sub-Prefect-Sofara Association

Commune of Madiama

CRRA/Mopti Mayor Station Village Head Village Head

Commune Council Village Head Village Head The State

Civil Society Syn Rice Growers Herders Association Association NRMAC of Nérékoro

Villagers Villagers Villagers Villagers

Other Associations Village Associations Women’s Associations Parents, Health, etc.

CARE/Djenné Other NGOs

Fig. 7.1. Network of relationships between the population, government and civil society. Building Social Infrastructure for Sustainable Development 91

irrigation works. The livestock service mon- tive documentation (births, deaths, etc.) and itors cattle health and the movement of the collection of local taxes. The mayor’s herds within the zone. The forestry ser- office also provides a more direct connec- vice polices the cutting of wood for fuel and tion to government services and advocacy timber to assure that over harvesting for improving those services. The Commune does not occur, frequently levying fines mayors in the Cercle of Djenné also have and collecting fees for wood cutting. The their own association that helps them to Centre Régional de Recherche Agronomique better understand their new roles. Into this (CRRA)/Mopti is charged with developing mixture, a new player, the NRMAC, has new and adapted technologies to increase been introduced to support decentralized production on the farm and in the river decision-making in NRM. waters. Rural civil society has been restricted for the most part to the village level. Village Establishment of the NRMAC associations are common among men and have provided a point of contact for state In September 1999, a SANREM CRSP-spon- development efforts. Women’s associations sored delegation of national agricultural have been largely avoided until recently. researchers, the newly elected mayor of the The Herders’ Association of Nérékoro was Commune, and representatives of the local established as a local organization to protect Mopti Rice Office and a World Bank NRM the interests of migrant (transhumant) Project visited all ten villages in the herders who have local bases within the Commune of Madiama. In each village, the region, but little or no administrative repre- chief and a group of his counsellors were sentation. informed about the objectives, value and Since the end of the Traoré regime role of participation in Village NRM User in 1991, the establishment of Rural Com- Groups and a commune level NRMAC. munes has brought government closer to The members of the delegation ex- the population while ostensibly empower- plained that the primary purpose of the ing the people to improve their live- NRMAC was to provide a forum for reflec- lihoods. Supporting this movement, non- tion on NRM in order to improve communal governmental organizations (NGOs) have resource management. Village NRM User provided an alternative source of resources Groups would provide an essential link for that are stimulating new development communicating technological innovations initiatives and organizational skills develop- to other groups in the Commune. The ment and providing incentives for self- NRMAC would provide a network through improvement. which SANREM CRSP researchers could The communal elections in June of learn about commune priorities, technology, 1999 brought the first elected officials to and information needs. In addition, it would live and work in the rural communes of help the local population improve the pre- Mali. Although these officials were elected vention, mitigation, and resolution of NRM locally, their names were drawn from conflicts and develop a plan for the rational national party lists and they are the govern- use of natural resources within the com- ment’s voice at the local level. Being more mune. Committee participation would lead closely in touch with the population, their to improved leadership skills in working representational experiences direct them to with technical assistance in planning and be more responsive to local concerns. Time implementing NRM activities. will tell whether these elected officials will In each village, the chief held a gather- be able to be truly responsive to the ing to select a diverse group of four or five expressed concerns of the populations they delegates to represent the village in a com- serve. mune-wide general assembly to elect the The most important services that the NRMAC. According to the relative impor- mayors now provide are simple administra- tance of the activity to village livelihoods, 92 K.M. Moore et al.

either two herders, two farmers, or one of and the Sub-Prefect of Sofara in attendance. each were selected. In addition, two or three The anticipated objectives, role, and struc- more villagers were selected to represent ture of the NRMAC were again described. women, hunters and crafts/forest gatherers. Translations into the two local languages User groups had already been devel- (Peular and Bambara) were provided. oped in three of the ten villages where an Participants were divided into four discus- earlier participatory appraisal had been sion groups (organizational and administra- conducted. Three World Bank NRM Project tive issues; dryland farming, rice farming (PGRN) villages had already developed their and fishing issues; livestock, hunting, gath- own Village NRM Committees that met and ering and craft issues; and the role of selected their representatives. Because the women) and enjoined to debate the con- four additional villages were completely cerns and priorities of the commune with new to the idea of village NRM committees, respect to each topic. After these groups their village meetings were more extensive. reported their conclusions at the plenary Two of these villages initially declined to session, all research, technical service, NGO participate, but one village later sent four and elected officials withdrew and elections representatives to the General Assembly. were held for positions on the NRMAC. This Each of the nine participating villages and elected committee was made up of 11 men the local irrigation management committee and two women. The village that had origi- sent three to five representatives (45 in all, nally declined to participate for political including seven women) to the General reasons sent a representative to join the Assembly meeting held on 10 October 1999. committee after it had been established. The Mayor of the Commune opened the General Assembly with the 45 village repre- sentatives and an additional 25 participants, Legitimacy of the NRMAC including researchers and representatives from an NGO, a PGRN, the deconcentrated Four factors contribute to the legitimacy government services, Commune Council of the NRMAC as a viable community

Fig. 7.2. The Natural Resource Management Advisory Committee of Madiama. Building Social Infrastructure for Sustainable Development 93

organization. They are all based on the shared understandings of customary prac- foundation of the participatory approach tice at the village level. developed between researchers and the com- The first step in developing formal munity for establishing and operating the agreements of co-operation/collaboration committee. The first involves gaining formal was the signing of a protocol with CARE/ recognition of the committee as a legal entity. Djenné, the NGO providing the NRMAC The second is the establishment of relation- with institutional development training and ships with stakeholders and partners in the assistance. Once formally recognized as a association’s environment. The third is serv- registered civil society organization, the ing a valued purpose for the community. The NRMAC also signed a partnership agree- fourth is the re-election of committee mem- ment with the Commune Council. This doc- bers after their first term had been served. ument, perhaps the most significant of the NRMAC’s formal protocols, provides a framework for the NRMAC: (i) to influence Formal recognition of the association NRM decisions within the commune; (ii) to be consulted concerning decisions of the In order to become a legally recognized asso- Commune Council; (iii) to be recognized ciation, the NRMAC was required to be throughout the commune as a significant in conformity with national laws concern- player in the resolution of conflicts linked to ing associations and cooperative NGOs. natural resources; and (iv) to actively partic- Because legal recognition is based on an ipate in the economic development of the association’s internal structure and gover- commune. nance, the General Assembly of the Village The NRMAC is pro-actively exploring NRM User Group members drafted, dis- other formal partnerships in order to better cussed, and passed by-laws that the judicial serve the population of Madiama. To authorities, in turn, formally approved. The develop collaborative relations, the NRMAC NRMAC then informed the Prefect in invited representatives of the technical ser- Djenné of its intent to form the association vices serving the Cercle of Djenné to recent and followed this with a formal request for meetings. For example, members of the legal recognition (including all official NRMAC are interested in building a rela- stamps). Initially, the Prefect refused the tionship with the Service Locale de la association’s request, because the domain of Réglementation et du Contrôle (SLRC), the NRMAC activities fell within the range which is charged with protecting the of authorities devolved to the Commune forestry resources of the Cercle. The SLRC, Council. Therefore, the NRMAC requested in turn, is looking for partners to interface that the Mayor, who had assisted in the with and help it implement its activities at process of associational development, send the local level. the Prefect a letter in support of formal The second type of partnership builds recognition. The Mayor obliged and, upon collaborative relationships with customary receipt of that letter, the Prefect approved authorities and villagers. Although not for- their request. The NRMAC was formally reg- mally documented as are those in modern istered as an association by the Cercle of civil society, these relationships have been Djenné on 23 October 2001. formed while conducting activities that involve the key village actor, the village Relations with other associations, technical chief. This kind of partnership began with services and villages the establishment of Village NRM User Groups under the direction of the village Partnerships have been developed to pro- chiefs. By sending representatives to the vide a framework for productive relation- General Assembly to elect the NRMAC, the ships. These partnerships are either formal chiefs have in effect confirmed the legiti- ones, conforming to the standards of national macy of the NRMAC. The reticence of cer- civil society, or informal ones, based on tain chiefs to designate village members to 94 K.M. Moore et al.

participate in the initial General Assembly uncement of the process for re-electing of the association bears witness to the NRMAC members was circulated through- validating role chiefs play. This form of out the commune, in each village, and on legitimacy is fragile and ad hoc (or arbitrary) rural radio. As when it was first constituted, in nature. Unlike formal, documented the process started in each village of the recognition, it may be withdrawn at any commune. The village chiefs re-established moment. the Village NRM User Groups, and five rep- Frequent communication with all part- resentatives from each of the ten villages ners is essential for effective organizational were sent to a General Assembly at the com- functioning. However, this communication mune seat of Madiama. At this Assembly, is more than a matter of transferring informa- presided over by the Mayor and the Sub- tion, because it involves continually renew- Prefect, the NRMAC President and the ing understandings between the NRMAC NRMAC Executive Bureau presented an and the village chiefs. NRMAC members activity and financial report of their accom- have regularly kept the village chiefs and plishments during the past 3 years. After Village NRM User Groups informed of their a question and answer session, all the activities, the training programmes they par- NRMAC members resigned. Following an ticipate in, and the research activities they open debate and consideration of their expe- monitor. In addition, NRMAC trainers have rience and training, all members of the com- conducted several training sessions, first ini- mittee were re-elected by acclamation, tiated as training-of-trainers by outside thereby renewing their mandate. As an addi- experts, at the village level. tional outcome of the debate, a commission of peers that includes a representative from the Commune Council, the village chiefs, A valued purpose and other customary/religious leaders, the CRRA/Mopti, and NGOs working in the area Unless the NRMAC does something valued will monitor this new term of office. With by the community, community members this renewed mandate, the NRMAC is ready have difficulty understanding why the asso- for another 3 years of activities. ciation exists and why it should be of any concern to them. Built on the priority con- cerns of villagers, the NRMAC’s mission has Training Received been to promote the management of natural resources in the commune by introducing The NRMAC has been provided with con- and adapting technologies to local condi- siderable training to accomplish its tasks. tions so that the population can improve The SANREM CRSP has provided this train- their standard of living. An essential ele- ing and the associated research activities ment in this mission involves the manage- through its partners in Mali, CARE/Djenné ment of conflicts generated in the use of and the CRRA/Mopti. Even before receiving natural resources by various community formal recognition as an association of the members. By providing such services (e.g. commune, the NRMAC visited each village protecting and planting trees, resolving con- to develop a list of priorities to help focus flicts, and introducing new technologies), the SANREM research activities. These lists, the NRMAC legitimizes itself in the eyes of and the committee’s consolidation of them, local leaders and villagers. were presented to CRRA/Mopti and USA researchers at a meeting on 1 February 2000 in Madiama. An extended discussion Re-election of committee members ensued during which the committee priori- tized two biophysical themes, improved soil Validation of this legitimacy was recen- fertility in croplands and improved pasture tly witnessed with the re-election of the management. The committee also stressed NRMAC for another 3-year mandate. Anno- the importance of reinforcing its organiza- Building Social Infrastructure for Sustainable Development 95

tional capacities. These priorities formed shops and informal assistance and exchange the basis for the initial research and out- visits. Institutional reinforcement train- reach relationship. ing began with an institutional analysis to The training that the NRMAC has identify organizational strengths and weak- received can be grouped into three broad nesses. Based on this analysis, an insti- categories: (i) holistic management (HM) of tutional development plan was drafted. natural resources; (ii) conflict management Functional literacy training was a core and resolution; and (iii) development of training element and focused on con- organizational and leadership skills. All cepts of democratic governance along with training activities were developed and basic lessons in financial management and extended in a training-of-trainers format. In accounting. Subsequent training workshops addition, a few key leaders of the NRMAC addressed strategic planning, understanding have profited from study tours. Four mem- NRM guiding texts, codes, and laws, decen- bers went to Chad to observe an ongoing tralization codes and laws, and lobbying. programme of HM of rotational grazing on The NRMAC benefited from the training, open rangelands; two members travelled to because a framework was created for Niger to observe a programme of conflict res- exchange and dialogue between the NRMAC olution through cattle trail delineation and and other local actors involved directly or maintenance; and one member attended a indirectly in NRM. The NRMAC also prof- meeting of regional non-governmental orga- ited from the support of CARE in the pro- nization representatives. motion of the inter-village negotiations for The first training introduced NRMAC establishing bourgoutière’s (seasonal wet- members and their local technical assis- land basins) management agreements. Addi- tance partners to the principles of HM. Six tional institutional coaching has included subsequent HM training workshops focused assistance in the drafting of by-laws and on applying these principles to evaluate on- registration of the association, financial farm research trials, establish wetlands monitoring, seeking external funds, and management, and develop an HM grazing building linkages with other local services system for open rangelands. HM training and NGOs. has been coupled with the conflict manage- ment training from the outset. The five con- flict management workshops held for the Services Provided committee focused on consensus building, managing power, managing change, and The NRMAC has served as an interface for adapting this training to the management of the commune with government services. wetlands and open rangeland grazing. Since However, the NRMAC is not limited to this the initial workshops, the lessons learned at function alone. During its first 3 years of these training sessions have been routinely existence, the committee has initiated on its communicated at the village level, at first own many additional activities that are under the supervision of the SANREM train- detailed below. ers. However, the burden of this communi- cation has increasingly shifted to NRMAC trainers. To date, NRMAC trainers have Monitoring research trials designed and implemented two HM and three conflict resolution workshops at the Improved soil fertility village level, in addition to hosting a work- shop for Commune level representatives During the first year, local chiefs and village across the Cercle of Djenné. residents in collaboration with researchers CARE/Djenné has provided the commit- set up user groups in three target villages. tee with institutional reinforcement in the Each group included the village chief and at form of organizational and leadership train- least two women. These Village NRM User ing in two formats, formal training work- Groups chose collaborating farmers for the 96 K.M. Moore et al.

field tests that included benchmark charac- Pasture improvement terization studies and soybean and fertilizer trials. Contact between researchers and Given the complexity of coordinating the farmers and herders was monitored during management of communal pasturelands, the season, and formal meetings were held pasture improvement research began more to discuss the achievements, expectations slowly, by building rapport and establishing and difficulties of NR users at these sites. common objectives within the community at Researchers had initially targeted the both the commune and village levels. The on-farm trials in a limited number of vil- NRMAC and collaborating researchers iden- lages, but at the outset of the second year, tified those management practices that led to the NRMAC made certain that each of the a general regression of plant communities to ten Village NRM User Groups participated. the level of annual grasses and other species On-farm trials involved the application of of low-forage quality that severely limit the rock phosphate (PNT), manure amend- forage potential. Biophysical research began ments, micro-dosing with inorganic fertil- with four rainy season exclosures (two in the izer, field stabling, and the intercropping of village of Siragourou and two in the village of millet and cowpeas or the rotation of millet Nérékoro) to demonstrate pasture regenera- or sorghum with cowpeas. Trial sites located tion capacities and measure the available in each village have allowed for participa- seed stock. This work was followed by the tion by all community members. Although establishment of two open-range rotational the farmer collaborators did not manage all grazing sites (in Torokoro and Siragourou) plots successfully, committee members and ensilage trials to optimize forage learned from this lesson and more closely resources for women based on Cassia tora,a monitored the quality of farmer participa- prolific and noxious weed. tion in subsequent years. These trials have Early on the NRMAC identified the provided a focus for addressing issues of regeneration of bourgou (a wetland grass increasing soil fertility within the highly valued as forage) as a high-payoff Commune. investment. Bourgoutière regeneration is

Fig. 7.3. NRMAC members and researchers planning an on-farm trial. Building Social Infrastructure for Sustainable Development 97

more complex than communal dryland pas- The NRMAC has also conducted four ture regeneration, because the water information and awareness building pro- resource on which it is founded is highly grammes on the local radio station (with prized for agriculture and for livestock. three rebroadcasts to date). Five committee Access to wetlands requires careful negotia- members, including two women, led each tion among all resource users. Of the programme, with a total of ten committee four villages that embarked on this process members participating. These programmes only two have advanced to the stage of phys- have described the NRMAC mission as well ical implementation and are attempting as its objectives, activities, accomplish- to master regeneration techniques (see ments and partners. Specific themes have ‘Agreement development’ below). Success- included NRMAC leadership of the Mayor’s ful regeneration is also dependent on annual campaign to protect the Acacia albida flooding, which did not occur last year. (Valanzan), local agreements for the promo- tion of bougoutière regeneration, and issues of decentralized administration that were Information exchange conducted with the Mayor and a representa- tive of the administration. Since the primary method of information The NRMAC recognizes that communi- exchange is through NRMAC meetings, the cation is the key to its continued viability as committee holds business meetings on the a functioning commune level organization. last Sunday of each month and less rou- NRMAC members are not equally proficient tinely animates training workshops at the at reporting back to their villages, due in village level. The Mayor or his representa- part to their educational levels, and conse- tive routinely participates as an ex-officio quently, the overall quality of these commu- member in these commune-level events. nications has suffered. Indeed, observers The village representative on the committee from the villages to the local administration reports the information and/or issues dis- in Djenné have noted that the poor flow of cussed in these meetings to the village chief information has severely hampered the and the Village NRM User Groups. Village entire decentralization process. To improve chiefs report being well informed of their capacity to address this issue, the NRMAC activities. NRMAC, with the assistance of CARE, has A considerable amount of information submitted, and had accepted, a proposal to filters through the NRMAC. Although much the Fondation de France for technical assis- of it involves the routine administrative tance in developing a communication strat- activities of the association, the NRMAC is egy to better disseminate their messages. also the interface between government ser- This strategy development was recently vices associated with NRM and the com- completed. mune. This privileged position allows the NRMAC to learn about new techniques and technologies, gain improved understanding Forestry services of the guiding documents, codes and laws concerning decentralization and NRM, and Early in the life of the NRMAC, the Mayor develop skills in the management of com- called upon it to assist him in the promotion munity affairs, including conflict preven- of a national campaign to protect the Acacia tion and management. albida. Drawing on the network of Village NRMAC members have been trained as NRM User Groups, the NRMAC was able to trainers in HM and conflict management disseminate the message quickly. This and resolution. They have appropriated this action resulted in a considerable decrease in training and developed their own training damage to this nitrogen-fixing tree. modules that are often based on a local the- The NRMAC has also taken the lead in atre format. This training programme has reforestation by purchasing and planting been conducted in four villages. trees. Six villages had sufficient water at 98 K.M. Moore et al.

the time of planting to assure tree estab- Conflict resolution lishment and consequently requested a total of 149 trees of three species (neem, Violent conflict in the Commune of baobab and néré) to be planted in their ter- Madiama has decreased over the past three roirs. The Village NRM User Groups were years. Although difficult to verify objec- responsible for tree planting and watering tively, community members claim that this until these trees were fully established. is, at least in part, due to the awareness The head of the Service de la Conservation building of the NRMAC. The local popula- de la Nature was impressed with the results tion deeply appreciates the ability to resolve of their independent action and assured conflicts locally (i.e. without recourse to the them that any assistance they might need in administration). the future for similar ventures would be On two occasions, the NRMAC has been available. called upon to intervene in local conflicts. The first occasion involved the early entry of cattle into the commune. This incident was Agreement development the result of a need for water for the cattle and did not actually involve cattle entering The NRMAC has initiated dialogues with unharvested fields. However, other herders selected villages in order to develop agree- in the commune were not happy about this ments for the regeneration of bourgou in cer- breaking of the agreement concerning the tain seasonal basins in the commune of date of entry, because their herds were still Madiama. These stakeholder negotiations forced to remain outside of the commune. were initiated in December 2000 in four vil- After informing the Mayor of the unautho- lages, but due to a lack of consensus in one rized entry into the commune, the other of the villages, only three villages were herders were ready to call the gendarmerie retained for the bourgou regeneration pro- (of the Cercle). However, NRMAC members gramme. Negotiations were also begun with spoke with the principals in the conflict and neighbouring villages and stakeholders in negotiated a resolution, thereby avoiding order to establish local agreements govern- the participation of the authorities. ing the sustainable exploitation of these basins. These dialogues proceeded under the aegis of the NRMAC, which sees this activity as a core component of its mission, and the Mayor with the assistance of CARE/Djenné and CRRA/Mopti. The objective of these agreements is to minimize conflicts between wetland users, improve the management of pastures, and develop the financial resources to maintain them. The agreements define the parameters of collaboration as well as the roles and responsibilities of each party. In the course of these discussions, it was determined that one of the bourgoutière, while constituting part of the customary terroir of the village, was in fact currently managed by the Casier de Bougoula (local water management asso- ciation), which only allowed for the pro- duction of rice. Two multi-village agreements have been drafted, but due to the lack of flooding in 2002, no progress in Fig. 7.4. NRMAC President Sao clarifies a point implementation has been made. during a meeting. Building Social Infrastructure for Sustainable Development 99

The second incident involved a fight accepted this proposal pending presenta- between a Peul and a Marka in the village of tion of further information. The NRMAC Promani. When one of them was seriously identified a local consultant and resubmit- wounded, the village chief called gen- ted their application with the additional darmerie and the aggressor was taken to jail information requested and the consultant’s in Djenné. It was only after this incident that CV. The Fondation then approved the pro- the NRMAC was called in. Although at this posal, and the consultancy has been com- point there was still considerable animosity pleted. between the two families, the NRMAC was able to negotiate an entente between them. This base enabled the NRMAC to convince Summary the family of the wounded participant to withdraw the charges against his assailant The NRMAC has begun to mature as an orga- and get him released from jail. nization at the service of civil society in the Commune of Madiama. It forms the core of a network linking socio-professional Resource mobilization groups, village leaders, locally elected offi- cials and government services. Built on a The NRMAC has benefited from both tech- foundation based in both modern legal tra- nical and financial assistance through the ditions and customary village law, the SANREM CRSP project. However, the com- NRMAC is on the cutting edge of the trans- mittee recognizes that this is really only formation in rural civil society in Mali start-up capital. It must be able to generate during this era of decentralization. It has its own resources if it is to continue provid- provided space for dialogue between vil- ing services to the community. lages and ethnic groups and is building the confidence to address sensitive issues invol- ving resource allocation during this period Internal of socio-economic transformation. NRMAC membership cards have been NRMAC members see communication designed and printed at the expense of the as key to NRMAC viability as a CSO. association to provide a credible presenta- Linkages with CRRA/Mopti research trials, tion of the association, and to generate funds other service providers, NGOs, the village through a one-time membership fee. By the chiefs and the Commune Council place it middle of 2003, over 250 persons had paid in the centre of an important network of the 500 FCFA fee. NRM decision makers. With its members trained not only in the management of their organization, but also in how to pro- External vide leadership for other community Learning of an opportunity to request groups at the village and commune funding from the Fondation de France, the levels, the committee has assumed a lead- NRMAC met to discuss what to propose. ership role in disseminating information Initially a set of income generating activi- concerning new technologies, innovative ties, including animal vaccinations, soap approaches to community based NRM, production by women and techniques for decentralization, tree planting and other seed multiplication, were suggested. After issues of natural resource conservation. further discussion, however, it was con- Members have dealt with conflict situa- cluded that its NRM mission would be best tions and facilitated their resolution. They served through assistance in the develop- have also coordinated the establishment of ment of a communication strategy for the multi-village resource management agree- association. With the assistance of CARE/ ments, although they have not as yet Djenné, the NRMAC submitted a proposal. achieved the stage of full rule compliance. The Fondation de France tentatively Nevertheless, from the villages to the 100 K.M. Moore et al.

Commune Council, community members References have been satisfied with NRMAC perfor- mance and have renewed their mandate. Dietz, T., Ostrom, E. and Stern, P.C. (2003) The In their article in Science, Dietz et al struggle to govern the commons. Science (2003) note that NRM governance is becom- 302, 1907–1911. ing more difficult since critical problems are Flora, J. (1998) Social capital and communities of place. Rural Sociology 63(4), 481–506. at increasingly larger scales, involve non- Pretty, J. and Buck, L. (2002) Social capital and local influences, and require adaptive social learning in the process of natural capacity. They cite five capabilities neces- resource management. In: Barrett, C.B., sary for adaptive governance: (i) providing Place, F. and Aboud, A.A. (eds) Natural information; (ii) dealing with conflicts; (iii) Resources Management in African Agri- inducing rule compliance; (iv) providing culture. CAB International, Wallingford, UK, infrastructure; and (v) being prepared for pp. 23–34. change. The NRMAC is well positioned to Vedeld, T. (1994) L’Etat et la Gestion des respond to these requirements. Chapter 16 pâturages: la création et l’érosion d’institu- evaluates the extent to which these expecta- tions pastorales au Mali (Dossier No. 46). International Institute for Environment and tions have been achieved. Development, London, UK. 8 Conflict Management Training

Jeff Goebel,1 Catherine Ampagoomian2 and Abdoulaye Touré3 1Goebel & Associates, PO Box 2503, Redmond, OR 97756, USA; 2Tulane School of Public Health and Tropical Medicine, Tulane University, New Orleans, Louisiana, USA; and 3CARE/Tombouctou, Mali

Since conflict over West Africa’s diminishing (NRMAC). This experiential training add- natural resources is impeding development resses specific NRM issues of contention efforts and perpetuating land degradation, in Madiama through a three-step process conflict resolution is necessary to immedi- designed to complement traditional conflict ately diffuse disputes and avert violent management and improve decision making. clashes. Perhaps, more important in the long This modular training programme estab- run, is ensuring that resource users have the lishes skills necessary for short-term conflict skills to improve consensus building and resolution and for long-term management. build the social infrastructure necessary Though conceived under the ACM umbrella, to manage conflict situations. Although cur- this approach differs from many other inter- rent decentralization laws have allowed for vention programmes, because it goes beyond natural resource management (NRM) deci- negotiating the resolution of a specific con- sion-making and conflict resolution at the flict. Rather, the training programme empow- community level, the local populations have ers local leaders by providing them with not received adequate training and are skills and a process for sustainable faci- unprepared for these new responsibilities. litation and management of a diversity of Assessing the francophone civil society orga- conflicts. nizations (CSOs), Fox et al. (2001) state that The SANREM Project did not imple- few CSOs have had any previous experience ment ACM to resolve specific conflicts, but in conflict resolution or in the related area of to build ‘people’ skills, training trainers and policy advocacy. Indeed, the successful man- training locals in the necessary skills. Key agement of conflicts requires re-empowering stakeholders were farmers, herders and local institutions or creating new ones, based other natural resource users from the com- on the traditional practices that allow com- mune, members of the NRMAC, local NGOs munities to address new challenges on their and extension agents, traditional leaders own (Pendzich, 1993). and government administrators. Since the In the Commune of Madiama, SANREM initial workshops, lessons from these train- CRSP has introduced an Alternative Con- ing sessions have been routinely transmit- flict Management (ACM) approach, ‘Conflict ted to the village level. Although initially Resolution and Consensus Building’ (CRCB), led by SANREM trainers, NRMAC trainers to the newly formed CSO, the Natural themselves have become increasingly Resource Management Advisory Committee involved. © CAB International 2005. Conflict, Social Capital and Managing Natural Resources (ed. K.M. Moore) 101 102 J. Goebel et al.

This chapter introduces the three com- intra-community conflicts also occur as nat- monly used approaches to conflict resolu- ural resources diminish. Conflicts are inter- tion in the rural communities of the Niger state as well, given that nomads’ and Inland Delta Area. In light of current conflict transhumants’ movements traverse national resolution and management theories, it borders. In these cases, traditional conflict explains how SANREM’s approach and management has not been able to cope with implementation of this conflict management the current conflicts, as conflicts and vio- training in Madiama reverses the cycle of lence over immediate survival and long- conflict and natural resource degradation. It term livelihood have escalated. includes building confidence among local leaders in their ability to mitigate conflict in the commune of Madiama, and establishing Conflict Resolution and Management a successful social infrastructure for long- term consensus building that can address In response to escalating conflict around the the underlying causes of NRM conflicts, world in recent years, there have been major issues associated with diversity, scarcity, developments in the management of com- power and change. peting interests of different natural resource stakeholders. Conflict resolution has mer- ged with NRM to create a more specific field Characterization of Conflicts of conflict management, adapted to NRM issues. The methodology falls into this new Messer et al. (1998) point out that food inse- field, generally referred to as ACM, which curity due to land degradation, scarcity and Pendzich et al. (1994) defines as ‘a multidis- poverty is a major cause and consequence of ciplinary field of research and action that conflict. In Madiama, this is certainly true, seeks to address the question of how people as the cycle of conflict and natural resource can make better decisions together, particu- degradation perpetuates itself. In the Sahel, larly on difficult, contentious issues’ (p. 5). the most serious conflicts have generally Current literature tends to focus on forest involved resource access issues between conflicts, where communities are pitted and among farmers and herders (Cissé, against outside stakeholders such as the 1999). government or private sector companies. The Inland Delta of the Niger is attract- The Food and Agriculture Organization ing a growing concentration of herders, (FAO) Division of Forestry (2000), in farmers and fisherman. Competition has reviewing this field of study, summarized come to replace more complementary and three viable avenues of conflict management cooperative practices of resource use. The for local populations: Customary Conflict conversion of pastureland into cropland Management Systems, National Legal along with the expansion of cultivation into Systems, and ACM. These are described irrigated agriculture along water points has below. limited water and dry-season pasture access, thus disturbing traditional pastoral- ist movement. Meanwhile, the need for Customary conflict resolution herder mobility instigates overt conflict involving the trampling and destruction of Despite Mali having a National Domain Law cultivated fields. and a judicial system, the Inland Delta There is conflict-creating competition region tends to follow traditional land allo- for access to wetlands, grazing areas and cation practices. If conflicting land claims crop residues. This NRM conflict occurs not arise after some years or in the next genera- only between these different types of tion, village elders or the village head may resource users (e.g. pastoralists and farmers) arbitrate. Only if village authorities cannot but also within user types as land degrada- resolve a dispute does litigation proceed to tion shrinks the resource base. Inter- and the Rural Council, and perhaps the Sous- Conflict Management Training 103

Prefet. Most NRM conflicts are currently interactions between parties who have resolved at the village level by the chief and common interests and goals, and who differ his counsellors, or if they proceed further in only on the means of achieving them’ the system the methods remain informal. (p. 47). Cousins states that their theory rests However, if the conflict is not settled at this on a typology of conflicts which determines level it moves up to the Chef d’arrondisse- the most effective resolution/management ment level (pre-commune context), into the approach. His classification system com- national legal system administered by the prises three groups: Justice Ministry at the Cercle level (Djenné) • Management problems, involving argu- or the regional authorities in Mopti. Yet, at ments or differences over the choice of the heart of the perennial nature of most of alternatives among persons having the these conflicts is the lack of a single same goals and interests. accepted authority for the resolution of nat- • Disputes, involving competing but ural resource rights, and consequently negotiable interests, and issues of gain tenure security (Touré, 1996; Lo et al., 1996; or loss. Maïga and Diallo, 1998; Ngiado, 1996). • Conflicts, involving the development and autonomy of the individual or iden- tity group that inherently involve non- National legal system negotiable human needs and questions of identity. The national legal system governing NRM is based on legislation and policy decisions, Cousins (1996) goes on to explain a including regulatory and judicial docu- second theory that deals with processes and mentations. Litigation and adjudication are procedures including fact finding, negotia- the main strategies for addressing con- tion, facilitation, conciliation, collaborative flicts (FAO, 2000). These systems tend to planning, arbitration and adjudication as disregard indigenous knowledge and long- defined by Pendzich et al. (1994) and term community needs, thereby creating a Anderson et al. (1996). These processes pro- winner–loser situation (FAO, 2000). While vide a primary alternative to the legal and access is universal to modern courts in Mali traditional systems of resolution that com- in principle, limits still exist due to costs of monly appear in the literature on Alterna- transportation and a lawyer to advocate a tive Conflict/Dispute Management. This case. National codification of French laws multidisciplinary field of research and has restricted the flexibility of agricultural action seeks to address the question of how and livestock producers to adjust their pro- people can make better decisions collabora- duction systems as needed to assure subsis- tively, particularly on difficult, contentious tence (Mortimore, 1997). In recent years, issues (Pendzich et al., 1994). ACM tech- Sahelian governments have been attempting niques are often used to resolve environ- to adapt village-level customary law to the mental disputes worldwide and encompass need for standardized modern law (Bohrer a range of lesser-known methodologies. and Hobbs, 1996). Similar methods, based on these premises include the International Model Forest Network, Adaptive Management Areas – Contemporary conflict management Western USA, Zimbabwe’s CAMPFIRE, alternatives India’s Joint Forest Management programme and Stakeholder Analysis. Reviewing two theories that support recent ACM refers to a variety of collaborative conflict management approaches, Cousins approaches that seek to reach a mutually (1996) notes that Burton and Dukes (1990) acceptable resolution of conflict issues define conflict management as ‘how to through a voluntary process that has been handle disagreements and arguments over developed as an alternative to adversarial or choices and preferences that result from non-consensual methods, such as judicial 104 J. Goebel et al.

or legal recourse, etc. (FAO, 2000). ACM’s respectful listening, while allowing oppor- three main premises are: (i) it is assumed tunities for confrontation and expression of that conflict is a normal process in society; concerns. It is the process itself that fosters (ii) successful alternative conflict manage- the desired outcomes. The training work- ment relies on constructive communication shops were designed around conflict man- between all legitimate parties or ‘stakehold- agement within the context of Madiama’s ers’ in a dispute; and (iii) power imbalances NRM disputes and were associated with var- are nearly always an issue in negotiations, ious situations such as resource scarcity and and problems that result from negotiating in issues associated with power, control, situations of unequal power can seriously change and diversity. Key stakeholders were undermine efforts at reaching a lasting farmers, herders and other natural resource accord (Pendzich, 1995). Within this gen- users from the commune, local NGOs and eral ACM approach lies the SANREM extension agents, traditional leaders and methodology used in training the Madiama government administrators. Commune stakeholders. This methodology This training model consists of a three framework is based on the fundamental step process that fosters conflict manage- elements of Alternative Conflict/Dispute ment and successful resolution. The first Management, including fact finding, faci- two steps incorporate different sets of exer- litation and collaborative planning, and cise tools to help the NRMAC become suc- negotiation. cessful in resolving internal and external The available literature on AMC and its conflicts. This approach complements tradi- approaches indicates a lack of case studies tional conflict management and is adapted that articulate the actual methods used. The to local circumstances. In Madiama, the SANREM ACM approach emphasizes long- tools used in the first step set the stage for term consensus building and addresses the the workshop and provide the lead facilita- conflict resolution process itself, instead of tor the information needed to choose a spe- facilitating resolution of a particular con- cific path of training. The second step’s tool flict. The following method incorporates set includes the basic conflict resolution these basic premises and fundamental ele- process. The third step is a series of modules ments of ACM, and was adapted to fit the chosen specifically to address the partici- dynamic context of natural resource con- pating community’s issues and conflicts. flicts in the Commune of Madiama. After explaining each technique, the tech- nique’s rationale and the exercises used to teach it are discussed. Training sessions Methodology follow a set pattern consisting of an overview of the training modules and their Alternative conflict resolution/management objectives, the teaching/training techniques training – creating facilitators and finally, a review of the exercises com- prising the consensus building and conflict This section describes a method for training resolution training modules. facilitators in ACM. The SANREM project did not implement ACM to resolve particu- lar conflicts; rather the team trained mem- Overview of workshop training objectives bers of the NRMAC in the basic elements of ACM, including communication, confi- Since SANREM’s conflict resolution training dence-building, and other ‘people’ skills in workshops were considered a training of order for them to engage in ACM, and in trainers (TOT) for ACM, the objectives were facilitation, thereby creating trainers to participatory in nature and focused on intro- implement ACM and further disseminate ducing new skills and strategies to Madiama the necessary skills. The conflict manage- that would complement traditional con- ment process engages all conflicting parties flict management. The primary objectives in a secure, power-equalizing atmosphere of were to: Conflict Management Training 105

1. Increase each participant’s capacity The lead facilitator is the external/out- to communicate effectively regarding emo- side implementer of the workshop and the tionally charged issues, and demonstrate training sessions. Early on in the workshop new communication skills. programme this facilitator hands over the 2. Increase participants’ confidence in task of routine daily facilitation, and acts as dealing with confrontation and develop a a ‘guide on the side, not a sage on the stage’. sense of empowerment. Nevertheless, he or she opens the training 3. Train participants in the art of facilita- session each day, determines which mod- tion and recognition of its value in conflict ules are needed to tailor the workshops to resolution. the local context, oversees the sessions, and 4. Become aware and understand the ensures that the workshops run smoothly physiological and behavioural responses in and remain focused. various conflict situations. The lead facilitator uses a training module called ‘The Interview Process’ in The participatory training process aims to order to gather information and contextual- increase participants’ knowledge of ACM, ize the workshop. Usually done on the first and to raise their self-confidence in address- day, this process helps the lead facilitator ing difficult issues (FAO, 2000). determine vital facts about the situation, such as what the issues are, who the players are, and what behaviours, powers, resources Teaching techniques – training of trainers and stakes are involved. A key responsibil- These TOT workshops used real life situa- ity of the lead facilitator is to ask effective tions to achieve the goal of transferring skills questions and observe the group processing directly to participants, while simultane- their understanding of the questions and ously building their capacity for future com- answers. The lead facilitator ensures there is munity dissemination and guidance of the the opportunity for 100% participation. conflict management process. Training par- ticipants to implement conflict management Facilitation: roles of the successful facilitator means training facilitators to guide the and recorder process. Facilitation training is more than an exercise in learning ‘people skills’ and con- Within the first day of workshops, the flict management practice itself, because it other two levels of facilitation are initiated. also teaches vital teaching techniques. Good Once the workshop commences each day, facilitation skills are best learned through a facilitator (not a mediator, but often one observation of other trained facilitators and of low positional power in the group) is through experience. During the first work- chosen to take on traditional facilitation shop in Madiama, participants observed the roles and lead activities. Facilitators are lead facilitator and rotated into the facilita- also chosen for all small group conflict res- tor position themselves, developing their olution exercises. empathy and building functional skills. At the beginning of a workshop, the par- ticipants are requested to define successful roles. Staying in a circle, one at a time, the Lead facilitator participants answer the following ques- The lead facilitator’s role is to guide the tions: (i) What is the role of a successful process towards the desired consensual out- facilitator? and (ii) What is the role of a suc- comes of the group. Three levels of facilita- cessful recorder? The recorder, who was the tion occur during these workshops. The lead facilitator for the previous activity, records facilitator, who is conducting the training their responses on newsprint. Throughout workshops, carries out the highest level of the workshop, all small group activities facilitation. The other two levels are dis- begin by selecting a group facilitator who cussed below as part of the teaching tech- then becomes the recorder for the next niques and workshop exercises. activity. 106 J. Goebel et al.

This process ingrains a sense of experiential process and through repeti- empowerment within the group as the tion. People will observe others in order to facilitators model power-sharing behav- learn what is and is not effective. iours throughout the workshop. Facilitator Consequently, the facilitator must model behaviours, such as allowing the partici- the desired behaviour. A great deal of pants to direct themselves, selecting partic- thought goes into the actions and interac- ipants to co-facilitate with the facilitator, tions of the facilitator with the group, and facilitating while not directly partici- because the participants are keenly observ- pating (from outside the circle), create a ing the actions and interactions. The pur- sense of equity within the group and further pose of every action should focus on demonstrate the importance of empower- fostering the desired outcomes of the work- ment. By defining the roles of facilitator and shop. Repetition allows participants to recorder, the participants gain a sense of observe the facilitator’s behaviour and thus self-direction. This importance of balance is experience a consistent pattern of conflict illustrated through shifting between the resolution and consensus building. roles of powerful facilitator and submissive recorder. The recorder’s role is to develop Small group organization the trust of groups in conflict by accurately listening and recording what each individ- People have a tendency to coalesce with ual says. others similar to themselves. The old adage Balance of power is encouraged thro- ‘birds of a feather flock together’, describes ughout the workshop and is modelled in this behaviour. One challenge with ‘like behaviours such as having a representative thinking’ or ‘group think’ is that new ideas from one ethnic group begin the workshop are seldom created. Small group activi- and a representative from another ethnic ties must include a diverse cross-section of group close the workshop, or having a participants. New ideas to facilitate old prob- farmer, a herder, Mayor or NRMAC Pre- lems can frequently come from the intro- sident serve on a panel. Panels are formed duction of new perspectives and new ways of to begin a dialogue about specific conflict seeing things. issues. Selecting powerful representatives When working with groups in conflict, or spokespeople from a variety of perspec- it is important for them to exchange per- tives allows for an open sharing of percep- spectives. A stalemate indicates that cur- tions. These activities develop listening rent thinking, or perception, is not skills and foster the importance of respectful providing the group with workable solu- listening and allow participants to have a tions for unresolved conflicts. The opportu- sense of being heard. Grounding, Adaptive nity for people to express their assessment Learning, Defining Successful Roles and of a present situation, their concerns and Defining the Worst/Best Possible Outcomes fears about addressing the situation, and provide ample opportunities to practice lis- their best hopes for the situation in a set- tening exercises. ting, where respectful listening occurs, allows new perspectives to be formed. This group organization gives people the oppor- Repetition and imitation tunity to change their ‘point of view’ to a Since repetition and imitation are fun- ‘viewing point’. It is more difficult to create damental training tools used throughout new solutions to gridlock if groups are training workshops, this participatory ins- using the same base of information, and if tructional model also uses these tools as the groups are made up of people who think primary pedagogic strategies for teaching alike. For this reason, diversity tends to techniques and allowing natural, success- foster a level of insights not produced by ful learning. Imitation ensures that skills homogeneous groups. are highly transferable. The process of imi- tation is most effectively learned through an Conflict Management Training 107

Alternative conflict management training (i) Who are you and what is your relation- exercises ship to the topic of the session?; (ii) What are your expectations of this workshop? and Step one: Tools for empowerment (iii) How do you feel about being here? Since the first step is designed to introduce Participants are free to say as much as or as the workshops, gather information for the little as they want, but usually follow the direction the workshop will take and lead of the first person to speak. develop a safe venue for conflict resolution, The rationale is that this participatory step one tools model effective behaviour process models listening with respect and when striving for conflict resolution by cre- establishes ‘verbal territory’ for all partici- ating a sense of potential equity and respect- pants. It engages the whole brain through ful listening. The process includes but is not thinking and feelings questions, gets people limited to fact finding, balancing power and rooted in the ‘here and now’, allows the facilitation. This step involves participants expression of hidden agendas, shared hopes in a generic set of activities in which self- and apprehensions for the meeting, and awareness and participant interaction create provides initial information to the facilita- an open-mindedness that makes them tor. Someone besides the lead trainer usu- receptive to the workshop’s next steps. In ally facilitates this, demonstrating a step one, they neither process issues, nor are willingness to share power as a facilitator they in the process of conflict resolution per and allowing participants to realize they can se; rather they are gaining confidence, facilitate the process themselves. mutual trust and facilitation skills to engage fully in the workshop. THE ‘GREETING CIRCLE’ The greeting circle, an American Indian adaptation exer- THE ‘GROUNDING’ In this opening tool, cise, allows participants to introduce them- participants sit in a circle and answer three selves individually to each other. The questions, one person at a time, going participants stand and the greeting circle around the circle. The three questions are: activity begins with a designated leader. The

Fig. 8.1. NRMAC trainers lead session on conflict resolution. 108 J. Goebel et al.

lead person moves into the centre of the what did not work in the session. This feed- circle, turns to the person next to him or her back allows participants to improve their and greets them. The lead person then abilities as facilitators and provides the lead moves to the next person and so on around facilitator with ideas for continual improve- the circle. Meanwhile, the person first ment. Finally, asking provides direction for greeted also steps into the centre of the upcoming experiences. The adaptive learn- circle, and, following the lead person, greets ing process at the session’s end provides the next person in the circular line and con- direction for future sessions because of the tinues greeting people around the circle. emphasis on addressing fears and focusing This continues until each person has gone on best outcomes. The human mind wants around the inside of the circle greeting to find a solution. others, and the outside of the circle, being greeted. ‘HONOURING’ Honouring is a process of The rationale is that the circle is a formally recognizing individuals who have design of shared power, a theme continually benefited the group in some way, such as reinforced throughout the conflict resolu- people who facilitated, prepared meals, tion process. It is a turn-taking exercise and made contributions to the group in other again, is usually led by someone in the com- venues, etc. The group is asked to stand in a munity. This activity breaks down intimida- circle and the individual(s) being honoured tions and levels power. Greeting is one of the are asked to come to the centre of the circle. oldest ways of connecting for humans. From Someone in the outside circle states why the past experience, many conflicts, particu- people in the centre are being honoured. larly ones dealing with power and intimida- The group applauds once the words are tion, are resolved through the greeting. The spoken. perspective of village participants was that The rationale is that honouring people the greeting circle allows conflicting parties builds community. After modelling the to build a friendship and that it allows them process of honouring, the facilitator often to put the conflict in the centre of the circle uses it to resolve conflicts between parties where it can be resolved. Malian custom has by asking one side to speak to the other by its own greeting circle, usually done at the saying something respectful or explaining end of a conflict resolution process. why one party appreciates the presence of the other in its community. The people AN ADAPTIVE LEARNING PROCESS Adaptive being honoured in the centre of the circle are learning occurs at the end of any enriching often very self-conscious. Again, conflict experience such as the greeting circle, or at resolution is about awareness; self-con- the end of the day. Again going around the sciousness, or self-awareness, is an impor- circle, one person at a time speaks, answer- tant first step towards change. When people ing two questions: (i) How do you feel about are aware of their internal feelings in a given the experience or situation? and (ii) What situation, they can then make a conscious did you learn from it that will make you suc- decision to take action and change behav- cessful, resolve the conflict, overcome the iour to what is consistent with their desired impasse, etc.? outcomes. The rationale is that there are three main reasons for asking these adaptive Step two: A process for coping with conflict learning process questions. First, asking reinforces topics covered during the session This is where the actual process of alterna- and brings out individual insights concern- tive conflict resolution begins. Step two tools ing these topics. Having others introduce continue to allow participants to explore and what was experienced during the session is understand the natural human response a strong method of reinforcing learning. to potentially threatening situations and Second, asking provides feedback to the issues. The progression of questions noted group and the facilitator about what did and below moves from developing a common Conflict Management Training 109

understanding of the situation and the asso- • What beliefs/behaviours/strategies/ac- ciated feelings to putting the situation in an tions will foster the best outcomes? emotional context. The second question explores the ‘worst possible outcomes’ imag- WORST/BEST POSSIBILITY The worst/best ined of confronting the situation. These possibility exercise allows the participants imagined results are why confronting the to explore and understand the importance of issue is ‘dangerous’, because they often worst and best outcomes as well as possibil- paralyse individuals and groups into inac- ities. The following two questions are asked: tion. Should this occur, it is important to ask (i) What are the worst possible outcomes of the other side of the question: What is the the workshop? and (ii) What are the best worst possible outcome of not confronting possible outcomes of the workshop? the situation?, because this question allows Understanding these possibilities is impor- groups to recognize the hopelessness of tant, particularly if the group is confronting either side of the question. a serious conflict. Also, allowing the group Step two tools allow participants to suc- to express the worst/best outcomes at the cessfully move from a focus on worst possi- workshop level helps the group begin to ble outcomes to a focus on best possible understand the physiological patterns of outcomes as a distinct possibility. Often, individuals who confront potentially threat- allowing the group to explore conflict reso- ening situations or issues. lution at a generic level, before confronting Worst outcomes are feared future out- real and potentially threatening issues, pro- comes, often based on past experience, with vides the group with an understanding of a presently experienced emotion and physi- the human behaviour surrounding conflict cal reaction. When believed, they affect resolution. As a result, they are more capa- people’s perceptions, beliefs, values, and ble of successfully focusing on the issue. strategies and, consequently, tend to be self- These tools also help in developing a more fulfilling prophecies. (Outcomes are auto- complete understanding of the issue, partic- matic and based in the lower brain stem). ularly if diverse perspectives are involved. In contrast, best outcomes are hoped for With this understanding and clarity of the future outcomes, sometimes not previously best possible outcomes, the human mind experienced, but intensely imagined, with a works to solve the problem of moving to the presently experienced emotion and physical desired outcomes of the group. response. When believed, they affect people’s perceptions, beliefs, values and THE BASIC PROCESS QUESTIONS The fol- strategies. And, consequently, tend to be lowing set of questions are asked of the par- self-fulfilling prophecies. (Unlike the fear of ticipants in order to develop an worst outcomes, the hope for best outcomes understanding of the conflict, gather con- is not automatic and originates in the neo- cerns the participants have about con- cortex part of the brain that is seeking the fronting the situation, define their best good of community.) outcomes and foster the best outcomes of Possibility thinking is the acknowledge- the situation. These questions are directed ment that both worst and best outcomes are to the individuals or groups. inherently present in each moment, up to, and often after the event. This balanced • What is the situation? (Define the con- view allows movement towards desired out- flict. What is the evidence of this con- comes, because it is based both on negative flict in your environment?) How do you and positive thinking. A simple analogy can feel about it? help with this explanation. It is a fact that • What are the worst outcomes of con- the sun will rise (or the earth will turn fronting/not confronting unresolved towards it). However, there is a possibility of conflict? not seeing the sun rise because of cloud • What are the best outcomes of con- cover in the sky or some other impediment fronting unresolved conflict? such as a windowless room, even though it 110 J. Goebel et al.

Fig. 8.2. Conflict resolution through the village theatre. has risen. The worst possible outcome is the participants to identify the best possible believing the sun has failed to rise, even outcome of confronting the conflicts though it always rises. Possibility thinking between herders and farmers. The answer in conflict resolution approach acknowl- was that people would learn to live in har- edges the balance of a focus on both worst mony with each other and foster actions and and best possible outcomes. behaviours that are consistent with that out- An example of this possibility thinking come. The group from Tatia Nouna liked occurred at a workshop in the village of this alternative and decided to focus their Tatia Nouna when the villagers and partici- actions towards this outcome. Throughout pants in a group were asked to identify the the process of moving towards peaceful co- worst possible outcomes that could result existence, any of the three outcomes are from confronting the conflicts between possible, and by acknowledging these possi- farmers and herders. The answer was vio- bilities, the community is able to choose to lent confrontation and death. If the mind move towards their desired outcome. imagined this as the future, then why would The process also changes behaviour anyone take any action to resolve the situa- concerning creating desired outcomes tion if any attempt to resolve the issue because beliefs change as people learn from would result in death? In a second worst each other in a respectful environment. possible outcomes question, the partici- Because the human brain is a problem solv- pants were asked to name the worst possible ing organ, when it is presented with a prob- outcome of not confronting the conflicts lem, as when the space between a present between farmers and herders. Again, the situation and best possible outcome are cre- answer was death. This created a difficult ated, the mind goes to work to seek resolu- problem for the mind to solve. If confronting tion to the problem. Sometimes, because the the problem led to death, and not con- answer is difficult to determine, the way fronting it also led to death, and no one the mind finds resolution is to say ‘I don’t wants to die, then what is to be done? At this know’ or ‘It is impossible to solve’. Once a point, there is a third possibility that asks person acknowledges the possibility of not Conflict Management Training 111

knowing or the impossibility of finding a focused on modification as change. Usually resolution, restating the question in terms of in conflict resolution, transformational possibility (i.e. ‘if you did know?’ or ‘if it change is required to move people to foster were possible, what would you say?’) the best possible outcomes of confronting recharges the mind to examine the problem conflict. This process is about modifications again, but with a different emphasis. The and transformations. The participants were mind has a very difficult time leaving an shown the Change model diagram (Fig. 8.3) unresolved issue alone and, therefore, con- with results being either the Worst Possible tinues to search for a solution so that it can Outcomes or Best Possible Outcomes. The be at peace. diagram demonstrates that if change at a Because of the power of the human basic level is desired, someone must modify brain to focus and the limitations of the strategies and actions. If change at a funda- brain to gather all relevant information for mental level is desired, someone must trans- problem solving, the mind often confuses form beliefs and behaviours. If natural fact with possibility. The mind tends to pro- resource users want to change the Sahel, ject potential outcomes as real due to past that change will have to be at the funda- experience, when, in fact, the future has not mental level, because the underlying cause happened yet. This sets the body up to take of these conflicts is the perception of a appropriate action based on the ‘perceived’ scarcity of degrading resources. Hence, in outcome. Human beings are often quite dif- order to resolve these conflicts, it is neces- ferent if there is a perceived positive out- sary to have more than conflict resolution come instead of a negative outcome. These strategies and good communication; rather, actions and behaviours can have the effect it is necessary to change the management of of creating the outcome most focused on. these resources. This is where conflict reso- This reaction is based on the automatically lution methodology becomes linked to the ‘hard-wired’ system for self-preservation underlying issues of these disputes and generated from the lower brainstem that brings in holistic management (HM). protects us from the physical and emotional damage created from worst outcomes. A less Step three: Module tools for contextual powerful force in the mind is the ability to adaptations imagine best outcomes generated by the neo-cortex part of the brain. The third step in the conflict resolution The ‘worst possible outcomes’ response workshop is tailored to the particular local allows the exploration of the possibilities of context. A series of modules are chosen and best possible outcomes in confronting a sit- adapted to the specific issues and conflicts uation or issue. Because most people are facing the participating community. Each paralysed with the fear of confronting module uses a set of questions, adapted to an issue, the potential for fostering best pos- the issues at stake, to guide people through sible outcomes of confronting a situation specific conflicts. The modules are: one-on- never arises. When people are allowed one conflicts, managing scarcity, managing the opportunity to explore best possible power, managing change, managing diver- outcomes, they begin to imagine ways to sity and conducting interviews. encourage these outcomes. Two questions arising from the determination of the best possible outcome engage the problem solv- Discussion ing nature of the human mind: (i) What beliefs and behaviours will foster the best The CRCB training workshops conducted in possible outcomes? and (ii) What strategies Madiama have demonstrated the success of and actions will foster the best possible an ACM programme in Mali. The methodol- outcomes? Beliefs and behaviours focus ogy was tailored to the Commune of on a fundamental transformational change Madiama and specific NRM conflicts that within people. Strategies and actions are are occurring across West Africa. As both a 112 J. Goebel et al.

Fig. 8.3. Change model. short- and long-term approach to manage phere of respectful listening that allows for conflict, the SANREM programme has led to the confrontation and expression of con- considerable changes in the ways that con- cerns to foster desired outcomes. The work- flicts and NRM decision-making are han- shops in Madiama were designed around dled within the community. In this sense, its conflict within the context of NRM disputes method has developed into a more partici- and included modules covering concepts patory approach to ACM by engaging partic- such as scarcity, power, control, change and ipants themselves to lead the conflict diversity. SANREM’s approach to CRCB has management process and ultimately resolve led to changes in the community’s percep- their own conflicts and build consensus in tions of conflict and local power struggles. management decisions. In training commu- nity members and the NRMAC in sustain- able ACM methods of conflict management, Equalizing power SANREM has built the communities’ capac- ity to manage conflict more successfully, Creating an atmosphere of respect and independent of outside intervention. equality is essential to this process and pro- Conflict resolution rests on a handful of motes potential power equalizing. Exercises premises about conflict, change and power like ‘greeting circles’ enable each individual (Pendzich, 1993). The training method is of differing status and power to interact on a flexible enough to be adapted to different level playing field, creating an increased cultural contexts in order to build sustain- level of confidence and a sense of empower- able community capacities to resolve con- ment for participating farmers and herders. flicts and reach consensus. The process of Power tends to equalize itself, especially conflict management engages all conflicting when using such techniques as rotating parties in a secure, power-equalizing atmos- facilitators, balancing gender and power in Conflict Management Training 113

group settings, and having the lead facilita- the community that any individual or group tor on the side rather than centre stage. of individuals can create a secure forum to When each person is expected to engage in discuss conflicts and their reasons, and the same exercises and interact as stake- serve as a mechanism for change. The train- holders during workshop exercises, individ- ing workshops enabled individuals to re- uals feel less intimidated and more think their views of resolution processes by empowered to collaborate as equal stake- using skills such as listening, recording and holders and resource users with diverse facilitating. Success was not measured in interests. These transferable techniques negotiated agreements and resolution of allow for a more open forum and an aware- specific conflicts, but was defined by enthu- ness of how to best cope with issues of siastic participants engaging in training ses- intimidation and power. sions and sharing their experience with the villagers. Most of the participants continue to use facilitation and conflict resolution New skills, perspectives and behaviours techniques in various roles and settings beyond major NRM conflicts. These settings Based on general principles of conventional include a women’s cooperative and individ- conflict resolution and ACM, the participa- ual households, where behaviour has been tory workshops helped to develop a sense of modified. This illustrates the adaptability trust vital to conflict management and the and sustainability of the approach, as well consensus building that is fundamental to as the skill-building dimension of the work- effectively applying newly formed human shops. Participating community members capital. Facilitation skills empowered every have learned to use ACM for larger commu- participant to become a part of the process. nity conflicts, and they have used individ- Each workshop had a purpose, and each ual ACM skills and approaches in other exercise was done with a purpose. Through settings, as well. repetition and imitation, the participants come to understand the purpose that the ‘adaptive learning process’, completed after From conflict to consensus the workshop exercises, reinforces. The skills demonstrated in each exercise are SANREM’s CRCB approach to helping com- transferred directly to the participants; con- munities cope with conflicts over NRM and sequently, while the group was creating, extend their learned skills beyond present individual consensus-building skills were conflict is an important addition to the field learned during each workshop. Ultimately of ACM. This ACM process complements participants learned to alter the ways in other aspects of the SANREM project, which they communicate and engage in thereby increasing the success of current social relationships. The conflict manage- research designed to transform the cumula- ment process created new perspectives and tive underlying biophysical and social brought diverse views to the table. By creat- causes of conflict. As previously mentioned, ing willingness to explore, communal solu- there is a void in the literature concerning tions became a more viable option. consensus building approaches and ACM Consequently, the learned and applied training methods, where the purpose is to CRCB skills have allowed best possible out- develop the community’s capacity to comes to emerge from NRM disputes in the manage conflicts as they arise, not merely to commune. solve one specific conflict. Disseminating A local observer noted that the conflict ACM approaches and different experiences management training workshops demon- will change community-level capacity to strated that conflict resolution was ‘every- manage natural resource conflicts in the one’s business’; someone did not need to be region and beyond. To bring change to their a village chief to have such facilitative skills land and community, NRMAC members are and capacities. The workshops have shown changing their views on confronting and 114 J. Goebel et al.

managing conflicts, and consequently are long-term sources and perpetrators of con- adapting their traditional methods of con- flict. One-time negotiated resolutions and flict resolution via alternative processes and static legal agreements are not always a newly learned individual ‘people’ skills. viable or sustainable option for rural com- Beyond the cases in which empowered par- munities given the dynamic, ever-changing ticipants have applied new skills, there is a nature of NRM. Initial facilitation along with broader recognition in the community of consensus building/decision-making skills behaviours that lead to and increase conflict and mechanisms, in sync with local tradi- and, as a result, behaviour is becoming more tions, will allow local Malian and West positive and respectful. This shift in behav- African communities to successfully manage iour creates a more peaceful tone in com- their resources within the context of decen- munities when confrontations between tralization. Consensus building skills are an disputing land users occur. essential form of human capital for reaching Movement is occurring in the region to agreements on NRM. Capacity building for expand conflict management training to conflict resolution creates a critical mass of more communes in surrounding areas. As human capital necessary to change social the process proves adaptable and effective behaviours and relationships that can lead to with the transhumant population, the more effective decision-making and manage- opportunity exists for this process to extend ment of natural resources. Efforts should be across national borders. The initiation of made from governments and NGOs to sup- activities to regenerate natural resources port ACM interventions, using consensus (i.e. holistically managed bourgoutière and building skills to improve decentralized, rangelands) exemplifies the consensus- community-based NRM. building aspect of these conflict manage- ment workshops aimed at natural resource decision-making. NRMAC members have References shared their CRCB workshop experiences with villagers. Using alternative methods to Anderson, J., Gauthier, M., Thomas, G. and confront disputes and resolve conflicts is Wondelleck, J. (1996) Addressing natural leading to a more open community with resource conflicts through community beliefs that confronting NRM disputes can forestry: setting the stage. E-conference on potentially create positive outcomes. There Addressing Natural Resource Conflicts is a greater sense of security regarding nego- through Community Forestry. Food and tiations and agreements. Compared to cus- Agriculture Organization. Rome, Italy. tomary and legal mechanisms of conflict Bohrer, K. and Hobbs, M. (1996) Post-Praia resolution, this ACM approach transcends Progress towards Tenure Security and Decentralization: Review of CILSS-member important gate-keeping issues found in County Legislative Reforms. University of both. CRCB creates a forum empowering all Wisconsin, Land Tenure Center, Madison, NRM actors, including herders and women, Wisconsin, pp. 1–41. which tend to become marginalized during Burton, J. and Dukes, F. (1990) Conflict: Practices traditional conflict resolution. CRCB builds in Management, Settlement and Resolution. relationships across traditionally divided St Martin’s Press, New York. ethnic and production system communities Cissé, S. (1999) Décentralisation et développe- working towards mutually beneficial out- ment dans le delta interieur du Niger (Mopti comes and avoiding winner-take-all legal Republique du Mali). In: Bourgeot, A. (ed.) system resolutions. Horizons Nomades en Afrique Sahélienne: Sociétés, Développement et Démocratie. National and regional governmental Edition Karthala, Paris, pp. 137–154. support is needed to continue the use of ACM Cousins, B. (1996) Conflict management for strategies to manage widespread NRM con- multiple resource users in pastoralist and flicts. The cyclical nature of conflict and nat- agro-pastoralist contexts. Institute of Deve- ural resource use requires a multi-levelled lopment Studies Bulletin 27(3), 41–54. approach in order to mitigate both short- and FAO (2000) Conflict and Natural Resource Mana- Conflict Management Training 115

gement. Food and Agriculture Organization, Mortimore, M. (1997) History and Evolution of Rome, Italy, pp. 1–20. Land Tenure and Administration in West Fox, L., Ouattara, R., Boly, A., Ba, M.B., Konaté, Africa (Drylands Programme Issue Paper No. Y., and Charlick, R. (2001) Francophone 71). International Institute for Environment Civil Society Organization Assessment: Mali and Development, London, UK, pp. 1–35. and Côte d’Ivorie Synthesis Report. West Ngaido, T. (1996) Redefining the boundaries of Africa: Civil Society Strengthening for control: post-colonial tenure policies and Conflict Preventions Study. Associates in dynamics of social and tenure change in Rural Development (ARD), US Agency for Western Niger. Ph.D. dissertation, University International Development, Broadening of Wisconsin, Madison, Wisconsin. Access and Strengthening Input Market Pendzich, C. (1993) Conflict management and Systems (BASIS), Burlington, Vermont. forest disputes: A path out of the woods? Lo, H., Sylla, C., Ndione, P. and Agne, A. (1996) Forests, Trees, and People Newsletter, 20, Ressources forestières et conflits au Sénégal: 4–9. quelles approches pour une typologie? Pendzich, C., Thomas, G., and Wohigent, T. (1994) Bulletin Arbres, Forêts et Communautés The role of alternative conflict management Rurales 8, 26–36. in community forestry (Community For- Maïga, I. and Diallo, G. (1998) Les conflits fanciers estry/FTPP Working Paper No. 1). Food et leurs gestion dans la 5ème Région du Mali and Agriculture Organization, Rome, Italy. (Dossier No. 76). GRAD. International http://www.fao.org/ DOCREP/005/X2102E/ Institute for Environment and Development, X2102E00.htm (accessed 4 May 2004). Programme Zones Arides, London, UK. Pendzich, C. (1995) Gestion des conflits sur Messer, E., Cohen, M.J. and D’Costa, J.D. (1998) les ressources forestières : est-on sorti de Food from Peace: Breaking the Links l’auberge? Arbres, Forêts et Communautés between Conflict and Hunger (Food, Rurales 7, 31–37. Agriculture and the Environment Discussion Touré, D. (1996) Prévention et gestion des conflits. Paper 24). International Food Policy Centre Canadien d’Etude et de Coopération Research Institute, Washington, DC, pp. Internationale, Ouagadougou, Burkina Faso, 1–44. pp. 1–50.

9 Holistic Management Applied to SANREM Research and Development in Madiama

Sam Bingham Allan Savory Center for Holistic Management, 394 Vanderbilt Road, Asheville, NC 28803, USA

The length of title indicated the breadth of Although no categorically successful the challenge. Each word in ‘Sustainable model of such a programme had yet been Agriculture and Natural Resource Mana- seen in West Africa, new techniques for gement Collaborative Research Support improving communication among parties of Program – West Africa (SANREM/CRSP – varied background and point of view gave WA)’ coded one of the Big Ideas of turn-of- grounds for optimism. The traditional millennium thinking about development anthropologist’s questionnaire, for example, and progress. If the dominance of northern, has been replaced by much more open developed-world, science had in the past ended and often graphically and symboli- promoted destructive exploitation of nature, cally enhanced dialogues structured accord- it would now exploit the synergy of local ing to methodologies bearing names such as and regional knowledge to discover, or Rapid Rural Appraisal, Participatory Land- rediscover, the science of managing a return scape/Lifescape Appraisal (PLLA), etc. to stable productivity of one of the most dev- Mali, like many other countries in Africa astated ecosystems of the world. today, now enjoys a strong contingent SANREM reflected a widespread acade- of home-grown scientists and technicians mic interest in studying traditional practices who can call on deep cultural know- for previously overlooked environmental ledge and communicate directly in local strengths and in conducting useful dialogue languages. with traditional practitioners about ways to Despite such assets, however, the grail preserve these in modern conditions. of managing resources in sustaining and Indigenous practitioners working in respect- sustainable ways spontaneously embraced ful partnership with American and African by the local population remains elusive. scientists would discover, or rediscover, a Holistic Management® (HM) was one packet way to reverse the degradation of the of seeds that SANREM broadcast on the real Sahelian landscape and its agropastoral and virtual soils of the Commune of economy. The programme would honour the Madiama in its attempt to find answers. As scientific traditions of the USA and Europe the copyright symbol suggests, Holistic as well as the conventions and practicalities Management, capitalized, refers to a rather of multi-ethnic West African communities strictly defined management model, in while assuring inclusion of gender, ethnic this case one developed through the non- and caste diversity. profit Allan Savory Center for Holistic © CAB International 2005. Conflict, Social Capital and Managing Natural Resources (ed. K.M. Moore) 117 118 S. Bingham

Management and largely embodying the West African scientists and development thinking of Allan Savory himself (Savory, professionals. Madiama itself brought 1999). together three ethnic groups, each interpret- Savory, a trained wildlife biologist born ing economic and social activity and natural in southern Africa, grew up amid the eco- phenomena according to its own culture. nomic, cultural, and environmental chal- For this reason, bringing the Holistic lenges facing his native continent and has Management point of view into SANREM’s spent half a century personally involved in Mali project represented a rather courageous the ever-evolving succession of European choice on the part of the project leaders. The and North American policies to address holistic approach took its place as one more them. He is known among scientists first for player in a marketplace of ideas governed by synthesizing from existing research and his acreative tension. Nevertheless, the Holistic own observation the radical scientific Management Model’s recognition of cultural hypothesis that the ecological vigour of cer- priorities and its emphasis on practical tain environments requires large herds of management choices accord well with large herbivores behaving in the manner SANREM’s mandate to involve local people induced in ‘wild’ conditions by the pres- in the scientific discussion. Its concern for ence of pack-hunting predators. only pushing action when social organiza- Savory’s efforts to incorporate this idea tion could support it mapped perfectly on to into a practical model for managing what he the conflict resolution component of the calls ‘brittle’ environments, of which the project and the development of the Natural Sahel offers a prime example, led him to a Resource Management Advisory Committee second, more generalized conclusion. Holis- (NRMAC) within the commune. tic Management reflects the view that In retrospect, the relevant questions are: the ecosystem, including human economic 1. What ideas did Holistic Management activity, cannot be studied or managed contribute to a general scientific under- piecemeal. Research, policy, and manage- standing of the problems of Madiama? ment risks miss the point by attempting to 2. How did it affect perceptions among isolate parts of the system – as in laws to the various constituencies within the com- protect a particular endangered species or mune? research on the productivity of a particular 3. How did it affect communications plant. between people in the commune and the Over the years, Savory and others of various researchers who worked there? similar persuasion have criticized much 4. Will it account for any significant and research, policy, and management for fail- lasting change in decision-making or man- ure to account for externalities of the social agement at any level? and environmental context. However, they have also contended that overemphasis on To answer all four questions requires some producing clean, quantifiable data often understanding of how the Holistic Mana- produces irrelevant data, or at best, data of gement Model evolved, and how it may be little use for informing practice. applied. The Holistic Management Model® pre- sumes to help the researcher, politician, or manager avoid these pitfalls, but the stri- The Gestation of Holistic Management dency of arguments among scientists over the charge reflects significant differences in Allan Savory (1999) credits the revolution- perception of how science should operate ary insight that overgrazing is a function of and the relationship between science and time management and animal behaviour, practice. As the project began, preconcep- not herd size, to the French scholar of pas- tions about what should be done were tures, André Voisin. Savory himself worked shaped in different disciplinary domains in the rather different environment of and intellectual cultures of American and Southern Rhodesia, now Zimbabwe, where Holistic Management in Madiama 119

he sought to explain why the ‘unmanaged’ weather, economics, culture and envi- grassland he knew from his youth sup- ronmental conditions. ported enormous herds of wild ungulates and recovered from even severe droughts Savory figured that the conjunction of without loss of biodiversity but degraded these principles would bring a new dawn to rapidly when grazed by much smaller num- livestock production, game management bers of domestic stock under human man- and most efforts to reverse desertification in agement. areas impacted by livestock. Sheep, goats, Inspired by Voisin to test the common cattle and horses would become instru- wisdom that the damage resulted from ments of restoration. Reducing livestock simply overstocking the range, he eventu- stock numbers, a policy that no herding ally found four keys to unlock the riddle. society has ever accepted, would fade as a policy imperative. The economic potential •Inso-called ‘brittle’ environments, of long-lost grassland would salvage the bal- characterized by low humidity, a pro- ance sheet where irrigation, fertilizer and longed dry season and erratic preci- supplementary feed had failed. The docu- pitation, plant communities stagnate mented success of individual managers over when rested, and bare ground increases, the past 30 years leaves no doubt that Savory whereas rest in stable, humid environ- had got something right, though the revolu- ments (such as France) results in cov- tion is taking its time. ered soil and rapid development of Anthropologists, meanwhile, followed diverse communities, typically culmi- a parallel path to similar conclusions. In nating in forest. their seminal paper ‘Rethinking range ecol- • Rest allows succession to advance in ogy: implications for rangeland management non-brittle environments, because mic- in Africa’ (1991), British-based scholars robes and other small organisms effi- Scoones and Behnke (1993) argued the ciently recycle the carbon accumulated genius of traditional nomadic societies in by plants. In brittle environments, how- managing land and animals for maximum ever, carbon is better recycled by grazing efficiency. They debunked the stereotype animals through both trampling and that traditional people acquired large herds digestion. Large herbivores are thus nec- for prestige only and attacked a brace of essary to maintaining diversity, produc- standard development policies such as vet- tivity and stability in brittle areas. erinary programmes, fencing, water deve- • Overgrazing and overtrampling are lopment, genetic improvement, confined principally functions of time, not num- feeding, sedentarization and matching sto- bers, as Voisin had determined. Dense cking rate to ‘carrying capacity’ – this con- herds of great size that move frequently cept being meaningless where conditions and allow plants to recover before fluctuated and herds moved so dramatically. regrazing them recycle nutrients with- Their findings rested on a decade of out weakening plants. Even single ani- research issuing from the same institutions mals that do not migrate overgraze the that did the original development of Rapid most palatable plants in the areas where Rural Appraisal techniques – the Inter- they linger. national Institute for Environment and •Onthe enormous unfenced ranges of Development in London, the Institute of pre-colonial Africa (and similar brittle Developmental Studies at the University of environments elsewhere), pack-hunting Sussex, and the British Government’s predators assured beneficial herd beha- Overseas Development Institute. A stream viour, and nomadic herders generally of studies in the same vein gave academic developed similar patterns. Limited support to recent policy initiatives to estab- lands, however, demand management lish (re-establish) the grazing rights of tran- that is ‘holistic’ to the extent that it sient herders in the brittle environments of responds to ever-shifting conditions of Africa (Behnke et al., 1993). 120 S. Bingham

Kerven and Alimaev (1998) published (Knowledge of hydrogen and oxygen similar findings from an extensive study of does not give the whole story on water. Central Asia following the Soviet collapse. Knowing the alphabet does not imply They also concluded that productive, envi- understanding of words, sentences or ronmentally sound, husbandry of brittle Shakespeare.) lands depended on large, moving, herds and • In complex, self-organizing systems, the enormous flexibility of herding soci- processes and the relationship patterns eties, including their willingness to take up of wholes are the only constants and other professions in time of drought. therefore become the focus of research This refrain of extreme flexibility, and management. common to Savory and the British anthro- • Management, being a human endeav- pologists, has made the claims of both diffi- our, must have a goal, an assumption of cult to document through methodologies a better state. based on controlled studies of short dura- tion (less than 20 years, depending on pre- Therefore, the shepherd/scientist-friendly vailing weather cycles). If there is a ‘science’ Holistic Model looks like this: of management, it is the study of how orga- The whole is at the top, followed by a nizations and their leaders respond to three-part goal describing the quality of life change in order to achieve goals. Good man- you expect, the production needed to sup- agement responds quickly and properly to port it, and the condition of your resources new opportunities, including conditions required to produce that. Next are the changed by management itself (feedback). processes that define the ecosystem – com- munity dynamics (succession), water cycle, mineral cycle and energy flow. Therein lies A Schematic Rendition the genius of nature – the interplay of untold variables boils down to measurable effect on A good model predicts. You tweak it here only four basic processes (Fig. 9.1). and see what wobbles there. It interprets. Since holistic managers posit that con- You can look at a wobble and deduce the ditions reflect management, they can assess tweak that produced it. It instructs. You can past practices and future propositions by tweak it this way and that until you learn looking at the effect of management tools on what makes the wobble you want. And best the four processes. Fortunately there are of all, you can do all this cheaply without only six direct tools to consider, plus wrecking anything important. Computers money, labour and creativity. If this seems allow you to poke around complicated oversimplified, consider that virtually all models, but the most powerful are simple, past policies and management practices and a participatory research/development have not included living organisms, grazing, programme requires one that equally ampli- or animal impact in the tool kit and fies the brainpower of the Peul shepherd do not recognize the differential effects of trying to supply goat milk for her baby sister long-term rest in brittle and non-brittle and the scientist sipping coffee at the stop environments. light on the way to her institute. When planning future actions, ethical, The Holistic Management Model rests economic, ecological and social considera- on four axioms: tions weight the value of a given tool, and these appear in seven testing questions that • The condition of the resources always are augmented by certain guidelines that reflects the management of them. No determine how the tools are applied. excuses. Finally, a feedback loop joins planning, • Management means management of monitoring (the four processes), controlling ‘wholes’ which have characteristics not and replanning into one process. present in, or predictable from, their Over the course of the project, a number constituent parts studied in isolation. of workshops and training sessions were Holistic Management in Madiama 121

Fig. 9.1. Schematic of the Holistic Management Model. conducted to promote this line of reasoning. Holistic Management Training in Generally these events were carried out in Madiama conjunction with conflict resolution train- ing and focused primarily on members of The NRMAC selected two women and three the NRMAC composed of representatives men participants to join ten IER scien- from all the villages in the commune. tists, three US scientists, seven community Institut d’Economie Rurale (IER) personnel service technicians, the mayor of Madiama attended some of these and others were and his assistant in a 5-day Holistic exposed to the Holistic Management Model Resource Management Workshop on Con- in separate sessions. sensus Building and Conflict Resolution 122 S. Bingham

(1–5 November 1999). After the workshop, experience in introducing Holistic Manage- NRMAC members (under the guidance of ment in his country through the World the HM trainers) reviewed the workshop for Bank’s West African Pastoral Perimeter the full NRMAC and then led the partners in Project. Consultations followed with the a transect tour and an HM-focused discus- NRMAC in March and again in July 2001 to sion of natural resource features of the assist in setting up managed grazing trials in Commune. The committee expressed an two sites (Siragourou and Torokoro) and interest in training for all NRMAC members training local environmental monitors to in the national language. track compliance. In September, a delega- From 2 to 17 May 2000, three consecu- tion from the NRMAC had an opportunity to tive 3-day workshops were conducted with visit the HM site in Chad and speak with vil- all members of the NRMAC, two in Holistic lagers who had applied it, albeit in circum- Management and one in Conflict Resolution. stances much less complicated than prevail These sessions included the new CARE/ in Madiama. Djenné SANREM Facilitator. Researchers of Three additional training programmes IER participated in some sessions. Com- in October 2001 were conducted by the mittee members learned to evaluate different Holistic Management and Conflict Reso- options for improving resource management lution trainers focusing on managing power in the Commune. Group discussions were conflicts, the management of conflicts asso- encouraged and the committee generated ciated with wetlands, and the management several proposals to improve soil fertility of pasturelands in general. The overriding and the productivity of wetlands. The con- objective of this training was to provide a flict resolution workshop included training structured opportunity for NRMAC leaders in group facilitation and consensus building, to gain experience and become confident as including many techniques that were imme- trainers. diately applied and continue to be used. In March 2002, an HM trainer provided NRMAC members also roughed out their training on how to develop a management holistic goal reflecting universal values – plan for timed grazing in open-range pas- a desire for food security, health, safety, edu- turelands. This workshop involved intro- cation, respect, free time and opportunity; a ducing the NRMAC to the timed grazing of stable income from small enterprise, crops, open-range pasturelands, identifying vil- livestock, fish and increased skills; and a lages ready to participate in the experiment, resource base of good water, fertile soil, lush evaluating the pastures to be managed, grazing and vigorous people. dividing the pastures into parcels, and pre- The autumn 2000 training session in paring village level environmental super- Holistic Management began with a review of visors to monitor the activity. This was the research trials on soil fertility and pas- followed up in July 2002 by a second HM ture management. Visits to farmer field sites training session to assist in the establish- facilitated these discussions. Small group ment of the timed grazing programme in two exercises reinforced the Holistic Manage- sites (Siragourou and Torokoro). The pri- ment themes and the application of tools mary objectives of this training were to and management techniques presented. The familiarize the environmental monitors NRMAC analysed the conditions for bour- with the terrain and the routines they must gou1 management (bourgou being an impor- follow, examine the rotational parcels to tant source of fodder that grows in flooded determine the initial grazing periods, estab- areas during the rainy season), identifying lish a schedule of tasks and designate res- stakeholders and their perspectives, as well ponsibilities for implementation. as technical issues of bourgou restoration. A conflict resolution workshop on man- They developed an action plan for imple- aging change in October 2002 built on the menting bourgou restoration. previous training in managing the grazing SANREM brought in an HM trainer/ trials and was designed to prepare NRMAC consultant from Chad, who had in-depth members and the environmental supervisors Holistic Management in Madiama 123

for the inevitable conflicts that would arise Training in Holistic Management began in making such a management change. The at the end of the first year of project activi- primary objectives of this training were to ties in Madiama. At an intellectual level strengthen local capacity to face and manage members of the NRMAC and other villagers conflict situations, develop techniques to participating in the training quickly mas- manage change, and develop and refine tered the principles of the Holistic Man- techniques for the mediation of conflict. agement Model. They had no difficulty in This provided an important opportunity for seeing erosion, drought and falling water NRMAC trainers to take the lead in the train- tables as manifestations of a damaged water ing exercise. cycle, weed infestations and pest outbreaks Holistic Management for open-range in terms of succession, and fertility loss as a pasturelands was reinforced in May 2003 failure of the mineral cycle. They frequently with a review of principles and discussion of returned comments on the order of, ‘Thank the results of the previous year, which were you for explaining what is going on. We see influenced by extremely poor rainfall. This things in our own way, but up to now out- also led to the final Conflict Resolution and side advisors just told us what to do but Consensus Building Workshop (October never why they thought it was right.’ 2003) in which NRMAC trainers led the Faith in theory has a long gestation training activity in the twinned village of period, however. The final step of turning Tatia Nouna where the build-up of herds faith into a community consensus strong during the descent into the bourgoutière rou- enough to risk complex changes in organi- tinely leads to confrontations and conflicts. zation and practice did not develop during The NRMAC brought together opposed the life of the project. Nevertheless, an interests within the village to confront and important change in perception did occur. discuss the natural resource management Acceptance of the Model did not come (NRM) issues related to these conflicts. as easily to many of the researchers as it did to the average villager, given that the former tended to have invested enormous effort in Holistic Management Contributions to mastering a particular speciality. Quite a Scientific Understanding number found the Model’s insistence on dis- tinguishing between cause and symptom Beginning with the initial Participatory confusing, because they had been trained to Landscape/Lifescape Appraisal (PLLA), seek remedies. Observing that the soil fertil- groups from Madiama Commune produced ity and tilth that previous generations lists of ‘constraints’ that isolated problems enjoyed did not depend on periodic doses of from their holistic context making them irre- rock phosphate does not automatically solvable. The issues included: wells are cause a fertilizer specialist to reflect on why drying up; the parasitic Striga plant destroys it might now, or to seek alternatives if chem- the grain; the cattle are starving; the soil ical fertilizer proves economically unprof- does not produce like it used to; the all-sus- itable or ecologically unsustainable. A real taining bourgou plant is vanishing from the dialogue from one of the training sessions lakes; firewood has disappeared; and the follows: transhumant herds come through before the harvest and ravage the crops. This constel- Question: ‘Seriously, why are Madiama’s lation of sorrows is typical of Sahelian com- soils so much worse today’? munities and well-documented across Response: ‘They lack phosphorous and semi-arid lands worldwide. Since the typi- nitrogen!’ cal response is a demand for technical reme- Question: ‘Why?’ dies for each problem, the introduction of a Response: ‘Because they are degraded!’ different line of reasoning into discussions Question: ‘Why are they degraded?’ in Madiama Commune is an important Response: ‘Because they lack necessary legacy of the project. nutrients!’ 124 S. Bingham

Nevertheless, at least one IER researcher and they would have been held in topsoil by did become intrigued by the suggestion that organic material. Now, evidently one or direct importation of soil nutrients and more of these processes had broken down. organic matter by transient livestock could The cycle no longer functioned as it play a large role in restoring and maintaining should, and nutrients had moved some- soil fertility, and preliminary experiments where else. Plant communities had shifted he conducted with village collaborators in towards non-palatable species and annuals. stabling herds for short periods on fields Perennial grasses had almost entirely disap- after the harvest produced very promising peared except for vestiges of the least desir- results. able. Larger game species had all but For input, the Holistic Management vanished, and domestic stock had shifted Model asks: towards small ruminants. A visiting biolo- gist noted that biodiversity is now actually • What happens to the water that falls on greater in cultivated areas than in grazing Madiama and the surrounding area (the areas, although in the former Striga and Water Cycle)? Historical data on well weed infestations have increased markedly. levels and rainfall may be interesting Energy too is an issue. If someone con- but are mostly useful to substantiate the ceived of agriculture as the harvesting of unanimous opinion that enormous solar energy, Madiama abounded in signs of amounts of water that once entered the decline. Crop production per hectare had water table through percolation into fri- fallen. Herd fertility and milk production able, protected soil, now leave the had fallen. Ponds yielded fewer fish and less system through runoff and evaporation. bourgou. Trees like the Acacia albida that • Where do soil nutrients come from in produced greenery in the dry season and Madiama, and where do they go fixed nitrogen in the soil were all in steep (Mineral Cycle)? The process of loss is decline. of more concern than the fact of it. What tools had been used? How had the • What changes in plant and animal com- use of tools changed? Going tool by tool, munities have occurred (Community what impact on the four processes could be Dynamics)? Concern for biodiversity expected? loss is a step in the direction of holism as opposed to the customary fixation on • Technology: Steel ploughs pulled by controlling pest and weed outbreaks. bullocks have largely replaced hand • What happens to the available solar cultivation, enabling exploitation of energy (Energy Flow)? much larger fields, and this has practi- Every one of these questions could form the cally eliminated the old practice of a basis of several research projects designed to 6–10 year fallow. Older inhabitants quantify the state of basic ecosystem note that today even land that is taken processes, but neither NRMAC members nor out of crop production does not recover, even outside observers had much difficulty but in fact degrades further. reaching consensus on general answers. In • Rest: Cultivated ground gets little rest Madiama, the almost complete loss of now; nor does grazing land, as increas- perennial grasses and the expanding extent ing numbers of small ruminants are of bare ground where nothing grows at all kept year round. Overrest is not a seri- mean a high percentage of rainfall never ous issue. enters the productive cycle. Consequently, • Fire: Burning has not been used as a tool chronic ‘virtual drought’ coexists with per- much in recent years, due to suppres- sistent flooding. sion by the Forestry Service, although Historically, nutrients in surface soils annual grassland is fire prone. had to have been brought from lower levels • Grazing: Although transhumant herders by deep rooted perennial plants or imported with cattle continue to pass through the through the manure of transient livestock, commune, more small ruminants and Holistic Management in Madiama 125

draft animals graze continuously year overwhelmed the capacity of the NRMAC, round than in the past, thereby assuring IER and American partners to effect signifi- severe overgrazing of plants. cant change in 5 years, the challenge to the • Animal impact: Passing herds continue future has been fairly communicated. to create significant trampling events The holistic analysis can also take some and deposit large amounts of manure, credit for planting the notion among vil- but much of the latter does not enhance lagers, researchers and project leaders that crop land. Wide areas around ponds lasting solutions to Madiama’s woes had to and other water points are severely involve people and land outside the com- degraded despite enormous deposits of mune. Most of the cattle and much of manure and moisture in the root zone of the smaller livestock actually spent more perennial plants, if any remained. time outside the commune than within it, • Living organisms: Little has been and much of the stock that passed through done to promote habitat for beneficial belonged to people living elsewhere. Exp- creatures. loitation of other resources such as fire- wood, fish and thatch also involved Such an analysis, crude as it may outsiders. Certainly much of the conflict has appear, allowed both villagers and research- had roots in this fact. ers to focus on probable causes of degrada- The experience in Chad highlighted for tion that they might be able to address, thus the Madiama delegation the importance of immediately raising the question of whether including external stakeholders in planning to attempt that or simply mitigate the symp- and negotiation. Unfortunately the logi- toms. In the case of Madiama, this discus- stical and administrative difficulties of this sion resulted in a conclusion that the soil challenge limited what could be done fertility question and the grazing land biodi- within the project period, but a start was versity loss are linked and that livestock and made. crop issues cannot be solved separately. This conclusion is reinforced by the fact that, in an erratic rainfall environment, the Impact of Holistic Management food security envisioned in the holistic goal will depend on maintaining multiple forms The Holistic Management approach proba- of production. bly did not precipitate much change in the A shift from looking for discrete solu- way IER and expatriate researchers worked tions to individual problems to correcting in Madiama, given the fact that most of the basic flaws in a complex system compels an research proposals were generated in admission that real progress might depend response to the PLLA and the list of con- on changes in social organization, personal straints provided by the NRMAC before they relationships, and improved management or participating researchers had been rather than a new technology. Unfortunately exposed to its principles. The NRMAC did hugely complicated and contentious land generate a proposal to attempt restoration of tenure issues and jurisdictional ambiguities a previously productive bourgou zone arise. Longstanding family and inter-ethnic around a wetland basin. The project failed conventions come into question. for multiple reasons, though the genesis of Even though no major changes in man- the idea is a fair illustration of the potential agement were achieved during the course of of a holistic discussion. the programme, the general consensus that The idea arose when a discussion of the the problems in crop and livestock produc- Holistic Management Model’s ‘marginal tion cannot be solved separately probably reaction’ guideline focused on rehabilita- stands as the Holistic Management Model’s tion of a bourgou site as the action that the most significant contribution to the fut- NRMAC felt promised the greatest return for ure of Madiama. If the social, political and the effort. Other tests in the Model raised research implications of this observation questions about the difficulty of controlling 126 S. Bingham

grazing access to a rehabilitation site and source for the rest of the area, and the com- social and cultural resistance to fences as a munity would see the benefit of managing means to control access. This, in turn, led the restored common land for the benefit of one of the women members of the NRMAC all. As this would be a revolutionary initia- to propose that resistance to fencing by the tive on a number of counts, it is hard to say several customary users of the bourgou how realistic it might be, but it does illus- might be overcome if the area were con- trate the variety of ideas that issue from the signed to a committee of women who would holistic process. use the forage to produce milk for malnour- ished children and elderly. This was tempered by the memory of a discussion that took place following the first Perceptions among constituencies within the training session when members of the commune NRMAC reported to a broader selection from the community. After they made a pre- The very notion of grouping villages into sentation of the four ecosystem processes, communes came to Mali only shortly before when asked to identify the causes for the SANREM did. It arrived in Madiama as part decline of the bourgou, an elderly man of a nationwide administrative reorganiza- opined that a lot of it succumbed to drown- tion aimed at decentralizing power, but ing, a counter-intuitive diagnosis for a plant many questions of jurisdiction and relation- that grows in water. But the man pointed ships with residual layers of authority going out that the water cycle had changed. He back to pre-colonial times remain open to argued that now, since the land around had this day. Madiama’s Natural Resource lost its cover, the runoff came in fierce sheets Management Advisory Committee was the that filled the lakes faster than even bourgou first and only one in the country. It had no could grow, and silt from the erosion statutory or traditional authority, but it did blocked sunlight. The bourgou would not open communications among ten villages come back, he predicted, as long as the adja- that had not had any institutional relation- cent land remained bare. He also remarked ship before. That in itself was a major that bourgou was not the first good plant to achievement, not directly the result of the vanish from Madiama’s ponds. Holistic Management, but certainly in the Unfortunately, neither American nor spirit of the approach. IER scientists were able to incorporate any Much of the conflict within the com- of these ideas into a proposal for rehabilitat- mune has an inter-ethnic character, but ing bourgou within a timeframe that fit most of it stems from competition over SANREM’s conditions. The weak link of resources. Different groups specialize in commitment and understanding had yet to farming, herding and fishing. The conflict support concrete community action. Lacking resolution component of the programme volunteer labour to plant bourgou, abroad- generated a great deal of healthy discussion cast seeding was done without a grazing of how human nature reacts under condi- management plan, but it was washed out tions of scarcity, power imbalance and altogether as the old man and the Model change. The Holistic Management Model, warned it might be. particularly its emphasis on the contribu- Nevertheless, the experience may yet tion of animals to soil fertility and the bear fruit. In the summer of 2002, one of the importance of watershed health to the pro- NRMAC members suggested that the com- ductivity of wetlands and rivers opened at mune ought to grant temporary rights to least the possibility of win–win solutions. someone to fence an area of potential bour- Bambara farmers and Peul stock raisers on gou production, plant it as a crop and sell the NRMAC at least began to discuss the it. The profit motive would assure that it possibility of mutual benefit in a way that was planted properly and protected. The could not have happened at the beginning re-established plants would provide a seed of the project. Holistic Management in Madiama 127

Lasting influence of Holistic Management? References

Since 1993, the Center for Holistic Manage- Behnke, R. and Scoones, I. (eds) (1991) Mana- ment has operated a demonstration site in gement and Sustainability of African Zimbabwe near Victoria Falls and trained Communal Rangelands (ODI/P59). Com- and supported local field workers in neigh- monwealth Secretariat Conference, Overseas Development Institute, London, UK. bouring villages. After 10 years several do Behnke, R., Scoones, I. and Kerven, C. (eds) (1993) have grazing programmes, but that only hap- Range Ecology at Disequilibrium: New pened after years of patient discussion and Models of Natural Variability and Pastoral organization-building. Success in other pro- Adaptation in African Savannas; Overseas jects, including water development and Development Institute and International women’s microbanks, and a training pro- Institute for Environment and Development, gramme for game guides came first, as com- London, UK. munity capacity grew. Kerven, C. and Alimaev, I.I. (1998) Mobility and It is unreasonable to lament the fact that the Market: Economic and Environmental Madiama’s ad hoc aggregation of villages Impacts of Privatisation on Pastoralists in Kazakstan, Paper presented at the confer- has not reorganized communal grazing, ence on Strategic Considerations on the integrated its crop and livestock production, Development of Central Asia, CODOCA and negotiated grazing and parking plans with Chinese Academy of Science, Urumchi, transhumant stock raisers, or sorted out the China. property rights around water points and Savory, A. (with Butterfield, J.) (1999) Holistic ponds. However, the idea that a different Management: a New Framework for approach to management could restore a Decision-Making, 2nd edn. Island Press, devastated landscape where many technical Covalo, California. solutions had failed was new and will be Scoones, I. and Behnke, R. (1993) Rethinking remembered. range ecology: implications for rangeland management in Africa. In Behnke, R., Scoones, I. and Kerven, C. (eds) Range Note Ecology at Disequilibrium. New Models of Natural Variability and Pastoral Adaptation 1 Bourgou, Echinochloa stagnina, is a highly in African Savannas. Overseas Development productive plant that grows underwater with Institute, London. the seasonal rise of the Niger River and the associated system of lakes and marshes. As the water falls the bourgou becomes the major source of fodder for a wide region.

10 Farmers’ Knowledge and Perceptions of Soil Fertility

Todd A. Crane1 and Boureima Traoré2 1Department of Anthropology, University of Georgia, Athens, Georgia, USA; and 2Centre Régional de la Recherche Agronomique, Institut d’Economie Rurale, BP 205, Mopti, Mali

According to participants in the Parti- nicated to farmers, and farmers can improve cipatory Landscape/Lifescape Appraisal production practices. (PLLA) (see Chapter 6), soil fertility is a Based on generations of practice and major constraint for villagers of Madiama. observation, farmers have developed their Scientists (Powell et al., 1996; Brouwer and own knowledge system and theories of soil Powell, 1998; Scoones and Toulmin, 1998; fertility maintenance that might or might Gobin et al., 2000) likewise cite soil fertility not match up with scientific knowledge and decline as the major limit to agricultural theory. For example, while the scientific production in the Sahel. However, despite approach to soil classification is typically this similarity of problem identification, it is based on the geological origin and chemical surprising how often farmers or scientists composition of soil, farmer classification question the usefulness of each other’s soil tends to be based on more utilitarian factors fertility science and practice. This problem such as texture, water retention and crop- raises several questions: What are the differ- ping uses. Regardless of scientific valida- ent assumptions and beliefs that they bring tion, it is these folk models that guide to the problem? Are they really talking about management behaviours at the local level. the same thing? Do they really have the same Therefore, efforts to provide scientific perception of the problem? research in support of local decision- Unsurprisingly, scientists and farmers making need to be integrated with folk often approach the problem of soil fertility knowledge systems for optimal efficacy. from vastly different theoretical and practi- This chapter examines the subject of cal perspectives. Due to their unique linking local knowledge with scientific approaches, both scientists and farmers research on soil fertility, in both theoretical have valuable insights to contribute to each and applied terms. The end goal of other. Consequently, explicit efforts to inte- this effort is to collaboratively develop grate scientific and folk knowledge of soils improved soil-management techniques via through participatory research methods epistemological integration of local and sci- have several benefits. Among these benefits entific knowledge. The chapter begins with is that folk knowledge can guide aspects a brief examination of some of the historical of the scientists’ research programme, scien- problems and failings of introduced agricul- tific research findings can be better commu- tural technologies. In an effort to explicate © CAB International 2005. Conflict, Social Capital and Managing Natural Resources (ed. K.M. Moore) 129 130 T.A. Crane and B. Traoré

the shared and divergent perspectives, local users from the beginning, participatory knowledge systems and institutional scien- approaches to technological development tific research are compared according to go a long way towards avoiding these pit- classificatory systems and fertility manage- falls and contribute to the ultimate goal of ment regimes. The scientific bases for soil developing feasible, sustainable and cultur- classification and soil fertility management ally appropriate production systems. are based on a review of the literature. The folk model of soil classification and soil fer- tility management developed from field Participatory approach in soil management interviews with producers. This parallel research presentation highlights where scientific and folk knowledge overlap and where they In the past 15 or so years, participatory diverge regarding the subject of soils. The approaches to natural resource management theoretical aspect of this relationship is, (NRM) research have become more preva- however, only part of the picture. The final lent. These research approaches incorporate part of this chapter examines how cultiva- local knowledge and local practice into the tors in the Commune of Madiama have development of improved technologies for received and perceived scientific research. peasant producers. The broader field of soil The chapter then concludes with a dis- science literature (Arrouays, 1987; Tabor, cussion of ways that local knowledge and 1993; Gobin et al., 2000) provides examples practice has related with the agronomic of integrating local knowledge and scientific experiments on the ground in Madiama. research of soils, and there are specific examples that focus on various topics of relevance to Mali, including seed varietal Problem Statement development (Adesina, 1992), striga control (Debrah et al., 1998) and soil fertility man- Historical weaknesses in top-down agement (Defoer et al., 1998). Many similar agricultural research cases can be found throughout the world. As described in Chapter 1, the SANREM Historically, development efforts in the agri- approach stresses participatory research, cultural sector have often relied on the and it was within this framework that soil introduction of technologies and techniques fertility trials were undertaken by the scientific research community and Following the PLLA and subsequent government bodies. However, while institu- follow-up by the NRM Advisory Committee tional science is a powerful way to under- (NRMAC) (see Chapter 6), several meetings stand soil nutrient cycles and develop focused around the theme of participa- technologies to influence these cycles tory technological development were held favourably, its weakness has often been that between researchers, NRMAC members and the techniques and technologies developed other peasant collaborators in an attempt to do not translate well into the realities of broadly identify technologies adapted to the agriculture as practised by the farmers them- local context. At the end of these meetings, selves. There have been reasons for this the peasants had chosen technologies to test weak translation, including lack of consid- as well as collaborators whose fields would eration for the socioeconomic realities of be used to test them. In an effort to under- peasant life (such as financial or labour lim- stand how they would evaluate the tests in itations), lack of consideration for the long- their own terms, local collaborators men- term environmental or social impacts of tioned four criteria: the vigour of the introduced technologies, and differences in seedlings, the colour of the leaves, the con- conceptualizations of how and why agricul- dition of the grain heads, and most impor- ture works the way it does (Rhoades, 1984, tantly the quantity of grain harvested. 1989; Chambers et al., 1989; Scoones and Visits to the fields were organized each Thompson, 1994). By involving the end- year under the direction of the NRMAC Farmers’ Knowledge and Perceptions of Soil Fertility 131

accompanied by senior researchers from ble to scientists (see Talawar and Rhoades, CRRA/Mopti in order to assess and discuss 1998 for summary of ethnopedology; and the season and the performance of the Furbee, 1989; Pawluk et al., 1992; Dialla, experiments. During these visits, the peas- 1993; Tabor, 1993; Zimmerer, 1994; Sillitoe, ant collaborator would discuss the perfor- 1996 for case studies). Both local and scien- mance of the field as well as the difficulties tific systems frequently recognize that soil is and the advantages related to technology. a foundational and unifying element in any Following the harvest, a researcher’s pre- landscape, domesticated or otherwise (see sentation of the quantitative experimental Chapter 3). This, however, is not to say that findings at a meeting with the NRMAC, the they always approach or even understand peasant collaborators, the field technicians soils in the same way. Because ethnopedo- and local dignitaries augmented the infor- logical research provides a clear and strong mal field assessments. This approach means by which the ideologies of research allowed for the qualitative and quantitative scientists and rural producers can be linked, evaluation of the experiments by the local a study of local knowledge of soils was con- collaborators and researchers together. ducted in association with the soil fertility trials.

Linking scientific and farmer knowledge Methods One of the challenges in participatory research is linking the sometimes disparate Characterization of scientific soil classifica- folk and scientific knowledge systems. The tion schemes are presented based on a liter- contrasting of ‘local’ and ‘scientific’ is not to ature review. The local knowledge aspects suggest that local knowledge is not, at some of this chapter are based on semi-structured level, scientific. In as much as local knowl- interviews with peasants across the Com- edge is based on systematic observation and mune of Madiama. Interviews were con- experimentation through time, it can be said ducted in 2001 and 2003–2004 and centred to be scientific. However, for the purposes of on knowledge of soils and soil management, this chapter, ‘scientific’ is used to indicate specifically principles of soil classification, knowledge that is institutionally supported, management practices associated with vari- informed primarily by formal education, ous soil types, and indicators of quality. The and inclined to universalistic scientific findings here do not represent a random laws. The scientific approach to improving sample, but this was an acknowledged con- soil fertility is based on the body of knowl- cession to the participatory approach to edge developed by institutional scientific research. Since NRMAC members served as research through the years, broadly referred contacts and gatekeepers within each vil- to as ‘the literature’. Peasants have, how- lage, they are disproportionately repre- ever, rarely read ‘the literature’. Conse- sented in the sample and they usually quently, their body of knowledge of soil decided who else in their villages would fertility is based on insights gained through make a good informant. This sampling practice, observation and informal informa- method was chosen because it is more likely tion flows. This includes knowledge devel- to actually represent channels of informa- oped and passed across generations as tion flow of SANREM research, which is well as knowledge based on personal typically presented to NRMAC members experience. first. They then take information to their vil- Over the last 15–20 years, the study of lages and share it with members of village- local knowledge of soils, ethnopedology, level NRM User Committees, after which the has become increasingly prevalent in an information circulates informally among effort to make scientific research on soils other villagers. This model of information more accessible to farmers and to make flow was the basis for presentation of find- farmers’ soil management practices accessi- ings from the soil fertility field trials. 132 T.A. Crane and B. Traoré

Scientific Model of Soils should be noted that the SANREM soil survey combines elements of the French and In order to contrast scientific and local American classification systems, which are knowledge, a brief review of the literature subsequently informed by local knowledge for scientific perspectives on principles of and practice (see Chapter 3). soil classification and prescribed manage- ment practices follows. Despite the uni- versalistic nature of science, there are Principles of classification numerous distinct systems of soil classifica- tion used by soil scientists around the The first step of any classificatory system is world, each based on different principles. In defining the domain of things to be classi- the late 19th century, the USA and Russia fied. Soil is defined as a natural formation of independently began to develop soil classi- variable thickness resulting from the com- fication systems in efforts to assess soil bined action of atmospheric agents, plants resources and rationalize their exploitation. and animals on stone, which constitutes the By the mid 20th century, many industrial- surface of the continents (Bockman, 1990). ized and agrarian nations had undertaken A strength of this definition is that it under- similar endeavours, the result being a large lines one of the essential characters of soil: variety of classification systems, some very it is a complex medium at the crossroads similar and some vastly different. of mineral matter and living organisms However, most scientific soil classifica- (Vannier, 1987; van Breemen, 1993). Indeed, tion systems have been constructed to deal soil is comprised of: specifically with types of soils found within the relevant national boundaries. The litera- • Mineral components (generally the ture provides an American system, a French majority by weight); characterized system, a Russian system, a Brazilian according to the mineral size: boulders, system, a Chinese system, a German system, stones, gravels, sands, silts, clays. a Japanese system, a Canadian system, an •Organic components (normally a minor Australian system, a New Zealand system, a component). One can find remains of South African system, and so forth (Finkl, this organic matter in humus. It is esti- 1982; Guillet, 1989; FAO, 1998; Spaargaren, mated by the quantity of carbon in the 2000). Inasmuch as they were all con- soil. structed to address a need that exists within •Living organisms, microorganisms and the geographic boundaries of a nation, each other animals. system could be said to be a local form of •Organo-mineral complexes; organic and knowledge in its own right. In the latter half mineral matter which are chemically of the 20th century, serious efforts were bonded to each other. undertaken to create a universally applica- ble system of soil classification under the American classification auspices of the FAO and the United Nations, but this is by no means complete or univer- There have been several different attempts sally accepted or used even to this day over the last 150 years at classifying soils (FAO, 1998). within the USA. The most recent American Some systems are based on perceptible soil classification system, originally pub- properties of the soils themselves, whereas lished in 1975, is hierarchically structured, others are based on geological origins that with processes of soil genesis forming the can only be inferred from extensive geologi- primary point of distinction. Below this, cal survey and laboratory analyses, and still within each grouping, soils are classified by others are based on soil–climate interac- observable ongoing processes that control tions. An extremely brief review of some of soil formation. Below this level are found the major systems is presented below to distinctions based on properties of the soils illustrate the diversity within the field. It themselves (Ahrens and Arnold, 2000). The Farmers’ Knowledge and Perceptions of Soil Fertility 133

Brazilian system originated from an early similar to the American system, except that American system, but was significantly it is not a hierarchically structured method modified in order to accommodate the types of analysis. Instead, it simply lists 28 major of soils found in Brazil that did not appear soil groups based on parent materials and in the American taxonomy (Spaargaren, geographical zones, followed by 153 dis- 2000). crete soil types within those groups. The contemporary system is a revision of an ear- lier effort that listed 24 major soil groups French or ecological classification and 106 types of soils. It should be noted An early French system, the Commission of that one of the goals of the FAO-UNESCO Pedology and Mapping of Soils was a hier- effort was to ‘promote the establishment of a archically based system of classification. generally accepted soil classification and The first level was distinguished by nomenclature’ (Spaargaren, 2000, p. E-138), processes of soil formation, and then subse- a first step towards a completely universal quently divided by ecological conditions system. and context, followed by soil properties. This system, however, was replaced in the early 1990s by the Pedological Reference World Reference Base for Soil Resources Base (PRB), a non-hierarchical classification that is instead intended as a general ‘refer- Despite the fact that it was co-developed by ence system’. The PRB typology is three FAO in conjunction with the International dimensional in character and recognizes 90 Society for Soil Science (ISSS), the World different major groups that are defined by Reference Base (WRB) was explicitly soil properties, soil origins and position in intended to provide a universally accepted the mantle. The French system is similar to system of soil classification that could be what is known as the Ecological classifica- integrated with national systems. This tion system because it considers a soil’s system focused on easily observable soil ecological context and processes in its cate- properties as the primary points of classifi- gorization (Spaargaren, 2000). cation, including the horizon in which a soil is found. The geological processes that led to the formation of the soil are secondary in Russian classification significance. The WRB identified 20 major soil groups before merging with the Legend The Russian classification system, pub- of the Soil Map of the World effort, wherein lished in 1977, is characterized by heavy it settled on 30 major soil groups reliance on soil-forming processes, such as (Spaargaren, 2000). weathering and biochemical processes, as In summary, American and FAO classi- the primary point of distinction. It recog- fications particularly take into account the nizes 71 different types of soils. At lower degree of evolution of the soil profiles. levels, it distinguishes soils based on their Western European classifications take into respective bedrocks, geographical posi- account a soil’s milieu, processes and the tion and degree of soil development. In the characteristics. Dominant modern classifi- Russian classification system, soils are cations (Russian, American and FAO, characterized according to their processes French) retain the following criteria that more than they are by their properties depend on the degree of development of the (Spaargaren, 2000). soil profile in relation to the degree of evo- lution: FAO-UNESCO Legend of the Soil Map of the • Weathering: Weathering is related to the World evolution of the mineral fraction of the The contemporary FAO-UNESCO classifica- soil, particularly with the formation of tion, finalized in 1988, is in some ways clays. 134 T.A. Crane and B. Traoré

• Movement of material: This criterion is the fertility of a soil represents, in a given cli- used to speak about eroded or alluvial mate, its aptitude to ensure crop growth and soils. harvests in a regular and repeated fashion • Pedoclimate: The pedoclimate makes it (Vilain, 1989). The problems of soil fertility possible to distinguish hydromorphic maintenance, while universal among agri- soils. culturalists, varies with climate, type of soil, • Time: There are soils whose formation cultivation practices and techniques and can be rapid and others whose forma- environmental change (Sebillotte, 1989). tion is slow, which means soils can be From a scientific point of view, in order spoken of as having short and long to give an opinion on a soil’s fertility, it is cycles. necessary to evaluate not only its capacity to • Organic matter: This plays a significant provide crops with the conditions and fac- role in the formation of the organo- tors they need, including light, water, min- metallic complexes upon which many eral elements, but also the costs and risks properties of the soil depend. associated with cultivation (Sebillotte, 1989). Indeed, the same soil would not The development of any classification allow the same levels of production if it system makes it possible for soil researchers were in a different climate or water regime. to have shared reference points. However, Similarly, a soil can be fertile for one species as anyone can see, even university trained and unsuitable for another (Bockman, research scientists approach the problem of 1990). Generally speaking, in order to soil classification from a number of angles manage soil fertility, it is necessary to use using appreciably different logics, depend- techniques that enable soil to produce abun- ing on their training. Is a soil best char- dantly in the short term, but also maintain acterized by its geological origins and or improve its long-term fertility. developmental history, or is it best charac- To fertilize soil is to improve its physi- terized by its component parts and physical cal, chemical and biological properties so characteristics? To what extent is the eco- that crops flourish. Four major processes logical context important in characterizing a can be identified as parts of soil fertility soil unit? All are equally valid methods in maintenance in an agriculturally productive terms of their ability to describe a soil accu- system. First, the soil must be managed so rately. However, the primary question in the that the plants gain the nutritive elements applied research context is their utility. they need. Second, the soil needs to be Which way of approaching soils is most managed in such a way that its structure valuable to the end user, whether researcher supports the rooting and the circulation of or farmer? This is a legitimate concern of air and water. Third, it must be managed so the local beneficiaries of research that that its physicochemical properties support makes the incorporation of local knowledge the biological activity necessary for good into soil research even more important. decomposition of organic matter. Fourth, Acknowledging that science has many ways the soil must be managed so that its bio- of approaching a single topic makes it easier logical processes limit the presence of to see local classificatory schemes as just pathogenic organisms and encourage advan- another way of approaching soils, a useful tageous ones. way that can be integrated with scientific Depending on the context, management schemes. of soil fertility can entail any number of techniques, but typical among them are the addition of organic matter (in the form of Description of soil fertility management manure, compost, etc.), the addition of min- erals necessary for plant health (nitrogen, One of the factors that any soil classificatory phosphorus, potassium, etc.), and the man- system will address is the question of soil agement of abiotic factors such as water, fertility. From an agricultural point of view, either by draining or irrigating. In this last Farmers’ Knowledge and Perceptions of Soil Fertility 135

case, the fertility of the soil itself (if mea- and thus soil fertility. It should be noted that sured by organic matter or mineral content) red and white clay and white sand are not is not improved, but the principal con- regarded as agricultural soils, but are sub- straint, water excess or shortage is changed soils identified by only a few informants so as to improve agricultural productivity. who had experience digging wells. The fine and very fine clay types are exclusively located in the river floodplain. These are Folk Model of Soils characterized by their stickiness and the degree to which a person can sink into them Principles of classification when they are inundated. Although not all informants presented the typology below in Where scientific approaches, such as those all of its detail, no one presented any con- used in SANREM, often employ geological tradiction to it. origins and processes of transformation to Even though peasants discuss soil in classify soil types, the folk system of soil terms of the categories presented above, classification found in Madiama is based they also quickly acknowledge that these primarily on the salient component materi- categories are not defined in absolute terms. als of the top horizon, which in this case are Instead, soil is understood as a constantly sand, clay and rock. The primary difference shifting gradient of these three component between these three broad categories that materials. For example, most sandy soil in three categories generally correlate with the commune has at least some measure of basic field management strategies is the clay in it and most clayey soil had some degree to which they contain and retain measure of sand. Soil categorized as rocky organic matter and water. Millet and soil always has either a sand or clay base, groundnuts do better in sandier soils while but sand is more common. Further ques- sorghum and okra favour clayey soils. tioning found that black sand is actually While sand, clay and rock provide the acknowledged to be a blend of sand and fundamental reference points when refer- clay, thereby revealing that these categories ring to soils, subcategories are divided out are not mutually exclusive or entirely dis- based on colour. As Fig. 10.1 shows, black is tinct. The widespread acknowledgement the most salient colour for agricultural soils of mixed types illustrates that although due to its association with organic fertilizers there are specific named categories for soil

Dugukóló (soil)

Bógó Céncén Bélé (clay) (sand) (rock)

Blen Fin Guema Blen Fin Guema (red) (black) (white) (red) (black) (white)

Manama Kirikiri (fine clay) (very fine clay)

Fig. 10.1. Folk soil typology resulting from semi-structured interviews. 136 T.A. Crane and B. Traoré

types, farmers understand soil as a three- to which it sticks together exhibits the rela- component continuum that is constantly tive quantity of water it contains, which is, varying through space rather than discrete of course, linked to quantity of organic plots of one type or another. matter, particularly in sandier soils. In as It is important to point out that focusing much as clayey soils tend to retain more exclusively on soils themselves misses a key water, they are also thought to contain more aspect of local knowledge. Local knowl- nógó and so are considered stronger by edge, being situated in the context of an nature than sandy soils. It is of note that active production system, does not entirely even though black clay is thought to be the distinguish between soil and landscape fea- strongest soil, it is not necessarily the most tures. Distinct from, though related to soil preferred. Black sand is widely agreed to be types, hydrology and elevation changes preferable because it is more friable and within fields also play into management therefore easier to work when wet. strategies. Clay is often associated with low The scientific analysis of soil fertility is spots and sand with high spots, but several typically expressed in terms of processes informants speak of how elevation can play that affect chemical composition and physi- a role in their field management regardless cal structure. The farmer’s point of view of soil type. Although conventional wisdom generally agrees with this, though its expres- is that millet is grown in sandy soils and sion differs. In Madiama, soil fertility is a sorghum is grown in clayey soils, these function of several processes, some socioe- ‘rules’ are frequently broken according to conomic and some ecological. Briefly stated, the presence of elevation changes that affect the basic process in maintaining soil fertil- microhydrology. For example, a low spot in ity is the application of organic fertilizer, a sandy field might be planted in sorghum nógó. Although fallowing is known to reju- due to the higher accumulation of water venate soil strength, it is very rarely prac- there during the course of the growing tised due to perceived inability to leave season. This is all to say that soil typology fields out of production, because there is a alone is insufficient to understand what shortage of land. As the application of nógó guides management behaviour. Landscape is the primary means of maintaining soil fer- level features also play a role, as does cogni- tility, farmers’ conceptualizations of nógó tion of how the component parts of the man- are essential to understanding their beliefs agement system relate to one another. regarding the management of soil fertility. Farmers describe nógó as being food for plants, food that is placed on the surface of Description of soil fertility management the fields and then ploughed in at the begin- ning of the agricultural season, where it When asked what shows that soils have begins to break down until it is consumed by ‘strength’ (fanga), peasants uniformly say the plants. that crops grow well there, because they are According to informants, all sorts of large, green, and produce lots of grains. This organic matter, including leaves, grasses, is to say that agricultural productivity is the manure and so forth, fall into the category primary indicator of soil fertility. When of nógó. But as with human foods, some asked specifically if looking at the soil itself plant foods give more strength than others. revealed if it was strong, the primary indi- Animal manure is the preferred fertilizer cator offered was consistently its colour. A and its acquisition, a socioeconomic pro- strong soil is a dark soil, because darkness cess, is of central importance to soil fertil- indicates that there is nógó, organic matter, ity maintenance. However, not all animal in it. Water retention is a secondary indica- manures are the same. Informants consis- tor and is recognized to be linked to quantity tently agree that small ruminant manure of nógó in a soil. Water retention is typically is ‘stronger’ than cow manure, because it tested by picking up a handful of inundated endures in the soil for up to 7 years whereas soil and squeezing it in the fist. The degree cow manure only lasts for 2 or 3 at best. Farmers’ Knowledge and Perceptions of Soil Fertility 137

There are three general means of acquir- that there is a difference between feeding ing manure for the fields: corralling animals crops and improving soils. in the actual agricultural fields at night, cor- A third quality which farmers use to ralling animals at night in household pens distinguish chemical fertilizer from nógó and transferring the deposited manure to the farafin is that chemical fertilizer requires fields, and gathering manure from open pas- sufficient rain to be effective and can actu- ture areas to haul to fields. Manure acquisi- ally kill crops if there is not enough rain. In tion strategies vary across households in contrast, nógó farafin works regardless of correlation with wealth, which often mani- the amount of rain, and actually increases fests itself in terms of animal ownership. water retention in the soil. The fourth dis- For all farmers interviewed, there is a tinction made by a few informants is that clear and significant difference between chemical fertilizer is made in factories and nógó and chemical fertilizer. The first indi- sold in markets whereas nógó farafin is pro- cation of this is that two primary categories duced by households and is not available for are commonly used in informal conversa- purchase. tion about fertilizer: nógó (farafin) (fertilizer Overall, the folk model of soils in [African]) and tubab nógó (European fertil- Madiama is not unlike a scientific model in izer). When asked to list all forms of nógó, any significant way. Motivation is provided without using the modifiers farafin or tubab, by the need to manage agricultural soils and farmers would frequently not mention the principle of categorization is based on chemical fertilizer. When subsequently observable soil properties such as texture, asked if chemical fertilizer is a form of nógó, colour, component materials, and water they would reluctantly tack it on to the list retention. The application of manure as an afterthought. This might be partially increases the quantity of organic matter in attributed to the fact that it is too expensive the soil and fosters beneficial biological for most farmers to buy and therefore not in processes in the soil. Nuanced knowledge of their regular cognitive repertoire of fertilizer abiotic aspects of soils is illustrated by options. However, the fact that chemical fer- planting strategies that are based in land- tilizer is widely used on watermelons, scape features and microhydrology, as well which many farmers grow, would contradict as by soil preferences based on physical this explanation. traits of soil such as friability and water When asked how chemical fertilizer retention. Finally, the distinction between and nógó farafin are similar, farmers all organic fertilizer and chemical fertilizer responded that they both help crops grow, indicates that there is a difference between which is the quality that appears to unify the biological processes of decomposition and overall category of nógó. When asked how the addition of chemical nutrients needed chemical fertilizer and nógó farafin are dif- for plant growth. While chemical fertilizers ferent, four major factors arise. The first and may make plants grow better, they do not universally expressed difference is that improve the quality of the soil itself in the chemical fertilizer, while powerful, is only same way organic fertilizers do. These find- good for one season. Its efficacy, its fanga, ings have contributed to the integration of does not endure in the soil. Conversely, local and scientific knowledge in SANREM nógó farafin is seen as effective for any- research. where from 2 to 8 years after application, depending on the type and intensity of application. This is connected to the second Linking Local and Scientific Knowledge observation that while chemical fertilizer helps plants grow, it does nothing to Presentation of findings to NRMAC and improve soil quality itself. This observation others is significant because although farmers all say healthy-looking crops are the best way This chapter has described how, in theoret- to tell the quality of soil, they still recognize ical terms, local knowledge of soils in 138 T.A. Crane and B. Traoré

Madiama can be linked with scientific two generations. Chemical fertilizer was not research on soils. However, it is more impor- used for grain crops. As explained by tant that they are well linked in the applied Madiamans, the primary reason it is not context. Consequently, after each harvest a used with grain crops is that chemical fertil- debriefing is organized for members of the izer is expensive and grains, being subsis- NRMAC, collaborating peasants, and other tence crops, do not generate income. persons in charge of the commune such as Consequently, it is difficult to justify spend- the mayor. The peasants describe to the ing that much money on chemical fertilizer. group the difficulties they have encountered Additionally, they explained that the tech- during the growing season, the state of the nique requires too much time considering crops during the vegetative cycle and the the size and number of grain fields. amounts harvested. The financial and Although microdosing with manure has labour requirements for each experimental been seen in both grain and calabash fields, technique are presented to the group along it is relatively rare. with the quantified totals of the harvest. Field observations indicate that the fre- After the third season of field tests, a final quency and intensity of intercropping is debriefing was presented to the NRMAC low. Even when intercropping and rotation members and participating peasants. As a of grains with legumes are practised, they part of this, NRMAC and other participants are not always done as a means of improv- were able to ask questions regarding the ing soils. One explanation offered for this by tests, keying in on important questions of two different farmers was that groundnuts labour inputs for the various techni- and cowpeas improve the soil, because they ques, scheduling problems and diminishing are low-growing and have lots of foliage returns on financial and labour investments. which is good for the soil. Although inter- cropping can be widely found in fields, the benefit of intercropping cowpeas with millet Discussion and sorghum was most often expressed in terms of labour and space efficiency rather There is a wide range of soil fertility main- than soil management. By mixing beans in tenance strategies already employed in with grain seeds, the planting and harvest- Madiama, and most correlate with tech- ing of beans requires little extra time and niques tested in SANREM field trials. In energy, as well as no additional space. fact, farmers in Madiama know and use all The use of cattle enclosures can be of the introduced techniques, though to found in Madiama, but it occurs only at low variable degrees, intensities and situations. intensity. Small ruminants are typically However, to say that the experimental tech- penned adjacent to the household and their niques were already known among local manure is collected and then spread on participants does not devalue the experi- fields. In the short season during which ments themselves. Rather than introducing transhumant herds are passing through the new techniques, the experiments served to commune, some herds pass the nights in scientifically measure and illustrate to the herd owners’ field (if he is local) or in a local community the variable utility of known host’s field, a practice that provided the techniques and their long-term impacts. The basis for the corralling experiments. Field restitution has already caused several par- proprietors already recognize the benefits of ticipants to revisit their soil fertility man- keeping animals in the fields, but there is agement regimes and reconsider the low systematicity of this practice, partly due application of certain techniques. to the short season and partly due to lack of Prior to SANREM, microdosing of sufficient grazing material in the fields. chemical fertilizer and manure was known In some instances even before the com- and practised. The cultivation of watermel- pletion of the soil fertility experiments, ons as a cash crop has been done with practices could be seen changing among microdoses of chemical fertilizer for at least NRMAC members and participants in the Farmers’ Knowledge and Perceptions of Soil Fertility 139

experiments. Preliminary interviews with ments, one peasant collaborator had already peasant collaborators who attended the sought out a passing Fulani herder to ask debriefing indicate particular interest in two him to let his herd pass the nights in his of the methods presented: microdosing and fields for a couple of weeks after the harvest corralling. Although they found the rotation was finished. No money was paid, though and association experiments useful, the food was delivered to the herder and his peasant collaborators did not express as family in exchange for this service. This much interest in the impact of these tech- mirrors traditional arrangements found else- nologies as they did in the microdosing where in areas of overlap between farming experiments that demonstrated even a rela- and pastoral systems of production (Painter, tively small amount of fertilizer, when 1994; Van den Brink, 1995). applied intensively, can make a marked dif- ference in crop productivity. One research participant had already begun utilizing this Conclusions technique in his other fields prior to the completion of the soil fertility experiments, Before the field tests were even finished, a because he saw the efficacy after the first certain interlinking of local knowledge and year and that was all the convincing he scientific practice was underway. Local needed. knowledge contributed to the soil typology In discussions with NRMAC members developed by SANREM through the inclu- and test collaborators, three caveats accom- sion of a locally salient soil property, water panied the enthusiasm for microdosing: (i) retention, which is not part of a typical sci- even microdosing requires what for the entific soil survey (see Chapter 3). Further- peasants is a significant amount of cash and more, local knowledge of soil management therefore is not as useful for the large seg- also shaped one of the field experiments. As ment of seriously cash-strapped house- previously mentioned, the results of the holds; (ii) chemical fertilizer is best used in ethnopedological research indicated that addition to traditional organic fertilizers, peasants widely agreed that small ruminant which aid in water retention, but not as a manure was significantly more powerful replacement; (iii) chemical fertilizer, due to than cow manure in terms of maintaining its great and fast strength, can actually kill soil fertility. This hypothesis was directly sprouting millet if the early rainfall is weak incorporated into the field tests the follow- and therefore carries some risk with it, even ing 2 years (see Chapters 11 and 12). Not though it is perceived as promising under only did scientific research validate the the proper conditions. local knowledge, but it was also able to elab- Corralling has also gained the interest of orate the efficacy of variable quantities of several peasant collaborators, who see two manure, showing the point of diminishing distinct benefits of this method. First of all, returns on the application of small ruminant it minimizes labour, because it requires no manure and quantifying its efficacy relative gathering, hauling or spreading of manure. to cow manure, a point which greatly inter- Instead, manure is deposited directly in the ested the local participants. This demon- fields by the cattle themselves. Furthermore, strates how local knowledge can inform the field gets the urine, which gathered scientific research that then bolsters it manure lacks. A second benefit is that the through experimentation. manure is more evenly spread within the Within the scientific literature, there are field than it is with the haul-and-spread numerous systems for characterizing and method, which tends to create clumps of classifying soils. There are undoubtedly manure. The primary constraint cited for vastly more folk systems. When conducting this method is that the period of availability participatory, in situ soil fertility experi- of large and numerous transhumant herds of ments and land management research, it is cattle is relatively short in Madiama. useful to elicit local knowledge on soils and Immediately after completion of the experi- incorporate it into the research as well as the 140 T.A. Crane and B. Traoré

extension of the findings. Scientists can ben- Netherlands Journal of Agricultural Science efit from farmers’ observations and insights. 46, 65–75. In the SANREM context, such knowledge Defoer, T., De Groote, H., Hilhorst, T., Kante, S. has guided experimentation, enhancing the and Budelman, A. (1998) Participatory sense of participation and ownership among action research and quantitative analysis for nutrient management in southern Mali: a local collaborators. Scientific experimenta- fruitful marriage? Agriculture, Ecosystems tion can bolster and expand local knowl- and Environment 71, 215–228. edge by systematically quantifying and Dialla, B.E. (1993) The Mossi indigenous soil clas- describing the results of various manage- sification in Burkina Faso. Indigenous ment practices, both introduced techniques Knowledge and Development Monitor 1, as well as those already known. While there 17–18. is no magic bullet solution to soil fertility FAO (1998) World Reference Base for Soil maintenance, participatory experimenta- Resources. FAO, Rome, Italy. tion and exchange of ideas between peas- Finkl, C.W. (1982) Soil Classification. Hutchinson ants and researchers has been found to Ross, Stroudsburg, Pennsylvania. Furbee, L. (1989) A folk expert system: soils clas- stimulate new approaches within scientific sification in the Colca Valley, Peru. research as well as the adoption and adapta- Anthropological Quarterly 62, 83–102. tion of scientifically developed techniques Gobin, A., Campling, P., Deckers, J. and Feyen, J. by peasant farmers themselves. (2000) Integrated toposequence analyses to combine local and scientific knowledge sys- tems. Geoderma 97, 103–123. References Guillet, D. (1989) A knowledge-based-systems model of native soil management. Anthro- Adesina, A. (1992) Village-level studies and pological Quarterly 62, 59–68. sorghum technology development in West Painter, T., Sumberg, J. and Price, T. (1994) Your Africa: case study in Mali. In: Moock, J.L. terroir and my ‘action space’: Implications of and Rhoades, R. (eds) Diversity, Farmer differentiation, mobility and diversification Knowledge and Sustainability. Cornell for the approche terroir in Sahelian West University Press, Ithaca, New York, pp. Africa. Africa 64, 447–463. 147–168. Pawluk, R.R., Sandor, J.A. and Tabor, J.A. (1992) Ahrens, R.J. and Arnold, R. (2000) Soil taxonomy. The role of indigenous soil knowledge in In: Sumner, M.E. (ed.) Handbook of Soil agricultural development. Journal of Soils Science. CRC Press, Boca Raton, Florida, pp. and Water Conservation 47, 298–302. E117–E135. Powell, J.M., Fernandez-Rivera, S., Hiernaux, P. Arrouays, D. (1987) Farmers’ knowledge and soil and Turner, M.D. (1996) Nutrient cycling in survey: a multivariate treatment of agrope- integrated rangeland/cropland systems of dologial data. Soil Survey and Land the Sahel. Agricultural Systems 52, 143–170. Evaluation 7, 81–86. Rhoades, R. (1984) Breaking New Ground: Bockman, O. (1990) Agriculture et fertilisation: Agricultural Anthropology. International Les engrais – leur avenir. Norsk Hydro, Oslo, Potato Center, Lima, Peru. Norway. Rhoades, R. (1989) The role of farmers in the cre- Brouwer, J. and Powell, J.M. (1998) Increasing ation of agricultural technology. In: nutrient use efficiency in West African agri- Chambers, R., Pacey, A. and Thrupp, L.A. culture: the impact of micro-topography (eds) Farmer First: Farmer Innovation and on nutrient leaching from cattle and sheep Agricultural Research. Intermediate Tech- manure. Agriculture, Ecosystems and Envi- nology Publications, London, pp. 3–9. ronment 71, 229–239. Scoones, I. and Thompson, J. (1994) Beyond Chambers, R., Pacey, A. and Thrupp, L.A. (1989) Farmer First: Rural People’s Knowledge, Farmer First: Farmer Innovation and Agri- Agricultural Research and Extension Prac- cultural Research. Intermediate Technology tice. Intermediate Technologies Publicat- Publications, London, UK. ions, London, UK. Debrah, S.K., Defoer, T. and Bengaly, M. (1998) Scoones, I. and Toulmin, C. (1998) Soil nutrient Integrating farmers’ knowledge, attitude and balances: what use for policy? Agri- practice in the development of sustainable culture, Ecosystems and Environment 71, striga control interventions in southern Mali. 255–267. Farmers’ Knowledge and Perceptions of Soil Fertility 141

Sebillotte, M. (1989) Fertilité et système de pro- structs favouring net primary productivity. duction ‘ecologie et aménagement rural’. Geoderma 57, 183–211. INRA, Paris. Van den Brink, R., Bromley, D. and Chavas, J.P. Sillitoe, P. (1996) A Place Against Time: Land and (1995) The economics of Cain and Abel: agro- Environment in the Papua New Guinea pastoral property rights in the Sahel. The Highlands. Harwood, Amsterdam. Journal of Development Studies 31, 373–399. Spaargaren, O.C. (2000) Other systems of soil Vannier, G. (1987) The porosphere as an ecologi- classification. In: Sumner, M.E. (ed.) cal medium emphasized in professor Handbook of Soil Science. CRC Press, Boca Ghilarov’s work on soil animal adaptations. Raton, Florida, pp. E137–E174. Biological Fertility of Soil 3, 39–44. Tabor, J.A. (1993) Soil survey and indigenous soil Vilain, M. (1989) La production végétale: classification. Indigenous Knowledge and Agriculture d’aujourd’hui. Lavosier TEC et Development Monitor 1, 28–29. DOC, Paris, France. Talawar, S. and Rhoades, R. (1998) Scientific and Zimmerer, K. (1994) Local soil knowledge: local classification and management of soils. answering basic questions in highland Agriculture and Human Values 15, 3–14. Bolivia. Journal of Soil and Water van Breemen, N. (1993) Soils as biotic con- Conservation 49, 29–34.

11 Biophysical Assessment of Alternative Soil Fertility Technologies Through Modelling

Oumarou Badini,1 Boureima Traoré2 and Mohamed Sidi Mohamed Touré2 1International Programs/Development Cooperation, Washington State University, Pullman, WA 99164–5121, USA; 2Centre Régional de la Recherche Agronomique, Institut d’Economie Rurale (IER), BP 205, Mopti, Mali

One of the objectives of the SANREM bio- find sustainable ways to improve crop pro- physical research in West Africa has been ductivity and reduce production risks due the assessment of the effects of land use pat- to irregular and low rainfall as well as low terns and agricultural practices on the land- fertility of soils. Technical options to scape and lifescape systems of smallholder improve the productivity of crops exist. farmers. The overall goal of the biophysical However, the identification of sustainable tools and methods development activity agricultural options requires analyses that conducted in the last 3 years in Mali is to can make assessments over long timeframes support researchers and natural resource of uncertain and changing biophysical fac- decision-makers at local and regional scales tors, and variable social and economic set- with appropriate data, tools and methods to tings (Stöckle, 1989). The selection of identify and evaluate alternative technolo- alternative technologies suitable to a given gies. This has led to the creation of a natural environment requires a long time series (30 resource baseline information and model- years or more) of crop responses to weather, ling system as a tool to support decision soil and management. Therefore, it is making, and as a platform for more focused impractical to rely solely upon empirical research. To reach such a goal, the strategy observations to demonstrate that a particu- on the biophysical side has been to integrate lar set of cropping systems is sustainable (or baseline database development, soil and not sustainable) over a 40–50 year time- land use cartography, geographical informa- frame. tion systems (GIS), and cropping systems New tools such as simulation models simulation modelling tools. The baseline that portray the response of crops to environ- data, soil and land use maps will increase mental, site and management variables offer natural resource stakeholders’ awareness the best opportunity, other than seemingly and knowledge about existing natural endless experimentation, to evaluate the resource sustainable production potentials long-term effects of farming practices on the and limitations. landscape and lifescape (Loomis and Connor, The challenge faced by farmers in the 1992). They can provide the information semi-arid region of West Africa in general, necessary to assess the long-term biophysical and in the study region in particular, is to viability of different crops and alternative © CAB International 2005. Conflict, Social Capital and Managing Natural Resources (ed. K.M. Moore) 143 144 O. Badini et al.

technologies suited to a given environment. A cropping system here is defined as a crop They can also provide coefficients and sto- (or pasture) community, together with man- chastic crop yield outputs to be used for risk agement practices such as tillage methods and socioeconomic analyses of different and rotation used in crop production management practices (see Chapter 12). (Loomis and Connor, 1992). At this field The main objective of the present scale it is possible to examine the impact of chapter is to share the experience in the management on crop productivity, sustain- assessment of the viability of existing ability, and overall biophysical viability. By and alternative technologies adaptable observing and/or simulating field condi- to the study landscape and similar areas, tions over years, effects of crop rotation, using cropping systems simulation model- tillage practices, soil conservation, soil ling. Cropping Systems Simulation Model amendments and subsequent yields can be (CropSyst) (Stöckle and Nelson, 1996) was seen, and analyses made of the use of developed with a focus on crop growth sim- resources such as nutrients and water. ulation in response to soil, weather and Economic analyses of costs and benefits can management, has been adapted to condi- also be conducted at the field scale. tions in the Madiama region, and used to At a higher level, fields are compo- simulate the long-term effects of selected nents of farms under the management of par- soil fertility management practices. Simu- ticular farmers. The principal crops and lation modelling is used here as a comple- management practices employed on a par- ment to field-testing to allow for long-term ticular farm constitute a farming system. evaluation and comparison of existing and Ruthenberg (1980) states that because farms alternative soil fertility management prac- are organized to produce a net economic tices. The present assessment of the simula- return they are the fundamental units for tion results allowed for the identification of economic and sociological analyses. The soil fertility technologies that are best suited regional organization of farming systems is to the region. the landscape level. It offers an additional An approach integrating biophysical venue for analysis, including matters such indicators of productivity and sustainability as drainage and water quality analyses. At generated from field monitoring and simula- this level, economic and sociological studies tion is used for this assessment. Yield vari- might include service roles of and impacts ability and production risk as well as soil on towns and villages within the region. losses due to water erosion were analysed Details on biophysical landscape char- over multiple seasons to assess the viability acteristics and farming systems of the study of existing and alternative cropping systems site are given in Chapter 3. and management practices. This approach provides a simple and more complete method for evaluating a set of complex CROPSYST: cropping systems simulation model interacting variables in an easy-to-under- Brief overview of model stand manner applicable to most cropping systems. It allows yield efficiency, variabil- CROPSYST is a multi-year, multi-crop, daily ity and environmental sustainability to be time-step generic crop growth simulator. advanced on an equal basis. The model is developed to help decision- makers analyse current and alternative cropping systems and predict their environ- Methodology mental and socioeconomic sustainability. It is a comprehensive cropping system model Scale considerations to be used as an analytical tool to assess the effects of weather, soils, crop rotations and Individual fields are the fundamental geo- management practices on complex cropping graphical units in the modelling and analy- systems and the environment. The model- sis of cropping systems in the present study. ling framework has a capacity to reflect site Modelling of Alternative Soil Fertility Technologies 145

specificity and integrates several different sites. A simulation control file combines the components and management options. This input files as desired to produce specific holistic approach permits users to assess the simulation runs. In addition, the control file sustainability of cropping systems from the determines the start and ending day for sim- point of view of constraints imposed by ulation, defines the crop rotations to be sim- productivity, resource conservation, protec- ulated, and sets the values of all parameters tion of water quality and socioeconomic requiring initialization. The CROPSYST considerations. model has been adapted for West African Details on management options and conditions. Database and model parameter model components can be found in the files (previously non-existent for the study model’s user’s manual (Stöckle and Nelson, area) were established for weather, soils, 1996) and elsewhere (Badini et al., 1997). crops and different management schemes in The model simulates the soil water budget, the present study. soil-plant nitrogen budget and crop growth and development. The amount of water LOCATION AND WEATHER DATABASE Para- used for crop growth can be determined by meters characterizing the site of interest are modelling the components of the soil water name (Madiama), latitude, longitude, eleva- balance. The water budget components con- tion and daily rainfall as well as mini- sidered are: irrigation, precipitation, rain- mum and maximum temperatures. Actual fall intercepted by the crop and surface weather data from 1970 to 2002 for Djenné residue, surface runoff, soil and residue (located at about 25 kilometres from the evaporation, infiltration through soil layers, study site Madiama) and Madiama were transpiration, deep percolation and water used with the weather generator CLIMGEN storage in the soil profile. to generate 50 years of daily data including Nitrogen is an important nutrient solar radiation, rainfall, air temperature and whose availability is primordial for crops to relative humidity. These compiled weather reach their growth potential. The compo- data were used for the simulation modelling. nents of the nitrogen budget in CROPSYST include transport, transformation, ammo- SOIL DATABASE Generally, the soils of the nium sorption, crop nitrogen uptake and commune are poor in nutrient elements. residue mineralization. Nitrogen types Baseline values have been derived that (nitrate and ammonium), form (organic or show very low values for total nitrogen (less inorganic, solid or liquid), and application than 0.1% in all soils), very low values for methods can be specified in each simula- organic carbon content (0.08 to 0.5%) and tion. CROPSYST simulates crop growth and deficient phosphorus content (2.5 to 8 ppm). development using a generic crop simulator. The soils range from strongly acid (pH 4.2) Species and cultivars are characterized by in the hydromorphic floodplains to slightly a set of parameters that determine crop res- acid (pH 6.2) in the loamy and sandy plains. ponse to the environment. The crop para- Using data provided by the Soil Survey meter file is structured in the following of Madiama Commune (Badini and Dioni, sections: phenology, morphology, growth, 2001) and seasonal biophysical monitoring residue, plant nitrogen, harvest index and data from the site, a soil parameter file was salinity tolerance. constructed for each of the eight land units identified in the commune (Chapter 3). The plains of sandy to loamy materials (unit t2 – Model inputs used in the study tropical ferruginous soils or USDA Ultic In order to run CROPSYST, input data describ- Haplustalfs), which cover about 25.5% of ing the location, weather, soils, crops and the commune and on which millet is grown, management from the study site are needed. were used for the present simulation, and Separation of files allows for an easier link the comparative biophysical assessment of of CROPSYST simulations with GIS allowing alternative and unimproved soil fertility the extrapolation of research results to other management technologies. 146 O. Badini et al.

CROP DATABASE From the existing data- term evaluation and comparison of existing base, a 90-day millet cultivar (local name: and alternative farming and NRM technolo- sagnori) was used in the present simulation. gies without long-term field tests to identify Calibration for the millet cultivar parame- crops and technologies that are best suited ters and yield estimate was performed using to the region. Technologies evaluated four years of field data on phenological include crop rotations and organic events (emergence, flowering and maturity (manure) and inorganic (chemical) fertiliza- dates) and crop yields from the biophysical tion (Table 11.1). Information and data for monitoring and field trials conducted in the management database were derived Madiama since 1999. These were adjusted from a review of soil fertility literature in accordance with results compiled from (Traoré, 2002) and soil fertility trials con- similar tests in the region. ducted in Madiama. After calibration of millet cultivars, the databases for soils, weather, crops and man- MANAGEMENT DATABASE In the course of agement practices were combined through a the SANREM West Africa project, a number simulation rotation table in CROPSYST to run of soil fertility management technologies a 30-year simulation of crop growth, yield (Traoré, 2002) were field tested to deter- and soil water erosion rates of selected soil mine which were the most viable and fertility management practices. The outputs adaptable to the research area and region. from the simulation were used to assess the Simulation modelling was used as a com- biophysical viability of selected soil fertility plement to field-testing to allow for long- management practices (Table 11.1).

Table 11.1. Soil fertility management simulated.

Fertilization practices Treatment N fertilizing unit (kg/ha)

Traditional management Control (no N input) 0 Crop rotations • Continuous cropping Millet–millet 0 • Continuous cropping Cowpea–cowpea 6.03 kg N every year • 2-year rotation Cowpea–millet 6.03 kg N every 2 years • 3-year rotation Cowpea–millet–millet 6.03 kg N every 3 years Organic fertilization • Cow manure 2 t/ha every year 19.6 kg N every year 5 t/ha every 3 years 49 kg N every 3 years • Small ruminant manure 2 t/ha every year 25 kg N every year 5 t/ha every 3 years 62.5 kg N every 3 years Inorganic fertilization • Recommended 100 kg diammonium phosphate 15 kg N at planting, 23 kg N (18-46-0) at planting, 50 kg urea at 10 days, and 23 kg N (46-0-0) each at 10 and 50 days at 50 days after emergence • Microdose complex 1 60 kg/ha Complex cereal NPK 9 kg N at planting (6 g/poquet) (15-15-15) at planting • Microdose complex 2 60 kg/ha Complex cereal NPK 9 kg N at 10 days (6 g/poquet) (15-15-15) at 10 days after emergence • Microdose DAP and urea 20 kg/ha of diammonium 3.6 kg N at planting and (2 g/poquet DAP and phosphate (18-46-0) at planting, 4.6 kg N at 30 days after 1 g/poquet urea) and 10 kg/ha of urea (46-0-0) at emergence 30 days after emergence Modelling of Alternative Soil Fertility Technologies 147

Assessment of crop rotation, organic and limitation set for water, nutrients and tech- inorganic soil fertility management practices nology. Attainable or optimum yield corre- sponds to the best yields achieved through From the many outputs CROPSYST allowed, skilful use of the best technology. High the present study used yield and erosion yields reported by experimental stations rates as the key indicators of productivity and the best farmers serve as measures of and sustainability. Crop yields are the dom- attainable yield. The concepts of actual, inant factors affecting, for example, farmer attainable and potential yields assist in adoption of alternative practices such as soil assessments of farming systems and help to fertility and conservation practices. Erosion identify opportunities for improvement. rates are determinants of soil fertility status They also serve to define the intensity of and water holding capacity, both of which farming. The average yield obtained from are critical to plant production. The amount each management scenario will be com- and quality of soil eroded or retained on a pared to the attainable (optimum fertiliza- plot of cropped land is a primary determi- tion) yields to determine their level of nant of sustainability for either food or cash efficiency. In this study, optimum fertiliza- crops. Quantification of yields and expected tion corresponds to the fertilization levels erosion rates over time is important in recommended by research stations in Mali deciding on the productivity and sustain- (Table 11.1). ability of a given natural resource base. To determine the overall biophysical STABILITY Stability is defined here as the viability of each technology, the variables of variation in yield over time, i.e. as relating productivity and sustainability were com- to repeatability and predictability in farm- bined into a unique viability factor deter- ing. Yields may vary over years with mined as a weighted value of yield efficiency, weather and other causes. Marten (1988) stability and soil sustainability. The best- used the term stability in reference to the adapted technologies are then determined as degree of such variation. Standard deviation the ones with the best yield efficiency, stabil- (SD) and coefficient of variation (CV%) are ity and sustainability. Details on each of used as measures of degree of variation or these indicators are given below. stability of yield. Comparisons are made between the different management practices to determine the most stable in terms of pro- Indicator of productivity: yield ductivity. Productivity is explicitly defined by yield, the amount of useful product per unit land YIELD PROBABILITIES Yield probabilities area. The concepts of efficiency, stability of exceeding a given level of yield were com- and stochastic dominance (riskiness) are puted and plotted to compare different man- used here to analyse and compare the yields agement practices, and to determine the obtained from the different management most risk-efficient and dominant manage- practices simulated in the present study. ment systems. For example, if the plots of yield probability level of different manage- EFFICIENCY Because yield is the resultant ment practices do not intersect, the most of various natural and human factors such risk-efficient management is the one whose as radiation, water, nutrients, and labour, it probability at all outcomes (e.g. yields) is has the property of also measuring effi- more than, or occasionally equal to, the ciency relative to such inputs (Loomis and probabilities of all other managements. This Connor, 1992). Observed yields may fall is known as first-degree stochastic domi- anywhere on a continuum between crop nance (FSD) and is based on the fact that failure and potential yield. In the Food and decision makers prefer more income to less Agriculture Organization (FAO) terminol- (Anderson et al., 1977). ogy, potential yield is considered as that In the common situation where more obtained under the ideal conditions with no than one probability function first degree 148 O. Badini et al.

dominates the others or if no yield function nance rank (from 1st to nth) that is then dominates, then, the second-degree stochas- combined into a final biophysical viability tic dominance (SSD) principle is applied to rank to determine the overall dominating produce a ranking. SSD can be established technologies. The best-adapted technolo- by measuring the cumulative area below two gies are then determined as the ones with functions (Anderson et al., 1977; Johnson the best yield efficiency, stability and sus- et al., 1994). For example, a yield probabil- tainability. ity distribution, I, dominates another, J, by SSD if the cumulative area under I exceeds the area under J. SSD is appropriate for a risk Results and Discussions averse decision maker, but is inappropriate for risk-neutral and risk-preferring individu- Productivity assessment of soil fertility als. SSD is appropriate in the Sahel, because management practices most farmers and most government policy Crop rotations makers cannot afford to be other than risk averse. It is assumed that farmers can avoid or reduce problems such as weeds, soil-borne diseases, low fertility and soil hardpans Indicator of sustainability through rotations. Rotations are also used to Sustainability is concerned with whether manage water availability by sequencing the production level can be maintained crops that have different patterns and over the years at the site. An overlap evi- amounts of water and nutrient use. In this dently exists between concepts of stability way, the system can fine tune to rainfall vari- and sustainability. Large yield variations ations over the years and give higher or more and persistence of crops both affect sus- efficient, stable, risk-efficient yields, and tainability and stability, but maintenance of allow a more sustainable production envi- soil resources is the key issue, because soil ronment than the existing systems. The degradation has the effect of gradually low- results from the field tests and the long-term ering the potential of site until farming is simulation confirm this assertion. no longer possible. Proper management of In the simulation runs, monoculture soil resources is the key to sustaining agri- (continuous cultivation) for millet was com- culture, because soil properties determine pared with a 2-year cowpea–millet rotation the type of farming that can be practised. and a 3-year cowpea–millet–millet rotation. Soil characteristics important to sustain- Figure 11.1 shows that long-term average ability include the ability to supply essen- millet yield was 583 kg/ha for the low tial plant nutrients and water and the ease input traditional practice compared with of tillage. Therefore, rate of soil loss and 650 kg/ha for the cowpea–millet–millet nutrient exhaustion serves to quantify sus- rotation (n-m-m), and 726 kg/ha for tainability. The soil water erosion rates cowpea–millet rotation (n-m), a difference (SWER) are used to compare the manage- of 67 kg and 143 kg/ha, respectively. Overall ment practices and technologies in the pre- yield efficiency (Table 11.2) compared to sent assessment. the attainable optimum production condi- tions is, respectively, 52% for the 2-year rotation and 47% for the 3-year rotation. Overall biophysical viability There is no difference in stability between To determine the overall biophysical viabil- the two rotations with a CV of 56% for both ity level of each technology, the variables of the 2-year and 3-year rotations. yield efficiency, stability and soil sustain- In addition, the long-term simulation ability were combined into a unique viabil- has allowed for the determination of yield ity factor. In reference to each of these probabilities and risks linked to production biophysical indicators (efficiency, stability, and the choice of the best technology among SWER), each technology was given a domi- the compared systems. Modelling of Alternative Soil Fertility Technologies 149

800 58

Average CV (%) 700 56 56 56

600

54 500

400 52 CV (%) Yield (kg/ha)

300 50 50 200

48 100

0 46 Millet mono. Millet in cowpea–millet Millet in cowpea–millet–millet Practices Fig. 11.1. Millet long-term average yield and variation (CV%) in rotation practices (monoculture, cowpea–millet, and cowpea–millet–millet).

Table 11.2. Long-term millet yields (mean, stability and efficiency) for soil fertility management practices.

Stability Efficiency recommended Fertilization practices Mean SD CV (%) yield (%)

Potential productivity 1949 694 38 140 Recommended (attainable) fertilization 1392 458 33 100 Traditional low input monoculture 583 298 51 42 Cow manure (2 t/year) 1011 277 27 73 Cow manure (5 t/3 years) 928 271 29 67 Small ruminants manure (2 t/year) 1078 293 27 77 Small ruminants manure (5 t/3 years) 993 279 28 71 Microdose complex 1 – NPK (6 g/hill 910 287 32 65 at planting) Microdose complex 2 – NPK (6 g/hill 924 280 30 66 at 10 days after emergence) Microdose DAP and urea (2 g/hill DAP 908 279 31 65 at planting and 1 g/hill urea 30 days after emergence) Millet in cowpea–millet (n-m) 726 409 56 52 Millet in cowpea–millet–millet (n-m-m) 650 367 56 47 150 O. Badini et al.

Given the erratic nature of the weather Considering yield efficiency level, sta- in arid and semiarid Africa and the subse- bility and stochastic dominance, the 2-year quent stochastic yields, yield probability cowpea–millet rotation is the most produc- levels represent another way of looking at tive, stable, and risk-efficient of all three production stability and the risk involved in rotations under comparison. The continu- growing a given crop in a particular envi- ous millet system is the least efficient. This ronment and management scheme. Figure analysis confirms the hypothesis that crop 11.2, which shows probability distributions rotations will give more risk-efficient yield for the two rotations and the traditional con- than mono-cropping systems. tinuous monoculture practice, indicates that in 1 year out of 10, the yield for millet in the cowpea–millet rotation will exceed Organic (manure) fertilization 1217 kg/ha compared with 1154 kg/ha and 950 kg/ha for millet in cowpea–millet– Organic fertilization from manure trans- millet rotation and continuous millet, ported to the field or corralling animals respectively. In 5 years out of 10, the millet directly in the field is a known traditional yield in a cowpea–millet rotation is practice in the Sahel. But given the difficul- expected to exceed 628 kg/ha compared ties in producing the required amount for with 550 kg/ha and 496 kg/ha for millet optimum crop growth, SANREM proposed a yield in cowpea–millet–millet and continu- number of alternative improved technolo- ous millet, respectively. Since Fig. 11.2 gies that basically bring in lower quantities shows the dominance of the 2-year cowpea– in manure. Manure from small ruminants millet rotation over the 3-year cowpea– (sheep and goats) and large ruminants millet–millet and the continuous cropping (cattle) at a rate of 2 t every year or 5 t every 3 of millet, this 2-year rotation appears to be years was simulated and their corresponding the most risk-efficient practice among the yield outputs compared to existing unim- three in the present comparison. proved management systems (Table 11.1).

10

9

8 Millet in rotation cowpea– millet 7 Millet in rotation cowpea– millet–millet 6 Millet monoculture 5

4

3

2

Probability of exceeding yield in years out of 10 Probability of exceeding yield in years out 10 1

0 300 500 700 900 1100 1300 Yield (kg/ha)

Fig. 11.2. Yield probability functions for rotation practices. Modelling of Alternative Soil Fertility Technologies 151

1200 60

1000 51 50

800 40

600 30 28

29 CV (%) 27 27 400 20 Average millet yield (kg/ha) 200 10

0 0 Traditional Cow manure Cow manure Small rum. Small rum. 2 t/yr 5 t/3 yr manure 2 t/yr manure 5 t/3 yr Organic fertilization practices

Fig. 11.3. Millet long-term average yield and variation (CV%) for organic fertilization practices.

Figure 11.3 shows that among the four 496 kg/ha for the traditional low input prac- manure type and rate combinations, small tice. It shows that from a biophysical stand- ruminant manure at 2 t every year (PR 2t) point, PR 2t is the dominant and most has the highest long-term average (1078 risk-efficient practice among the four kg/ha) compared with yields between 1011 organic fertilization technologies screened kg/ha and 928 kg/ha for the other organic in the present work. fertilization practices. Yield efficiency for PR 2t is 77% of recommended attainable Inorganic fertilization yield compared to 73% or less for the other manure practices (Table 11.2). Although not It is assumed that crops that are not limited statistically different from each other, by the availability of nutrients will grow all manure practices showed a significant best and will have the highest yields. But difference in yield from the traditional prac- given the fact that local farmers often cannot tice (only 583 kg/ha) as well as lower vari- afford the recommended doses, a number of ability (27–29% of CV) compared to the improved technologies that basically bring traditional practice with 51% of variability. in lower levels of fertilizers (microdosing) Also, organic fertilization has exhibited the and different dates of placement are pro- lowest yield variability of all management posed and compared to existing (unim- practices simulated. proved) and recommended fertilization Overall, Fig. 11.4 shows that the rate of systems (Table 11.1). 2 t of small ruminant manure applied each The long-term 30-year average yield year stochastically dominated all other outputs from the three fertilizer microdos- organic fertilization simulated in the analy- ing management practices (Fig. 11.5) show sis. In 1 year out of 10, yield from PR 2t will no significant yield difference. But the aver- exceed 1424 kg/ha compared to 1279 kg/ha age yields from all microdosing practices for cattle manure at a rate of 5 tonnes every were significantly higher than the control 3 years (Bov 5t). Similarly, in 5 years out of (traditional practice with low input) with a 10, PR 2t allows a productivity of 1061 kg/ha difference in yield between 325 and 341 compared to 890 kg/ha for Bov 5t, and kg/ha. Microdose complex 2 (with 60 kg/ha 152 O. Badini et al.

10

9 Cow manure 2 t/yr Cow manure 5 t/3 yr 8 Traditional Small rum. manure 2 t/yr Small rum. manure 5 t/3 yr 7

6

5

4

3

2 Probability of exceeding yield in years out of 10 Probability of exceeding yield in years out 10

1

0 300 500 700 900 1100 1300 1500 Yield (kg/ha)

Fig. 11.4. Yield probability functions for organic fertilization practices.

1600 60

1400 50 50

1200

40 1000

32 32 31 800 30 30 CV (%) 600 20 Average millet yield (kg/ha) 400

10 200

0 0 Recommended Traditional Microdose 1 Microdose 2 Microdose DAP Inorganic fertilization practices

Fig. 11.5. Millet long-term average yield and variation (CV%) for inorganic fertilization practices.

Complex cereal NPK (15-15-15) at 10 days and Microdose DAP (20 kg/ha of diammo- after emergence) has an average yield only nium phosphate at planting, and 10 kg/ha of slightly higher than Microdose complex 1 urea at 30 days after emergence). All three (60 kg/ha Complex cereal NPK at planting) practices have a similar CV of about 31% Modelling of Alternative Soil Fertility Technologies 153

10 Microdose 2 – NPK at 15 days 9 Microdose 1 – NPK at planting Millet monoculture 8 Millet recommended 7 Microdose DAP-urea

6

5

4

3

2 Probability of exceeding yield in years out of 10 Probability of exceeding yield in years out 10

1

0 300 500 700 900 1100 1300 1500 1700 1900 2100 Yield (kg/ha)

Fig. 11.6. Yield probability functions for inorganic fertilization practices.

and an efficiency level compared to attain- Sustainability and biophysical viability able yield of about 65% (Table 11.2). This assessment of soil fertility management represents a yield improvement of about practices 64% over the traditional practice. Sustainability analysis of all management The yield probability distribution func- practices tions (Fig. 11.6) give a small edge to Microdose 2, which slightly dominated the Erosion rates are determinants of soil fertil- other two Microdose practices at most prob- ity status and water holding capacity that ability levels. Although Microdose 1 and 2 are critical to plant production. The amount have the same total amount of fertilizer of soil eroded from a plot of cropped land input (60 kg/ha), the difference is that is a primary determinant of sustainability Microdose 2 was implemented at 10 days of crop production. Sustainability is con- after emergence instead of at planting cerned with whether the production level as done with Microdose 1. Despite a lower can be maintained over the years at the site. total input in fertilizer (30 kg/ha), which is Although some erosion during cropping only half of both Microdose 1 and 2, cycles is inescapable, all sites have some tol- Microdose DAP has an average yield similar erance to erosion. A strict tolerance with the to Microdose 1 and only a difference of 16 soil loss rate equal to or less than the soil for- kg/ha from Microdose 2. Furthermore, even mation rate would allow production to con- with lower fertilizer input it has shown a tinue indefinitely. Generally, estimates of slight SSD to Microdose 1. This result soil formation rates at undisturbed sites are strengthens the importance of date of fertil- near 0.1–0.2 mm depth per year (i.e. about izer placement and type, which was picked 1.3–2.6 t/ha/year). Rates increase with rain- up by the model. Economic analyses as pre- fall (Loomis and Connor, 1992), and many sented in Chapter 12 should establish the other soils in Madiama have soil loss above financial viability of this finding for deci- these figures. The soil type Ultic Haplustalfs sion makers. (tropical ferruginous soils) used in the 154 O. Badini et al.

present analysis has exhibited one of the The results show that simulation model- lowest soil loss rates with an overall average ling can be a good tool to help show the long- of about 1 t/ha under the parameters used in term impact of management on soil loss and the modelling. sustainability. The ranking of the practices as In the present analysis, the rate of soil related to the level of soil loss shows that loss is used as an indicator of sustainability. improvement of both soil physical and nutri- Cumulative (30-year) soil loss was com- ent status is paramount to conserving the soil puted for each management practice and resource over the long-term. Practices that comparisons were made to determine the yield more crop biomass and residue, such as least prone to soil loss. The most sustainable the recommended optimum fertilization management is considered as the one practice, and improve soil binding and water that would have less total soil loss over holding capacity, such as organic fertiliza- the 30 years and would therefore maintain tion through manuring of fields, are better in production level over years at the site. sustaining the resource production base, Figure 11.7 shows that in the course of 30 which is the soil. In general, manure, which years the soil fertility management of provid- allows binding of soil particles and an ing 2 t of small ruminant manure every year improvement of soil water holding capacity (PR 2t) has the least cumulative soil loss (9 and fertility status, has shown superiority in t/ha on soil t2), compared to all other prac- maintaining the soil resource. tices. It is followed by Bov 2t (2 tonnes cow manure every year), PR 5t, and in fourth place, Overall ranking of the biophysical viability of the recommended fertilization. The least soil fertility management practices effective practices are the traditional continu- ous cropping and the rotations. Figure 11.8 Combining yield efficiency, stability and shows the long-term trend of cumulative soil sustainability levels of all compared man- loss, and Fig. 11.9 shows the soil loss proba- agement practices, the technologies have bility levels. Overall, during the course of 30 been ranked from most viable to least viable years manure application at 2 t/year has as individual practices on biophysical shown less soil loss at all probability levels. grounds (Table 11.3). Although optimal The least sustainable (i.e. with higher soil inorganic fertilization allows higher yields, loss) are the rotations and monoculture. organic fertilization appears to be the most

50 46.0 47.0 46.4 45 40 35 30 25 21.6 22.0 22.9 23.2 22.8 19.5 20 15 10.6 8.9 10

Cumulative soil loss (t/ha) 5 0

1 2 DAP Tradit. Recom.

Microdose Microdose Cowpea–millet Small rum. 2 t/yr Small rum. Cow5 t/3 yrmanure 2 t/yr Cow manure 5 t/3 yr Cowpea–millet–millet Management practices

Fig. 11.7. 30-year cumulative soil loss for different management practices on Ultic Haplustalfs soil in Madiama commune. Modelling of Alternative Soil Fertility Technologies 155

50 Recom. 45 Traditional

40 Small rum. 2 t/yr Small rum. 5 t/3 yr 35 Cow manure 2 t/yr

30 Cow manure 5 t/3 yr DAP 25 Microdose 1

20 Microdose 2

Cumulative soil loss (t/ha) Cowpea–millet 15 Cowpea–millet–millet

10

5

0 12345678910111213 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 Years Fig. 11.8. Long-term cumulative soil loss for all management practices on Ultic Haplustalfs soil in Madiama commune.

100 Recom. Traditional 90 Small rum. 2 t/yr 80 Small rum. 5 t/3 yr

70 Cow manure 2 t/yr Cow manure 5 t/3 yr 60 DAP

50 Microdose 1 Microdose 2 40 Cowpea–millet

30 Cowpea–millet–millet

20 Probability of exceeding value on abscissa 10

0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 Soil loss (t/ha) Fig. 11.9. Soil loss probability functions for all management practices on Ultic Haplustalfs soil in Madiama commune. 156 O. Badini et al.

Table 11.3. Biophysical viability ranking of soil fertility management practices effect on millet yield in Madiama commune.

Efficiency Stability Sustainability rank rank rank Total Viability Fertilization practices (1) (2) (3) (1ϩ2ϩ3) rank

Small R. manure (2 t/year) 2 2 1 5 1 Cow manure (2 t/year) 3 1 2 6 2 Small R. manure (5 t/3 years) 4 6 3 13 3 Recommended 1 8 4 13 4 Microdose 2 6 3 6 15 5 Cow manure (5 t/3 years) 5 7 5 17 6 Microdose DAP 8 4 7 19 7 Microdose 1 7 5 8 20 8 Traditional (monoculture) 11 9 9 29 9 Millet in cowpea–millet 9 10 10 29 10 Millet in cowpea–millet–millet 10 11 11 32 11 viable practice. The 2 t/year of small rumi- ual researchers and community decision nants’ manure, followed by the 2 t/year of makers in Madiama and in the region in large ruminants’ manure applications and reaching decisions about how to best plan the 5 t every 3 years of small ruminants’ their land use and manage their resources. manure have demonstrated higher overall Also, information generated from these long-term viability than the optimum rec- analyses provides improved insight into ommended fertilization. In all criteria, rota- crop production responses to various agri- tion practices together with low input cereal cultural practices and soil management monoculture appear to be the least viable technologies across the Sahel. The end technologies in terms of yield efficiency, sta- result is identification of suitable practices bility and soil sustainability. based on performance evaluation through field monitoring and simulation modelling. The short-term field results and long- Conclusions term simulation outputs presented here have shown that all alternative management tech- One of the greatest challenges in the semi- nologies contributed to substantial increase arid Sahel region is to develop cropping in yield. Long-term yield efficiency levels strategies that provide the most efficient were between 42% for low input traditional level of nitrogen uptake for long-term yield practice and 77% for small ruminants’ man- stability and soil resource base sustainabil- ure at an application rate of 2 t/ha every year. ity. The SANREM experience in Mali shows Overall, by considering both productivity that soil, weather, crop and management (yield efficiency and stability) and sustain- databases, in combination with cropping ability(soillossrates),the2t/yearofsmalland systems simulation modelling, offer better large ruminants’ manure applications, the 5 t insights about what crop, technology or every3yearsofsmallruminants’manure,and cropping system would best fit a specified the optimum recommended fertilization are landscape. Through long term screening the most viable technologies in the long-term. and analysis plant breeders can be provided The least viable individual technologies that with a better representation of the actual sit- showed both lower yields and higher rates of uation faced by farmers for whom breeding soil loss were the cowpea–millet rotations programmes are undertaken. In addition, and the low input traditional practices. the extrapolation of findings to other simi- Under the variable conditions of rain- lar environments can be facilitated. It is fed agriculture, nitrogen management is an expected that such tools will assist individ- essential component of the production Modelling of Alternative Soil Fertility Technologies 157

process. This can be easily accomplished Badini, O. (1993) Propositions de technologie en with external fertilizer inputs, because it is milieu reel: Conditonnement de boeufs de possible to control their availability in rela- trait et production de fumier en saison seche tion to crop requirements and weather. au Campement de Dogoma (région Ouest du However, because of high costs or lack of Burkina Faso). In: Lawrence, P.R., Lawrence, K., Djikman, J.T. and Starkey, P.H. (eds) availability of inorganic fertilization, a Proceedings of the Fourth Workshop of the number of soil fertility management alterna- West Africa Animal Traction Network, Kano, tives were compared to the unimproved low Nigeria. The International Livestock Center input monocropping practice. All alterna- for Africa (ILCA), Addis Ababa, Ethiopia. tive technologies including the rotations Badini, O. and Dioni, L. (2001) Etude exhibited higher overall biophysical viabil- Morphopédologique de la Commune de ity than the unimproved practice. Organic Madiama, Cercle de Djenné (Technical and inorganic fertilization were the techni- report SANREM West Africa). Washington cally superior alternatives, because legumes State University, Intenational Programs, in rotations provide uncertain additions of Pullman, Washington. Badini, O., Stöckle, C.O. and Franz, E.H. (1997) nitrogen relative to crop requirements due Application of Crop Simulation Modeling to the variable impact of weather on their and GIS to agroclimatic assessment in growth. However, manuring in general, and Burkina Faso. Agriculture, Ecosystems & especially small ruminants’ manure applied Environment 64, 233–244. even at reduced rate of 2 t/year, offers a sus- Johnson, A.K.L., Cramb, R.A. and McAlpine, J.R. tainable and viable alternative to inorganic (1994) Integrated land evaluation as an aid to fertilization. Although each of these tech- land use planning in Northern Australia. nologies was tested individually higher Journal of Environmental Management 40, returns would be anticipated by combining 139–154. them in more integrated packages. Loomis, R.S. and Connor, D.J. (1992) Crop Ecology: Productivity and Management in Organic fertilization is a traditional Agricultural Systems, 1st edn. Cambridge practice in the Sahel, but the difficulties University Press, Cambridge, UK. in producing the required amount for Marten, G.G. (1988) Productivity, stability, sus- optimum crop growth limits its availability tainability, equitability and autonomy as and use. Research and development efforts properties for agroecosystem assessment. must emphasize strategies and methods to Agricultural Systems 26, 291–316. produce more manure and other organic fer- Ruthenberg, H. (1980) Farming Systems in the tilizers. The successful application of tech- Tropics. Clarendon Press, Oxford, UK. niques for the mass production of organic Stöckle, C.O. (1989) Simulation models in agri- fertilizers and manure using ‘composting culture: From cellular level to field scale. In: Clema, J.K. (ed.) Proceedings of the 1989 peats’ and ‘animal stabulation’ in other sim- Summer Computer Simulation Conference, ilar environments such as Burkina Faso Austin, Texas (24–27 July), pp. 639–644. (Badini, 1993) indicate that future efforts Stöckle, C.O and Nelson, R.L. (1996) CropSyst in soil fertility improvement should be (Cropping Systems Simulation Model) directed towards the technologies of produc- Users’ manual (Version 2.0). Biological tion of organic fertilizers. The long-term Systems Engineering Dept, Washington State impact analysis performed here confirms University, Pullman, Washington. that organic fertilization is good for crop pro- Traoré, B. (2002) Techniques de conservation et ductivity as well as soil sustainability. d’amélioration de la fertilité des terres cul- tivées en zone sahélienne du Mali: Synthese bibliographique SANREM CRSP West Africa (Working Paper No. 03–01). Virginia Tech, References Office of International Research, Education and Development, Blacksburg, Virginia. Anderson, J.R., Dillon, J.L. and Hardaker, J.B. (1977) Agricultural Decisions Analysis. Iowa State University Press, Ames, Iowa.

12 An Economic Evaluation of Alternative Soil Management Technologies

Peter Wyeth,1 Boureima Traoré,2 Oumarou Badini,1 Mohamed Sidi Mohamed Touré2 and Charlene Brewster3 1International Programs-Development Cooperation, Washington State University, Pullman, WA 99164–5121, USA; 2Centre Régional de Recherche Agricole, Institut d’Economie Rurale, BP 205, Mopti, Mali; and 3Office of International Research, Education and Development, Virginia Tech, Blacksburg, VA 24061–0378, USA

This chapter evaluates alternative soil man- weighing the benefits farmers derive from agement techniques that were tested in the practices against the costs they incur in Madiama. Trials of these techniques were implementing them. developed following the identification of Although we compare profits among declining soil fertility and crop yields as practices, we do not conclude that farmers matters of special concern among farmers. should necessarily all adopt the practice As Chapter 10 makes clear, farmers already that brings the greatest net returns. Different had a sound grasp, in general qualitative farmers have different means at their dis- terms, of the relative benefits of different posal, including cash for buying chemical techniques for conserving soil fertility. fertilizer or access to animal manure. They What they wanted from the research was also have different personal preferences for some level of quantification of their relative one technique or another. For instance, an returns and costs, and providing this is the interest in cowpeas, because their families objective of this chapter. can eat the peas and they can feed the leaves The evaluation employs a three- to their livestock, may lead farmers to grow pronged approach. First, it takes account of them in rotation with millet, even though the outcomes of on-farm trials. Second, it this is one of the least effective techniques. relates them to the output of biosystems Although different farmers will continue to modelling reported in Chapter 11 so that the follow different practices, we hope that the two sets of results can be compared. Third, information in this chapter will allow their it provides an assessment of the economic decisions to be better informed. viability of the various techniques by deriv- This chapter is a truly multi-disciplinary ing a notional profit for each technique. effort. The farmers in Madiama not only Most households consume much more of worked with the research agronomist in deter- their crop production than they sell, but mining what trials should be run; they also did attributing values to outputs (crop yields) all the field work and assisted in evaluating and inputs (labour and materials such as fer- the results. The Institut de l’Economie Rurale tilizer or manure) provides a means of (IER) research agronomist and technicians © CAB International 2005. Conflict, Social Capital and Managing Natural Resources (ed. K.M. Moore) 159 160 P. Wyeth et al.

oversaw the implementation of the trials and research station trials do not. However, it is collected the data. The biosystems engineer also true that the greater realism comes at a provided the modelling input, and the econo- cost. Because uncontrolled differences in mists contributed the economic elements of conditions are far from equally dispersed the analysis. The Natural Resources Man- among the trial plots, variables other than agement Advisory Committee (NRMAC) in the treatments are not held constant and Madiama served as the mechanism of com- inter-farm comparisons are often difficult. munication between the researchers and the One advantage of biosystems modelling farmers. It was NRMAC members who is that control over all variables is perfect, selected the participating farmers, in several and contamination of results by random instances volunteering themselves. Besides events such as floods and invasions by ani- communicating individually with the farm- mals is unknown. Further, the model can ers, researchers also reported their analytical provide as many observations as the findings to the NRMAC. Its members actively researcher wants, and this chapter makes responded with their own assessments and use of yields simulated over a period of considerable insights and this chapter 30 years, using actual historical data from includes their views in its analysis. Madiama on the volume and timing of rain- fall. Such advantages, however, do not allow modelling to stand on its own, independent Analytical Method of field observations of some kind. As a matter of principle, the modellers need farm The results from the on-farm trials and the trial results to calibrate the model. As a modelling complement one another in matter of practice, farmers will be sceptical important ways, while each on its own of modelling results that they cannot clearly would have serious limitations. Regarding relate to field experience. the on-farm data, they are available for, at Unfortunately, both farm trial and mod- best, only the 3 years from 2001 to 2003. For elling results are not available for all treat- some practices there are fewer years than ments. One of the commonest treatments in that. Given the kind of variations that occur the farm trials was the application of PNT, a in the Sahel due to rainfall, and in on-farm natural phosphate rock from Tilemsi in Mali, trials due to the difficulty of holding con- as a fertilizer. However, the model could not stant everything other than the treatments, accommodate this because not enough is 3 years worth of data are not enough. The known either about the rock’s chemical rainfall in Madiama in 2001, 2002 and 2003 make-up or how it breaks down and makes was 671 mm, 396 mm and 825 mm, respec- its nutrients available to plants. Therefore, tively. This was a wide variation (unfortu- trial results alone must be relied on for PNT. nately so for the farmers, who suffered crop The same applies to the intercropping of losses due to the drought in the second year) cowpeas and millet, which the model saw but still a small sample of the possibilities. only as side-by-side pure stands. A benefit of Although everything is done to try to intercropping might be the reduction of keep the variations in on-farm conditions to Striga,aparasitic weed which devastates a minimum, Striga hermonthica (a parasitic millet but cannot survive on cowpeas, but weed) can unexpectedly affect one plot, an the model looks only at soil nutrients and unaccustomed amount of rain can flood water, not the impact of weeds. another, and animals can graze another On the other hand, only modelling before harvest. However, trials are con- results are available to assess full application ducted on-farm because conditions are rates of chemical fertilizer and a 2 year more realistic than they are on research sta- cowpea–millet rotation. The first was not tions. Because of this, and because farmers included in farm trials because it would have can see the results for themselves and know involved an expensive cash outlay beyond the first-hand what went into obtaining them, means of most farmers. The second was omit- on-farm trials have a credibility that ted because farmers and researchers preferred Economic Evaluation of Alternative Soil Management Technologies 161

the 3 year rotation and there was a limit to the • Applications of animal manure to crops, number of treatments that could be tested. either by collecting it and transporting it In Chapter 11, Badini and Traoré refer to the fields, or by agreeing with herders to efficiency, stability and sustainability as that they may graze their animals on criteria for evaluating techniques. The infor- crop residue in the farmers’ fields if they mation focused on in this chapter can easily leave the manure there. be related to the same concepts. Badini and • Applications of small amounts (micro- Traoré take yield as one measure of effi- doses) of chemical fertilizer. To mini- ciency, and we look at that here also, but the mize costs while increasing yields, more comprehensive measure of efficiency is farmers apply a few grammes of fertil- profits because, in one concise figure, it takes izer right next to the seed or the plant. account of not only benefits but also costs. • Application of PNT, a natural phos- Stability, or keeping risk down, is crucial phate rock from Tilemsi available in for any farmer, but especially for those with local markets, as a low-cost fertilizer. limited means. They cannot afford to follow • Rotating and intercropping millet and practices that do very well in some years and cowpeas (niébé) to diversify farm diets, very poorly in others, because storage facili- increase production of better quality ties are inadequate and no one can know how forage that the cowpea residues (les many bad harvests there might be in a row. For fanes de niébé) supply, and rest the soil the trials, we illustrate stability by reporting between millet crops, thereby taking average yield data for each treatment for each advantage of the nitrogen fixation of the 3 years that they were run. The consid- capacity of cowpeas. erable variation from one year to the next in these results will be evident. However, the The trials were conducted on small plots in long series of results from the modelling farmers’ fields (although the results are output allow a different view of stability. reported on a per hectare basis), and the farm- Because it is helpful for farmers to know how ers provided the labour and kept the harvest. badly they are likely to do in years when rain- The project provided the chemical inputs fall is poor, average and good, the minimum (fertilizers and insecticides) and plant mate- yield and profit that farmers can expect 75%, rial, and IER researchers and technicians 50% and 25% of the time, is reported assum- were present both to advise on techniques of ing that the weather pattern of the last 30 years implementation and to record labour time. remains more or less typical is reported. There were ten ‘unimproved’ control The modelling also permits an assess- plots that received no applications of ment of sustainability. Whereas Chapter 11 manure and chemical fertilizer and used no reported on this from the point of view of intercropping and no rotations. This treat- soil fertility, with modelling it is possible to ment, or lack of treatment, of the control graph profits over the 30 year period, which plots closely resembles the farmers’ current brings certain trends out clearly. At the end practices, except that the farmers generally of the chapter, the average discounted prof- spread some amount of manure on their its for the period are reported to take account fields. However, in practice the amounts are of the fact that poorer people, of whom there ill defined and highly variable. Using a con- are many in Madiama, generally discount trol plot with no soil fertility practices at all the future more heavily than those who are provides a consistent benchmark that better off because the needs of the poor are, all farmers readily understand. Table 12.1 by definition, more immediate. describes in detail the trials analysed in this report. All quantities are per hectare, and there are four replications in each of 3 years, Techniques Tested except where otherwise noted. One advantage the fourth treatment set, Researchers and farmers decided on the fol- corralling animals on a field, has over trans- lowing four main types of intervention: porting manure from elsewhere is that the 162 P. Wyeth et al.

Table 12.1. Treatment sets.

Treatment set Treatment

1. Millet with animal manure – • 2 tonnes cattle manure every year farm trials and modelling •5 tonnes cattle manure every 3 years •2 tonnes small ruminant manure every year •5 tonnes small ruminant manure every 3 years 2. Sorghum with animal manure – • 2 tonnes cattle manure every year (2 farmers) farm trials only •5 tonnes cattle manure every 3 years •2 tonnes small ruminant manure every year 3. Millet with animal manure and PNT – • As for Set 1, but with 300 kg of PNT farm trials only in the first year for each treatment. 4. Millet after corralling animals on the • 1667 cattle/ha (25/150 m2) for 2, 5 and 10 nights field before planting, first year only – • 3333 small ruminants (50/150 m2) for 2, 5, and farm trials only 10 nights (2002 and 2003 only) 5. (a) Millet with microdoses of DAP • 2 grammes DAP/pocket each year at planting (2003 only) (diammonium phosphate, 18-46-0) • 2 grammes DAP/pocket each year, planting time and/or urea (46-0-0) – farm trials only plus 1 gramme urea at tillering (2003 only) (b) Millet with microdoses of NPK • 6 grammes NPK/pocket each year at planting (15-15-15) – modelling only •6 grammes NPK/pocket each year at tillering (c) Millet with recommended full doses • 100 kg of DAP at planting, 50 kg of of DAP and urea – modelling only urea at 10 days and 50 days after emergence 6. Sorghum with PNT or microdosing • 6 grammes NPK/pocket each year at planting with NPK (15-15-15) or DAP – • 6 grammes NPK/pocket at planting (2001) or farm trials only (3 trials with PNT 2 grammes DAP/pocket at planting (2002, 2003) not reported) •6 grammes NPK/pocket at tillering (2001) or 2 grammes DAP/pocket at tillering (2002, 2003) 7. Cowpeas–millet intercropping – • No PNT modelling only • 300 kg PNT in the first year • 600 kg PNT in the first year 8. (a) Cowpeas–millet–millet 3 year • No PNT rotation – farm trials and modelling • 300 kg PNT in the first year • 600 kg PNT in the first year (b) Cowpeas–millet 2 year rotation – • No PNT modelling only

field benefits not only from the animals’ every 6 square metres and for small rumi- faeces but also from their nitrogen-rich nants one every 3 square metres.) Second, urine. Another benefit is that the manure the same result can be had by reducing the does not have to be collected and trans- number of animals and increasing the ported, thereby eliminating the labour costs number of nights they are on the field. and losses entailed. In the calculations, a (Three hundred cattle on a plot for 2 nights charge for a herder is deducted as a labour is the equivalent of 150 for 4 nights.) cost, but in practice compensation to the animals’ owners may take the form of crop residues that the animals are allowed to con- Assumptions sume. There are two points to make regard- ing the large numbers of animals per hectare. Tables 12.2–12.5 show the assumptions First, because the densities that occur in regarding labour per hectare, prices of crops practice are large, these may not be unrealis- and inputs, and input quantities per hectare. tic. (For cattle the figure translates to one The field technicians who supervised Economic Evaluation of Alternative Soil Management Technologies 163

Table 12.2. Person-days per hectare.

Millet Intercrop cp-m Cowpeas Sorghum

Labour to plough, weed, guard Ploughing (every year) 2 2 2 2 Planting (every year) 1 1.5 1 1 Weeding 1 (every year) 3.75 4.2 4 4 Weeding 2 (every year) 2.25 3.4 1 2.5 Guard corral 2 nights (1st year) 2 Guard corral 5 nights (1st year) 5 Guard corral 10 nights (1st year) 10 Labour to spread 2 t animal manure (every year) 2 2 5 t animal manure (1st year) 3 3 300 PNT (1st year) 1.5 1.5 1.5 1.5 600 PNT (1st year) 2 2 Microdoses NPK/DAP (every year) 1 1 Microdoses urea (every year) 1 1 Recommended rate (every year) 1 Pesticide treatment (every year) 0.5 1 the on-farm trials kept records of the assumed that the prices for the year 2000 amounts of labour used as the tasks were per- prevailed throughout the period. formed. These records did not directly yield There were no good market prices in days of labour per task because often a set Madiama for animal manure. In Table 12.4, number of individuals did not perform the the estimated values derived for it are based tasks. More typically, the farmer might go to on the content in the manure of nitrogen (N), his field with two of his younger children in phosphorus (in the form P2O5) and potas- the morning. After a period of time an older sium (in the form K2O) and the prices of child might join them and the four of them these chemical fertilizers. The nutritive con- work together before the younger ones return tent of the manure was determined in IER home. Later an older member of the house- laboratories. (Although this is quite vari- hold might also come out to the field for a able, the content assumed here is the same while. Every day could see different individ- as that assumed by the model in Chapter 11.) uals appearing for different periods of time, While the prices of manure arrived at in this and each farmer’s situation differed from way might reflect the nutritive value of the every other’s. We looked carefully at this data manure, they almost certainly overstate the and then, in consultation with the techni- opportunity cost of the manure to farmers by cians and the farmers, derived reasonable a considerable margin, as its value in other assumptions of person-days per task, shown uses (e.g. in plaster for house construction) in Table 12.2. Harvest labour is typically paid would be much less. Therefore, in addition a share of the crop, here assumed to be 10%. to the full value calculated in Table 12.4, we There were similar variations in the have also calculated the profits using half records regarding the prices at which farm- the value. This table shows both sets of ers bought and sold inputs and crops. Crop results. Table 12.5 shows the quantities per prices also varied through the year, being hectare of seed and fertilizer. lowest just after harvest and higher later. Again, working with the technicians and farmers, we took the data and derived Results the representative figures in Table 12.3. However, when deriving profits over 30 We developed budgets for all treatments and years using the modelling results, we controls based on the assumptions above 164 P. Wyeth et al.

Table 12.3. Prices for inputs and crops.

Unit 2000–1 2001–2 2002–3 Assumptions

Millet (selling) kg 110 150 90 Millet (buy seed) kg 165 165 225 Sorghum (selling) kg 110 150 90 Sorghum (buy seed) kg 165 165 225 Cowpeas (selling) kg 300 no harvest 250 Cowpeas (buy seed) kg 450 450 675 Cowpea residue (sell) kg 125 no harvest 120 Cattle manure (full) kg 15.99 15.99 15.99 Cattle manure (half) kg 7.99 7.99 7.99 Small ruminant man. (full) kg 22.13 22.13 22.13 Small ruminant man. (half) kg 11.06 11.06 11.06 NPK kg 225 not used 225 DAP kg Not used 260 260 Urea kg 200 200 PNT kg 75 75 75 Pesticide treatment hectare 12,000 12,000 12,000

1 Ploughing ⁄2 hectare 4,000 4,000 4,000 (2 days for 1 ha) Harvest mil/sorghum % harvest 10% 10% 10% Harvest cowpeas % harvest 10% 10% 10% Other labour Pers. day 750 750 750 Transport manure kg 250 250 250 200 kg donkey cart Transport production kg 250 250 250 200 kg donkey cart Sack 1 275 275 275 yieldϫ90%Ϭ80 kg @250–300 FCFA/sack

Table 12.4. Pricing manure.

Coefficient Fertilizer Price Value Element in the (kg fertilizer/ in the fertilizer manure Fertilizer manure (%) kg elem) manure (%) (FCFA/kg) (FCFA/kg)

Cattle manure kg. Urea 0.980 2.18 0.021 220 4.70 kg. P2O5 0.580 2.27 0.013 220 2.90 kg. K2O 1.680 2.27 0.038 220 8.39 15.99 Small ruminant manure kg. N 1.250 2.18 0.027 220 6.00 kg. P2O5 0.830 2.27 0.018 220 4.15 kg. K2O 2.400 2.27 0.054 220 11.99 22.13

together with the results of the farm trials of labour (excluding harvesting) in order to and the modelling. The tables that follow indicate the investment farmers need to are taken from much larger spreadsheets make to obtain the profits shown. The crop that laid out prices and quantities of yields yields include the 10% assumed to go to and both labour and material inputs in harvest labour, but the monetary profit or detail. They show the input costs and days loss is the net of this payment. Economic Evaluation of Alternative Soil Management Technologies 165

Table 12.5. Input quantities per hectare.

Input Quantity Units

Cattle manure 2–5 mt/ha Small ruminant manure 2–5 mt/ha Seed, millet 5 kg/ha Seed, cowpeas 20 kg/ha NPK 15-15-15, microdoses 60 (6 g/pocket x 10,000 pockets/ha) DAP, microdoses 20 (2 g/pocket x 10,000 pockets/ha) NPK 15-15-15, recommended 100 kg/ha DAP, recommended 100 kg/ha Urea, microdoses 10 (1 g/pocket x 10,000 pockets/ha) Urea, DAP 100 kg/ha PNT 300–600 kg/ha

In the tables for the farm trials, the aver- labour costs incurred before harvest as well age yields for each year’s replications only as seed and manure costs. They provide a take account of the middle two results, leav- fairly complete accounting of the resources ing aside the highest and the lowest, farmers must put into each treatment. because variations in yield among replica- tions are commonly very great indeed. We wanted to reduce the possibility that the Treatment sets 1, 2 and 3: millet with animal averages would be distorted by a single manure, with and without PNT in the first unrepresentative result, but did not want to year; sorghum with animal manure substitute a different distortion by deciding on an ad hoc basis which results were out- The trials using animal manure taken out to liers and which were not. As the tables the fields were both with and without 300 show, even after trimming the means, plenty PNT applied in the first year. The figures in of variation among yields remains. Table 12.6 suggest that for millet the results In the tables for the model results, we with PNT are considerably better than those also report three results for each treatment. without. Also, the results without PNT are As the model took account of 30 years of real sometimes worse than the no-fertilizer con- rainfall data, it was able to generate a fre- trol, but it is not clear how much weight quency distribution of yields. Farmers can should be put on this outcome. There were expect the low yield or more for 23 or 24 two farmers involved, one who carried out years out of 30, the middle yield or more for all the trials with PNT (accounting for the 15 years out of 30 and the high yield or more cash costs in 2001) and another who imple- for only 7 or 8 years out of 30. mented them without PNT. Differences in In the tables for the farm trials we report soils or farm practices could account for annual costs of inputs for each year, as they some of the yield differences. Furthermore, too vary significantly from one year to the the negative results of using manure without next for some treatments. The costs of pur- PNT are not borne out either by the field chased inputs are those that cannot be pro- trials for sorghum or the modelling, results vided on farm. Here this means chemical for which are shown in Tables 12.7 and 12.8. fertilizer, PNT and pesticide treatment for Because of the cost of PNT and the value cowpeas. Some farmers might purchase attributed to manure, the overall pre-harvest seed and animal manure, but typically they costs noted here are very high in the first obtain these out of their own resources. year for several of the treatments, especially Cash costs are important because some where manure is priced at the value arrived farmers are cash poor, particularly at plant- at in Table 12.4. In fact, in some instances ing time. Costs, excluding harvest, include total pre-harvest costs are higher than those 166 P. Wyeth et al. ϩ 5 t (20,955) ϩ 2 t ϩ 5 t ϩ C man. C man. SR man. SR man. 120 360120 560458 390 (63,946) 10,258 21,174 21,184 61,365 58,488 33,019 30,966 Unim- C man. C man. SR man.man. SR Unim- 2 t proved 2 tt 5 t 2 t 5 proved 300PNT 300PNT 300PNT 300PNT 2003Avg. 36,897 (2,653) 9,727 (7,245) 30,936 (6,584) 16,627 (3,474)36,897 23,781 (18,080) 68,095 16,239 104,068 28,711 73,305 87,375 61,384 41,055 37,343 Avg. 9,727 8,742 29,949 18,651 358 16,239 44,698 74,706 63,181 55,781 20022003Avg. 0 0 0 625 672 1,552594 0 734 672 0 500 2003 296 495 677 672 350 411 973 1,55250,348 14,375 14,375 62,626 14,375 14,375 2003 1,2251,333 1,781 50,3481,077 1,244 14,375 0 0 0 0 0 0 0 62,626 14,375 0 0 0 Millet: animal manure – trials. manure Millet: animal ield (kg/ha) 2001 217 859 1,437 1,281 734 258 1,006 1,562 1,562 1,441 able 12.6. otal cost excl. 2001 14,075 50,048 102,508 62,326 133,203 14,075 73,673 126,133 85,951 156,828 (FCFA) 2002 (14,075) (50,048) (14,075) (62,326) (14,075) 1,563 (3,136) 58,900 (11,504) 45,608 Full cost manure Half cost manure with profitable 208% more manure, practice with full cost than the unimproved profitable is on average 71% more treatment the most profitable Without PNT, half priced manure. with half profitable more 359% practice with full cost manure, unimproved than the profitable is on average 278% more treatment the most profitable With PNT, priced manure. T Y T harvest (FCFA) Cost purchased 2002inputs (FCFA) 2001 14,075 Avg. days labour 2002 2001Profit/loss 50,048 p/day 0 14,075 6,359 0 62,326 9 0 14,075 0 14,075 11 0 50,048 0 10 14,075 0 62,326 11 0 14,075 0 10 0 0 9 0 22,500 22,500 12 0 22,500 11 22,500 0 12 0 11 0 Economic Evaluation of Alternative Soil Management Technologies 167

Table 12.7. Sorghum: animal manure – trials.

Unim- 2 t CM 5 t CM Unim- 2 t CM 5 t SRM proved each yr 3 yrs proved each yr each yr

Yield (kg/ha) 2001 688 1,469 1,375 2002 56 220 280 280 1,040 1,140 2003 594 828 819 594 1,313 1,221 Average 446 839 825 437 1,176 1,180 Total cost excl. 2001 14,450 50,423 102,883 harvest (FCFA) 2002 14,450 50,423 15,450 14,450 50,423 62,701 2003 14,375 50,348 14,375 14,375 50,348 62,626 Cost purchased 2001 0 0 0 0 0 0 inputs (FCFA) 2002 0 0 0 0 0 0 2003 0 0 0 0 0 0 Avg. days labour p/day 10 12 11 10 12 12 Profit/loss 2001 50,390 88,098 26,797 (FCFA) 2002 (7,153) (21,754) 22,038 22,038 85,102 85,855 2003 30,936 12,848 48,106 30,936 49,812 30,532 Full cost manure Average 24,724 26,397 32,313 26,487 67,457 58,194 Half cost manure Average 24,724 42,384 45,636 26,487 83,444 80,319

for the recommended rates of chemical fer- or level of profit, or more, for the stated per- tilizer (see below). centage of seasons. The differences between Among the manure treatments for using manure and not using it increase from millet, and taking the overall average of the the 75th to the 50th to the 25th percentile as three years, the most favourable results in the rainfall that these percentiles are based the trials were obtained with 5 t of cattle on increases. manure spread at the beginning of the three Figure 12.1 illustrates the profits over year period. (Yields were as good with 2 t of 30 years for the unimproved control and for small ruminant manure per year, but profits small ruminant manure. (The cattle manure were not as good because of poor rainfall in results were omitted so as not to further clut- the second year, when the 2 t were spread ter the figure, but the results are similar to and the 5 t were not.) Farmers expected the those for small ruminant manure. The prof- results from small ruminant manure to be its graphed are those calculated using better than those from cattle manure, manure at half its full nutrient value.) For because they are aware that small ruminant the treatments requiring 5 t of manure every manure is richer in nutrients. This advan- 3 years, the downward spike in profits at tage does not show up consistently in trial those times is very evident. However, the yields, and although it did in the modelling general trend for each of the treatments is (see below), the higher cost attributed to the similar. The trend for ‘unimproved’ millet, small ruminant manure because of its nutri- on the other hand, is clearly downward. ent superiority reduced or wiped out any superiority in profits. Table 12.8 shows the modelling results. Treatment set 4: millet after animals have These are consistent with those from the been corralled on the field before trials in that results of all four treatments planting in the first year with manure are of the same magnitude and better than the no-manure control. As men- The first conclusion in this treatment set is tioned earlier, the way to interpret the per- well in line with expectations and holds centiles is that farmers can expect that yield every year: the more nights the animals 168 P. Wyeth et al.

Table 12.8. Millet: animal manure – model results.

Unimpoved C man. 2t C man. 5t SR man. 2t SR man. 5t

Yield (kg/ha) 75% 418 756 720 818 765 50% 505 991 875 1,057 959 25% 642 1,196 1,050 1,303 1,123 Total cost excl. harvest 825 32,798 27,469 45,076 37,701 Cost purchased inputs 0 0 0 0 0 Avg. days of labour p/day 9 11 10 11 10 Profit/loss (FCFA) 75% 25,303 21,258 24,314 14,790 18,333 Full cost manure 50% 33,574 43,381 38,942 37,319 36,627 25% 46,476 62,758 55,441 60,535 52,091 Half cost manure 75% 25,303 37,244 37,636 36,915 36,771 50% 33,574 59,367 52,264 59,445 55,065 25% 46,476 78,744 68,763 82,660 70,529

The most profitable treatment with median rainfall is 29% more profitable than the unimproved practice with full cost manure, 77% more profitable with half cost manure.

140,000

120,000

100,000

80,000

60,000 FCFA

40,000

20,000

0 0 5 10 15 Years 20 25 30 35 (20,000) Unimproved SRM 2t SRM 5t/3

Log. (Unimproved) Log. (SRM 2t) Log. (SRM 5t/3) (40,000) Fig. 12.1. Millet profits with organic fertilizer.

spend on the fields the higher the yields and presumably because of the higher rainfall in the profits. The second more interesting the third year. Profits were lower that third conclusion is that this result holds even 3 year only because the price of millet was years after the treatment was applied, which FCFA 90 rather than FCFA 150. Once again, considerably reinforces the value of the prac- there is not much to choose between the tice. On the fields that had cattle, yields were treatment with cattle and that with small even higher in the third year than the second, ruminants. (There are only 2 years’ worth of Economic Evaluation of Alternative Soil Management Technologies 169 (Small rum. 3333 head) Unim- 64,667 108,073 (Cattle 1667 head) 10 11 12 12 10 9 14 28,626 9 Unim- proved 2 nights nights 5 10 nights proved 2 nights 5 nights 10 nights 2003 14,242 23,781 47,62914,242 53,113 81,016 42,859 57,168 Average 26,563 43,184 61,40928,790 51,385 102,537 106,988 77,711 20022003Average1,300 1,000 440 560 700 500 1,120 2003 660 1,119 408 556 385 692 1,248 875 782 375 375 500 884 1,250 750 813 938 2003 14,375 14,375 14,37514,375 14,375 14,375 14,375 14,375 0 0 0 0 0 0 0 0 Millet: night corralling – trials. Millet: night corralling ield (kg/ha) 2001 219 469 875 1,375 able 12.9. otal cost excl. 2001 14,075 15,575 17,825 21,575 Profit/loss 2001Profit/loss (FCFA) 6,548 2002 58,900 77,144 71,931147,906 112,563 43,338 49,656 131,875 For cattle, the most profitable treatment is on average 303% more profitable than the unimproved practice. than the unimproved profitable 303% more is on average treatment For cattle, the most profitable practice. than the unimproved profitable is on average 256% more treatment For small ruminants, the most profitable Y T harvest (FCFA) Cost purchased 2002inputs (FCFA) 2001Avg. days labour 2002 14,075 p/day 14,075 0 0 14,075 0 14,075 0 14,000 0 15,500 0 17,750 0 0 21,500 0 0 0 0 0 0 0 0 T 170 P. Wyeth et al.

Table 12.10. Millet: microdoses of chemical fertilizer – trials.

Unimproved DAP DAPϩUrea

Yield (kg/ha) 2003 463 1,190 1,257 Total cost excl. harvest 2003 14,375 20,325 23,075 Cost purchased inputs 2003 0 5,200 7,200 Avg. days of labour p/day 9 10 11 Profit/loss (FCFA) 2003 20,983 70,487 72,824

The most profitable treatment is on average 247% more profitable than the unimproved practice.

Table 12.11. Sorghum: microdoses of chemical fertilizer – trials.

Unimproved NPK Unimproved NPK/ DAP 1 NPK/ DAP 2

Yield (kg/ha) 2001 688 1,256 442 1,031 906 2002 56 520 56 510 20 2003 594 875 594 1,063 641 Avg. 446 884 364 868 522 Total cost excl. 2001 14,450 28,700 14,450 28,700 28,700 harvest (FCFA) 2002 14,450 30,800 14,450 20,400 20,400 2003 14,375 30,725 14,375 20,325 20,325 Cost purchased 2001 0 13,500 0 13,500 13,500 inputs (FCFA) 2002 0 15,600 0 5,200 5,200 2003 0 15,600 0 5,200 5,200 Avg. days labour p/day 10 11 10 11 11 Profit/loss 2001 50,390 89,780 27,236 68,536 56,747 (FCFA) 2002 (7,153) 36,963 (7,153) 46,059 (17,794) 2003 30,936 36,048 30,936 60,757 28,563 Avg. 24,724 54,264 17,006 58,451 22,505

For the trials on the right, DAP was substituted for NPK in the second and third years because it is cheaper. The NPK/DAP 1 treatment was at applied at planting, the NPK/DAP 2 treatment was applied at tillering.

data on the latter because this treatment was shows that microdosing also paid off for first applied in the second year.) sorghum, though not as consistently. In fact, the application at tillage only did better than the unimproved control in 1 year. Treatment sets 5 and 6: millet and sorghum The modelling results include not with microdoses of chemical fertilizer only DAP and urea but also two treatments with microdoses of NPK (15-15-15), the There was only one year of trial results for first being applied at planting and the millet using microdoses of chemical fertil- second at tillage. According to the results in izer. (It was tried in 2002, but there was a Table 12.12 and Fig. 12.2 there is little to complete crop failure due to low rainfall.) choose between the three. NPK is cheaper The result was that microdosing clearly did than DAP and slightly more expensive than well in terms of both physical yield and urea (see Table 12.3), but much more is used profit, and better than the control. The yield (6 g of NPK per pocket instead of 2 g of DAP using urea (46-0-0) as well as DAP (18-46-0) and 1 g of urea), so the DAP–urea combina- was sufficiently higher than that using DAP tion is slightly more profitable. The model alone to pay for the urea and the labour to calculated yields for the recommended spread it, but not much more. Table 12.11 quantities of DAP and urea (100 kg of each, Economic Evaluation of Alternative Soil Management Technologies 171

Table. 12.12. Millet: chemical fertilizer – model results.

Microdoses Recom. rate Unimproved NPK1 NPK2 DAPϩurea DAPϩurea

Yield (kg/ha) 75% 418 707 745 733 1,000 50% 505 817 840 809 1,358 25% 642 975 978 966 1,603 Total cost excl. harvest 14,075 28,325 28,325 22,775 57,325 Cost purchased inputs 0 13,500 13,500 7,200 42,500 Avg. days of labour p/day 9 10 10 11 10 Profit/loss (FCFA) 75% 25,303 38,401 41,923 46,353 37,032 50% 33,574 48,718 50,852 53,549 70,749 25% 46,476 63,613 63,867 68,317 93,837

DAPϩurea microdosing (with median rainfall) is 59% more profitable than the unimproved practice.

180,000

160,000

140,000

120,000

100,000 FCFA

80,000

60,000

40,000

20,000

0

0 5 10 15 Years 20 25 30 35 Unimproved Micro DAP urea Recommended Log. (unimproved) Log. (Micro DAP urea) Log. (recommended)

Fig. 12.2. Millet profits with chemical fertilizer. the first at planting and the second at tillage). still downward. Microdosing is effective at This was not tried in the fields, on the raising short term profits, but not to sustain grounds that few farmers could afford the yields and profits over a long period. cash outlay but, if any could, the model shows it would be clearly the most profitable. Similar conclusions are implicit in Treatment set 7: cowpeas and millet Fig. 12.2. The curves and the trends for intercropped microdosing and for recommended rates are clearly above that for the unimproved Table 12.13 shows the results from the inter- control. What is interesting and significant cropping trials. The control was millet on is that the microdosing trend is clearly its own and the treatments intercropped 172 P. Wyeth et al. Without PNT 300 kg PNT 600 kg PNT 380 0 0380 0 0 660 0 660 0 0 Unim- Total C’pea Cow- C’pea Total Cow- C’pea Total proved Millet C’peas residue grain Millet peas residue grain Millet peas residue grain 20022003 460 53203 22 750 244 216 47 175 313 57 2003 14,375 7,975 14,688 2003 22,663 7,975 20,538 0 0 6,000 2003 28,513 7,975 20,538 42,859 7,526 (9,868) 6,000(2,342) 5,488 (10,210) 8,480 28,513 (1,730) 25,182 0 6,000 15,873 (8,090) 18,080 7,783 6,000 0 6,000 6,000 Average 47,995 32,474 (3,765)28,709 23,164 45,416 2,370 40,947 47,787 36,209 (4,538)31,671 39,712 Average 648 392 47 216 537 98 383 486 87 371 Millet and cowpeas: intercropping – trials. intercropping Millet and cowpeas: excl. 2001 14,075 14,888 7,825 22,713 19,638 30,100 31,075 41,538 49,738 72,613 (FCFA) 2002 14,000 7,788 18,288 26,075 7,825 18,288 26,113 7,825 18,288 26,113 (FCFA) 2002 0 0 6,000 6,000 0 6,000 6,000 0 6,000 6,000 (kg/ha) 2001 734 594 120 594 734 248 906 484 205 937 cost purchased purchased 2001 0 0 6,000 6,000 11,250 17,250 22,500 28,500 28,500 51,000 ield able 12.13. otal esidues for fodder. Even the most profitable treatment is not, on average, more profitable than the unimproved practice, unless account is taken of the value of the cowpea practice, unless account is taken of the than the unimproved profitable is not, on average, more treatment Even the most profitable r Y T T harvest Cost inputs Avg. days labour p/day Profit/loss (FCFA) 2001 965,027 55,182 48,165 16,862 64,005 49,588 27,335 35,609 97,660 85,197 14,572 12,763 101,056 2002 6 45,944 41,731 (18,288) 6 23,444 0 78,181 (18,288) 0 12 59,894 78,181 (18,288) 6 0 59,894 6 12 6 6 12 Economic Evaluation of Alternative Soil Management Technologies 173

millet with cowpeas in alternating rows. roughly the same magnitude as those from One treatment was without PNT, one with the control, but in the third year the treat- 300 kg spread the first year, and one with ment results were much better than the con- 600 kg the first year. These results are the trol. The delayed reaction might have arisen least satisfactory among the treatment sets. because the PNT takes time to break down The millet half of the treatment performed into a form useable by the crop, but what- reasonably well in the sense that the millet ever the reason, the consequence is that yield was always greater on a per hectare the three year rotation with 300 kg of PNT basis than the pure millet yield in the con- performed best. It was better than either the trol. (In the table, the millet yields in the unimproved millet on millet or the rotation treatments are for half a hectare while the without PNT and, as was the case with control is a full hectare.) However, the cow- the intercropping techniques, it also did peas failed completely in the second year, better than the treatment with 600 kg. Again, presumably because of low rainfall, and adding the value of the cowpea residues were very poor in the third, in spite of excel- adds greatly to revenues and profits. lent rain. If profits are reckoned on the basis Although the model did not consider the of millet grain and cowpeas alone, the prof- application of PNT, it was able to look at not its are less for every treatment than they are only the 3 year cowpeas–millet– millet rota- for the control. The treatment with 300 kg of tion but also the 2 year cowpeas–millet–cow- PNT performed better than that for 600 kg, peas–millet variation. In Table 12.15, the left though it is not clear why. hand block of results shows millet under no If account is also taken of the value of rotation, 3 year rotation and 2 year rotation. the cowpea leaves or residues, the profit sit- The second block of results shows the uation for intercropping is rescued, as these same for cowpeas: cowpeas every year, every are quite valuable as fodder, around FCFA 3 years and every 2 years. The last two 125/kg, in monetary terms. Not a great deal columns on the right show the average of weight should be put on the precise fig- annual profits for the two rotations. ures for the residues, as the ratio of residue Millet yields are best with the 2 year to grain varies greatly among replications rotation and cowpea yields with the 3 year and treatments, but the general point rotation. Consumption preferences for remains that the value of the residues is an millet may lead farmers to prefer the 3 year important element in determining the value rotation, but the 2 year rotation is the more of the crop. profitable, not only because millet yields are better but also because cowpeas are more profitable as a crop than millet. Treatment set 8: cowpeas and millet In fact, according to the model, growing in rotation, with and without PNT cowpeas all the time is as good as any rota- in the first year tion from a cash return point of view. This is the one set of treatments where results from As with intercropping, rotation treatments the farm trials and modelling might be in included one without PNT, one with 300 kg conflict. In the intercropping trials cowpeas the first year, and one with 600 kg the first failed outright in the second year, and in the year. The trials shown in Table 12.14 used a third did poorly in spite of good rain. By 3 year, cowpea–millet–millet rotation. As contrast, the model suggests that year by the intercropping results just showed, the year yields of cowpeas are actually less vari- first year might have been the best year of able than yields of millet. It gives coeffi- the three to grow them. This was fortunate cients of variation of 19–22% for cowpeas for experimental purposes, but the treat- and of 40–48% for millet. If the opportunity ment might be risky, because if yields are presents itself, this discrepancy would be poor when the cowpeas are grown the tech- worth investigating to find out whether it is nique will be a loss maker. In the second the farmers who are too easily discouraged year, the yields from the treatments were of by cowpeas or the model is overlooking an 174 P. Wyeth et al. Total 15,382 28,551 600 PNT 191,367 ϩ 28,551 Total C’pea Total 61,427 15,382 90,555 300 PNT Rotation ϩ 187,755 244 280 1,375 563 58,521 61,427 150,938 C-M-M rotation without PNTC-M-M rotation C-M-M rotation Unim- Cowpea C’pea Total proved Millet Cowpeas residue grain Millet C’peas residue grain Millet residue C’peas grain 2003 14,375 14,3752003 0 0 14,375 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2003 (3,644) 5,896 5,896 90,555 20022003 244 214 266 141 Average240 321 1,400 350 1,742 540 450 1,775 281 362 Average 18,172 9,854 58,521 150,938 26,076191,367 22,115 56,568 25,482 54,138 61,427 187,755 15,382 Cowpeas and millet: rotation – trials. millet: rotation Cowpeas and ield (kg/ha) 2001 578 350 1,400 450 1,742 362 1,775 able 12.14. otal cost excl. 2001 14,075 34,250 34,250 0 34,500 0 34,500 0 57,000 0 57,000 (FCFA) 2002 17,721 13,812 13,812 17,721 17,721 22,413 22,413 T harvest (FCFA) 2002 Cost purchased inputs (FCFA) 14,075 2001 14,075Avg. days labour 2002 p/day 2001Profit/loss 0 12,000 40,438 0 58,521 9 0 0 9 14,075 0 9 0 12,000 0 0 0 0 0 9 0 0 9 0 11 0 0 0 0 0 0 11 0 0 0 9 0 0 0 11 0 0 0 11 0 The most profitable treatment is on average 211% more profitable than the unimproved practice. than the unimproved profitable on average 211% more is treatment The most profitable Y T Economic Evaluation of Alternative Soil Management Technologies 175

Table 12.15. Cowpeas and millet: rotation – model results.

(Cowpea (Cowpea (Cowpea (Cowpea (Cowpea (Cowpea –mil–mil) –mil) –mil–mil) –mil) –mil–mil) –mil) Millet Millet Millet C’peas C’peas C’peas All grain All grain

Yield (kg/ha) 75% 418 419 460 331 366 331 50% 505 550 628 379 430 375 25% 642 721 851 440 481 452 Total cost excl. 825 825 825 21,000 21,000 21,000 7,483 10,913 harvest Cost purchased 0 0 0 12,000 12,000 12,000 3,960 6,000 inputs Avg. days of p/day 9 9 9 10 10 10 910 labour Profit/loss 75% 25,303 25,442 29,317 52,790 62,114 52,790 37,544 45,715 (FCFA) 50% 33,574 37,798 45,179 65,441 79,090 64,388 51,425 62,135 25% 46,476 53,928 66,196 81,624 92,699 84,813 66,722 79,448

Difference between most profitable treatment and the profit of the unimproved practice (with median rainfall): 53% with the 3 year rotation, 85% with the 2 year rotation.

120,000

100,000

80,000

FCFA FCFA 60,000

40,000

20,000

0 0 5 10 15 Years 20 25 30 35

Unimproved Millet (cp-m-m) Log. (unimproved) Log. (millet (cp-m-m)) Fig. 12.3. Millet profits with rotations. aspect of the Madiama environment that is year rotation produces similar results after 30 not favourable to this crop. years. The upward spikes in the curve for the The most significant point about the rotation are generally, though not always, in rotations is evident in Fig. 12.3, namely that, those years when the higher priced cowpeas like microdosing, rotating cowpeas is not a are produced. Any positive impact on millet practice that sustains profits over time. To might be understated in that the model does keep it simple, the figure shows only the not take account of any beneficial reductions curves for millet-on-millet and the cow- in striga that the rotation might produce, peas–millet–millet 3 year rotation, but the 2 but such an impact would only eliminate 176 P. Wyeth et al.

occasional crop failures due to striga infesta- above in Table 12.7, as well as the modelling tions. It would not prevent millet yields from results presented in Table 12.8, animal steadily declining. manure in the absence of PNT gives clearly better returns than the control. The earlier analysis in this chapter and Comparisons Among Treatment Sets Figs 12.4 and 12.5 also suggest the following conclusions. In this section, the results of the various treat- ment sets are compared to see if one or another clearly stands out from the others, Profits generally follow yields either as especially profitable or as particu- larly inferior. Figure 12.4 shows the 3-year This means that higher yields generally pay averages for millet obtained from the trials. for the increased costs of obtaining them. Yields are measured in bars against the left (However, since this analysis has not shown hand y axis and profits are represented by the marginal returns to input use, there is no line graphs measured against the y axis on the indication of whether the amounts of inputs right. For the manure treatments, the solid used in the treatments here are economi- line shows profits calculated using the lower cally optimal.) manure cost while the dotted line shows profits figured at the higher cost per kilo. The only treatments that are clearly Comparisons between treatments worse than the average unimproved control hold up across low, medium and are those using animal manure without high rainfall years for most practices PNT. However, as already noted, the results for these treatments might not be trustwor- That is, if a treatment is better than another thy. In the trial results for sorghum shown in a low rainfall year, it is also likely to be so

1,400 120,000

1,200 100,000

1,000 80,000

800

Yield (kg) Yield (kg) 60,000 Profit (FCFA) Profit (FCFA) 600

40,000 400

20,000 200

0 0 M CM 2t SRM 2t CM 5t/3 SRM 5t/3 Intcrp C- Micro DAP Micro DAP Rot C-M-M Cor C 5 nts Cor C 2 nts CM 2t+PNT Unimproved Cor C 10 nts Cor C 10 nts Cor SR 5 nts Cor SR 5 nts Cor SR 2 nts CM 5t/3+PNT Cor SR 10 nts Cor SR 10 nts SRM 5t/3+PNT SRM 5t/3+PNT Micro DAP+urea Micro DAP+urea Intcrp C-M 300PNTIntcrp C-M 600PNTIntcrp C-M

Yield Profit 1 Profit 2 Rot C-M-M+300PNT Rot C-M-M+600PNT Fig. 12.4. Millet average profits and yields – trials. Economic Evaluation of Alternative Soil Management Technologies 177

1600 90,000

1400 80,000

70,000

1200 ) 60,000 1000 FCFA 50,000 ( 800 Yield (kg)

40,000 Profit 600 30,000 400 20,000

200 10,000

0 0

CM 2t SRM 2t Urea CM 5t/3 SRM 5t/3 Micro DAP Micro NPK 1 Micro NPK Unimproved Micro NPK 2 Micro NPK Millet (cp-m) Recommended Millet (cp-m-m)

Yield 1 Yield 2 Profit 1 Profit 2 Fig. 12.5. Millet yield and profit – first 10 and last 10 years.

in medium and high rainfall years as well. of the small amounts involved. Costs are (This conclusion derives from the tables much higher for animal manure, at FCFA shown earlier in the chapter, not from the 22,126 for 2 tonnes of small ruminant figures in this section.) manure (using half the full values attributed to fertilizer in Table 12.4) and FCFA 22,500 for 300 kg of PNT. However, the true oppor- The best results in terms of profit were tunity cost for the manure could be lower obtained from three treatments and the farmer would have to buy the PNT only every 3 years, usually for much less From the trials, the best results in terms of than a full hectare. It is also important to profit obtained from the three treatments (in note that although microdosing is more descending order of profitability) were: profitable than applying manure, modelling results suggest that the difference is not sus- 1. Corralling animals on the fields at night tainable. prior to planting. 2. Microdoses of chemical fertilizer. 3. Spreading animal manure on fields From the modelling, spreading animal with PNT. manure (without PNT) or the recommended Overall costs are low for the first two. For rates of fertilizer produce the most the corralling treatment there need be no sustainable results cash costs at all if the farmer can use his own animals or come to some arrangement with Figure 12.5 illustrates these conclusions. As a livestock owner to pay in some way other in Fig. 12.4, yield is measured against the y than a cash fee, for instance by allowing the axis on the left and profit against the one on animals to feed on crop residues. There are the right. The left-hand, darker shaded bar some unavoidable cash costs for the micro- for each treatment shows the average yield dosing treatments, but the cost per hectare is for the first 10 years of the 30 year period, only FCFA 7200 for DAP and urea because and the right hand, lighter shaded bar 178 P. Wyeth et al.

shows yield for the last 10 years. The solid own, it still results in more organic matter line shows average profits for the first 10 being incorporated into the soil. As empha- years and the dashed line that for the sized earlier, management of crop residues is second 10. an important part of farming. If farmers grow millet year after year It remains true that applying recom- without doing anything to conserve soil mended rates of chemical fertilizer is expen- fertility, the model says that yields and sive. Although the model takes account of profits in the last 10 years will be half those the high risk due to variable rainfall, apply- in the first ten. In contrast, if farmers use any ing chemical fertilizer at recommended or of the manure treatments or the recom- microdosing rates is a higher risk practice mended rates of fertilizer, their yields and than applying manure. In a drought year the profits will fall very little or not at all. There nitrogen in chemical fertilizer is largely lost, is not a lot to choose between 2 t every year whereas more of it remains tied down in or 5 tonnes every 3 years, nor between the manure, to the benefit of crops in subse- manure from cattle and small ruminants. quent years when rainfall is better. The model is clear that the full, recom- The lack of sustainability of microdos- mended rate of fertilizer performs the best. ing chemical fertilizer is clear. Yields start There is a common belief that animal out about the same as for spreading manure manure is better than chemical manure and profits start out higher, but over the last because of its high organic matter content 10 years both yields and profits are lower and, at higher rates of application than than those for the manure treatments. assumed here, no doubt that is true. The cowpea–millet crop rotations show However, the model assumes that 30% of the a pattern similar to that for microdosing, vegetable matter in the crop is returned to the except that yields do not start out as high and soil every year. Therefore, although chemi- both yields and profits fall to lower levels. cal fertilizer contains no organic matter of its Millet yields are no higher with rotations

Table 12.16. Millet: differences in average discounted profits between improved and unimproved practices.

Discount rate

0% 15% 25% 50%

FCFA Ratio FCFA Ratio FCFA Ratio FCFA Ratio

Unimproved 35,888 1.0 11,394 1.0 7,686 1.0 4,249 1.0 Cattle manure 2 Mt a year 58,468 1.6 13,666 1.2 8,709 1.1 4,619 1.1 Small ruminant manure 58,988 1.6 13,406 1.2 8,510 1.1 4,503 1.1 2 Mt a year Small ruminant manure 55,173 1.5 12,556 1.1 7,827 1.0 3,925 0.9 5 t every 3 years Cattle manure 5 Mt 53,854 1.5 12,776 1.1 8,056 1.0 4,117 1.0 every 3 years Recommended rates 72,653 2.0 16,805 1.5 10,943 1.4 6,054 1.4 chemical fertilizer Micro DAP urea 58,936 1.6 15,454 1.4 10,082 1.3 5,452 1.3 Micro NPK 2 54,973 1.5 14,529 1.3 9,523 1.2 5,180 1.2 Micro NPK 1 53,579 1.5 14,229 1.2 9,369 1.2 5,125 1.2 Millet (cp-m) 59,393 1.7 16,510 1.4 11,036 1.4 6,175 1.5 Millet (cp-m-m) 51,299 1.4 14,470 1.3 9,563 1.2 5,257 1.2 Economic Evaluation of Alternative Soil Management Technologies 179

than with unimproved monocropping, and practices clearly benefit the land whether it profits are only higher because the cowpeas belongs to a rich or a poor farmer, the per- are worth more. In other words, cowpeas ceived benefits are distinctly lower for the contribute to conserving soil fertility chiefly latter. This makes it especially important to through slowing down its decline by keep- educate poor farmers in the returns they can ing millet off the fields every so often. get from those improved practices that require the lowest outlays, not only in cash but also in kind and in labour. If poorer farmers discount the future at higher rates, they have less incentive to adopt improved practices Conclusion

Every producer discounts future profits to The combination of results from field trials some degree, meaning they value profits and computer modelling has yielded useful earned sooner more highly than the same insights into what farmers can do to profits earned later. Poorer producers are improve their soil fertility and raise generally supposed to discount the future at incomes. Without the trial results there higher rates, because their immediate needs would be doubt about the applicability of are more urgent. A more sustainable prac- the modelling and without the modelling tice, by definition, is one that shows good there could be some question about how returns in the future rather than the present robust the trial results are. and therefore, by its nature, not as attractive It is notable that the modelling results to a poor farmer as to a rich one. are less erratic than the field trial results. At Table 12.16 illustrates this phenome- least part of the explanation is that the non for the practices discussed in this chap- model does not include certain elements of ter. The leftmost pair of columns refers to risk, such as Striga infestations, floods, the average undiscounted profits over 30 shortages of labour, and invasions by ani- years for each practice. Since there is always mals that farmers face every year. We do not discussion about what the appropriate dis- wish to underestimate the harshness of count rate is, three, 15%, 25% and 50%, are farming conditions in Madiama, or the risk- applied in the next three pairs of columns. iness due to the unpredictable variations in At least the 25% and 50% rates would be rainfall. Nevertheless, the results of both the very high for a western farmer, but they are on-farm trials and the modelling indicate not unrealistic for poor farmers in the Sahel. that techniques such as corralling animals As the discount rate increases, the dis- on fields, spreading animal manure, or counted returns decline rapidly, and there is microdosing with chemical fertilizer pay also some reduction in the relative differ- off. These practices are within the means of ences between the returns to the treatments most farmers, and we hope that the docu- as shown by the columns of ratios. This mentation of relative costs and returns will table illustrates the common observation help farmers to be better able to choose that, while the actual benefits of sustainable among the alternatives.

13 Economic Linkages Among Occupational Groups in the Madiama Commune

Charlene Brewster,1 Daniel Kaboré,2 Michael Bertelsen3 and Peter Wyeth4 1Office of International Research, Education and Development, Virginia Tech, Blacksburg, VA 24061, USA; 2Institut de l’environnement et des recherches agricoles, Burkina Faso, 04 BP 7192, Ouagadougou, Burkina Faso; 3College of Agricultural and Life Sciences, Virginia Tech, Blacksburg, VA 24061–0334, USA; and 4Washington State University, Pullman, WA 99164–5121, USA

In the Niger Delta Region of Mali there is a resources. Therefore, resource management long history of shared resource use among and poverty alleviation and economic farmers, fishermen and pastoralists. Like growth must be addressed hand in hand. most areas of sub-Saharan Africa, the region Economic growth within the commune is is characterized by a combination of grow- one way of relieving the pressure on natural ing population pressure, low and declining resources, but this can only happen if such agricultural productivity and low soil fertil- growth does not preclude the alternatives of ity. With increasing natural resource deple- future generations. If it is to be sustainable, tion, the area has witnessed heightened economic growth must lead to sustainable conflict between affected groups over access use of natural resources. to cropping and grazing lands and water. In It is difficult to design and implement the Madiama Commune, the major river pro-growth policies that will positively (the Bani) supplies less water and cropping impact natural resource users unless link- activities suffer each year because of dec- ages within the local economies are properly reasing rainfall in the area. Crop yields have understood. The objective of this study is to declined significantly over the years, and model these linkages and analyse the poten- farmers have expanded cultivated lands to tial impact of targeted growth strategies on compensate for yield losses. The increased different groups of natural resource users. pressure on the natural resource base has Using information from the Participatory led to greater potential for conflict as fami- Landscape/Lifescape Appraisal (PLLA) and lies try to achieve better health, food secu- an extensive data set collected in the study rity and education for their members in area, a Social Accounting Matrix (SAM) the face of grinding poverty. This poverty model for the Madiama commune was itself is one of the greatest contributors to developed. The SAM analysis was used to: resource degradation. The poor do not (i) analyse the linkages among stakeholders invest in soil- and water-conserving tech- and institutions and (ii) trace the path nologies not only because they cannot afford through which growth strategies may affect to, but also because they discount the future different sectors and institutions within the more heavily than those with monetary commune. © CAB International 2005. Conflict, Social Capital and Managing Natural Resources (ed. K.M. Moore) 181 182 C. Brewster et al.

Economic Actors farmers have had to expand cultivated lands not only to compensate for the drastic yield The commune, the newly created and decline, but also to increase total production lowest decentralized administrative unit in in an attempt to keep pace with the increas- Mali, was chosen as the relevant geographi- ing population. This extensive agriculture cal survey unit. Within this study area, farm- (made possible by the availability of animal ers (Marka, Dogon and Bambara), herders traction) resulted in the virtual disappear- (Peul) and fishermen (Bozos) are now in per- ance of fallow, the traditional means of rest- manent or semi-permanent residence (Sada ing land and restoring soil fertility. Sy, 1994). In the past, most of the land in the Over time, unsuccessful fishermen have commune was flooded by the end of the become farmers (Lae and Weigel, 1994), rainy season, greatly benefiting agriculture increasing the pressure on agricultural (rice, sorghum) due to the enhancement of resources, especially cropland. Herders also soil fertility and provision of adequate mois- have progressively had more difficulty find- ture for plants. The government built a ing grazing land for animals during the crop- system of canals to capture floodwaters as ping season along the river and/or on the part of the effort to promote wetland reces- farmers’ fields. In addition, they have had sional rice cropping. Fishermen were able to less chance to prevent crop damage caused fish for nearly the whole year, because water by their animals because of larger fields that in the river never reached a level that pre- have developed along their migration vented it (Daget, 1994). Peul herdsmen routes. The depleted natural resource base followed the rains south during the rainy has resulted in high rates of livestock mor- season and from harvest time to early in the tality during drought with little opportunity cropping season moved north to take advan- for herd reconstitution. For these reasons, tage of crop residues (rice, sorghum in many herders have also become farmers, farmers’ fields), the bourgou (Echinocloa keeping all of their livestock manure for stignina – a wetland grass) and the pasture themselves. Similarly, bad crop yields over along the river when the water level was time have led farmers to a diversification low. Farmers cultivated their crops during strategy, adding livestock that remains on the rainy season and raised animals. Their farm year round to their enterprise mix. cattle were mostly entrusted to Peul herders Based on the history and social structure of who removed them from the area during the the region, four groups of natural resource rainy season. During the dry season, farmers users were characterized: benefited from manure deposited during the herds’ stay on their fields. A barter system 1. Farmers: agricultural production sys- usually existed between the different tems are the major activities of this group of classes of resource users: farmers trading individuals. cereals for milk or manure from the Peul 2. Agropastoralists: these individuals rear who gained water and pasture access while animals in conjunction with farming activi- the Bozos traded fish for the products of the ties, as part of a risk reduction strategy. others. 3. Sedentary pastoralists: their animals This pattern of natural resource use graze both within and outside of the com- began to change a few decades ago as a con- mune, but the household is sedentary. In sequence of climate variability and human addition, these individuals usually carry out activities that combined to progressively some type of farming activities. deplete the resource base. The Niger River, 4. Transhumant pastoralists: these indi- as well as one of its main tributaries, the Bani viduals travel inside and outside of the com- River, supplies less and less water each year mune seeking water and pasture for their because of decreasing rainfall and up-stream animals. Generally, they move around the dam activities. Rice farmers often experi- Niger River Delta region’s water sources ence severe water shortages that prevent rice during the dry season and then move south cropping. Crop yields have decreased and or east during the rainy season. Economic Linkages Among Occupational Groups 183

A 2-week PLLA in the commune pro- first group of data was related to household vided preliminary data that were used to characteristics, production and consump- select villages for more formal production tion, factors exchanged (labour, equipment, and income-expenditure structural surveys. land, money). The origin and source of fac- In-depth structural surveys were carried out tors exchanged were also recorded. Four in February–March and in September 1999. enumerators and two supervisors adminis- Information from the PLLA, as well as tered the questionnaire. Both the enumera- secondary data (Mali 1996 Census data), tors and the IER researcher attended a 2-day was used to select five study villages. These training session in Madiama in order to five of the ten villages in the area for study learn the objectives of the study, the SAM represent the demographic distribution of approach and the requirements of this kind stakeholders: Madiama, the headquarters of of survey. The survey was conducted over a the commune; Nerokoro, a pastoralist vil- 2-month period, took 3–4 hours for each lage; Promani, a village of sedentary and respondent, and was scheduled with a transhumant pastoralists and farmers; break after 2 hours for Muslim prayer or Tombonkan, a farmers’ village; and Tatia- lunchtime. Each enumerator surveyed a Nouna, a village of farmers and agropas- maximum of two households per day in toralists. The list of the households in the order to avoid weariness and preserve data commune was obtained for each village quality. from the 1996 census data available from The second group of data was collected government offices (Dept of Sofara, 1996). on microenterprise activities in the com- The sample size was 120 households ran- mune that are income generating activities, domly selected and representing about 10% such as food processing, handcrafts, retail of the commune households in 1996. The trade, livestock trading, cereal trading and so distribution of the four groups of stakehold- forth. This data was collected in September ers in the sample is presented in Table 13.1. 1999 on a sub-sample of 60 households Survey participants were randomly drawn from the larger sample of 120 house- selected from the village list. A meeting was holds. A pre-test was implemented with a held in each village with the farmers/ few respondents in order to: (i) correct/ herders in order to explain the objectives of improve the questionnaire and (ii) give enu- the study and to facilitate their patience and merators survey experience. The training, openness to the questions. the pre-test and the questionnaire improve- Two types of data were collected. The ment took 1 week (Kaboré et al., 2000).

Table 13.1. Sample distribution by activity and village, Madiama Commune, Mopti Region, 1999.

Group by main activity

Pastoralists Agro- Village Farmers pastoralists Sedentary Transhumant Total

Madiama 12 20 0 0 32 Promani 08 05 10 0 23 Tombonkan 09 01 0 0 10 Tatia-Nouna 14 06 04 0 24 Nerekoro 0 01 09 21 31 Total 43 33 23 21 120 (36) (27) (19) (18) (100)

Note: sample size as a percentage of total population is expressed in parentheses. 184 C. Brewster et al.

The SAM Framework and Multiplier assumed that expenditures are set indepen- Decomposition dently of income. These consist of: (i) gov- ernment, (ii) capital, and (iii) the rest of the A SAM model was developed to meet the world. By convention, receipts (incoming objectives of the study. The SAM model is a value added) to a sector are listed in the modified input/output model that shows rows while expenditures (outgoing) appear the flow of income and expenditure among in the columns (see Table 13.2). Table 13.2 production activities and economic actors. contains transformation matrices involving

It is a very flexible and powerful tool that the endogenous accounts: A11 gives the can be used to analyse economies in diverse intermediate input requirements, i.e. the

social and cultural settings (Taylor and input–output transaction matrix; A13 reflects Adelman, 1996). It has been used in analy- the expenditure pattern of the institutions sis of village economies (Subramanian, (households) for different commodities they

1988) as well as for nationwide studies consume from the production sector. A21 (Dorosh et al. 1991; Arndt et al. 1998; dispatches the value added generated by

Pradham et al., 1999). Given its flexibility, production activities by factor, while A32 the SAM can also be developed to address captures the income distribution to the var-

specific topics such as environmental issues ious institutions. A33 maps the income trans- (see Miller and Blair, 1985) or migration fers within and among households (Adelman et al., 1988). (institutions). The row totals received by

The SAM is organized as an accounting endogenous accounts are given by yn,which matrix of modeller-selected endogenous and consist of: (i) expenditures by endogenous

exogenous sectors’ inflows and outflows. It accounts (Ann) summed up as n and (ii) is based on the assumption that production expenditures by the exogenous accounts

activities are endogenous and demand- (Anx) summed up as x and referred to as ؉ ؍ driven. Endogenous accounts are those for injections. Finally yn n x and similarly, which changes in the level of expenditure income received by exogenous accounts yx ؉ ؍ directly follow any changes in income. The is yx l t. Since Table 3.2 is a SAM, rows ؅ ؍ endogenous accounts typically include: (i) and columns should balance; that is: yn yn ؅ ؅ ؅ ؍ production activities (the input–output sub- and yx yx where yn and yx are column ana- matrix), (ii) factors (labour) and (iii) institu- logues of row sums yn and yx, respectively. tions (households and firms). Exogenous For the purposes of the SAM analysis, accounts are typically those for which it is the commune economy is assumed to consist

Table 13.2. Simplified SAM structure.

Expenditures (outflows)

Endogenous Exogenous

Sum of other Receipts (inflows) Production Factors Institutions Accounts Totals

Endogenous

Production A11 – A13 x1 y1 Factors A21 –– x2 y2 Institutions – A32 A33 x3 y3 Exogenous ؅ ؅ ؅ Sum of other l1 l2 l3 tyx Accounts ؅ ؅ ؅ Totals y1 y2 y3 yx Economic Linkages Among Occupational Groups 185

of five account categories: production activi- tions in one sector or industry. It can pro- ties, commodities, factors, institutions, vide decision-makers (at the commune capital account and the rest of the world level) with information about the differen- (see Table 13.7). The community’s economy tial impacts of economic growth on various has been organized around ten main sectors within the commune. Three types of activities/commodities. These are cereals multipliers are calculated using the method other than rice (millet, sorghum, maize and developed by Pyatt and Round (1979) and fonio), rice; vegetables, legumes, livestock, illustrated by Holland and Wyeth (1993): fish, natural resources, retail trade, durables and government services (schools, dispen- 1. Direct effects multipliers (transfer saries and so forth). Activities/commodities effects): these capture the interactions result from an aggregation of agricultural within each category of accounts, e.g. inter- activities and their associated microenter- sectoral input–output elements. For exam- prises. For example, the livestock activity ple in the case of an exogenous shock such includes not only livestock production, but as an increase in the demand for a commod- also traded livestock and produced cereals ity (e.g. small ruminants), the direct effect and traded cereals are combined. Factors multiplier demonstrates how the shock include land, capital (equipment), hired affects other production activities in the labour, and family labour. Institutions con- commune. sist of farmers, agropastoralists, sedentary 2. Indirect or open loop multipliers: when pastoralists, transhumants and government an exogenous shock affects one sector, indi- services. The capital account includes rect multipliers illustrate how the impacts investment/saving in physical assets (agri- are transmitted to the other types of cultural and non-agricultural equipment) accounts; for example, they may show how and livestock investment. Table 13.7 outlines a shock in production affects factor the basic SAM structure for the commune. incomes. These are one-way effects, and any Exogenous shocks or changes intro- impacts on the originating sector are duced into the commune economy will alter excluded. For example, indirect multipliers the circular flows of resources among sec- may be used to illustrate the impact of tors within the commune. Examples of changes in production activities on factors exogenous variables include government such as labour, or changes in incomes of the input subsidies or taxes. The magnitudes of different occupational groups within the these impacts (measured by multipliers) commune. will depend upon the strength of the link- 3. Closed loop multipliers: this group of ages between the sectors. Similarly, the multipliers shows the full circular effect of impact (and multipliers) of a decision made a shock. They demonstrate how a shock to a within the Madiama commune will be sector travels outward to other sectors and related to endogenous variables such as then back to the original point of origin. changing land distribution among sectors or changing transhumance scheduling through commune taxation of grazing lands. The Economic Interdependencies within multipliers provide important information the Commune to commune level decision-makers on the prospects for, and differential impacts of, When describing the economic interactions economic growth on the various sectors. within the commune, two sectors can be dis- This has direct implications for who bene- tinguished. The tradable sector, as the name fits and loses from policy changes and, by suggests, consists of goods and services that extension, how conflicting sectors benefit are traded with regions external to the com- and how they might accordingly react to mune. The price of tradable goods is deter- changes. mined outside of the commune where there Multiplier analysis allows decision is a larger effective demand for these com- makers to fully trace the effects of interven- modities. The commune can sell all of its 186 C. Brewster et al.

production at this price. However, as pro- confined to the commune. Expansion of the duction rises, the cost of producing addi- non-tradable sector is constrained by low tional units of production also rises (upward levels of demand, rather than problems of sloping supply curve). If these costs rise supply. More of these goods and services can above the market price, then the commune be provided (their supply is elastic), but the will no longer be able to sell this commodity level of demand does not justify this. to the external world without sustaining An examination of the sales data for economic losses. Growth strategies aimed at primary agricultural products reveals the expanding the tradable sub-sector must, importance of exports in the trade of pri- therefore, focus on improving technologies mary livestock products (Fig. 13.1). The and management systems so that production livestock trade is particularly important to costs are lowered. The non-tradable sector sedentary pastoralists, and for this group consists of goods and services that are not over 50% of sales is to areas outside of the traded outside the wider area, although commune. The major tradable crop prod- there may be some trade within the com- ucts include rice and vegetable crops, pri- mune itself. In Madiama, the non-tradable marily gumbo (okra). goods include meat, fish, millet, milk, Figure 13.2 describes the pattern of sorghum and legumes. Labour and manure microenterprise activity in the commune. are non-tradable inputs, because their use is Sales in the tradable sector account for 20%

60 50 40

FCFA 30 Outside Commune 20 Million Different Village/Same 10 Commune Same Village 0 Farmer Ag Farmer AnimalAg-Pa Ag-Past AnimalSed-Past AgSed-Past AnimalTrans-Past Trans-Past Animal

st Ag

Ag

Group

Fig. 13.1. Sales of primary crop and animal products.

Same Village

Different Village/Same Commune Outside Commune

Fig. 13.2. Distribution of microenterprise sales volume in Madiama commune. Economic Linkages Among Occupational Groups 187

Table 13.3. Microenterprise activity for Madiama commune (’000 CFAF).

Type of activity Farmers Agro-pastoralists Sedentary pastoralists Transhumants

Same village sales 3025.6 21098.7 24525.9 Sales within commune – 2570.4 4541.0 18132.0 different village Sales outside commune 10710.0 3641.4 2934.5 1299.9 Totals 16306.0 29281.1 455924.4 1299.9

Table 13.4. Selected direct effect multipliers for Madiama commune.

Introduction of shock Sector of impact Own effect multiplier

Rice Rice 1.17 Small ruminants Small ruminants 0.16 Microenterprise Small ruminants 1.17 Microenterprise Large ruminants Large ruminants 0.19 Microenterprise Large ruminants 1.17 Microenterprise Poultry microenterprise Poultry microenterprise 1.37 of all microenterprise activity, with seden- Results and Discussion tary pastoralists accounting for the largest share of activity and transhumants the least The direct effect multipliers describe the (Table 13.3). impacts of a shock to one sector or industry, This pattern of trade suggests a direc- on the other sectors. Table 13.4 illustrates tion for growth strategy. Over the years, the effect of exogenous increases in demand development agencies have increasingly for selected commodities. Only impacts on looked towards the microenterprise sector sectors that are affected are shown and the as a vehicle for growth. The existence of a impacts on the other sectors are small. The tradable microenterprise sector (though rel- largest multipliers are those associated with atively small) suggests a starting point. One rice and livestock. Interpretation of these way of increasing incomes is to reduce the multipliers is straightforward. For example, supply constraints associated with the trad- if there was an increase in rice demand of able sector so that more commodities and 1 million FCAF, then interindustry trans- services can be produced and traded. Thus, actions alone would induce increases in research investments that focus on lowering output of 1.17 million FCAF. Increases in livestock production costs through, for demand (or investment) in the livestock and example, improved pasture management livestock microenterprise sectors would will be instrumental in raising incomes. also induce large increases in output. This highlights the importance of the Open loop effects are shown in Table SANREM pasture productivity research (see 13.5. This group of multipliers describes the Chapters 14 and 15). However, since much impact of a shock to one sector on other sec- of the commune economy still revolves tors outside of that block. The open loop around the non-tradable sector (millet, multipliers in Table 13.5 demonstrate the sorghum and legumes such as cowpea) impact of exogenous changes on incomes of research is needed to increase productivity stakeholders within the commune. For in these areas also. Much of the SANREM example, a 1 million CFAF increase soil fertility work is focused on the non- in demand for millet will result in increases tradable sector (see Chapters 11 and 12). in income of 216,000 CFAF for farmers, 188 C. Brewster et al.

Table 13.5. Selected open loop effects for Madiama commune.

Small Large Natural resources Millet Rice ruminants ruminants and food products

Farmers 0.216 0.199 0.230 0.221 0.203 Agropastoralists 0.355 0.340 0.395 0.380 0.269 Sedentary pastoralists 0.224 0.207 0.238 0.230 0.132 Transhumants 0.125 0.108 0.122 0.117 0.068

Table 13.6. Closed loop multipliers for institutions within Madiama commune.

Sedentary Farmers Agropastoralists pastoralists Transhumants

Farmers 1.34 0.34 0.34 0.38 Agropastoralists 0.58 1.58 0.58 0.66 Sedentary pastoralists 0.35 0.35 1.35 0.40 Transhumants 0.18 0.18 0.18 1.20

355,000 FCFA for agropastoralists, 224,000 13.6) and these results show that in general FCFA for sedentary pastoralists and 125,000 transhumants benefit the least from injec- FCFA for transhumant pastoralists. It can be tions of income into the system. seen that in general, the multipliers for tran- shumant pastoralists are significantly less than those for the other groups of stakehold- Conclusions ers. In addition, agropastoralists would ben- efit the most from increases in commodity The study illustrated the linkages between demand, followed by sedentary pastoralists, socioeconomic groups in the commune and farmers and then transhumants. It is also the potential to use SAM analysis for com- noteworthy that the open loop multipliers munity level economic studies. An analysis for the natural resources and food products of production and sales activity demon- category is large, illustrating the potential strated the relatively closed nature of the for economic growth through sustainable commune; i.e. the tradable sector is rela- exploitation of natural resources. tively small compared to the non-tradable Selected closed loop effects are shown sector. Impacts from exogenous changes in Table 13.6. These multipliers show how a in demand for the commune’s agricul- shock in one sector is transmitted through all tural and livestock production, as measured the other blocks and back to where it started. by multipliers, are shared differentially The multipliers in Table 13.6 show how the among socioeconomic groups with the tran- effects of changes in income of one group shumant group always benefiting least. will work through the system and impact This reinforces earlier conclusions that incomes of all groups in the commune. For interventions need to specifically target this example, an exogenous increase of 1 million group. FCAF in the income of farmers, will stimu- These results were discussed at face-to- late an additional increase of 34,000 FCAF to face meetings with community representa- farmers, 54,000 FCAF for agro-pastoralists, tives of the occupational groups. Leaders 35,000 FCAF for sedentary pastoralists and were intuitively aware of these results, since only 18,000 FCAF for transhumants. they already knew that the transhumant This lower level of impact for the tran- group is the least integrated into the shumants group holds in every case (Table commune economy. Given this knowledge, Economic Linkages Among Occupational Groups 189 ACTIVITIES Cereals Rice Veget Legumes Livestock Fish Retail NatRes Gvt Services Durables Social accounting matrix outline, Madiama commune, Mopti, Mali, 1999 – million FCFA. Mopti, Mali, 1999 – million Madiama commune, matrix outline, Social accounting eget 0 0 1.1493 0 0.074 0 0 0 0 0 egetCereals 0RiceV LegumesLivestockFish 0RetailNatRes 8.21Gvt ServicesDurables 0 0 0Family Labour 0 0 LabourHired 0 3.8535Capital(Eqpt) 28.05Land 0 0 324.2573 0 0 13.16603 0 0 472.4437 0 1.5 2.331826 65.53844 0 0 13.4241 0 0 0.27765 11.0417 0 0.05156 0 4.19 0 0 546.459586 0.63998 0 0.00125 0.3923 0 24.31131 0 0 26.468702 0 0 11.38 0 0 2.278 110.6296 0 2.4454 0.33446 0.907 0 3.3009 44.6285 2.576114 0 0 3.5513 0 0 0 0 0 0.179311 0 0 0 10.0975 0.135302 0 0 0 0 0 0 0 0 0.31598 0 77.2578 0.3 0 0 0 0.6771 0 0 0 0 0 0 0 0 55.6915 0 0 0 0 2.37549 0 0 1.107 0 0 0.5536 1.845 0 0 0 2.214 0 0 0 1.107 0.5535 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1.059 0 0 able 13.7. Activities CerealsRiceV LegumesLivestockMeatFishRetailNatResGvt ServicesDurables 0Commodities 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0Factors 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 T 190 C. Brewster et al. 0 0 0 0 0 0 1.059 8 3 . 7 ACTIVITIES 0000 0 0000 0 0000 0 0000 0 0000 0 2000 0 0000 COMMODITIES 00 00 Cereals Rice Veget Legumes Livestock Fish Retail NatRes Gvt Services Durables Cereals Rice Veget Legumes Livestock Fish Retail NatRes Gvt Services Durables (continued) rables ranshum 0 0 0 0 0 0 0 0 0 0 otal Institutions SavingsROW Imports 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Farmers 0Farmers AgroPastSedPastT Central Govt 0T 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 eget 0 0 70.9293 0 able 13.7. otals 352.458636 517.97345 70.9293 11.76971 648.588102 25.483701 243.714202 5.97639 7.3801 1.059 Capital Institutions T T Activities CerealsRiceV 195.9733 0 0 205.0028 0 0 0 0 Du LegumesLivestockFishRetailNatResGvt Services 0 0 0 0 0 0 0 0 0 0 0 0 0 0 11.7706 0 0 0 0 0 410.252 0 0 0 0 0 0 0 25.4837 0 0 0 68.4864 0 0 0 5.6764 0 0 0 0 0 0 Economic Linkages Among Occupational Groups 191 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0000 0 0000 0 0000 0 0000 0 0000 0 0000 0 0000 0 0000 0 0000 0 0000 0 0000 0 0000 0 0000 0 0000 0 0000 0 0000 0 0000 0 0000 0 0000 0 0000 0 eget 0 0 0 0 ranshum 0 0 0 0 otal Institutions Cereals 0 0 0 0 Rice 0 0 0 0 V Legumes 0 0 0 0 Livestock 0 0 0 0 Fish 0 0 0 0 Retail 0 0 0 0 NatRes 0 0 0 0 Gvt Services 0 0 0 0 Durables 0 0 0 0 Family Labour 0 0 0 0 Hired LabourHired 0 0 0 0 Capital(Eqpt) 0 0 0 0 Land 0 0 0 0 Farmers 0 0 0 0 AgroPast 0 0 0 0 SedPast 0 0 0 0 T Central Govt 0 0 0 0 T Savings 0 0 0 0 ROW Imports 313.1809 20.76969 3.795578 59.96981 276.26701 102.675597 315.431425 48.3300977 0 59.45546 otals 509.1542 225.77249 74.724878 71.74041 686.51921 128.159297 383.917825 54.0064977 7.38 60.51446 T Commodities Factors Institutions Capital 192 C. Brewster et al. 8 8 3 3 . . 7 7 1.059 5.9764 70.9293 11.7706 25.4837 2.91798 6.134663 2201.8893 1885.3334 1557.9825 53.111642 3 . 0 156.4853 352.4586 312.9707 517.9735 238.3359 648.5881 175.2278 243.7142 0 0 00 00 0 00 0 0 00 00 00 00 00 00 00 59.4555 0 60.5145 0 0 0 0 0 0 0 0 0 0 0 8 0 0 0 0 3 . 7 Central Institutions Capital ROW Total 00000 00000 00000 00000 00000 00000 00000 00000 00000 00000 00000 00000 00000 00000 00000 00000 Capital ACTORS F 000 Family Hired Hired Family Labour Labour (Eqpt) Land Farmers AgroPast SedPast Transhum Gvt Exports (continued) rables eget 0 0 0 0 23.1158 42.0647 1.07559 7.245488 0 0 0 74.724878 eget 0 0 0 Family Labour 0 0 0 Du CerealsRiceV LegumesLivestockMeatFishRetailNatRes 0Gvt Services 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 226.5395 0 0 204.3053 0 0 0 98.40711 58.83713 0 0 47.2603 75.19646 49.6067 33.76504 5.3408 10.35212 0 15.75832 134.2907 0 25.20564 16.644515 3.75848 96.7902 148.525 8.810493 5.2396 0 17.6468 71.15771 3.456024 0.4127 76.248 0 3.9249 62.2226 0 0 0 3.0013127 0 278.2808 15.382225 0 0 0 0 0 0 509.15422 0 686.51921 0 0 225.77249 0 0 0 128.1593 71.74041 0 0 54.006498 383.91783 Hired LabourHired 0 0 0 Capital(Eqpt) 0 0 0 Land 0 0 0 able 13.7. Activities Cereals 0 0 0 T Rice 0 0 0 V Legumes 0 0 0 Livestock 0 0 0 Fish 0 0 0 Retail 0 0 0 NatRes 0 0 0 Gvt Services 0 0 0 Durables 0 0 0 Factors Commodities Economic Linkages Among Occupational Groups 193 1620.1468 2139.2328 Institutions ranshum 158.032308 4.4614 0.4044 0 2.52381 3.566856 3.91458 5.78819 0 0 50.06093 228.75247 otal Farmers 516.117Farmers 16.1903AgroPast 2.6165SedPastT 1.1Central GovtT 570.503 31.4773 313.3308 17.775 14.6849 2.7143 0.3995 2.89231 0.51798 0.25796 1.3 0 0.97222 0.22309 2.319956 2.7675 0 19.68656 32.56083 0 1.336477 0.59809 0 0.22309 0 0 0 143.7609 714.63536 0 3.31 0 4.201837 74.7116 0 427.22161 129.3527 1.914634 758.5329 0.664 0 0 0 10.090471 SavingsROW Imports 0 0 0 0 0 0 0 60.9544 0 131.90265 102.71844 8.9242 104.4368 0.790075 2.7105 6.05726 0.572105 0 0 64.9865 0 402.72279 0 1281.2057 otals 1557.983108 53.1116 6.1347 2.91798 714.635425 758.532914 427.221594 228.752545 10.0905 402.7228 1281.2058 9530.5309 T Institutions Capital 194 C. Brewster et al.

however, they argued that perhaps the more District-level Socio-economic Linkages and important concept at hand was the notion of Growth: Towards Sustainable Natural Re- inter-relationships between the group, and source Management under Environmental the knowledge that ‘if one group benefits, Stress and Conflict in the Niger Delta Region then we all will benefit’ as stated by the tran- of Mali. Selected paper presented at AAEA Annual Meeting, 30 July–2 August 2000 at shumants representative. Tampa, Florida. Lae, R. and Weigel, J.-Y. (1994) Adaptabilité des pêcheurs aux changements environnemen- References taux et économiques. In: Quensière, J. (ed.) La pêche dans le Delta Central du Niger- Adelman, I., Taylor, J.E. and Vogel, S. (1988) Life Approche Pluridisciplinaire d’un Système in a Mexican Village: a SAM perspective. de Production Halieutique. I.E.R.-ORSTOM, Journal of Development Studies 25, 5–24. Karthala, Paris, France, pp. 295–310. Arndt, C., Jensen, T.H and Tarp, F. (1998) Struc- Miller, R.E. and Blair, P.D. (1985) Input–Output tural Characteristics of the Economy of Analysis: Foundations and Extensions. Mozambique: a SAM-based analysis. DERG, Prentice Hall, Englewood Cliffs, New Jersey. University of Copenhagen, Copenhagen, Pradham, B.K., Sahoo, A. and Saluja, M.R. (1999) Denmark. A Social Accounting Matrix for India, Daget, J. (1994) La pêche dans le Delta Central 1994–95. Economic and Political Weekly du Niger-Approche pluridisciplinaire d’un 34(48), 3378–3394. système de production halieutique, Avant- Pyatt, G. and Round, J.I. (1985) Accounting and propos, Jacques Quensière ed. IER-ORSTOM, fixed-price multipliers in a Social Account- Karthala, Paris, France. ing Matrix framework. In: Pyatt, G. and Department of Sofara. 1996 Census Unpublished, Round, J.I. (eds) Social Accounting Matrices: Djenné Cercle, Mali. a Basis for Planning. World Bank, Dorosh, P.A., Bernier, R., Randrianarivony, A.R. Washington, DC, pp. 186–206. and Rasolomanana, C. (1991) A Social Sada Sy, B. (1994) Préface. In: Quensière, J. (ed.) Accounting Matrix for Madagascar: La pêche dans le Delta Central du Niger- Methodology and Results (Working Paper Approche Pluridisciplinaire d’un Système #6). Cornell Food and Nutrition Policy de Production Halieutique. IER – ORSTOM, Program, Ithaca, New York. Karthala, Paris, France. Holland, D and Wyeth, P. (1993) SAM Multipliers: Subramanian, S. (1988) Production and Distri- Their Decomposition, Interpretation and bution in a Dry-land Village Economy Relationship to Input–Output Multipliers in the West Indian Deccan. Ph.D. disserta- (Research Bulletin XB1027). Washington tion. University of California, Berkeley, State University, College of Agriculture and California. Home Economics Research Center, Pullman, Taylor, J.E. and Adelman, I. (1996) Village Washington. Economies: the Design, Estimation, and Use Kabore, P.D., Traoré, B., Taylor, B.D., Bertelsen, of Village Wide Economic Models. Cambridge K.D. and Holland, D. (2000) Modeling University Press, Cambridge, UK. 14 Dry Season Feed Supplements: the Potential Role of Cassia tora

Meriem El Hadj,1 Ozzie Abaye,1 Amadou Kodio2 and Moussa Keïta2 1Department of Crop and Soil Environmental Science, Virginia Tech, Blacksburg, VA 24061–0404, USA; 2Centre Régional de la Recherche Agronomique, Institut d’Economie Rurale, B.P. 205, Mopti, Mali

During the summer of 2002, we sat down Cassia tora (C. tora, Cassia obtusifolia with women from three villages (Madiama, (Senna obtusifolia)) is a leguminous plant Siragourou and Nérékoro) within the that fixes atmospheric nitrogen. Hoveland et Madiama Commune, Mali. The objectives al. (1976) found that among other tropical of this meeting were to pursue questions on legumes C. occidentalis L. and C. obtusifolia pasture resources, a priority concern of the L. (sicklepod) were the most tolerant to low community, and to understand women’s available soil P. For this reason C. tora has an food production constraints. Almost all the important role to play in soil fertility women expressed interest in resolving live- enhancement. However, due to its aggres- stock feed problems, particularly for small sive nature that excludes grasses from being ruminants during the dry season. This was grown with it or in the surrounding areas, most important to the women in Nérékoro, and its unpalatability for animals, it imposes who are from a different ethnic group (Peul) more problems than potential benefits. In and, unlike the women from Madiama and most cases, grazing animals are used to con- Siragourou who help plant as well as har- trol undesired species in pasture by varying vest field crops, do not participate in field- the stocking rate. In this case, however, this work. As the Nérékoro women explained, option is not available since animals do not their responsibilities include raising chil- graze on fresh C. tora as it is unpalatable to dren, and buying and fattening small all livestock. Since the use of herbicide to ruminant animals for market and milk pro- control this plant is unlikely, some alterna- duction. In terms of fattening animals as tives are necessary to reduce its propagation well as increasing milk production, they or improve its quality for animal feed. said the critical period of feed shortage is Keeping in mind that the research goal during the dry season (November–May). was to increase forage resources and provide With this began an exploration of ways to a source of feed for the dry season, we preserve some of the forage plants for later decided to investigate the hypothesis that use in the dry season. When asked about ensiling C. tora could provide a nutritious available plant species that remain green source of feed. The objectives of the experi- late after the rainy season has passed that ment were: might be improved through ensiling, the women pointed to the Cassia tora that filled 1. To evaluate the chemical characteris- their degraded pastures. tics (crude protein, in vitro dry matter © CAB International 2005. Conflict, Social Capital and Managing Natural Resources (ed. K.M. Moore) 195 196 M. El Hadj et al. digestibility, fibre, etc.) of ensiled vs. fresh Invasive Species and Biological C. tora. Significance 2. To examine the effect of additives (water or honey) and other forages (grasses) Most plant species have evolved with preda- on the quality of the ensiled material. tors that eat them. This raises the question of what happens if someone takes a plant and moves it to a place where its predator does Identifying Characteristics and not exist. To keep its population under con- Distribution of C. tora trol, the plant needs its predator. When a plant is reproducing without control and Cassia (Senna) tora L., foetid cassia, C. (S.) pushing out native plant species, it has obtusifolia L., sicklepod, coffee weed, become a weed or an invasive exotic plant. etc., are very closely related annual Ehrenfeld (2003) sees the invasive species weedy shrubs belonging to the family not as a threat to biodiversity and ecosystem Caesalpiniaceae in the order Leguminosae stability, but rather as a potential for nutri- (Cock and Evans, 1984). Cassia is a weed ent cycling processes in the soil. Ehrenfeld of 26 crops in 67 countries and is more reviewed data on invasive species and soil prevalent in soybeans, groundnuts, pas- nutrient cycling and suggested that invasive tures, cotton and sugarcane than other plant species frequently increase N avail- crops. Cassia obtusifolia, commonly ability, alter N fixation, and produce litter known as coffeeweed, sicklepod, or coffee- with higher decomposition rates than coex- pod, is the most prevalent species of isting native species. Most of the data was Cassia, can reach heights of 1.8–2.1 m, and based on observational studies, and there is produces sickle-shaped seedpods. The very little information on the time since teardrop-shaped leaves grow in clusters of invasion, the causes of invasion and plant four to six leaflets. It is often confused with densities. Davis (2003) states that extinc- C. tora.Infact, it is difficult to distinguish tions often result from the combined effects between the two (Brenan, 1958). Some of multiple processes, such as over-harvest- investigators claim to have seen plants ing and habitat loss, and therefore, although intermediate between the two species. competition from introduced species might Bhandari (1978) treated C. tora as the syn- not be the primary cause of native species onym of C. obtusifolia. Others distin- extinction, it is a contributing factor. Davis guished the two on the basis of epidermal (2003) sustains the idea of dispersal by structure and seed analysis (Mall, 1957; deriving from Hubbell (2001) the theory of Pandey, 1970; Mathur, 1985). Randell ‘ecological drift’ as a positive factor of global (1995) discussed the possibility that C. tora biodiversity and that competition from evolved in Asia from C. obtusifolia. Cassia introduced species is not likely to be a obtusifolia is well documented as a weed common cause of extinctions of long- of groundnuts and soybeans in the south- term resident species at global, metacom- ern USA, and is widespread throughout munity, and even community levels. the tropical world. C. tora is becoming Chandrasekaran and Swamy (2002) studied increasingly prominent as a pasture weed the influence of an herbaceous weed com- in the southwestern Pacific, Asia and munity on natural secondary forest and Africa. De Wit (1955) characterized C. tora man-modified ecosystems. Man-modified as a weed species throughout the tropics, ecosystems showed significantly greater probably of South American origin. Singh weed density than the natural ecosystem; et al. (1970) characterized it as a minor the researchers attribute this difference to a weed of various crops but of greatest higher incidence of exotic weed species in importance as a pasture weed and sug- man-modified ecosystems due to a higher gested that it became established in degree of disturbance and open environ- response to overgrazing. ment. In addition, the sites invaded by the exotic weeds showed higher herbaceous Dry Season Feed Supplements: C. tora 197

litter production. Uhl and Jordan (1984) lower the toxicity. Cassia tora could be fed attribute the production of successional up to 10% level. Other researchers sug- forests to the nutrients released from the gested the presence of tannins and saponins decomposition of herbaceous litter. These in the seed of C. tora. Lower feed intake, studies suggest that exotic plant invasions poor growth response coupled with influence the ecosystem structure, species enlarged and fatty livers, reduced spleens composition and aboveground biomass. and haemolysed erythrocytes (destruction Cassia tora, if managed, can improve of red blood cells) suggested the presence of pasture productivity by providing more these toxins in Cassia seeds (Katoch et al., litter and increasing organic matter in the 1978; Katoch and Bhowmik, 1983). Chand soil as well as recycling essential nutrients. and Katoch (1983) reported that 0.1 M It can also provide a source of feed during sodium hydroxide treatments to C. tora the dry season when used as silage. meal resulted in 77.27% tannin extraction and 25% saponin extraction, and treatment with rectified spirit eliminated 35.61% Cassia as Forage and 58.33%, respectively. However, more research needs to be done in order to char- Although Cassia tora is considered a poiso- acterize the exact cause of Cassia toxicity, nous plant, the toxic element in Cassia has particularly because, as some herders in not been clearly defined. The seeds appear Madiama pointed out, the small ruminants to exert their toxicity upon the skeletal mus- eat C. tora in the pasture fresh or dry if cles, kidney and liver. The leaves and stem nothing better is available. also contain toxin, whether green or dry. Cassia tora, known as Chakwar in India The plant can poison animals if they con- and Jue-ming-zi (seeds of the legume) in sume it in the field, in green chop, in hay or China, has many virtues, attributed to the in grain containing the seed. Perkins and anthraquinones present within the plant, Payne (1985) suggest that anthraquinones and is used as a medicinal plant in these found in the leaves are the toxic com- countries as well as in Africa. The seed of pounds of this plant. Others report that the C. tora has physiological functions as an seeds of the plant contain emodin, other antiseptic, diuretic, diarrhoeal, antioxidant, anthraquinones and xanthones (Manjunath and antimutagen (Wu et al., 2001). The and Subbajois, 1930; Gupta and Sharma, water extract of C. tora has been used as a 1965; Rangaswami, 1963). Tiwari and health beverage in China. In India, C. tora is Behari (1972) have found anthraquinones in also utilized to feed livestock, because the roots of C. tora. Pal et al. (1977) have also Chakwar (C. tora) seeds have been found to found emodin from the leaves of C. tora. be very rich in protein. Patel et al. (1971) Toxicity has been observed in cattle and evaluated Cassia seeds in partial replace- broilers, and other animals are also suscep- ment of concentrate mixture from the ration tible to the effects of this plant. In cattle, of milking cows. Cassia seeds were boiled in diarrhoea is usually the observable first water, dried and ground. The results symptom of poisoning. Later, the animals go showed that 15% boiled Cassia seeds could off feed, appear lethargic, and tremors be included in the ration of cows without appear in the hind legs, indicating muscle any adverse effect, and a net profit per degeneration. As the symptoms progress, animal and per year was observed when the urine darkens to the colour of coffee, and using this ration. the animal becomes recumbent and is Singh et al. (2001) conducted an exper- unable to rise. Death often occurs within iment on broilers looking at the impact of 12 h after the animal goes down. C. tora seed on their growth and meat pro- In their study, Sood et al. (1990) used duction. The results indicated that C. tora C. tora seeds as a potential poultry feed. seed could be safely incorporated up to the The seeds were treated with sodium hydrox- 10% level in broiler mash for maximum ide followed by rectified spirit (alcohol) to return on least expenditure. This study 198 M. El Hadj et al.

confirmed results found previously by C. tora hay was found to contain 12.70, Pandit et al. (1979). These reports indicate 26.80, 1.76, 46.95, 2.41 and 0.64% of crude that Cassia seeds could be incorporated in protein, crude fibre, ether extract, nitrogen animal feed, but it would be even better if free extract, calcium and phosphorous, the entire plant could be utilized as a source respectively. The digestibility coefficients of feed. for organic nutrients in C. tora hay were sig- Previous studies by Gupta et al. (1970, nificantly more than the silage Gupta et al. as cited in Ranjhan et al., 1971) have shown (as cited in Ranjhan et al., 1971) reported that a neglected summer legume that is not earlier. The data for the chemical composi- accepted in the green stage, when conserved tion and digestibility of C. tora hay as into silage with or without molasses was compared to silage are given in Tables 14.1 readily accepted by the livestock and was and 14.2. quite nutritious. Ranjhan et al. reviewed the The silage concept is more prevalent in chemical composition and nutritive value of temperate regions, with their distinct sea- C. tora fed as hay and compared it to the C. sons, than in the tropics. Nevertheless, tora fed as a green plant and/or as silage. silage production in the tropics has become The green plant was collected at flowering more relevant to fulfil the forage needs of stage, air dried for a day, and stored in the smallholder farmers (FAO, 2000). Silage shed for a day until the dry matter content of making is less dependent on weather condi- the fodder reached approximately 75%. The tions than haymaking. More research needs hay was evaluated after a month and fed to to be done to determine the potential of C. adult sheep. Palatability was determined tora conservation as silage and to evaluate using four yearling rams and four 3-year-old Cassia nutrition and toxicity in silage. male buffalo calves. Gupta et al. (as cited in Ranjhan et al., 1971) observed that the fodder was palatable and could be com- Ensiling Cassia tora pared to the dry matter consumption observed on feeding C. tora silage. The The goal of silage preservation is conversion chemical composition and nutritive value of of moist, chopped forage with a short storage

Table 14.1. Per cent chemical composition of Chakunda hay as compared to Chakunda silage (from Ranjhan et al., 1971).

Dry Crude Ether Crude Nitrogen Treatment matter protein extract fibre Total free extract Calcium Phosphorus

Chakunda hay 84.50 12.70 1.76 26.80 11.79 46.95 2.41 0.64 Chakunda green 20.90 12.92 4.12 24.35 12.17 46.44 2.19 0.39 Chakunda silage 23.75 10.51 3.43 30.96 11.06 44.05 2.22 0.43 (without molasses) Chakunda silage 28.82 12.72 2.41 23.55 10.10 51.22 2.33 0.40 (with molasses)

Table 14.2. Digestibility coefficients of various nutrients in Chakunda (from Ranjhan et al., 1971).

Treatment Dry matter Crude protein Crude fibre Ether extract Nitrogen free extract

Chakunda hay 60.66Ϯ1.96 57.09Ϯ3.33 68.52Ϯ2.37 34.83Ϯ1.46 57.00Ϯ3.20 Chakunda silage 59.20Ϯ1.34 39.74Ϯ1.11 49.66Ϯ1.21 15.36Ϯ3.15 71.13Ϯ2.89 (without molasses) Chakunda silage 55.80Ϯ1.01 40.69Ϯ0.86 59.41Ϯ1.96 7.51Ϯ3.50 68.62Ϯ1.82 (with molasses) Dry Season Feed Supplements: C. tora 199

life to preserved forage that can be fed to in shallow layers, adequate compaction is livestock as needed. Most legume and grass obtained when the loose green material crops can be ensiled successfully. The key to under the person’s foot or animal’s hoof good silage management is air exclusion to does not exceed 2 cm. With wilted forages, obtain the best preservation of forage as adequate expelling of air can be expected silage. The production process of silage can only if the particles are less than 2 cm long. be divided in four stages: forage harvesting, The recommendation for longer pieces is to wilting, chopping, compacting and sealing add a final layer (10–15 cm) of fresh green (air-tightness). The proper execution of forage to weigh down the material before these stages has a big impact on the success closing the container. or failure of the fermentation and the quality Although harvesting forages at the right of the silage. stage of maturity and moisture content max- The most important treatment after har- imizes the nutritive value of the forage, sev- vesting the forage is chopping, followed by eral types of additives for enhancing the wilting and conditioning. Chopping is nec- quality of hay crop silage are available. The essary to obtain good compaction to exclude addition of honey, which was used instead air in order to promote a rapid initiation of of molasses, because it was readily available the microbiological processes of fermenta- in the village, can improve silage quality by tion. Chopping to between 2–4 cm lengths providing fermentable sugars. Fermentable has the additional benefit of ease of inges- sugars are an important factor in fermenta- tion, regurgitation and posterior rumination. tion, because forages, such as most legumes, Wilting and conditioning forage before including C. tora, are typically low in fer- ensiling is also highly beneficial to avoid mentable sugars and higher in buffering production of effluents1 or development of capacity. As a result, they have slower rates undesirable microorganisms. Dry matter of pH decline during ensiling, which trans- (DM) levels between 30% and 35% will pro- lates into poor fermentation of the fresh mote a better fermentation and an increased material. intake by animals. Anti-nutritional metabo- lites (e.g. tannins and alkaloids) in certain forages (e.g. herbaceous and woody legumes Materials and Methods and cassava leaves) will be eliminated or reduced. The same should be expected with Summer 2002 C. tora. The field-drying time required to reach an optimum DM content depends on Cassia tora was collected from three loca- the species and on the weather conditions. tions (Siragourou, Nérékoro and Madiama The time may vary from 4 to 24 hours villages) within Madiama Commune. Most depending on the thickness of the stems. of the C. tora plants were at an advanced The method of compaction depends on stage of maturity (more stem and fruits than the silo (cylindrical structure, pit or bag) leaves). Prior to ensiling the plant, samples dimensions. In vertical silos of 2 t or less, a were collected and separated into stem, leaf person walking over the successive layers of and seedpod, and dried to determine the forage can achieve compaction. In pit (hori- percentage dry weight of each component. zontal) silos, less than 4 metres wide, ani- Per cent stem, leaf and seedpod were 80%, mals or people walking over the material 19.6% and 0.4%, respectively (Table 14.3). can achieve compaction. Mud or water In a ‘normal’ growing season (in terms of accumulation around the silos must be rainfall and temperature), C. tora would be avoided to prevent contamination of the at an early to late flowering stage at the end forage, but water also provides weight for of October. However, during the 2002 grow- more effective packing. The main problem ing season only 430 mm of rainfall was with a small-scale container is to know recorded in Madiama, which was 350 mm when adequate compaction has been less than the previous year. Due to the obtained. In well-chopped forages, placed extreme drought conditions, the plants were 200 M. El Hadj et al.

Table 14.3. Percentage stem, leaf, and seedpod much lower for the Siragourou and Nérékoro of Cassia tora prior to ensiling, Madiama, October locations than for those of Madiama. To 2002. facilitate chopping for proper cutting length, which was done with a long sharp machete, Fresh Dry Dry Plant Plant weight weight matter parts the plant was placed on a wooden block. In parts (g) (g) (%) (%) Madiama, the women chopped the plant to a length of 0.5–1.5 cm; in the other locations, Stem 86.4 70.1 81 80 the villagers participated in the chopping Leaves 24.8 18.4 74 19.6 and the length varied between 1 and 3 cm. To Seedpod 5.0 0.3 60 0.4 enhance the quality of the silage and (fruit) improve its nutritional value, two treat- ments were incorporated prior to ensiling. The treatments consisted of fresh material forced to enter the reproductive stage earlier only (control), and fresh material with water in the season. or honey added. Since the fresh material was Wilting was not necessary, because the drier than desired for ensiling (less than plants had an adequate dry matter level. 70% moisture), the added water might help Prior to placing the chopped material in the facilitate the ensiling process (fermentation) bags that were used as silos, a sub-sample of and improve silage quality. fresh C. tora was taken from each location for Chopped samples from each location dry matter determination and chemical were placed in a 31ϫ51 cm, 4-mil polyeth- analysis. The moisture content of the ensiled ylene bag. Air was excluded from samples material varied slightly among locations. In in the bag by exerting weight on the plastic general, the moisture content of the fresh bags, and the top of the bags were folded, chopped plant (no treatment added) was gathered and wrapped with masking tape.

Fig. 14.1. Women chopping Cassia tora for ensiling. Dry Season Feed Supplements: C. tora 201

All samples were grouped by location and bags. The chopped materials were treated placed in a larger plastic bag to start the fer- and ensiled as they were at the Nérékoro mentation process for either 60 or 90 days. location. Like the Siragourou women, the After 60 or 90 days, the ensiled samples Madiama women came prepared with their were opened and immediately given visual, sharp machetes and asked a number of ques- colour and mould ratings, as well as sensory tions, such as why were they ensiling Cassia scores for sharpness (acidity) or off odours. tora, why was it necessary to chop it at this length, and why add additives. The women chopped the Cassia per instruction and Summer 2003 socialized for a while before they left. Since no difference in silage quality As in 2002, C. tora was harvested from between the 60 and 90 day ensiling period three locations (Madiama, Siragourou and was observed in 2002, in 2003 only a 60 day Nérékoro). In Nérékoro, the C. tora sample period was used to ensile the C. tora. After was brought to a central location for pro- 60 days, bags were opened and visually cessing. Most of the C. tora was at vegeta- assessed. Sub-samples were obtained from tive/early flower bud stage. Two men from each bag for quality analysis. the village did the initial chopping while the women reduced the size of the chopped material to a more desired size for ensiling. Laboratory analysis The length of the chopped material for the Nérékoro location ranged from 1 to 3 cm. Samples in 2002 were dried in a forced-air Prior to placing the chopped material in oven at 60°C, ground (1-mm screen) in a the bags, sub-samples of fresh C. tora stainless steel Wiley Mill,2 and analysed for were taken from each location for dry acid detergent fibre (ADF) (Van Soest, 1963), matter determination and chemical analy- neutral detergent fibre (NDF) (Van Soest and sis. Cassia was ensiled as is, with honey, or Wine, 1967; Goering and Van Soest, 1970), in with 20% chopped grasses. The grass vitro dry matter digestibility (IVDMD) components for this location were 40% (Tilley and Terry, 1963), and crude protein Dactyloctenium aegyptium, 10% Digitaria (CP).3 In 2003, samples were analysed for CP, longitudinalis, 20% Elucine indica and 30% ADF and total digestible nutrients (TDN). Panicum laetum. For the Siragourou loca- tion, the grass components were 50% Bracaria sp., 30% Elucine indica, and 20% Statistical analysis Dactyloctenium aegyptium, while at the Madiama village location, the components Three treatments (control, honey and water were 10% Elucine, 10% Panicum laetum, or grass) were applied at three levels. Level 0 40% Dactyloctenium aegyptium and 40% represents the fresh samples (not ensiled), Digitaria horizontalis. level 60 represents the samples ensiled for 60 The same treatments, ensiling and sam- days, and level 90 represents the samples pling procedures used at Nérékoro were ensiled for 90 days, done only in the first year used for both the Madiama and Siragourou (2002). A randomized complete block design locations. At the Siragourou location, ensil- (RCBD) was used to perform the analysis.4 ing and preparation of C. tora was done effi- ciently and effectively, because the women came prepared to do the job. They not only Results and Discussion chopped the C. tora at a 0.5 cm length (which is the most desired size for ensiling) Crude protein (CP) but also asked about what the experiment was expected to accomplish. After they fin- In 2002, C. tora CP varied from 9% to ished chopping, they volunteered to pack 10%. The addition of honey prior to ensil- the chopped materials into the small plastic ing slightly increased CP compared to the 202 M. El Hadj et al.

control or the water treatments (Fig. 14.2). the plant matures, the CP content declines, There was no significant difference between while the fibre components of the plant fresh and ensiled samples. The percentage of increase proportionally. In addition, the total CP in the August 2003 fresh and ensiled C. amount of rainfall for the 2003 growing tora was up to 50% higher than the October season exceeded the 1999 to 2002 average 2002 samples (Fig. 14.3). The average CP in seasonal rainfall. The total annual rainfall for the 2003 Cassia samples ranged from 1999, 2000, 2001, 2002 and 2003 were 602.1, 15–22%. This significantly higher CP con- 507.4, 678.1, 374.5 and 1130 mm, respec- tent in the 2003 C. tora was due to the fact tively. There was no difference in CP among that the fresh material ensiled in August was treatments in 2003, and fresh Cassia had sig- mostly leafy compared to the October 2002 nificantly higher CP than ensiled Cassia. The samples that were mainly stem (80%, 19.6%, CP values reported for 2002 are consistent 0.4% stem, leaf, seed pod, respectively) (see with the slightly higher values Ranjhan et al. Table 14.3). The 2003 C. tora prior to ensiling (1971) previously reported. Differences in was approximately 50% stem, 40% leaf and CP between 2002 and 2003 (significantly 10% immature seedpod. Leaves of grasses or higher values) may be attributed to the time legumes have more CP and less fibre com- the Cassia was harvested and the stage of pared with the lignified part of the plants. As maturity of the plant prior to ensiling.

12 10 8 Control 6 Honey

% CP 4 Water 2 0 06090 Period (in days) Fig. 14.2. Percentage CP in fresh vs. ensiled Cassia tora, October 2002.

25

20

15 Control Honey

% CP 10 With grass

5

0 Fresh Ensiled Fig. 14.3. Percentage CP in fresh vs. ensiled Cassia tora, August 2003. Dry Season Feed Supplements: C. tora 203

Neutral detergent fibre and acid detergent The percentage ADF of the ensiled C. tora fibre was much higher than the fresh C. tora (Fig. 14.5). The addition of honey to the Neutral detergent fibre (NDF) and acid deter- ensiled C. tora decreased ADF by three per- gent fibre (ADF) are two laboratory methods centage points (38% for the grass treated vs. that chemically distinguish the readily 34.6% for the honey treated silage). Overall, available and soluble cell contents from the for both 2002 and 2003, the ADF values of less digestible portion of the cell walls. NDF the ensiled C. tora were acceptable values to consists of cell walls while ADF represents maintain a reasonable production level. lignified cellulose (indigestible) fibres. NDF Ensiling C. tora decreased the amount is negatively related to forage intake by the of NDF compared with the fresh C. tora (Fig. animal, while ADF is negatively correlated 14.6). The ensiled Cassia with the addition to digestibility. NDF and ADF values of less of honey was significantly lower in NDF than 40% and 30%, respectively are consid- than the control and the water-treated C. ered prime values. As the plant matures tora. The percentage NDF was 57%, 50% from vegetative stages to flowers and seed and 52% for the untreated control, honey head production, the ADF and NDF values and water treated Cassia, respectively. NDF will increase proportionally. values of grass or grass legume hay, depend- In 2002, the ADF values ranged from ing on stage of maturity, range from 40 to 34% to 42% (Fig. 14.4). The combined effect 65%. As the plant matures from vegetative of ensiling and honey treatment reduced stages to reproductive stages, NDF values ADF values (Fig. 14.4), a trend also observed increase significantly. by Ranjhan et al. (1971). The addition of When compared to the fresh C. tora, honey to C. tora prior to ensiling decreased ensiling increased IVDMD (Fig. 14.7). ADF values by eight percentage points (42% Averaged over all treatments, the IVDMD of for the control vs. 34% for honey treated the ensiled Cassia was six percentage points silage). The ADF values observed for the higher than the fresh C. tora. The addition of ensiled C. tora are comparable to mature honey as an additive significantly improved lucerne hay. Generally, the ADF values of the IVDMD of the ensiled Cassia. The the fresh C. tora were much lower in 2003 IVDMD of the honey treated C. tora was 63% than in 2002 (Fig. 14.5). This was not sur- compared with 54% for Cassia tora ensiled prising, because the Cassia sample in fresh (with no additives). These values are August of 2003 was much higher in leaf (low comparable to typical warm and cool-season in ADF) than stem and seedpod content. forage crops grown in temperate regions.

45 40 35 30 Control 25 Honey 20 aaa a' b' a'b' a''b'' a''

% ADF Water 15 10 5 0 06090 Period (in days) Fig. 14.4. Percentage ADF in fresh vs. ensiled Cassia tora, October 2002. 204 M. El Hadj et al.

40

35

30

25 Control 20 Honey With grass 15 % ADF 10

5

0 Fresh Ensiled Fig. 14.5. Percentage ADF in fresh vs. ensiled Cassia tora, August 2003.

70 60 50 40 Control Honey 30 aaa a' b' b' a'' b'' a'' Water % NDF 20 10 0 06090 Period (in days) Fig. 14.6. Percentage NDF in fresh vs. ensiled Cassia tora, October 2002.

70 60 50

40 Control Honey Water 30 aaa a'b'a'b' a'' b'' a''b'' % IVDMD 20

10 AABB 0 06090 Period (in days) Fig. 14.7. Percentage IVDMD in fresh vs. ensiled Cassia tora, October 2002. Dry Season Feed Supplements: C. tora 205

Total digestible nutrient (TDN) is an Australia and other tropical regions have approximate measure of feed value based on demonstrated (Muhlback, 2000). In general, feed energy content. Energy is the most the August 2003 ensiled C. tora had a sub- yield- (milk, meat, wool) limiting factor in stantially higher percentage CP and lower livestock with mainly forage-based diet. The ADF. The August 2003 fresh Cassia was fresh Cassia was higher in TDN than the high in leaf content compared to the ensiled C. tora (Fig. 14.8). This might have October 2002 fresh C. tora, but unless the been due to the fact that the fresh material raw material (the fresh sample) is of a rea- was of a higher quality as indicated by its sonable quality, the ensiling process per se high leaf to stem ratio and some energy was would not enhance the quality of the end lost during the fermentation process. Within product. However, if the fresh material is the ensiled Cassia, the honey treated Cassia excessively dry and low in sugar, the addi- was higher in TDN than the control and tion of water or a carbohydrate source can grass treated Cassia. Although the addition enhance the quality of the materials by of grass slightly reduced the TDN values of improving the ensiling processes (such as the ensiled C. tora, this could be explained fermentation). by the advanced stage of maturity of the The goal for this project was to help grass added. Madiama Commune find a feed source they can tap into during the dry season when livestock feed is in short supply. In a much Summary drier year like 2002, the entire feed source for the livestock was gone at least 2 months In terms of the nutrient values, as indicated before the end of the ‘normal’ production by acceptable crude protein, fibre, rela- season, and C. tora was the only live and vis- tively high in vitro dry matter digestibility ible plant in the entire Commune. Cassia and total digestible energy values, ensiled tora even at its late stage of maturity could C. tora has a great deal of potential as a provide a good feed source with reasonably livestock feed. Although the October 2002 high nutrient content. ensiled C. tora had relatively low nutrient As for the question of toxicity, it is not values due to a low leaf to stem ratio of the clear how animals respond to a Cassia diet fresh C. tora,itwould still make a reason- as evidenced by the various studies. ably good feed for livestock. However, the Although anthraquinones are the major addition of honey, as the results in this toxic compounds found in C. tora, other study have shown, molasses or sugar toxins have been identified as well. improves the quality of the ensiled mater- However, both Gupta et al. (as cited in ial, as the results of similar studies in Ranjhan et al., 1971) and Ranjhan et al.

70 60 50 40 Control Honey 30 With grass % TDN 20 10 0 Fresh Ensiled Fig. 14.8. Percentage TDN in fresh vs. ensiled Cassia tora, August 2003. 206 M. El Hadj et al.

(1971) suggested that C. tora conserved as over-shade other species would reduce hay or silage was readily accepted by live- competition. If the plants were cut at differ- stock. Nevertheless, toxicity analyses of ent stages of maturity, it would allow some C. tora ensiled are needed, and a Brazilian to seed for regeneration. However, Hien scientist is exploring this question. (1997) states that C. tora produces heteroge- neous seeds that can extend the germination process over several weeks. Thus, if the Biological impact plants germinate and grow at different times, it would be hard to exterminate If making C. tora into silage was to be Cassia, even if it is cut at optimal times for adopted, it could affect the ecosystem within making silage this region. As discussed earlier, exotic species can bring both positive and negative effects into an existing ecosystem. Cassia Socio-economical impact tora may have positive effects in the recy- cling of nutrients; however it is also aggres- Livestock are central to the livelihoods and sive in its growth and excludes grasses and existing production systems in Madiama. other species from growing. In order to con- Field crops and livestock are the dominant trol C. tora without eliminating the plant, a production systems, with each generating cutting regime would be adequate. Ensiling 46.2% and 34.4% of the community income, C. tora could be a solution to two problems. respectively (Kabore, 2001). Livestock pro- If cut at the beginning of the growing season vides the commune as well as the country when still at the vegetative and leafy stage, with meat, milk, leather and wool products. the quality of the plant would be optimum Livestock exports to neighbouring coastal for conservation and feeding to small rumi- countries where livestock production costs nants during the dry season. In addition, it are much higher are an important source of would allow other species to emerge and national income. In Madiama, goats and persist early in the season. Because Cassia sheep are mostly kept in the villages and fed has a prolific seed production, not allowing by women. In Nérékoro, women are respon- all the plants to go to the reproductive stage sible for the fattening of animals and for milk would control its growth and would make it production. Increasing the feed supply, espe- ecologically more acceptable. In turn, a shift- cially during the dry season, would increase ing in plant population with an increase and meat, milk, skins and fibre production. coexistence of various species might occur. Brewster et al. (2001) have stated that Some might raise the question of the risk of any development strategy for Madiama exterminating C. tora over time if the plant is economy should focus on the livestock not allowed to enter the reproductive stage sector. Such a focus points to the impor- and regenerate. In order to prevent any loss tance of pasture management. Improving in biodiversity, stakeholders need to under- pasture management improves pasture pro- stand that cutting or grazing any plant has to ductivity and provides higher quantity and be done in a controlled and sustainable way, quality feed for livestock, thereby lowering and like any other plant, C. tora must be livestock production costs and raising the managed in a sustainable way. women’s income. The ensiling of C. tora Stage of maturity is critical to the man- requires few inputs and complicated tech- agement of C. tora both for reproduction and nologies, because labour is the most impor- ensilage. The results of the silage experi- tant input in this process. The use of locally ment show that even at a late stage of matu- available equipment and the reliability and rity, Cassia silage was of acceptable quality. repeatability of the ensiling process makes However, in order to allow other plants to Cassia silage a sustainable source of feed. A germinate and establish within the pasture study shows that crossbred cattle can at the plant community, cutting Cassia at a vege- very least be maintained and possibly gain tative stage when it is not tall enough to weight during the dry period in Guinea Dry Season Feed Supplements: C. tora 207

savanna zone by being fed ensiled cereal Technicon Autoanalyzer (Technicon Indus- crop residues (Olayiwole and Olorunju, trial Systems, Tarrytown, NY; 1976). unknown date). 4 Analysis of variance (ANOVA) procedures An economic analysis is necessary to were performed on all data to test the treat- characterize the potential economic benefits ment effects on various measured parameters using the GLM procedure of SAS (SAS Inst., of Cassia silage. This feeding option would 2001). The treatment means were compared be a supplementary feed that could give using least significant difference (LSD) at rapid and significant returns on invest- Pϭ0.05. ments. In addition it is a reliable and repeat- able process that could be done with any forage of relatively good quality. In conclu- References sion, Cassia silage could be a sustainable and economically feasible solution to feed Bhandari, M.M. (1978) Flora of the Indian Desert shortages during the dry season in Scientific (United Book Traders, 4, 471p) Madiama. The social and economic impacts Scientific Publishers, Jodhpur, India. of using a simple tool could far exceed the Brenan, J.P.M. (1958) New and noteworthy cassias inputs for Cassia silage. If the net benefits of from tropical Africa. Kew Bulletin 13, using a production system with Cassia 231–252. silage exceed the net benefits of using a pro- Brewster, C., Kabore, D., Bertelsen, M., Wyeth, P. (2001) Economic interdependence among duction system without Cassia silage, then socio-economic groups in Madiama com- Cassia silage is economically beneficial. mune of the Niger delta of Mali. SANREM This experiment is the first step. The next CRSP Research Scientific Synthesis step would be to conduct animal feeding Conference. trials to evaluate animal performance on Chand, S. and Katoch, B.S. (1983) Studies on the Cassia silage during the dry season and per- removal of tannins and saponins of some form an economic analysis. forest byproducts by chemical treatments. Indian Journal of Poultry Science 18, 9–12. Chandrasekaran, S. and Swamy, P.S. (2002) Recommendations on future research Biomass, litterfall and aboveground net primary productivity of herbaceous commu- nities in varied ecosystems at Kodayar in More research needs to be done to examine the western ghats of Tamil Nadu. Agricul- the toxicity of C. tora and to look at the effect ture, Ecosystems and Environment 88(1), of the ensiling process on these toxins. 61–71. Research should not be limited to one Charles, O.W. and Muller, H.D. (1975) Coffeeweed species, because other weedy species could (Cassia obtusifolia) toxicity response of provide with a good source of feed as well. laying hens. Poultry Science 54, 1745 (Abstr.) Animal trials and economic analysis are Cheeke P.R. (1998) Natural Toxicants in Feeds, necessary to complete this research and to Forages, and Poisonous Plants. Interstate apply it to a larger scale. Publishers, Danville, Illinois. Cock, M.J.W. and Evans, H.C. (1984) Possibilities for biological control of C. tora and C. obtusi- folia. Tropical Pest Management 30(4), Notes 339–350. Davis, M.A. (2003) Biotic globalization: does com- 1 The most common problem results when high petition from introduced species threaten moisture-silage forms a liquid effluent. Liquid biodiversity? BioScience 53(5), 481–489. effluents from silage are typically highly acidic De Wit, C.T. (1981) The energy spectrum: food and rich in organic matter, nutrients and salts and energy. Food security and a hungry and can be harmful to the environment. world. pp. 66–70. Proceedings of Wageningen 2 Thomas-Willey Mill, Model ED-5, Arthur H. Agricultural University Conference, San Thomas Co., Philadelphia, PA, USA. Francisco, CA (4–6 March 1981). Wageningen 3 Total N was determined colorimetrically Agricultural University, Wageningen, The (McKenzie and Wallace, 1954) with a Netherlands. 208 M. El Hadj et al.

De Wit, H.C.D. (1955) A revision of the genus Cassia species. Geobios New Reports 4(1), ‘Cassia’ (Caesalp.) as occurring in Malaysia. 96–97. Webbia 11, 197–292. Mckenzie, H.A. and Wallace, H.S. (1954) The kjel- Ehrenfeld, J.G. (2003) Effects of exotic plant inva- dahl determination of nitrogen: a critical sions on soil nutrient cycling processes. study of digestion conditions – temperature, Ecosystems 6(6), 503–523. catalyst, and oxidizing agents. Australian FAO (1971) Improving soil fertility in Africa (FAO Journal of Chemistry 7, 55–70. Soils Bull. 14). Food and Agriculture Mokwunye, A.U., de Jager, A. and Smaling, E.M.A Organization of the United Nations, Rome. (eds) (1996) Restoring and Maintaining the FAO (2000) Silage making in the tropics with par- Productivity of West African Soils: Key to ticular emphasis on smallholders (FAO Plant Sustainable Development (Miscellaneous Production and Protection Paper 161). Fertility Studies 14). International Fertility Proceedings of the FAO Electronic Development Center Africa, Lome, Togo. Conference on Tropical Silage, 1 Muhlback, P.R.F. (2000) Additives to improve the September–15 December 1999. silage making process with tropical forages. Goering, H.K. and Van Soest, P.J. (1970) Forage In: L.‘t Mannetje (ed.) Silage making in the fibre analysis [microform]: Apparatus, tropics with particular emphasis on small- reagents, procedures, and some applica- holders (FAO Plant Production and tions. Agriculture Handbook 379 (20), ill. Protection, Paper 161, 151–164). Gupta, P.C. and Sharma, R.K. (1965) Proceedings Olayiwole, M.B. and Olorunju, S.A.S. (no date) of the National Academy of Science India, Effects of ensilage, added nitrogen and Sec. A, 26: 398. energy on the utilization of grain sorghum Hien, F.G. (1997) Soil variability and effectiveness stover by beef cattle. National Animal of soil and water conservation in the Sahel. Production Research Institute, Zaria, Nigeria. Arid Soil Research and Rehabilitation 11, Pal, M., Roy, D.K. and Pal, P.R. (1977) Emodin 1–8. from the leaves of Cassia tora Linn. Indian Hoveland, C.S., Buchanan, G.A. and Harris, M.C. Journal of Pharmacology 39(5), 116–117. (1976) Response of weeds to soil phospho- Pandey, Y.N. (1970) Journal of the Indian rous and potassium. Weed Science 24(2), Botanical Society 49, 151–157. 194–201. Pandit, N.N., Singh, J. and Bhattacharjee, D.K. Hubbell, S.P. (2001) The Unified Neutral Theory (1979) Impact of feeding chakwar (Cassia of Biodiversity and Biogeography. Princeton tora) seed on the growth of broilers. Indian University Press, Princeton, New Jersey. Journal of Poultry Science 14, 176–178. Kabore, D. (2001) Etudes socio-economiques Patel, B.M., Patel, C.A., and Shukla, P.C. (1971) menees dans la commune de Madiama. Evaluation of C. tora seeds in partial replace- Projet IER/SANREM-CRSP. ment of concentrate mixture from the ration Katoch, B.S., Katoch, R.C., Mahajan, N.C. and of milch cows. Indian Journal of Animal Bhowmik, K.B.D. (1978) Utilization of Sciences 41(11), 1031–1034. Indian horse-chesnut (Aesculus indica) ker- Penning de Vries, F.W.T. and Djiteye, M.A. (eds) nals and wild isbgol (Plantago lanceolata) (1982) La productivite des paturages sahe- plant-meal as poultry feed ingredients. liens: Une etude des sols, des vegetations et Indian Veterinary Journal 55, 781–787. de l’exploitation de cette resource naturelle. Katoch, B.S. and Bhowmik. K.B.D. (1983) PUDOC, Wageningen, The Netherlands. Biological evaluation of Cassia seeds (C. tora Perkins, K.D. and Payne, W.W. (1985) Guide to the and Cassia floribunda) as poultry feed ingre- Poisonous and Irritant Plants of Florida. dients. Indian Veterinary Journal 60, University of Florida, Florida Crop Exten- 390–395. sion Service, Institute of Food Agricultural Mall, L.P. and Raina, D.K. (1957) Autoecological Science, Gainesville, Florida. studies on Tridax procumbens Linn. Bulletin Rangaswami, S. (1963) Crystalline chemical com- of the Botanic Society University of Sagar 9, ponents of the seeds of Cassia tora Linn. 35–54. Identity of tora substance C with Manjunath, B.L. and Subbajois, H. (1930) Rubrofusarin and tora substance B with Nor- Chemical examination of the seeds of Cassia Rubrofusarin. Proceedings of Indian tora Linn. Journal of the Indian Chemical Academy of Science 57, 88–93. Society 7, 521. Randell, B.R. (1995) Taxonomy and evolution of Mathur, M. (1985) Comparative study of epider- Senna obtusifolia and S. tora. Adelaide mal structures and seed analysis of two Botanic Gardens 16, 55–58. Dry Season Feed Supplements: C. tora 209

Ranjhan, S.K., Gupta, B.S. and Chabbra, S.S. (Industrial Method No. 334–74A/A). (1971) Chemical composition and nutritive Tarrytown, New York 10591. value of a summer legume, chakunda (C. tora Tilley, J.M.A. and Terry, R.A. (1963) A two stage Linn) hay with special reference to technique for the in vitro digestion of forage metabolisable energy for sheep. Indian crops. Journal of the British Grassland Journal of Animal Health 10(2), 217–221. Society 18, 104–111. Sanchez, P.A. (2002) Soil fertility and hunger in Tiwari, R.D. and Behari, J.R. (1972) Chemical Africa. Science 295, 2019–2020. examination of the roots of Cassia tora. SAS Institute (2001) SAS User’s Guide: Statistics Planta Medica 21(4), 393–397. (Version 8.2). SAS Institute, Cary, North Uhl, C. and Jordan, C.F. (1984) Succession and Carolina. nutrient dynamics and burning in Singh, J., Kumar, R. and Yadav, R.K. (2001) Amazonia. Ecology 65, 1476–1490. Impact of feeding chakwar (C. tora) seed on Van der Pol, F. (1992) Soil Mining, an Unseen growth and meat production in broilers. Contributor to Farm Income in Southern Indian Journal of Poultry Science 36(3), Mali (Bulletin 325). Royal Tropical Institute 308–310. (KIT), Amsterdam, p. 47. Singh, M., Panday, R.K. and Shankarnarayan, Van Soest, P.J. (1963) Use of detergents in the K.A. (1970) Problems of grassland weeds and analysis of fibrous feeds. II. A rapid method their control in India. In: International for the determination of fibre and lignin. Grassland Congress: Proceedings of the 11th Journal of the Association of Official International Grassland Conference. Surfer’s Analytical Chemists 46, 829. Paradise, Australia, pp. 71–74. Van Soest, P.J. and Wine, R.H. (1967) Use of deter- Sood, A.K., Sharma, M. and Katoch, B.S. (1990) gents in the analysis of fibrous feeds. IV. Improvement in the nutritive value of Indian Determination of plant cell wall con- horse-chesnut kernel and cassia seeds by stituents. Journal of the Association of chemical treatments for feeding of chicks. Official Analytical Chemists 50, 50. Indian Journal of Animal Sciences 60(11), Van Soest, P.J. and Wine, R.H. (1968) Determina- 1364–1369. tion lignin and cellulose in acid detergent Steele, K.W. and Vallis, I. (1988) The nitrogen fibre with permanganate. Journal of the cycle in pastures. In: Wilson, J. (ed.) Association of Official Analytical Chemists Advances in Nutrient Cycling in Agricultural 51, 780–785. Ecosystems. CAB International, Wallingford, Wu, C.-H., Hsieh, C.-L., Song, T.-Y. and Yen, G.-C. UK, pp. 274–291. (2001) Inhibitory effects of C. tora L. on Technicon Industrial Systems (1976) Individual/ Benzo[a]pyrene-mediated DNA damage simultaneous determination of nitrogen towards HepG2 cells. Journal of Agricultural and/or phosphorus in BD acid digests and Food Chemistry 49 (5), 2579–2585.

15 Controlled Grazing: Botanical Response and Animal Performance

Ozzie Abaye,1 Meriem El Hadj,1 Amadou Kodio2 and Moussa Keïta2 1Department of Crop and Soil Environmental Science, Virginia Tech, Blacksburg, VA 24061–0404, USA; and 2Centre Regional de la Recherche Agronomique, Institut d’Economie Rurale, BP 205, Mopti, Mali

The lack of pasture and forage resources has lation of animals and overgrazing. The exist- been a growing problem in the Sahel region ing low level of plant succession under the of sub-Saharan Africa. Population growth, current management system has resulted in extensive row cropping, increasing herd a severe botanical shift from dominantly size, high levels of exploitation of woody perennial grasses to annual grasses and from resources, and climatic fluctuations have all a grass-legume mixture to almost pure grass contributed to serious losses of plant biodi- sward. versity within this region. In addition, uti- To persist by reseeding or/and regrowth lization of greatly reduced pastureland from existing plants, forage grasses and resources has become a major source of con- legumes must be given sufficient time flict between farmers and herders. This has between defoliations, because the time certainly been the case in the Commune of between defoliation helps the plant to pro- Madiama. As a consequence, the Natural duce and store enough energy for regrowth Resource Management Advisory Committee (Smith, 1962). Plant response to defoliation (NRMAC) requested SANREM researchers depends on plant species, the extent of to focus their efforts on developing tech- removal of leaf area and growing points, and nologies to improve pasture management. the amount of carbohydrate available for The unimodal annual rainfall pattern regrowth. Therefore, good management limits grazing potentials in Madiama, with (such as rotational grazing) of these grasses respect to plant community composition, should allow sufficient time between defoli- yield and quality. The perennial species, ations to restore carbohydrate reserves or herbaceous as well as woody, have become leave enough leaf area for carbohydrate pro- scarce over the years. The less dependable duction. Because plants differ in their mor- annual grasses and other low forage quality phological and physiological responses to species severely limit potential animal different management strategies, in a mixed production. As perennial forages in these species pasture it is possible to see the dom- pastures have become nearly non-existent, inance of one or two species at any given annual grasses and some legumes have time. In a grazing situation, livestock prefer- become the dominant species in the ence for one species or plant part over Madiama Commune. Seasonal movement of another depends on relative palatability, animals (transhumant practices) is also acceptability and overall availability of pas- increasingly limited, resulting in overpopu- ture species or parts. © CAB International 2005. Conflict, Social Capital and Managing Natural Resources (ed. K.M. Moore) 211 212 O. Abaye et al.

Any plant that can adapt to defoliation objectives are: (i) to determine the influence in a pasture setting needs meristematic of the tethered grazing method on plant tissue close to the ground, out of danger of diversity; and (ii) to evaluate animal perfor- removal by livestock (Harris, 1978), because mance in terms of weight gains under teth- after defoliation leaves regrow from meris- ered grazing. tematic tissue at the base of each leaf and new leaves arise from auxiliary buds also located near the base of leaves (Smith, 1973). Materials and Methods These tissues must be close to the soil sur- face to escape damage during defoliation in The study was conducted at Siragourou and order to allow subsequent regrowth. Plants Nérékoro, two villages within Madiama with upright growth patterns are more sen- Commune. At both locations, an area was sitive to defoliation, and recover more fenced to keep animals and people out of slowly than plants with a horizontal growth the experimental plots. Siragourou and pattern (Butler et al., 1959). Livestock graz- Nérékoro villagers donated young sheep ing can influence grass/legume composition weighing 18–24 kg and 18–21 kg, respec- of a pasture (Schwinning and Parsons, tively. Since the purpose of the experiment 1996), because livestock will selectively was to determine indicators for optimal pas- graze more palatable legumes in preference ture management, the experiment involved to less palatable grasses. This selectivity can two treatments: 3 cm vs. 6 cm forage resid- lead to grass domination in pastures, espe- ual heights, each replicated four times (see cially when continuously grazed (Watkin Fig. 15.1). At each village, animals were and Clements, 1974). Thornley et al. (1995) weighed and tagged according to treatments state that persistence of unfertilized mix- and treated for internal parasites. Animals tures of rye grass and white clover is not were randomized according to weight and achievable, except when grazing eliminates treatment. Each animal in its grazing area the dominance of white clover. If a legume is represented an experimental unit; a total of highly competitive with grasses, the grass eight animals per site were used. Residual can often be maintained in the mixture plant height was determined based on mea- only by the presence of grazing animals surements taken prior to the initiation of the (Schwinning and Parsons, 1996). Depending experiment. The initial height of the forages on the severity of defoliation, root mass and at each site varied from 5 cm to 20 cm and depth can be reduced by grazing. 8 cm to 14 cm for the Siragourou and Root morphology is an important factor Nérékoro locations, respectively (see Table in successful competition for water and 15.1). Each animal was tethered separately nutrients in a pasture (Evans, 1977). A more to its own peg using a 6-m length of rope that developed and deeper root system generally allowed the animal to consume all forage gives one pasture species an advantage over within its range for the period of time deter- another (Davidson, 1978). In general, grasses mined by available forage and canopy resid- have longer, thinner, more finely branched ual height. Once each animal grazed its area roots, and more root hairs, than legumes to the 3 cm or 6 cm residual plant heights, (Evans, 1978). These differences in root they were then rotated to the next peg (see structure and root depth could give grasses Fig. 15.1). Water was provided to the ani- a competitive advantage over legumes in mals daily at the grazing location. At night, nutrient and water uptake (Haynes, 1980), the animals were kept in a kraal. No addi- and could partially explain why grasses are tional feed was provided for the night. better able to absorb limiting nutrients. Although the plan was to weigh the ani- The importance of managing for natural mals following each grazing cycle (after each succession in order to restore pastureland animal had grazed at all eight pegs in their resources and determine indicators for paddock), due to excessive forage produc- improving pasture management practices is tion during the 2003 growing season no more the purpose of this research. The specific than seven pegs within each paddock were Controlled Grazing 213

Fig. 15.1. Sheep pegged within paddocks.

needed during the season. Therefore, the ani- a rare species is one that is present only once mals were weighed only at the beginning and or twice. Brodie (1985) suggests that prob- the end of the grazing season, instead of at the lems with this method include under-asses- end of each grazing cycle. At both locations, sment of smaller species, over-assessment of guards were trained to watch the animals, conspicuous species and misidentification provide water and take the animals home at of species. Initially (August) the entire pas- night. The guards’ knowledge of animals and ture was visually evaluated for both ground their understanding of the process was cover and species diversity. At the end of the impressive. After their training, they placed grazing season (November) the grazed area all the animals in their correct experimental (all paddocks and all treatments) and the plots without any further guidance. Animals ungrazed area (outside of the 6 m circled remained on pasture for a total of 93 days. area grazed by animals) were evaluated using the DAFOR Scale.

Forage mass and botanical composition assessments Soil fertility

The botanical composition of pastures was Soil samples to a depth of 10 cm were determined by a non-destructive visual eval- obtained prior to the start of grazing by taking uation method using the Double DAFOR six to ten soil cores from each treatment Scale (Brodie, 1985; Abaye et al., 1995), replication (areas with tethered grazing and where Dϭdominant, Aϭabundant, Fϭfre- ungrazed areas). Samples were analysed for quent, Oϭoccasional, and Rϭrare are used N, P, K and pH in the IER Soils Laboratory at to evaluate the relative abundance of the CRRA/Sotuba. species. Brodie (1985) describes a species as dominant if it covers most or all of the area. A ranking of abundant is given to species Results and Discussion that cover approximately half to three-quar- ters of an area. Frequent ranking refers to Weather data for the Madiama Commune species well scattered throughout a site but covering less than half the area. A species The total amount of rainfall for the 2003 ranked as occasional occurs a few times, and growing season exceeded the 1999 to 2002 214 O. Abaye et al. verage A → measurements ← verage Replication Treatment Plot A → measurements ← 10.5, 8.2, 13, 6, 148.5, 5.5, 15.5, 10.5, 3718, 31.5, 8, 6, 9 15.411, 9.5, 10, 5.5, 7 10.3411.5, 10, 8, 39, 7.518.5, 31.5, 14, 19, 198, 8, 4, 12, 6.5 1 1 14.53.5, 6, 4, 8.5, 4 8.6 15.2 20.4 2 2 3 3 7.7 1 2 5.2 4 2 4 1 1 2 1 1 10.5, 8, 5, 16, 10.5 27, 8, 7, 10, 28 2 1 1 1 1 1 10.0 10.5, 8, 5, 10, 7 8, 10.5, 8, 12, 4 12, 6, 7, 6 8, 6.5, 17, 9.5, 17, 17 16.0 1 1 13.4 8.1 15.5, 8.2, 27.5, 14.5 8, 8.5 16, 16.5, 7, 8, 11, 38 14.7 7.8 16.0 Plant height taken prior to initiation of the grazing experiment for the Siragourou and Nérékoro villages, August, 2003. Nérékoro and Siragourou to initiation of the grazing experiment for the Plant height taken prior Siragourou Plant height (cm) Nérékoro Plant height (cm) able 15.1. 411 T Replication111 Treatment121 221 Plot 211 311 321 421 Controlled Grazing 215 Rainfall (mm)

Fig. 15.2. Annual rainfall recorded at the Madiama Commune, Mali: 1999–2003.

average annual rainfall. The total annual of these species was less frequent in parts rainfall for 1999, 2000, 2001, 2002 and 2003 of the same grazing areas. Among the few was 602.1 mm, 507.4 mm, 678.1 mm, legumes found at this site, Cassia tora (Ct) 374.5 mm and 1130 mm, respectively. The (sickle pod) was predominant. This annual monthly rainfalls for 2003 exceeded all but legume, widely known in the USA as a the May and June rainfalls for 1999–2003 noxious weed mostly found in soybean (Fig. 15.2). The rainfall for July, August, fields, is extremely aggressive. Most September and October of 2003 was grasses do not compete well with Cassia higher than the previous 4 years. Rainfall tora (Ct) because of its aggressive and inva- usually begins declining in September and sive nature. To a lesser extent, also present stops completely by the end of October. were other legumes including Zornia sp. In 2003, 31.5 mm rain was recorded for (Zsp) and Alysicarpus sp. (Asp) (see Table November. 15.2). The final visual assessment of ground cover and species diversity is shown in Table Forage diversity 15.3a and 15.3b. In most cases, ground cover was at an acceptable level (40–100%) regard- Forage ground cover, species diversity and less of treatments. In general, grazing regeneration assessments appeared to increase species diversity when compared to initial assessments (Table 15.2 NÉRÉKORO LOCATION Prior to grazing, vs. Table 15.3a and 15.3b). Dactyloctenium ground cover ranged from 45% to 100% at aegyptium (Da) and Panicum laetum (Pl) the Nérékoro location, where the predomi- were the dominant grass species. The domi- nant plant species were annual grasses. In nance of Dactyloctenium aegyptium (Da) the grazing area, the dominant grass after severe grazing can be attributed to species were Dactyloctenium aegyptium its morphological characteristic of having (Da), Panicum laetum (Pl) and Bracaria sp. meristematic tissue (areas of rapid cell (Bsp), which are all are highly palatable to division) located close to the ground where it livestock. However, the relative abundance is protected from being removed by the 216 O. Abaye et al.

Table 15.2. Initial visual evaluation of ground cover and botanical composition, Double DAFOR scale, for the Nérékoro site, August 2003.

Composition (%) Forages

Repa Trt b Prc Grd Grass Legume Weed De Af Fg Oh Ri

11196100 0 0 Da Pl 11170100 0 0 Pl Da 12185950 5Da, Pl Ct 12175100 0 0 Da Pl 22198100 0 0 Da, Pl Csp Pl 22150100 0 0 Da, Bsp 21198100 0 0 Da Pl Esp 21198100 0 0 Da Pl 31199100 0 0 Da Pl Zsp 31145100 0 0 Da Pl Esp 32199100 0 0 Da Bsp Pl 3219899 1 0Da Bsp Pl 4216599 1 0Pl Da Asp 42150100 0 0 Bsp Pl, Da 41170100 0 0 Csp Pl Bsp Da 411100 100 0 0 Da Pl

Repaϭreplication; Trtbϭtreatment (1ϭgrazed to 3 cm stubble height; 2ϭgrazed to 6 cm stubble height); Prcϭpasture; Grdϭground cover; Deϭdominant; Afϭabundant; Fgϭfrequent; Ohϭoccasional; Riϭrare. Grasses: Dactyloctenium aegyptium: Da; Panicum laetum: Pl; Eragrostis sp.: Esp; Bracaria sp.: Bsp. Legumes: Zornia: Zsp.; Alysicarpus sp.: Asp.; Cassia tora: Ct.

grazing animal. The Bracaria sp. (Bsp), dom- paddock. The ground cover and the type of inant in part of the pasture prior to grazing, species were similar to those shown in was less abundant at the end of the grazing Tables 15.2, 15.3a and 15.3b. However, as season (Table 15.3a and 15.3b). The disap- shown particularly in replications 3 and 4, pearance of this species was more evident the diversity of plant species was greater where paddocks were grazed to 3 cm, but than the grazed areas (Table 15.3a and where they were grazed to 6 cm residual 15.3b). The plants observed in these pad- height some was left, an indication that graz- docks included both new seedlings from ing preference for this plant could be higher reseeding and first growth after dry season than for other species. In general, more grass dormancy. In most cases, the presence of species were observed where grazing legume species was more evident in the occurred to 6 cm residual height than where ungrazed areas than in the grazed areas, an it occurred to 3 cm residual height. indication of animal preference for legumes Additionally, weedy types of plants such as rather than grasses. Cassia tora (Ct) were more frequently observed where pastures were grazed to 3 cm SIRAGOUROU LOCATION The initial ground residual height than to the 6 cm residual cover of the grazing area at this location, height. This indicates that grazing to 3 cm which was much less than at the Nérékoro residual height is less damaging to invasive location, ranged from bare to 80% (see Table plants than to native species that are more 15.5). In most cases, the legume Zornia sp. desirable to animals. (Zsp) was the most dominant species, fol- Table 15.4a and 15.4b shows the end of lowed by the grass Bracaria sp. (Bsp), which season ground cover and species diversity appeared in mixturewiththelegumesorpure assessment of the ungrazed area of each stands. Among the weedy types of plant Controlled Grazing 217

Table 15.3a. Final visual evaluation of ground cover and botanical composition of the grazed area: Nérékoro site, November 2003, Replications 1–3.

Composition (%) Forages Ground Repa Trt b Plc cover Grass Legume Weed Dd Ae Ff Og Rh

1119080 20 0Pl Da Cor Isp 1129990 10 0DaDl 1138599 1 0DaDl 114100 100 0 0 Da Dl, pp 1155525 0 75Ct Dl Da Cor Sid 12198981 1Pl, Da Cor 1228597 3 0Pl Da Bsp Cor, Isp 1238598 2 0DaDl Pl 12498100 0 0 Da Dl 12560908 2 Bsp Da Ct 21180991 0Pl, Da Bsp Isp 2126590 10 0Pl Da Cor Isp 21390100 0 0 Da Dl 21485100 0 0 Da Pl Bsp 21598100 0 0 Da Dl Ct 22190955 0Pl, Da Isp 22265982 0DaPl Isp, Cor 22398990 1Pl, Da Esp 22499100 0 0 Da 225100 100 0 0 Da Dl Pl 3118099 1 0DaPl Bsp Sg Ct 3129998 2 0Da Pl Isp 313100 90 10 0 Da Pl Bro Ct 31470100 0 0 Da Bsp 3219990 5 5DaBsp Isp Cor 322100 90 10 0 Da Bsp Isp Dl 32340941 5Da Esp, Pl Ct 32490100 0 0 Da Pl

Table 15.3b. Final visual evaluation of ground cover and botanical composition of the grazed area; Nérékoro site, November 2003, Replication 4.

Composition (%) Forages Ground Repa Trt b Plc cover Grass Legume Weed Dd Ae Ff Og Rh

41185 9 5 2PlDa Dal Cor, Isp 41275100 0 0 Da Pl Bsp 41385 98 1 1DaPl 41498 95 5 0Esp Da Isp 42140 98 2 0DaBsp Pl Cor 42250 98 1 1DaPl Ct, Isp 42399 88 2 10DaPl Bsp Cor, Isp 42415 0100 0 Isp Notes for Tables 15.3a and 15.3b. Repaϭreplication; Trtbϭtreatment; Plcϭpaddock; Ddϭdominant; Aeϭabundant; Ffϭfrequent; Ogϭoccasional; Rhϭrare. Grasses: Dactyloctenium aegyptium: Da; Panicum laetum: Pl; Eragrostis sp.: Esp.; Bracaria sp.: Bsp.; Pennisetum pedicellatum: pp; Brachiaria ramose: Bro; Schoenfeldia gracilis: Sg; Digitaria longitudinalis: Dl. Forbs: Ipomea sp.: Isp.; Cassia tora: Ct; Corchorus tridens: cor; Sida alba: Sida. 218 O. Abaye et al.

Table 15.4a. Final visual evaluation of ground cover and botanical composition of the ungrazed area of the Nérékoro site, November 2003, Replications 1 and 2.

Composition (%) Forages

Repa Trt b Plc Gr. Grass Legume Weed Dd Ae Ff Og Rh

111100 100 0 0 Pl Da 1129598 2 0PlDaCor Isp 1139510 0 0Da Dl Esp 114100 100 0 0 Da Pl, Dl 1159990100Bsp Dl Da Cor, Et, St 1169550 49 1Dl Ct Isp Cor, St 1179888 1 0Bsp Da Esp Cor 1189090100Da Dl Bsp Cor, Isp, St 12199100 0 0 Dl Da Esp, Dl 12280100 0 0 Da Dl 1239998 1 1Pl DaDl 12455100 0 0 Da 125100 80 20 0 Da Dl Tp 1267585 150Da Esp Bsp 1279580200Bsp Dl Da, Tp Cor 1289990 10 0Da Dl Esp Cor Cyp 21180100 0 0 Da, Pl 21299100 0 0 Da Dl Esp Cl 213100 99 1 0 Da Bsp Pl, Dl Isp 214100 100 0 0 Da Dl Esp 215100 100 0 0 Da 2169875 20 5Dl Bsp Da St 217100 100 0 0 Da Dl, Et 218100 99 0 1 Dl, Da Bsp Esp 221909721Pl, Da Zsp, Cor 22255100 0 0 Esp Da, Pl 223999910Pl, Da Dl Isp 224100 94 5 1 Da Pl Isp 22598100 0 0 Da Bsp Esp, Cyp 2267560373Dl Pl Tp Isp, Cor 227100 98 2 0 Da Esp Sg Ct, Isp 228100 98 2 0 Da Esp Sg Ct, Isp

species, Cassia tora (Ct) was most abundant Alysicarpus sp (Asp) had frequent to rare (Table 15.5). As mentioned above, Cassia appearances in these pastures. tora (Ct) is only marginally palatable to live- At the end of the grazing season, the stock. Animals consume Cassia tora (Ct) pasture had significantly less ground cover when it is at vegetative stage and in the than at the beginning of the grazing season. absence of more desirable forage. Other The legume Zornia sp. (Zsp), observed early grasses found in abundance to rare in this area in the grazing season, remained dominant at were Dactyloctenium aegyptium (Da), the end of the season. However, a shift in Microchloa indica (Mi), Panicum laetum (Pl) botanical composition to grasses in place of and Tribulus terrestris (Tt). Dactyloctenium the legume Zornia (Zsp) was evident (see aegyptium (Da) and Panicum laetum (Pl) are Table 15.6a and 15.6b). characterized as high quality palatable for- The grasses Schoenfeldia gracilis (Sg) ages while Microchloa indica (Mi) is less and Eragrostis sp. (Esp) were most dominant desired by the grazing animals. The legume compared with the Bracaria sp. (Bsp) that Controlled Grazing 219

Table 15.4b. Final visual evaluation of ground cover and botanical composition of the ungrazed area of the Nérékoro site, November 2003, Replications 3 and 4.

Composition (%) Forages

Repa Trt b Plc Gr Grass Legume Weed Dd Ae Ff Og Rh

31180100 0 0 Da Pl Bsp Sg Esp 31299982 0PlDa Bsp Isp, Cor 313100 95 5 0 Da, Pl Dl Sg Za, St 314100 90 0 20 Da Dl, Pl Esp 315100 80 15 5 Dl, Da Pl, Bsp Isp,Cyp Coc 316100 60 20 20 Bsp Dl, Da Isp 3179889 1 10Da Bsp Cl Cor 3189550 0 50 Da Bsp Pl 321100 95 5 0 Da Pl Bsp, Dl Isp, Zsp 32299801010DaPl Isp, Dl Esp, Gl 323100 82 3 15 Dl Da, Pl Bsp Isp 32495100 0 0 Da Bsp, Pl Dl 3259930 10 60Da Bsp Dl Ipo 326100 93 2 5 Da Bsp, Dl Cor, Isp 3279788 2 10Da Dl Pl Isp 3289998 2 0DaDl Pl Esp Isp 411100 95 5 0 Da Pl Dl Cor Cl 412100 85 5 10 Dl, Da Pl Isp, Gl 413100 99 1 0 Da Dl Bsp Tp, Esp Isp 41420 0 15 85 Isp 4159585510 Bsp Pl Esp Isp Ct 41699991 0Bsp Dl Da Pl Isp 417100 95 5 0 Da Pl, Bsp Isp Tp 418100 95 5 0 Da Dl, Bsp Tp Isp, St 4219980155Pl Da Tp, Bsp Isp Z, Ct 4229875 10 15Dl Da, Pl Isp Cor, St 42398982 0Da Bsp Pl, Isp Sg 4249030700Isp Bsp Da Cl, Cor Sg 425100 90 10 0 Ei Da Ipo 42699945 1DlDa Cl, Esp Bsp, Cor, Pl 42799964 0Pl, Dl Bsp Cl, Da Zsp, Ct 4289898 1 1DaDa Gl Isp

Notes for Tables 15.4a and 15.4b. Grasses: Dactyloctenium aegyptium: Da; Panicum laetum: Pl; Eragrostis sp.: Esp; Bracaria sp.: Bsp; Pennisetum pedicellatum: pp; Brachiaria ramose: Bro; Schoenfeldia gracilis: Sg; Setaria sp.: St; Cloris sp.: Cl; Digitaria longitudinalis: Dl. Forbs: Zornia: Zsp; Ipomea sp.: Isp; Cassia tora: Ct; Corchorus tridens: cor; Sida alba: Sida; Cyperus sp.: Cyp; grasslike weed: Gl; Tephrosia sp.: Tp.

was observed in abundance earlier in the abundant available forage, the result of season. Both Schoenfeldia gracilis (Sg) and excessive moisture during the 2003 grazing Eragrostis sp (Esp) are highly desired, good season, forage production exceeded animal quality forage crops. However, these grasses need, thereby causing under-utilization of are late maturing compared to the other plant species that are highly desirable at plant species adapted to the area. Also, vegetative/less matured stages, such as excessive primary growth and new growth Schoenfeldia gracilis (Sg) and Eragrostis sp. from seeds could have caused the relative (Esp). At a mature stage (plants with seed- abundance of these two grasses. Due to head), animals do not graze these two 220 O. Abaye et al.

Table 15.5. Initial visual evaluation of ground cover and botanical composition, DAFOR scale, for the Siragourou.

Composition (%) Forages

Ground Repa Trt b Plc cover Grass Legume Weed Dd Ae Ff Og Rh

111 60 40 60 0 Zsp Da Mi, Asp 11170 15 805Zsp Mi 121 50 40 60 0 Zsp Dsp Dsp 1217515805 Zsp, Mi Pl 221 30 10 90 0 Zsp Bsp Mi Asp 22165305020Dsp Bsp, Ct, Cn Da, Zsp Esp 211 75 5 95 0 Zsp Bsp Mi, Asp 2118020755Zsp Mi Bsp, Asp 311 70 20 80 70 Bsp Zsp Asp Mi 31170306070 Zsp, Bsp Tt, Isp 321 95 10 90 0 Bsp Zsp Asp, Ul 32185503020Bsp Pl, Zsp Asp 42141 33 670Bsp Zsp 4214060355 Zsp, Mi 411Bare 10 90 0 Zsp Bsp Mi 411Bare 10 90 0 Zsp Mi

Repaϭreplication; Trtbϭtreatment; Plcϭpaddock; Ddϭdominant; Aeϭabundant; Ffϭfrequent; Ogϭoccasional; Rhϭrare. Grasses: Dactyloctenium aegyptium: Da; Panicum laetum: Pl; Tribulus terrestris: Tt; Bracaria sp.: Bsp; Microchloa indica: Mi; Digitaria sp.: Dsp; Digitaria longitudinalis: Dl; Tribulus terrestris: Tt. Forbs: Zornia: Zsp; Alysicarpus sp.: Asp; Urena lobata: Ul; Ipomea sp.: Isp; Cassia tora: Ct; Cassia nigricans: Cn.

species, which explains the relative abun- cover. In most cases, regardless of treat- dance of these plants at the end of the 2003 ments, both grasses and legumes appeared grazing season. at similar frequencies in the ungrazed areas. The cause of the botanical shift at In addition to grass/legume compositions, this location could be attributed to grazing there were more plant species than in the preference and forage abundance, which grazed areas (Table 15.7a and 15.7b). At both exceeded animal need, rather than to locations, due to the unseasonably wet treatment effects (i.e. residual height). At the growing season, forage production (but not end of the growing season, all the grasses quality) was maintained at the highest level turned brown with the exception of the throughout the growing season. legume Zornia (Zsp) and a few broadleaf weeds. Regardless of soil moisture and tem- perature, annual grasses in this climate Animal performance undergo senescence, the process of plant Nérékoro and Siragourou locations degeneration that generally occurs at the end of the growing season in order to pro- During the first week of November, the duce seeds for the following season. animals were removed from the pastures Visual evaluation of ungrazed areas and weighed. The animals gained 1–3 kg revealed differences from the grazed areas over the season. For the Nérékoro loca- (Tables 15.6a and 15.6b vs. 15.7a and 15.7b). tion, there was only a slight difference in Most of the grazing areas included both weight gain between treatments (see Table grass and forbs (legumes and broadleaf 15.8). According to the field technicians, weeds), ranging from 15% to 95% ground the animals were heavier in August to Controlled Grazing 221

Table 15.6a. Final visual evaluation of ground cover and botanical composition of the grazed area for the Siragourou site, November 2003, Replications 1–3.

Composition (%) Forages

Ground Repa Trt b Plc cover Grass Legume Weed Dd Ae Ff Og Rh

11145100 0 0 Sg Bsp 11215 40 600Zsp Sg Esp 11315 2 98 0Zsp Sg 11425 10 900Zsp Sg Bsp 11520 40 600Zsp Sg 11670 30 700Zsp Et Isp 12110100 0 0 Sg 12230 80 200Bsp Zsp Ao 1233540600Zsp, Bsp Sg Ao, Ms 1245 0100 0 Zsp 12515 10 900Zsp Sg 12630 15 850Zsp Sg 1276060400Isp, Da Esp, Sg Sb 2111580200Zsp, Sg Esp, Ao Da 21210199 0 Zsp Sg, Ao 21310 1 99 0Zsp Sg 21410 10 900Zsp Bsp 21510 20 800Sg Sg 22130 40 60 0Zsp Sg Bsp Esp 22210 0100 0 Zsp 22320 5 95 0 Zsp 2242050500 Sg, Esp 22520 60 400Sg Zsp 22640 95 5 0Sg Isp 3115085150Bsp Pl, Zsp Sg Ct 31280 80020Bsp 3131515850Zsp Sg, Esp 31415 10 900Zsp Sg 3151515850Zsp Esp, Sg Bsp 3216060400Bsp Sg, Ao Zsp 3221050500 Zsp, Sg

mid-October, but might have lost weight better. Although the original plan was to due to increasingly poor forage quality as weigh the animals at the end of each cycle the season progressed. Although forage (after animals had gone through cycle of biomass was high, the forage quality dec- eight pegs/paddocks per treatment), due lined with the decrease in leaf to stem to the excessive forage growth during the ratio and increase in seed head produc- 2003 growing season (driven by excessive tion. Generally, such unpalatable forages rainfall), none of the animals went are high in fibre and low in crude protein through all the eight grazing areas/pad- and soluble sugars. If animals were docks. By the end of the grazing season weighed on a monthly basis instead of at two or three pegs (grazing areas/paddock) the end of the grazing season, the assess- remained ungrazed for treatment one ment of treatment effects and overall (where animals grazed to 3 cm residual animal performance would have been height). Animals grazing to the 3 cm 222 O. Abaye et al.

Table 15.6b. Final visual evaluation of ground cover and botanical composition of the grazed area Siragourou site, November 2003, Replications 3 and 4.

Composition (%) Forages

Ground Repa Trt b Plc cover Grass Legume Weed Dd Ae Ff Og Rh

323 5 100 0 0 Sg 324 5 80 20 0 Sg Sg 32520 60 400Bsp Sg Sg 411 5 10 90 0 Zsp Sg Bsp 41210 100 0 0 Sg 41315 100 0 0 Sg 4142015850Zsp Ao Pl Isp 415 5 10 90 0 Iv Sg 416 5 100 0 0 Esp 4171540600Zsp Bsp Esp Isp, Cb 4182055450Zsp Esp, Sg 42110 100 0 0 Sg 42215 100 0 0 Sg Ct 4231050500Zsp Bsp, Sg 42420 95 5 0Sg Tp 42515 98 2 0Sg Esp Zsp 42615 80 200SgZsp Esp 42760 40 600Zsp Sg Cb

Notes for Tables 15.6a and 15.6b. Repaϭreplication; Trtbϭtreatment; Plcϭpaddock; Ddϭdominant; Aeϭabundant; Ffϭfrequent; Ogϭoccasional; Rhϭrare. Grasses: Dactyloctenium aegyptium: Da; Panicum laetum: Pl; Eragrostis sp.: Esp; Bracaria sp.: Bsp; Cenchrus biflorus: Cb; Schoenfeldia gracilis: Sg; Agrostis sp.: Ao. Forbs: Zornia: Zsp; Ipomea sp.: Isp; Cassia tora: Ct; Tephrosia sp.: Tp; Ipomia vegans: Iv.

residual height moved much more slowly Summary through the cycle than those grazing to 6 cm residual height. As evidenced by the In areas with unpredictable rainfall and poor excessive forage mass left ungrazed this soil fertility (for information on soil fertility, growing season, each pasture paddock see Chapter 3), maintaining vegetative cover could have carried two animals instead of and the regrowth/persistence of forages is one animal/peg. Also, grazing on the 3 cm highly critical. The growing season is 80–90 residual height treatment paddocks days, with an average of 41 wet days and appeared to be much more severe on most rainfall totals varying between 203 mm and plant species (less regrowth and more 610 mm. The lack of pasture and forage bare ground) than it did on the 6 cm resources has been a growing problem in the residual height paddocks. The effect of region. With rapidly disappearing perennial the 3 cm grazing height on animal per- forages, annual grasses and some legumes formance was more evident at the have become the dominant species in the Siragourou location than at the Nérékoro region. The overall objective of this study location, as shown by weight loss of one was to improve pasture resources and animal (see Table 15.9). In contrast, ani- increase feed production. mals grazing to 6 cm residual height The research indicated that grazing gained the most compared to the residual intensity has a profound impact on species height of 3 cm at this location. diversity. The response to intensive grazing Controlled Grazing 223

Table 15.7a. Final visual evaluation of ground cover and botanical composition of the ungrazed area of the Siragourou site, November 2003, Replications 1 and 2.

Composition (%) Forages

Ground Repa Trt b Plc cover Grass Legume Weed Dd Ae Ff Og Rh

111 80 100 0 0 Sg Mi Zsp 112859820Sg, Mi Bsp Esp Ao, Zsp 113 25 35 650Zsp Sg 114 45 20 800Zsp Esp Sg 1155545550Zsp Bsp Esp 1167510900Zsp Sg Esp, Isp 1178030700Zsp Isp Esp Da Ao, Cb 1189540600Zsp, Da Cb, Esp 1216040591Zsp, Mi Sg Esp Ao 1224540600Zsp Sg Bsp Ind, Sp 1237050500Zsp Sg, Bsp Ao, Esp 124 65 45 550Zsp Sg Esp 1259030700Zsp, Isp Sg Bsp Esp, Ao, Dl 1268570300Isp, Bsp, Cb Zsp Ao, Br Dl, Pl 1276060400Isp, Da Zsp, Dl Tp 1288540600Isp Dl Esp, Sg, Da 211 75 40 60 0 Zsp Sg Esp Da 2124520800Sg Bsp Esp, Da 21345595 0 Zsp Esp, Sg, Iv 2148020800Zsp Bsp Ao Sg Cb 2158060400Bsp Zsp Isp Cb, Sg Iv 2163055450Bsp, Zsp Esp 2175055450Zsp Pl Esp, Sg 2188560400Da Zsp, Pl Dl Esp, Cb, Iv 221 80 98 2 0 Mi Bsp Sg Sg Esp 2224550500Zsp, Sg Bsp Esp 223 70 35 650Zsp Sg 2245020800Zsp Bsp Sg Esp, Sb 2253585150 Esp, Zsp Cb 226 40 95 5 0Sg Isp 2275090100Sg Cb, Bsp Sg 2286545550Zsp Esp, Sg Sg, Da Isp

was highly dependent on the morphological habit are more sensitive to defoliation, and characteristics of the plant species. Plants are slower to recover than plants with a hor- with a horizontal growth pattern were able izontal growth patterns. At the end of the to regrow at a faster rate after being defoli- growing season, the less frequent plant ated to 3 cm stable height than more upright species found were those with upright types of plants. This was mainly because growth habits that also happened to be the growing point (the site of rapid cell divi- highly desired (selectively grazed) by the sion) on the horizontal plants is close to the animals. ground, out of danger of being removed by The short- and long-term productivity the grazing animal. It is critical that these and persistence of annual plants under tissues be close to the soil surface to escape natural condition are far more challenging damage during defoliation and thus permit than the management of perennial forbs, regrowth. Plants with an upright growth grasses or shrubs. The seasonal survival and 224 O. Abaye et al.

Table 15.7b. Final visual evaluation of ground cover and botanical composition of the ungrazed area of the Siragourou site, November 2003, Replications 3 and 4.

Composition (%) Forages

Ground Repa Trt b Plc cover Grass Legume Weed Dd Ae Ff Og Rh

3116560400Sg, Zsp Bsp Esp 31215 40 600Zsp Bsp 31345 20 80 0Zsp Sg Bsp Iv 31440 5 90 0Zsp Sg 31540 15 850Zsp Et Sg 3164064351Sg Sg Cb, Esp Isp 3176020800Zsp Sg Cb, Esp 3185535632Zsp Esp, Bsp Pl Isp Cb, Tp 3213550500Sg Sg, Pl Ind Iv 32240 40 600Zsp Sg Bsp 32330 40 600Zsp Sg Vi 3246040600Zsp Sg Bsp Cb, Ao 3255030700Zsp Bsp Ct, Sb 3265065350Zsp Esp Cn Pl, Isp, Bsp 3276010900Zsp Sg Esp, Sg Ao, Da 32845595 0 Zsp Esp, Bsp Sb, Da Sg 41110 94 5 1Sg Sg 41225 98 1 1Mi Sg 4134050500Sg, Zsp Pl Esp 41485 90 100Pl Iv Sg 415509550Sg Zsp, Iv 4164050500Zsp Esp, Sg Sb Ao 41715 40 600Zsp Bsp 4182055450Zsp Esp, Sg 42120100 0 0 Cr Sg Esp 42240 80 200Sg Zsp Iv 4238560400Pl Da, Bsp Zsp, Ao 4246055450Bsp Tp Sg, Pl 4255060400Esp, Sg Sg Bsp Ao 4267550500Zsp, Dl Sg Ao Cb, Tp 42740 0 100 0 Zsp 4284520800Zsp Sg, Da Dl

Notes for Tables 15.7a and 15.7b. Repaϭreplication; Trtbϭtreatment; Plcϭpaddock; Ddϭdominant; Aeϭabundant; Ffϭfrequent; Ogϭoccasional; Rhϭrare. Grasses: Dactyloctenium aegyptium: Da; Panicum laetum: Pl; Eragrostis sp.: Esp.; Bracaria sp.: Bsp.; Microchloa indica: Mi; Digitaria longitudinalis: Dl; Cenchrus biflorus: Cb; Schoenfeldia gracilis: Sg; Brachiaria ramosa: Br. Forbs: Zornia: Zsp.; Ipomea sp.: Isp; Tephrosia sp.: Tp; Agrostis sp.: Ao; Ipomia vegans: Iv; Indigofera sp.: Ind; Valteria indica: Vi. Controlled Grazing 225

Table 15.8. The influence of tethered sheep grazing on animal performance: Nérékoro.

Initial animal weight (kg) Final animal weight (kg) Seasonal weight Average daily Animal number 31 July 2003 4 November 2003 gain (kg) gain (g)

Treatment I 2 21.0 24.0 3.0 32.0 3 18.0 19.5 1.5 16.0 4 18.0 20.5 2.5 26.9 7 18.0 19.0 1.0 11.0 Average ϭ 18.75 20.75 2.00 Treatment II 8 20.0 21.0 1.0 11.0 5 19.0 22.0 3.0 32.0 6 18.0 19.5 1.5 16.0 1 18.0 20.0 2.0 22.0 Average ϭ 18.75 20.60 1.85

Table 15.9. The influence of tethered sheep grazing on animal performance: Siragourou.

Initial animal weight (kg) Final animal weight (kg) Seasonal weight Average daily Animal number 31 July 2003 4 November 2003 gain (kg) gain (g)

Treatment I 224 25.4 1.4 15.05 318 22.0 2.0 22.0 421 22.5 1.5 16.0 720 19.0 Ϫ1.0 – Average ϭ 20.75 22.10 0.98 Treatment II 824 26.0 2.0 22.0 518 20.0 2.0 22.0 620 24.0 4.0 43.0 119 22.0 3.0 32.0 Average ϭ 20.50 23.00 2.75

regeneration of these annual grasses are plants will cause a shift in botanical com- solely dependent on self-reseeding. There- position that consists of redistribution and fore, the management of these types of grass replacement of the more desirable grasses swards needs to focus on the importance of by more weedy types of plants. Based on the the rest period and defoliation heights. The in-season observations, the 6 cm compared rest period between defoliation needs to with the 3 cm grazing height would promote be long enough to allow sufficient regrowth regrowth within the same season and possi- for grazing as well as reseeding. This time bly allow the plant to reseed for the follow- period can be shorter when rainfall is in ing growing season. excess, as it was in 2003, because it allows Regardless of the dominant effect of the rapid regrowth compared to under a highly variable environmental condition of drought condition where plant growth and the Sahel regions of Africa, the basic knowl- persistence is severely limited. The data edge of the response of annual grasses to showed that close grazing of these annual defoliation can make a big difference in 226 O. Abaye et al.

Table 15.10. Soil test results obtained from sites with different vegetations in pastures at two sites (Siragourou and Nérékoro) within the Madiama Commune.

Siragourou Nérékoro

Areas Areas Areas Areas Typical values mostly mostly Entire mostly mostly Entire for Malian Items legumes Cassia tora pasture legumes Cassia tora pasture soils pH (in water) 5.10 5.36 4.74 6.25 7.20 6.50 5.0–5.5 pH (KCl) 4.29 4.70 4.15 5.83 6.55 5.98 Organic matter % C Nitrogen (Azote) %N 0.02 0.03 0.01 0.03 0.06 0.05 0.05 Total phosphorus, ppm 13.00 15.42 13.00 14.73 41.41* 28.24* 7.0 CEC, meq/100 g 1.76 2.67 2.02 3.71 3.58* 3.71 Sodium (Na), ppm 0.09 0.18* 0.01 0.21* 0.09 0.17 0.01 Potassium (K), ppm 0.14 0.17 0.11 0.20 1.17 0.37 0.20 Calcium (Ca), ppm 0.59 0.97 0.51 1.43 1.97 1.34 0.70 Magnesium (Mg), ppm 0.37 0.52 0.29 0.87 1.13 0.32 0.40 Iron (Fe), ppm 30.76* 17.72 10.04 7.40 16.60 30.08* 20 Copper (Cu), ppm 0.52 0.20 0.28 – 0.12 0.48 – Zinc (Zn), ppm 0.12 0.12 – 0.48 0.76 0.96 – Manganese (Mn), ppm 8.80 18.40 9.40 14.36 15.52 21.56* 20

*Questionable values being re-run for accuracy.

Table 15.11. Particle size analysis of soil samples obtained from sites with different vegetations within pastures at two sites (Siragourou and Nérékoro) within the Madiama Commune.

Siragourou Nérékoro

Areas mostly Areas mostly Entire Areas mostly Areas mostly Entire Items legumes Cassia tora pasture legumes Cassia tora pasture

Sanda, % 87.2 83.9 86.8 89.3 75.8 78.3 Siltb, % 7.5 11.2 9.0 7.8 20.6 16.9 Clayc, % 5.3 4.8 4.1 2.9 3.6 4.8 aSand sizeϾ0.05 mm; bsilt size 0.05–0.002 mm; cclay sizeϽ0.002 mm.

maintaining in-season productivity and on the turnover of root and nodule tissue of long term persistence of these grasses. plants of Trifolium repens, Trifolium pratense, and Lotus uliginosus. New Zealand Journal of Agricultural Research 2, 415–426. References Davidson, R.L. (1978) Root systems – the forgotten component of pastures. In: Wilson, J.R. (ed.) Abaye, A.O., Allen, V.G. and Fontenot, J.P. (1995) Plant Relations in Pasture. Commonwealth Influence of grazing cattle and sheep toge- Scientific and Industrial Research Organiza- ther and separately on animal perform- tion (CSIRO), East Melbourne, Australia, ance and forage quality. Journal of Animal pp. 86–94. Science 72, 1013–1022. Evans, P.S. (1977) Comparative root morphology Brodie, J. (1985) Grassland Studies. George Allen of some pasture grasses and clovers. New and Unwin, Boston, Massachusetts. Zealand Journal of Agricultural Research 20, Butler, G.W., Greenwood, R.M. and Soper, 331–335. K. (1959) Effects of shading and defoliation Evans, P.S. (1978) Plant root distribution and Controlled Grazing 227

water use patterns of some pasture and crop several management schedules. Crop Science species. New Zealand Journal of Agricultural 2, 75–78. Research 21, 261–265. Smith, D. (1973) Physiological considerations Harris, W. (1978) Plant relations in pastures. in forage management. In: Forages: the In: Wilson, J.R. (ed.) Plant Relations in Science of Grassland Agriculture,3rd edn. Pasture. Commonwealth Scientific and Iowa State University Press, Ames, Iowa, Industrial Research Organization (CSIRO), pp. 425–436. East Melbourne, Australia, pp. 67–85. Thornley, J.H.M., Bergelson, J. and Parsons, A.J. Haynes, R.J. (1980) Competitive aspects of the (1995) Complex dynamics in a carbon-nitro- grass-legume association. Advanced Agri- gen model of a grass-legume pasture. Annales culture 33, 227–261. Botanicae 75, 79–94. Schwinning, S. and Parsons, A.J. (1996) Analysis Watkin, B.R. and Clements, R.J. (1974) The effects of the coexistence mechanisms for grasses ofgrazinganimalsonpastures.In:Wilson,J.R. and legumes in grazing systems. Journal of (ed.) Plant Relations in Pastures. Common- Ecology 84, 799–813. wealth Scientific and Industrial Research Smith, D. (1962) Carbohydrate root reserves in Organization (CSIRO), East Melbourne, lucerne, red clover birdsfoot trefoil under Australia, pp. 273–289.

16 Social Capital and Improved NRM

Keith M. Moore1 and Salmana Cissé2 1Office of International Research, Education and Development, Virginia Tech, Blacksburg, VA 24061–0378, USA; 2Centre Régional de la Recherche Agronomique, Institut d’Economie Rurale, B.P. 205, Mopti, Mali

As noted in Chapter 2, two competing systems local participation for some time, began of governance at the local level complicate the promoting decentralization as a necessary governance of natural resource management mechanism for the promotion of economic (NRM), particularly rule making and admin- growth, improvement of NRM and govern- istration with respect to resource tenure. The ment services, reduction in social inclusion dominant trend in African governance since and poverty, and development of civil soci- independence in the 1960s has been the ety (Manor, 1999; Blair, 2000). Indeed, centralization of the state (Wunsch and decentralization in various forms, whether Olowu, 1990). While centralization estab- political devolution to local government, lished state authority over land and other nat- deconcentration of government services to ural resources, for the most part the state did local level offices, or privatization of state little to interfere with customary land tenure assets and powers to private individuals, arrangements. National tenure or enclosure NGOs or corporations, has been increasing systems have only been invoked when in West Africa. The NRM objective of this external demand substantially increased land movement has been to increase local values (Woodhouse, 2003). Consequently, participation and thereby improve the qual- rural economic development has not been ity of sustainable NRM decision making. dependent on the willingness of local popu- In response to these local democratization lations. Furthermore, centralization substan- efforts, customary authorities have begun to tially reduced the civic capacity of local re-emerge as a political force (Ribot, 2002). populations to act coherently in their own The immediate consequence of this promo- interests. In cases of rival land tenure claims, tion of local governance has been the rein- this legal pluralism led to arbitrary solutions forcement of legal pluralism aggravating where the relative power of the litigants was conflict among resource users. determinant. The average resource user, man, In this chapter, we investigate the con- woman or youth, had little security of tenure tribution of the concept of social capital to if the value of his or her land increased understanding the development of civil significantly. society and establishment of enabling con- In response to this trend, the interna- ditions in which all stakeholders can have a tional donor community, non-government voice in decentralized decision making. organizations (NGOs), and even national This analysis highlights the quality of social governments, having pushed for increased relations and their positive and negative © CAB International 2005. Conflict, Social Capital and Managing Natural Resources (ed. K.M. Moore) 229 230 K.M. Moore and S. Cissé

impact on local NRM governance. The case has led to experimentation in the develop- of the Natural Resource Management ment of CSOs and, consequently, a clearer Advisory Committee (NRMAC) in Madiama definition of what constitutes them. The fol- demonstrates the extent to which decentral- lowing contrasts rural associational life ization has generated the conditions for the with criteria Davis (2000) specified. Civil growth of civil society’s role in local deci- society organizations must necessarily be sion making. The chapter begins by clarify- autonomous from the state and customary ing the concepts civil society and social authorities. An individual must have the capital first introduced in Chapters 2 and 7. right of entry into and exit from associations After introducing how the concept of social (i.e. membership must be voluntary). For capital helps to elucidate social relations in democratic (normative) purposes it is civil society, we examine various dimen- desired that the internal procedures of civil sions and affiliations of social capital to society associations be inclusive and trans- determine the extent to which broad-based parent (as demonstrated by leadership consensus over NRM decision making can selection and decision-making processes). be established at the commune level. The According to these standards, custom- analysis focuses on the social capital of ary village associations in rural Mali so often locally relevant referents: the local State invoked by development agents do not stand (Commune authorities), customary authori- up well. However, they are the most impor- ties (or the subordinate customary state), tant organizational entities in a village and and rural civil society organizations (CSOs). village residence is sufficient for, indeed An examination of the structure of rural obliges, membership, even for those of non- civil society follows. dominant ethnic groups who may reside in the village. Membership is not voluntary and essentially includes household heads for State and Civil Society in Africa – a Rural the primary adult association or tonba. Clarification Although leaders are chosen more for their personal efficacy than their clan or social Before going into the analysis of social capi- origin, Kassibo (1997) notes that democratic tal, it is necessary to clarify the relationships elections have never been a part of Malian between rural CSOs, customary association, tradition. Because traditional decision and the state. Blair (2000) and Woolcock making is consensual and combined with the (1998) give a useful generic definition of civil regime of force de jure, minority or less pow- society as an associational sphere intermedi- erful voices are often not heard. Civil society, ary between the individual/family/house- such as it is at the village level, is insular and hold and the state. Put into practice, however, perhaps only meaningful within the context this generic definition often masks quite of face-to-face relations of village life and the different perspectives on its existence and authority structures (the customary local qualities. In this regard, Mamdani (1996) state) governing them. This is not to say that highlighted the bifurcated state within social capital does not reside within these African societies, while Docking (1999) and relationships (more about this below). Guyer (1994) have noted an international Although there has been a tendency to dimension to civil society in Africa. The romanticize past associational forms and empirical reality of civil society in Africa is the extent to which civil society has existed not the same as that which evolved out of the in Mali, as elsewhere in Africa, drawing on European experience several centuries ago. the work of Mamdani (1996), we argue that Despite substantial growth of NGOs at customary governance practices have been the national level and sustained associa- confused with the characteristics of an tional life in rural villages, it is difficult to autonomous civil society. Customary associ- conclude that civil society as conceived in ational structures must be distinguished the West has advanced very far in rural Mali. from CSOs. These customary structures are However, the work of international NGOs characterized by village and clan solidarity, Social Capital and Improved NRM 231

defined within the context of and in opposi- since the very beginnings of the discipline tion to colonial and post-colonial states, and [sociology]’ (p. 2). According to Woolcock serve as an interface with local chieftain- (1998), it is an attempt to restore the role of cies. Chieftaincy is a holistic concept with- norms and institutions lost in the evolution out the European distinction between of economic theory. While increasingly pop- legislative, executive and judiciary roles ular among development specialists as a (Alexandre, 1968), whether the particular multidisciplinary integrating concept, it has domain be crop or pasture land or fishing been used in many different ways leading to waters. Indeed, ‘traditional’ chiefs and the considerable confusion. It has been used to apparatus of chieftaincy have both consti- describe both individual and group charac- tuted a form of authority of their own, and teristics, as well to conflate sources and con- have served to mediate relations between sequences and traverse disciplinary lines the state and society (van Dijk and Rouveroy (Portes, 1998; Woolcock, 1998; Castle, 2002). van Nieuwaal, 1999). In order to understand the quality of This customary state was based on a social relations with which social capital is regime of force in its external relations and associated, it is useful to first draw on the provided identity and security for those work of Coleman (1988). In seeking to com- living within its domain. In the past (and bine the economist’s principle of rational perhaps into the present) force has been the action for the analysis of social systems with- recognized modality on which production out discarding social organization, Coleman of surpluses for the larger society was based defines social capital as the function of a set (Ribot, 1999). Tax collection and corvée of resources within the social structure avail- labour (state required labour services) are able to the actor. These resources include the most obvious examples, but state-run trustworthiness, role expectations and oblig- production cooperatives have a similar rep- ations, information channels and effective utation. On one hand, village chiefs are seen sanctions. In this conception, social capital as the bulwark for the reproduction of vil- refers to individualized resources that can be lage life, clan/ethnic solidarity, and protec- used to achieve the actor’s interests. Putnam tion from the state (van Dijk and Rouveroy (1993) explains the collective character of van Nieuwaal, 1999). On the other, chiefs the concept. It allows citizens to solve col- are representatives of the national state call- lective problems, cuts transaction costs, and ing on the gendarmerie (i.e. the force fur- improves social tolerance and public health. nished by the nation-state) when a situation It is often measured by three indicative char- surpasses their negotiating capacities. acteristics focusing on the strength of social relations themselves: association member- ships, trust and the expectation of reciproc- Social Capital: the Quality of ity. While intuitively reasonable, both Associational Relations approaches have been criticized for their tau- tological nature, conflating the sources and Often linked with civil society, social capi- consequences of social capital, and not tal is a complementary concept that can be taking into account the negative dynamics applied to increase precision in the analysis associated with too much internal solidarity of the transition to new social forms of asso- (Portes, 1998). ciational life (local government, NGOs, etc.) In his analysis of the problem of embed- and inform the understanding of the role of dedness, Granovetter (1985) provides the various associational forms for socioeco- initial key to unravelling the confusion nomic growth and development. generated by the individual versus collec- Social capital provides a way to under- tive perspectives. In attempting to develop stand the relationships and social forces at theories of action integrating economics work in a development setting (Narayan, and sociology, theorists have tended to 2002). As Portes (1998) notes, ‘the term social posit ‘atomized actors’ that have been capital simply recaptures an insight present either under- or over-socialized into action. 232 K.M. Moore and S. Cissé

Under-socialized ‘atomized actors’ have common relationships allowing individuals been placed in a context of self-regulating and groups to draw on resources not found economic structures where social relations within the primary group. These ‘weak ties’ between them have no meaning. In over- provide the bridges for coalition building and socialized conceptions, ‘atomized actors’ establishment of broad based alliances for have been restricted from acting as individ- community integration and development. uals making choices within a context of The model thus constructed (see Fig. social relationships and roles. However, as 16.1) allows for measurement and analysis Granovetter notes, not only of social capital in local com- munities, but also of social capital at the level Actors do not behave or decide as atoms of the nation-state. At the micro-level, outside a social context, nor do they adhere Woolcock (1998) characterizes the dimen- slavishly to a script written for them by the sion of embeddedness as integration, refer- particular intersection of social categories ring to intra-community bonding through that they happen to occupy. Their attempts ties of solidarity and trust, and the dimen- at purposive action are instead embedded sion of autonomy as linkage, referring to in concrete, ongoing systems of social bridges between primary groups forming relations. (p. 487) intra- and inter-community networks. At the macro-level, these dimensions are described His solution to this dichotomy was to take a as synergy (see also Evans, 1996), referring to middle road and ‘embed’ actors within state–society relations, and organizational structures of social relations where choice of integrity, referring to corporate coherence action is an option. and capacity at the national level, respec- For Granovetter (1985), development is tively. Figure 16.1 presents common charac- a matter of a change in kind, not degree, of terizations of the conditions generated by the embeddedness that is the social ties, cul- intersection of these dimensions of social tural practices and political relations which capital. The subsequent analyses explore form the basis for solidarity, trust, member- these dimensions. ship, etc. Building on this insight, Woolcock At the micro level, high levels of both (1998) argues that it is necessary to take into embeddedness (integration) and autonomy account the negative effects of social capital (linkage) yield conditions of social opportu- identified by Portes and Landolt (1996) and nity conducive to healthy community devel- others. There are costs and benefits to social opment. On the other end, low levels of capital and they need to be measured in integration and linkage yield conditions of order to capture the multidimensionality of amoral individualism, where each individ- the concept. The issue is not a matter of ual looks out for himself or herself and trusts more or less social capital, but of what kind no one and extra-local social relations are not of social capital. Woolcock proposes a com- developed. High levels of linkage with low plementary concept to embeddedness, that levels of integration generate conditions of of autonomy (often referred to as ‘bridging anomie, where extensive extra-local social social capital’, Warren et al., 2001; or ‘weak relations exist, but social bonding to a pri- ties’, Granovetter, 1973). Social relations are mary community is minimal. Opposite this, not only embedded in various networks in is the condition of amoral familism where the social structure, but social relations have individuals are well integrated within their varying degrees of autonomy across those communities, but not well linked with the networks. exterior world. Autonomy refers to levels of exclusivity At the macro level, developmental and universality, the capacity to reach out states are generated by high levels of synergy and access resources beyond the immediate and organizational integrity, whereas anar- group. In an earlier work, Granovetter (1973) chy results from low levels of each. High had highlighted the importance of network levels of organizational integrity combined connections to others who have few other with low levels of synergy yield only Social Capital and Improved NRM 233

Embeddedness Autonomy Linkage Micro Level Low High

Low Amoral individualism Anomie Integration High Amoral familism Social opportunity

Organizational Integrity Macro Level Low High Anarchy Inefficiency, ineffectiveness Low (Collapsed states) (Weak states) Synergy Predation, corruption Cooperation, accountability High (Rogue states) (Developmental states)

Fig. 16.1. Consequences of social capital, applying the concepts of embeddedness and autonomy at the micro and macro levels (from Woolcock, 1998). weak ineffective conditions for state/society Commune of Madiama. The initial sample development. High synergy, but low organi- design involved a two-stage process. In the zational integrity yields conditions ripe for first stage, five of the ten villages in the predation and corruption. commune were selected as clusters to These dimensions of social capital pro- represent a broad cross section of the com- vide the conceptual tools with which to munity, assuring that villages of pastoralists, analyse the evolution of civil society within sedentary farmers, and agro-pastoralists rural Mali and evaluate both the national would be included. Households were ran- policy of decentralization, as well as the domly selected from 1996 village census strength of local institutions. Trust and the lists obtained from the local administrative expectation of reciprocity have been most authorities. These households were initially frequently identified in the functioning of interviewed in March of 1999 for an eco- small groups, families or communities. In the nomic analysis of household resources and case of rural Mali, trust and the expectation production activities (Brewster et al., of reciprocity are structured ethnically Chapter 13 this volume). A socio-institu- within the context of village clans and orga- tional follow-up survey with both house- nized by village elders as represented by the hold heads (male) and their leading wives village chief. Embeddedness in rural Mali is was conducted later that year in September. manifested through family and clan solidar- After a pre-test of the questionnaires, men ities, and autonomy by the density of bridges and women interviewers were trained in between these primary groups, as manifested both the content and objectives of the survey, by membership in CSOs. as well as in how to elicit standardized responses. Of the initial 120 households, the first round of the socio-institutional survey Analysis of Social Capital Formation participants included 118 household heads and 120 leading wives for a total of 238 inter- The panel survey sample views. In January 2003, the panel of house- holds was re-surveyed with some minor Data for this analysis come from panel sur- attrition, with 115 household heads and 114 veys of sample households resident in the leading wives completing the interviews. 234 K.M. Moore and S. Cissé

This sample represents approximately 10% become proactive in addressing major NRM of the household population in the com- issues. In this analysis we investigate factors mune. The total population of the commune critical to the social mobilization necessary was 7771 in 1999, rising to 7973 in 2001. for that endeavour. This panel was initially designed to Chapter 7 described the formation of the analyse differences between four different NRMAC as an institutional vehicle for com- production systems: those of farmers (pre- munity deliberation and management of nat- dominantly involved in crop production ural resources. The NRMAC initiated and with only a few animals), agro-pastoralists has been involved in many NRM activities. (substantially involved in both crop and This living experiment allows for addressing livestock production), sedentary pastoralists a number of questions. Three issues of inter- (predominantly herder, but staying in the est here are: (i) the relative strength of civil village to produce crops), and transhumant society in Madiama; (ii) the extent of social herders (technically resident, but predomi- capital that civil society organizations can nantly involved in transhumant livestock draw upon and its distribution within the management). In each village, the village social structure of the community; and (iii) chief was contacted and a meeting held with the extent to which the NRMAC established farmers and herders informing them about confidence and credibility in the eyes of the survey and the information that was community members and thereby enhanced being sought. In the process, the villagers the strength of rural civil society. were asked to determine the category into which each randomly selected household fell and the consensus category thus elicited Social capital was recorded on each questionnaire. Farmers, whether agro-pastoralist or The survey provides a solid data set to inves- not, are predominantly Marka. Herders, tigate the level and quality of connectedness whether sedentary or transhumant, are of social relations between and among predominantly Peul. However, over time groups in the population. Association mem- it appears that herders are building a bership (networks) and confidence in local sedentary perspective, with more of them institutions (trust) provide indicators to self-identifying simply as farmers or agro- interpret the level of embeddedness and pastoralists in the 2003 survey. The few autonomy of social relations in order to remaining Bozo or Dogon in the sample characterize the quality of social capital in have blended their identities within their the Commune of Madiama. Autonomy is particular village settings. One of the five measured by the level of respondent’s mem- official villages is composed of two hamlets bership in local organizations. Integration is settled at different times by Marka and indicated by the degree of confidence (trust) Peul, respectively. Because these villagers respondents have in village and community see themselves as distinct from one another institutions and their leaders. and they demonstrate different response sets they are treated separately in the fol- Membership lowing analysis. There are a number of different types of formal and informal associations to which someone Framing the analysis can belong. The most common membership for adult men is the tonba (or large village The primary hypothesis of the SANREM association). Other village-level associations CRSP Project in Mali has been that: When a include the Hunters’ Association (often with local population is provided with a method- occult restrictions on membership) and the ology for resource management and conflict Foresters’ Association. Although women’s resolution and an institutional vehicle for membership in the tonba has always existed inter-village, inter-ethnic dialogue, it can in principle, only recently has recognition of Social Capital and Improved NRM 235

women’s roles beyond the household entered bership. Men’s participation is slightly into village discourse and independent higher with 45% participating in one associ- women’s associations have formed. At the ation versus 24% without any membership. commune or inter-village level, associations Increasing membership indicates a higher have been formed around productive activi- level of connectedness, but the depth of par- ties, such as the Association of the Herders of ticipation was not determined. Nérékoro and the Rice Producers Association Association membership also varies by of the Casier du Syn.ASchool Parents village and ethnicity. The numerically supe- Association and a Health Association (linked rior Marka clearly dominate organizational to the government health service delivery memberships, composing nearly the entire system) also have formed over the past membership of commune level organizations decade. The NRMAC at the commune level such as the Association of the Casier du Syn with its constituent Village NRM User Groups and the NRMAC, as well as village level asso- is the most recently formed. ciations such as Hunters’Associations (cul- Among these primary associations, turally a Bambara/Marka tradition). Although men’s association membership has increased the Herders’ Association of Nérékoro is minimally between 1999 and 2003 (Table largely Peul, it still has a significant number of 16.1). However, women’s participation in Marka members. Participation in village and associational life has increased significantly. women’s associations is more balanced, In 1999, women did not actively manifest reflecting the Marka/Peul composition of their membership in the village association. each village, although minority groups (e.g. In 2003, 42% of women respondents claimed Bozo, Bambara, Dogon) tend not to be mem- such membership. Membership in women’s bers. One exception should be noted here. In associations has also increased from 33% to the administratively twinned village of Tatia 50% since 1999. Overall, 40% of the women Nouna, the Marka (Tatia) clans are more likely respondents claimed membership in one to be association members than the Peul association versus 33% without any mem- (Nouna) clans (more about this below).

Table 16.1. Changes in association membership by village.

Men’s average number Women’s average number of of association memberships association memberships Marka in Village 1999 2003 1999 2003 Village (%)

M 1.10 1.12 0.30 1.06 Madiama SD (1.03) (0.96) (0.47) (0.95) 97 n30 33 33 34 M 1.20 1.90 0.50 1.60 Tombonkan SD (1.32) (1.52) (0.53) (1.08) 60 n10 10 10 10 M 0.83 1.13 0.39 0.96 Promani SD (0.72) (0.87) (0.58) (0.93) 35 n23 23 23 23 M 1.12 1.04 0.43 0.70 Nérékoro SD (0.78) (0.98) (0.51) (0.64) 0 n25 25 23 23 M 1.62 2.69 0.62 2.00 Tatia SD (1.19) (2.21) (0.65) (1.08) 92 n13 13 13 13 M 1.27 1.00 0.09 0.09 Nouna SD (1.01) (1.18) (0.30) (0.30) 9 n11 11 11 11 236 K.M. Moore and S. Cissé

Trust Tombonkan (at the 0.01 level). For men, this difference carries over between the village In order to measure levels of trust in institu- of Nérékoro and those of Madiama and tions, respondents to the 2003 survey were Tombonkan (at the 0.01 level) as well. asked to identify their level of confidence in Women in Promani are more confident a list of local institutions over the past 3 in customary authorities than those of years on a five point scale (1ϭincreased a Nérékoro (at the 0.05 level). Promani and lot; 2ϭincreased some; 3ϭno change; Nérékoro are predominantly Peul villages 4ϭdecreased some; and 5ϭdecreased a lot). with the lowest levels of Marka residence in The list included the Commune Council and the sample. However, as the analysis of the the Mayor, representing local government; twinned village of Tatia Nouna demon- the village chief and religious leaders, rep- strates, confidence in customary authorities resenting customary authorities; and the is not necessarily a function of ethnicity. rice growers association (Comité de gestion This divided village exhibits extreme de Casier de Syn), the herders’ association confidence levels, with the Marka village (Association des Eleveurs de Nérékoro), and clans of Tatia expressing the greatest degrees the NRMAC, representing associations of of confidence. In fact, this pattern holds civil society. A factor analysis confirmed the across all measures of confidence in local separate grouping of local government, cus- institutions (local government, customary tomary authorities, and CSOs. society, and civil society at the 0.01 level). Overall, Peul villages have lower rates The Peul of Nouna are the least embedded of association membership. This may be group in their village. due to a tradition of greater individualism among the Peul (Wilson-Fall, 2000), and consequently, they are less likely to be join- ers. When the confidence of members in Table 16.2. Differences in mean scores for each of their associations is compared by confidence in customary society by village. ethnicity, Peul are statistically (at the 0.01 level) more likely to express confidence in Confidence in village level leadership. There is relatively customary society Village of more confidence among Marka in inter- residence Men Women village associations (with the exception of the Herders’ Association of Nérékoro). M 0.55ab 0.46e Nevertheless, there are no significant sta- Madiama SD (0.98) (0.98) tistical differences in confidence found n31 30 between Marka and Peul. This suggests a M 0.39cd 0.17f higher level of autonomy or linkage across Tombonkan SD (0.36) (0.48) networks for Marka than Peul. n10 10 M Ϫ0.47ac Ϫ0.68efg Comparisons of these perceptions of Promani SD (0.82) (0.88) trust by village reveals the extent to which n23 20 the population is embedded in their primary M Ϫ0.28bd 0.03g associations. The key indicator of embed- Nérékoro SD (1.06) (1.11) dedness is the expressed level of confidence n25 21 in customary society. Difference in means M Ϫ0.63h Ϫ0.68i tests applied to standardized village level Tatia SD (0.68) (0.59) scores for confidence in customary society n13 12 h i highlighted a distinctive pattern of statisti- M0.66 0.57 cally significant differences (Table 16.2). For Nouna SD (0.89) (0.69) n8 10 both men and women, greater confidence in customary society was expressed in the a, b, c, d, e, f, h, i All significant at the 0.01 level. village of Promani than in the central village g Significant at the 0.05 level. of Madiama and the neighbouring village of Positiveϭless trust; negativeϭmore trust. Social Capital and Improved NRM 237

Figure 16.2 graphically presents the Perhaps more instructive for the region two dimensions of social capital for each of are the other villages. The village of Madiama these villages. While the axis mid-points is the most populous and houses the seat of have been arbitrarily set at the standardized Commune governance. Factionalism divides mean values, this presentation provides the the different quartiers of the village. One opportunity to enrich the discussion with quartier, where recent in-migrants reside, is some qualitative insights from the typology. clearly being marginalized by the grand In the first instance, it can be noted that vil- families among the original settlers who lages are not homogeneous with respect to clearly dominated village life, and now are social capital, despite the fact that there are expanding their dominance to the commune. few differences in association memberships Relations between the NRMAC President between villages. Tatia is the only outlier, and the Commune Mayor, who are from which places it squarely in the midst of different quartiers of Madiama, were origi- social opportunity. This is most likely an nally tenuous. However, they have found anomaly driven by the forced integration of that they can work together for mutual two distinct primary groups in one official benefit. Nevertheless, this improvement of village. Nouna is at the other extreme of the relations does not appear to have filtered twofold characterization, amoral individu- down to others in their respective quartiers. alism. Such differences in self-perception Furthermore, the proximity to local gov- do not bode well for scaling up social capi- ernment resources and almost urbanized tal in Tatia Nouna. In response to this find- conditions appears to lead to low levels of ing, the NRMAC recently held a conflict confidence in customary authorities, but not resolution and consensus building work- necessarily significant network building shop for the combined village. through associational memberships. Indeed,

Fig. 16.2. Men’s social capital characterization of villages according to indicators (2003) of embeddedness (confidence in customary society) and autonomy (association membership). 238 K.M. Moore and S. Cissé

the characterization of amoral individualism capital between villages (the level of pri- is consistent with local gossip. Everyone is mary group loyalty), network building has out for his or her own good and coopera- increased, and for the most part, the strength tion/collaboration for the mutual achieve- of primary social ties has remained stable if ment of goals is difficult at best. not improved as well. The next stage of the Tombonkan provides an interesting analysis investigates the impact and impli- contrast. Although a small village with a cations of this positive movement in social predominantly Marka and farmer popula- capital. The working hypothesis is that tion whose lands are encroached on by the improved social relations along both dimen- neighbouring village of Madiama, it has ben- sions of social capital should have a direct efited from straddling a paved road, thereby and immediate impact on conflict and coop- gaining the attention of passing NGOs, and eration in the commune. has built up a relatively higher level of asso- ciational memberships. However, the same individual holds key positions in each asso- Conflict and cooperation ciation so that the benefits of building bridges to others are not reinforced by a Overall, from 1999 to 2003 the seriousness dense network. There have been several with which a list of 23 socioeconomic and occasions on which he could not participate environmental problems was perceived has in some activities, because he was busy on decreased, even though the year 2002 had another association’s business. Also, link- been one of the worst in the past decade with ages outside of the village are primarily lim- both low rainfall for rainfed agriculture and ited to him. a very weak inundation of the rice fields and The predominantly Peul villages of bourgoutière. Both men and women per- Promani and Nérékoro, although both char- ceived four problems that had significantly acterized in terms of amoral familism, show increased and shared three of them: the high promise in that only a little more par- drought, poor soils, and the lack of delta ticipation in secondary associations could flooding. Men also noted the insufficiency of generate a successful spiral into social pastureland, and women the lack of training opportunity. However, the strong, but insu- opportunities. Those aspects perceived as lar, solidarity of these communities and the improving for men included: reduced con- distrust among the dominant Marka popu- flict between farmers and herders, improved lation may inhibit such a shift. Promani was inter-community cooperation, education an old administrative centre of the area. Its opportunities, means of communication and central position has clearly shifted as labour availability, and reduced soil ero- Madiama has grown with its weekly market sion. Perceived improvements for women and siting of various services (school, included improved inter-community coop- health centre, rice association warehouse, eration and access to credit and firewood. IER outpost, other offices). However, the The interest here is in the improvement more inclusive membership of the Nérékoro in perceptions of farmer–herder conflict and Herders’ Association may have a counter- inter-community cooperation (Table 16.3). vailing effect and with its multiethnic mem- The pattern of decreased seriousness of bership is clearly the type of network farmer–herder conflict and improved inter- building that can strengthen social capital community cooperation was consistent within the Commune as a whole. across all villages between 1999 and 2003 with the following exceptions: the serious- ness of farmer–herder conflict increased for Analysis of NRM Conflict and Decision men in Nérékoro, and for women in Making Madiama and Nouna; the problem of inter- community cooperation increased for men Although the preceding analysis demon- in Tatia and Nouna, and for women, in strates variability in the quality of social Promani and Tatia. None of these differences Social Capital and Improved NRM 239

Table 16.3. Community perceptions concerning the same indicators as above to measure the seriousness of farmer–herder conflict and trust/confidence in civil society organiza- inter-community collaboration. tions and in local government. Perception of Respondents were asked who should be problems the responsible authority for local NRM con- regarding 1999 2003 flict resolution and decision making (Tables 16.4 and 16.5). Their responses were col- Conflict between Mena 3.50 3.24 lapsed into four broad categories to facilitate farmers and Women 3.19 3.06 the following analyses. The plurality of men herders and women believe that village institutions Inter-community Mena 2.31 1.95 (particularly the village chief) are best for cooperation Womenb 2.33 1.83 conflict resolution and decision making, but many are beginning to see a role for local a Difference in means significant at the 0.05 level. government. Only rarely did respondents b Difference in means significant at the 0.01 level. suggest that the state should still be respon- sible, and a few more felt that civil society was best situated to influence such decision was statistically significant. Significant dif- making and arbitration. ferences in perceived levels between vil- To assess community support for these lages, however, are notable. Overall, men institutions as makers of these decisions, the and women in Madiama and Tombonkan levels of institutional trust were compared. perceived that conflict between farmers and Among men, it is those who believe that herders and inter-community cooperation local government should be responsible for are less serious problems than their counter- NRM issues who consistently expressed parts in the herder village of Nérékoro per- high confidence in all types of local institu- ceived them to be, despite the perception of tions. Those men favouring village institu- improvements by all. tions expressed less confidence in each set of institutions. These differences are statis- tically significant for levels of trust in local Consensus over decision making government and in civil society. For women, responsibilities however, it is those few who believe that civil society should be responsible who Given that there has been an increase in tended to express the highest levels of con- social capital in the commune and it appears fidence overall. Only those women who to have reduced inter-community tensions, believe that local government should be the question shifts to how this translates into responsible tended to express higher levels the community’s perceptions of how to deal of confidence in local government as an with specific natural resource conflicts and institution. Although a clear plurality of management issues. Here the investigation both men and women believe that village focuses on how the social capital associated institutions should decide on NRM matters, with different local institutions is related their expression of confidence in village to perspectives on NRM decision-making institutions is weakest. responsibility. The working hypotheses are: Only one difference was noted with (i) for embeddedness, those who believed respect to the measures of autonomy. Men that customary authorities should decide preferring that local government be respon- natural resource issues will express high sible for NRM conflict resolution and deci- levels of confidence in those same authori- sion making were more likely to be members ties; and (ii) for autonomy, the linkage of of their village associations (at the 0.01 respondents to multiple local organizations level). There were no differences in levels of (networks) will be positively related to civil association membership for either men or society control over natural resource deci- women. The expectation was that participa- sion making. This comparative analysis uses tion in civil society organizations would 240 K.M. Moore and S. Cissé

Table 16.4. Men’s trust in local institutions according to who should be responsible for conflict and NRM.

Who should be responsible for conflict and NRM decisions

Level of trust in The state Local government Civil society Village institutions

Local government M a1.08a Ϫ0.39ab Ϫ0.09Ϫ b0.26b SD (1.35) (0.99) (0.99) (0.87) n4 35 14 53 Civil society M 0.16 Ϫ0.36c Ϫ0.04Ϫ c0.21c SD (0.80) (1.04) (0.88) (0.97) n4 35 14 53 Customary society M Ϫ0.70def Ϫ0.08f d0.23d e0.06e SD (0.32) (0.98) (0.99) (1.04) n4 35 14 53 a, b, c, d, e t-test for difference in means significant at the 0.01 level. f t-test for difference in means significant at the 0.05 level. Positiveϭless trust; negativeϭmore trust.

Table 16.5. Women’s trust in local institutions according to who should be responsible for conflict and NRM

Who should be responsible for conflict and NRM decisions

Level of trust in The state Local government Civil society Village institutions

Local government M abc0.81abc Ϫ0.25b Ϫ0.08a c0.04c SD (0.49) (0.99) (0.58) (1.00) n5 35 11 46 Civil society M d0.66d 0.02 Ϫ0.58d 0.06 SD (0.63) (0.97) (0.87) (1.09) n5 35 11 46 Customary society M 0.00 0.07 Ϫ0.42Ϫ Ϫ0.02Ϫ SD (0.84) (0.96) (0.99) (0.97) n5 35 11 46 a t-test for difference in means significant at the 0.01 level. b, c, d t-test for difference in means significant at the 0.05 level. Positiveϭless trust; negativeϭmore trust. have differentiated confidence levels, but it 16.6 and 16.7 present measures of the two did not. This could be because associational dimensions of social capital for men and membership might be qualitatively different women, respectively. Essentially, the same for some members, particularly in the case observations can be made for both men and of non-voluntary membership in village women. Indicators of autonomy, that is, link- associations. ages external to the primary group (number of associations an individual is a member of and confidence in civil society organizations), are Perceptions of the NRMAC significantly higher for those having partici- pated in NRMAC activities. This is also so for The NRMAC has played a central role in the measure of confidence in local govern- diminishing inter-group conflict and enhanc- ment for women. ing conditions for improved community Table 16.8 shows that while the confi- (multi-village) NRM decision making. Tables dence inspired by the village chief and reli- Social Capital and Improved NRM 241

Table 16.6. Measures of social capital according to men’s participation in NRMAC activities.

Level of trust in Number of associations Local government Customary society Civil society

No knowledge of M a0.95a 0.10 Ϫ0.051 c0.33c NRMAC SD (0.77) (1.05) (1.10) (1.03) n41 37 37 37 Awareness of M b0.97b Ϫ0.00Ϫ Ϫ0.04Ϫ Ϫ0.07Ϫ NRMAC SD (0.81) (1.05) (1.01) (0.99) n39 38 38 38 Participation in M ab2.20ab Ϫ0.11Ϫ 0.10 cϪ0.27cϪ NRMAC SD (1.80) (0.91) (0.89) (0.91) n35 35 35 35 a, b, c Difference in means significant at the 0.01 level.

Table 16.7. Measures of social capital according to women’s participation in NRMAC activities.

Level of trust in Number of associations Local government Customary society Civil society

No knowledge of M a0.77a c0.19c 0.11 b0.27b NRMAC SD (0.76) (1.00) (1.09) (0.96) n71 64 64 64 Awareness and M a1.48a cϪ0.30cϪ Ϫ0.18Ϫ bϪ0.44bϪ participation in SD (1.17) (0.94) (0.81) (0.91) NRMAC n 42 39 39 39 a, b Difference in means significant at the 0.01 level. c Difference in means significant at the 0.05 level. gious leaders is significantly superior to that their capacity to take advantage of opportu- of the NRMAC, the NRMAC inspires greater nities to enhance their well-being. People confidence than other local institutions of who trust each other and cooperate for a spe- civil society or local government. This is cific purpose may have a general resource clearly so for men. A parallel survey of vil- available for other cooperative endeavours. lage chiefs, elected officials, development However, unlike money, social capital is not agents and CSO leaders confirms this find- a universal resource, anonymous and fungi- ing (Touré, 2003). ble; it is tied to specific organizational forms for specific purposes (Warren et al., 2001; Evans, 1996). However, societal well-being Discussion is dependent on a shift from exclusive loy- alty to primary groups to networks of sec- Although social capital exists to some ondary associations (Narayan, 2002). The degree in all communities, its strength and problem of social capital for development breadth can vary. In addition, it can be either purposes is in ‘scaling up’, that is, how to viewed positively or negatively for poverty combine the strength of social solidarities at alleviation, NRM or economic development the village level across a broad set of vil- according to how the community fits into lages. the greater socioeconomic context. Through One of the constraints to this transition collective action, poor groups can increase can be ethnolinguistic fragmentation (Rodrik, 242 K.M. Moore and S. Cissé

Table 16.8. Comparison of levels of confidence in higher for those seeking NRM leadership various local institutions. from local government or civil society than for those seeking it from customary authori- Men Women ties. This shift in confidence in local gov- Religious leaders 1.89a 2.20 ernment and civil society appears to be (0.78) (1.00) associated with the NRMAC. Village chief 1.93b 2.00g This suggests that the population is (0.98) (1.02) divided, or possibly simply confused in their abcdef ghi NRMAC 2.17 2.38 primary loyalties. Civil society is emerging (0.84) (0.82) built on strong ties embedded at the village Mayor 2.44c 2.39 level and new bridges linking villages. But (1.08) (1.02) Rice Growers’ 2.54d 2.83h not all villagers are equal participants, and Association (0.85) (0.71) some villages have more positive social cap- Herders’Association 2.56e 2.78i ital than others. Whether this new rural civil (0.87) (0.79) society will be more inclusive of minority Commune council 2.61f 2.49 interests is unclear. Women are still tied to (1.06) (0.98) the village level loyalties, because these are N 110 103 most accessible to them. Herders, repre- sented by the villages of Nérékoro and a, d, e, f, g, h, i Difference in means significant at the 0.01 level. Promani, appear to have a strong solidarity b,c Differences in means significant at the 0.05 base at the village level as well, but could level. easily be isolated by the farmer majority at the commune level. Although we have taken a bottom-up approach in this analysis and in practice, it 1997, cited in Narayan, 2002) that can also is also worthwhile to discuss top-down lead to income inequalities. Brewster et al. issues for a moment. Without State support (Chapter 13 this volume) document the dis- for innovation in social relationships across parities between household production sys- civil society, many of the types of changes tems leading to such social differentiation. beginning to occur would not be possible. Under these conditions, if the institutions of Discussing relations between public and pri- conflict management are weak, exogenous vate actors, Evans (1996) notes that ‘effective shocks can easily trigger conflicts. The analy- states deliver rule-governed environments sis of social capital suggests that the relatively which strengthen and increase efficiency low level of confidence expressed in custom- of local organizations and institutions’ ary authorities is indicative of the need for (p. 1120). The State can contribute to social new mechanisms for conflict resolution to capital formation by providing complemen- emerge. tary resources of rules, information and Most of the Madiama population still infrastructure. It can most effectively do so perceives NRM decision-making as neces- when (at least some) State actors are embed- sarily resting at the village level. However, it ded in the local community in which they is these same people who express less con- work and actively assist in the co-production fidence in customary authorities than those of collective goods. He goes on to cite numer- who would have the local government regu- ous examples of such synergetic relations late NRM issues. Although it is not possible and how social capital may inhere in state- to determine from such cross-sectional data, private sector relations. Interestingly, the there appears to be increased confidence in only negative example is the single African local government, a product of decentraliza- case cited by Ostrom (1996), who describes tion. However, it is possible that increased the over-centralization of the Nigerian state inter-village networking has begun a shift bureaucracy as inhibiting the potential for towards increased confidence in local public–private sector synergy. Although the government. Certainly, overall confidence is rigidity of over-centralization is problematic Social Capital and Improved NRM 243

for African development, the present analy- all the village chiefs of the Commune of sis highlights the countervailing force Madiama questioned concerning their per- which, in part incites it, but most certainly ception of the creation of the rural com- complicates the situation. To be successful, munes (decentralization) noted that it had decentralization must come to terms with led to the breaking of relations with the the local customary state whose insular envi- Commandant (sous-prefet), that is, with the ronment left on its own will not transcend administration (Touré, 2003). Those same local loyalties. village chiefs also noted the positive con- Can social capital be constructed or tributions of the NRMAC to commune life. reinvigorated? We have demonstrated the This analysis of social capital demon- potential to do so. Evans (1996) suggests strates the importance of building on tradi- three elements for consideration in this tionally valued social relationships and regard: people’s perceptions of themselves combining these with linkages across groups shape the social structure and these are mal- for the management of various conflict situa- leable; the soft technologies of organiza- tions (involving land tenure or not). In this tional design can have a major influence; and way, viable negotiated solutions can be collective difficulties can be addressed achieved and a new social contract realized, through redefining the problem in ways con- thus scaling up social capital. ducive to collective action (see Chapter 8). We conclude that donor/NGO emphasis The experience of the NRMAC (Chapter 7) on building village level associations, while shows that it indeed constitutes a new ‘soft successfully mobilizing local resources for technology’. Shifting from exclusive loyalty development has three drawbacks: (i) the to primary groups to participation in sec- scale is too small for the costs of extended ondary associations requires the develop- replication; (ii) village social capital is too ment of new skills and the learning of new insular for these associations to have a trans- behaviours. It has allowed villagers to adopt formative effect on rural social structures new identities at the Commune level, and dynamics; and consequently, (iii) these thereby reshaping their self-perceptions. associations can easily revert to neo-neo- Training in conflict resolution has provided colonial mechanisms of divide and control the committee with self-confidence, a means and for the extraction of surpluses from the to collectively reframe NRM problems, and rural population. For rural civil society to the capacity to share this message with vil- grow, linkages between villages must be lage groups. developed, and citizen networks estab- lished. In particular, we recommend the reinforcement of all commune-wide associ- Conclusion ations which multiply the ties between agriculturalists and pastoralists. We must Has a broad-based local consensus over qualify this in regard to the development of resource use been established? No. However, women’s role in rural society. Often con- the building blocks for an autonomous civil strained to remain in the village by tradition, society are beginning to emerge. On this basis women’s village associations are serving to an inclusive approach can be developed mobilize women in their own struggles for locally to resolve the impasse of legal plural- improved quality of life. These opportuni- ism. Through a sector-specific (NRM) initia- ties for women to formally associate should tive, disparate groups are coalescing around be encouraged. broad-based, but narrowly defined organiza- Changes at the national level supportive tional themes. of independent CSOs are necessary in order Previously social capital was never that rural civil society can prosper and they explicitly mobilized due to the ignorance or are beginning to occur. Decentralization has contempt of the administration or the rigid- created the opportunity for the NRMAC to ity of its rules and procedures. It can now build traditional bonds at the village level be seen to have possibilities. For example, into a network of relations creating a modern 244 K.M. Moore and S. Cissé

tool at the commune level. The association is Mamdani, M. (1996) Citizen and Subject: the manifestation of a public good whose Contemporary Africa and the Legacy of Late bridging social capital allows it the auton- Colonialism. Princeton University Press, omy to function successfully, drawing the Princeton, New Jersey. strengths of customary associations into Manor, J. (1999) The Political Economy of Democratic Decentralization. World Bank. rural civil society. Washington, DC. Narayan, D. (2002) Bonds and bridges: social cap- ital and poverty. In: Isham, J., Kelly, T. and References Ramaswamy, S. (eds), Social Capital and Economic Development. Edward Elgar Alexandre, P. (1968) The problem of chieftaincies Publishing Limited, Northampton, in French Africa. In: West African Chiefs: Massachusetts, pp. 58–81. Their Changing Status Under Colonial Rule Ostrom, E. (1996) Crossing the great divide: and Independence (Institute of African coproduction, synergy and development. Studies, No. 7). University of Ife, Ife, Nigeria, World Development 24(6), 1073–1087. pp. 1–42. Portes, A. (1998) Social capital: its origins and Blair, H. (2000) Participation and accountability applications in modern sociology. Annual at the periphery: democratic local gover- Review of Sociology 24, 1–24. nance in six countries. World Development Portes, A. and Landolt, P. (1996) The downside of 28(1), 21–39. social capital. The American Prospect 26, Castle, E.N. (2002) Social capital: an interdisci- 18–22. plinary concept. Rural Sociology 67(3), Putnam, R.D. (1993) Making Democracy Work: 331–349. Civic Traditions in Modern Italy. Princeton Coleman, J.S. (1988) Social capital in the creation University Press, Princeton, New Jersey. of human capital. American Journal of Ribot, J.C. (1999) Decentralization, participation Sociology 94(Suppl.), 95–120. and accountability in Sahelian forestry: legal Davis, J.U. (2000) Classrooms of democracy? The instruments of political–administrative con- educational prospects of Malian civil soci- trol. Africa 69(1), 23–65. ety. In: Bingen, J.R., Robinson, D. and Staatz, Ribot, J. (2002) African Decentralization: Local J.M. (eds) Democracy and Development in Actors, Powers and Accountability Mali. Michigan State University Press, East (Programme on Democracy, Governance and Lansing, Michigan, pp. 289–320. Human Rights, Paper No. 8). United Nations Docking, T.W. (1999) International influence on Research Institute for Social Development, civil society in Mali: the case of the cotton Geneva, Switzerland. farmers’ union, SYCOV. Ph.D. dissertation, Touré, B. (2003) Enquêtes Leadership Local. Boston University, Boston, Massachusetts. Sustainable Agriculture and Natural Evans, P. (1996) Government action, social capital Resource Management Collaborative and development: reviewing the evidence Resource Support Program, Institut de on synergy. World Development 24(6), l’Economie Rurale, Centre Régional de la 1119–1132. Recherche Agronomique, Mopti, Mali. Granovetter, M. (1973) The strength of weak ties. van Dijk, R. and van Rouveroy van Nieuwaal, The American Journal of Sociology 78(6), E.A.B. (1999) Introduction: the domestica- 1360–1380. tion of chieftaincy: the imposed and the imag- Granovetter, M. (1985) Economic action and ined. In: van Rouveroy van Nieuwaal, E.A.B. social structure: the problem of embedded- and van Rijk, D. (eds) African Chieftaincy ness. American Journal of Sociology 91(3), in a New Socio-Political Landscape. Lit 481–510. Verlag/Transaction Publishers, Hamburg, Guyer, J. (1994) The spatial dimensions of civil Germany. society in Africa: an anthropologist looks at Warren, M.R., Thompson, J.P. and Saegert, S. Nigeria. In: Harbeson, J.W., Rothchild, D. and (2001) The role of social capital in combating Chazan, N. (eds), Civil Society and the State poverty. In: Saegert, S., Thompson, J.P. and in Africa. Lynne Reinner Publishers, Warren, M.R. (eds) Social Capital and Poor Boulder, Colorado, pp. 215–229. Communities. Russell Sage Foundation, Kassibo, B. (ed.) (1997) La Decentralisation au New York, New York, pp. 1–28. Mali: État des Lieux (Bulletin No. 14). Wilson-Fall, W. (2000) Conflict prevention and LitVerlag, Hamburg, Germany. resolution among the Fulbe. In: Zartman, I.W. Social Capital and Improved NRM 245

(ed.) Traditional Cures for Modern Conflicts: and policy framework. Theory and Society African Conflict ‘Medicine’. Lynne Rienner 27(2), 151–208. Publishers, Boulder, Colorado, pp. 49–66. Wunsch, J.S. and Olowu, D. (1990) The failure of Woodhouse, P. (2003) African enclosures: a the centralized African State – Introductory default mode of development. World chapter. In: Wunsch, J.S. and Olowu, D. (eds) Development 31(10), 1705–1720. The Failure of the Centralized African State. Woolcock, M. (1998) Social capital and economic Westview Special Studies on Africa, development: toward a theoretical synthesis Boulder, Colorado, pp. 1–22.

17 What We Have Learned

Keith M. Moore,1 Charlene Brewster1 and Michael Bertelsen2 1Office of International Research, Education and Development, Virginia Tech, Blacksburg, VA 24061–0378, USA; and 2College of Agriculture and Life Sciences, Virginia Tech, Blacksburg, VA 24061–0334, USA

We have seen how resource allocations have of many natural resource management shifted from a largely extensive pastoral (NRM) responsibilities. However, the newly system to an expanding, but still extensive elected officials have been poorly prepared agricultural system. The expansion of crop for these new responsibilities. Recognizing production in the wake of the introduction this weakness, the SANREM project adapted of animal traction has had a degrading effect its programme to closely articulate with this on the natural resource base of Madiama. process by establishing a multi-village com- Continuous cropping with little or no fal- mittee at the commune level, the Natural lowing has depleted soil fertility and forced Resource Management Advisory Committee over-grazing of fragile range resources. (NRMAC). The initial mandate of the Furthermore, legal pluralism with respect to NRMAC was to act as a local interlocutor for land tenure has thwarted the transition to NRM research and development services more intensified production systems, except at the Commune level. Having matured in a few cases where powerful actors could and become an officially recognized civil easily realize marketable returns. For the society organization (CSO), the NRMAC majority, a new consensus concerning the now: (i) bridges the rural knowledge divide rules of the game has to be achieved for linking rural civil society with local gov- sustainable development to occur. The con- ernment and deconcentrated technical catenation of production systems, once com- services; (ii) provides advice and support plementary, has become contentious with to the Commune Council on sustainable increasingly frequent conflicts. Meanwhile, NRM issues; and (iii) builds social relations social capital has remained strong at the pri- and consensus between and among com- mary group level (villages and clans), but munities, researchers, and other relevant weak at the inter-community scale of rural stakeholders. civil society where new rules need to be Social infrastructure alone, however, established. cannot resolve the crisis of sustainable The introduction of the SANREM pro- growth; intensification of crop and livestock ject in Mali coincided with the initiation of production will also be required to assure major steps in government decentralization continued livelihoods for the rural popula- in the late 1990s. These steps involved tion. Examples of this intensification can establishment of Rural Communes, election be seen in the cash cropping of water- of Commune Councils, and the devolution melons and in cattle fattening operations. © CAB International 2005. Conflict, Social Capital and Managing Natural Resources (ed. K.M. Moore) 247 248 K.M. Moore et al.

Intensification is also evident in: selected CROPSYST Model simulating the long-run soil fertility enhancements; greater efforts to viability of potential new technologies now develop manure exchange relations; the provides information for agroeconomic transformation of Cassia tora into a valued decision-making from the commune to the forage crop through ensiling; and the devel- regional level. By integrating agronomic opment of more intensively managed research and economic analysis, Institut de grazing regimes. Although the transition l’Ecomomie Rurale (IER) researchers are also will be challenging to all, it will be particu- able to more quickly identify promising new larly hard on those pastoral populations technologies and farmers are able to under- who have lost access to the primary resource stand the economic implications of new on which their management strategy was cropping practices. The Social Accounting based – free movement through extensive Matrix (SAM) developed for Madiama pasturelands. Commune provides guidance to researchers During the 1999 Participatory Land- and commune leadership about strengths scape/Lifescape Appraisal (PLLA) farmers and weaknesses in the local economy, poten- identified soil fertility and water retention tial impact of new technologies, and how capacity as primary constraints to sustain- these impacts would be distributed between able livelihoods. Through extensive litera- different groups within the community. The ture review and consultation with farmers, socioeconomic database has also provided SANREM researchers identified a wide data to evaluate the extent of social capital range of low-cost soil fertility-enhancing formation and the impact of the NRMAC. technologies that have been tested and refined on farmers’ fields. Various combina- tions of crop rotations and soil amendments, Adaptive Learning through the SANREM incorporating farmers’ perceptions and pref- Approach erences (for example, including the use of local, rather than research-generated The four pillars of SANREM, participation, cultivars, and inclusion of small ruminant landscape/lifescape scale, multiple stake- manure which farmers believe to be the most holders and interdisciplinarity, provide the potent) were examined. Farm trials and com- frame of reference structuring this work in puter modelling evaluated the yield effect Madiama. The working hypothesis was that and long-term sustainability of different when a local population is provided with: applications of organic manure, natural (i) methods for natural resource and conflict phosphate, and chemical fertilizers, includ- management, and (ii) an institutional vehi- ing a locally adapted International Crops cle for inter-village, inter-ethnic dialogue, Research Institute for the Semi-Arid Tropics the population can become proactive in (ICRISAT) protocol for micro-dosing inor- addressing major agricultural productivity ganic fertilizer that minimizes farmer input and natural resource management issues. expenses. One consequence of these inter- There were three points of intervention in ventions is that the traditional manure- this multi-pronged approach: (i) addressing fodder exchange has been revitalized the manifestation of NRM problems through through overnight paddocking of animals on directly developing the capacities to con- fields. front and resolve conflicts and the underly- The SANREM project also established ing factors which shape the conditions data and information products and decision generating those conflicts; (ii) introducing support tools, including a household level new technologies and strategies to increase socioeconomic panel survey database and the efficiency and sustainability of the nat- a multi-year, geo-referenced database of ural resource base; and (iii) developing weather, soil, cropping system, land use and social networks to increase local power in water point data. This latter database has the governance of natural resources. been used for constructing maps and model- In addition to being a symptom of under- ling at the commune level. The calibrated lying causes, conflict has an independent What We Have Learned 249

impact on both biophysical resources and the decision-making process. . . On the one the social relations governing resource use. hand, the technicians acted in accordance In the first instance, resources are over- with a conventional technology-transfer exploited in a competition to gain some model, whereby they presented themselves benefits before or instead of others leading to as providers of know-how to local people. On the other, local people, constantly seeing rapid degradation. More importantly, latent their own frame of reference, knowledge conflict inhibits groups from coming and know-how disregarded, came to suffer together to resolve their collective problems. from a lack of self-esteem, and adopted in By providing tools and capacities to manage some places what might be referred to as a conflict, we hoped to increase the confi- ‘welfare mentality’. In other words, they dence of the population in itself and in its tended to turn more and more to outside neighbours. agencies for solutions to local problems. The same tools used to resolve conflicts (Guèye, n.d., p.1) also helped to develop organizational skills for consensus building (human capital This paternalistic approach placed the local investment) and have led to the establish- population in the role of passive receivers of ment of a new form of social infrastruc- a project’s largesse. To the extent that par- ture bridging relationships between local ticipation existed, local people simply took communities (social capital investment). on the role of middlemen for the capture Through inter-village networking, sharing and/or distribution of project benefits. The of knowledge, tree planting and developing idea that an independent local leadership co-management accords, social capital was could be capable of deciding priorities was built and this led to further consensus build- never considered. ing and the trust necessary for wider dis- Into this local social, economic, and cussions throughout rural civil society. In political context the SANREM CRSP–West addition, we linked the communities with Africa Project introduced its approach to scientific research services to identify and locally controlled NRM. SANREM sought introduce technologies and strategies that to be inclusive of everyone in the have the potential to increase the productive Commune, as individuals regardless of capacity of the natural resource base. occupation, gender, age or other characteris- The process of empowering local people tics. SANREM also included the associa- to take charge of their own economic and tions that had been formed to represent social development is slow and complex. various socio-professional groups, such as Participation, the core of the SANREM rice cultivators, hunters, and pastoralists. approach, is not only inherent in many of the Representatives of local government, tech- tools introduced to assist local populations nical services and NGOs were all present in realizing their goals, but is also a develop- during the creation of the new civil society ment outcome in its own right. Let us con- organization, the NRMAC. Through real par- sider what we learned over the short ticipation in the decision making structures duration of the SANREM project in Mali. of local governance, local people would learn to behave in new ways that would empower them to take control over the man- Participation: towards a new vision agement of their natural resources as pro- vided for in the new decentralization laws. The first pillar addresses an inherited prob- lem, participation. Traditionally, develop- Social learning ment projects have approached local populations in a mode similar to that of Learning new roles does not unfold in a national governing structure: simple linear fashion. While villages enthu- siastically received the project into their A centralized and vertical approach . . . left communities, their previous experiences no room for local people to participate in with development projects shaped their 250 K.M. Moore et al. expectations. In other words, they expected Scale Issues to become traditional project beneficiaries. But this kind of passivity was counter to The second pillar of SANREM is landscape/ project objectives. During the relatively long lifescape scale. Scaling-up has two dimen- training process that followed, new skills sions: one social and one biophysical. The were developed to strengthen local leader- landscape/lifescape scale requires going ship capacities, but local enthusiasm beyond the village based gestion de terroir waned. While the cohort of local leaders approach. As scale increases, more stake- was being taught new skills, the necessary holders become involved and ecosystem reinforcement for the new roles they were limits expand as well. Gestion de terroir expected to fill was slow in coming. succeeded on the village level where face-to- The process of learning new roles is face relations between actors occur on a separate from the process of learning the daily basis. The decisions to be made at the skills to accomplish those roles. New roles landscape scale not only concern individual are only truly learned when the practice of adoption of production or conservation tech- those skills is reinforced through validation nologies, but also collective choices about by others and through the realization of land/resource use, whether on common material benefits. In rural societies where property or private lands. All the actors agricultural seasons shape life, the value using the same space must be taken into and meaning of these new roles cannot account. Here, accountability is an impor- materialize until at least a full annual cycle tant issue. Including stakeholders on a has been completed. Inter-village discus- lifescape scale higher than the village sions and networking to manage herder– involves gathering together those who do not farmer conflicts can only be tested with the have routine daily contacts, even though annual harvest season livestock passage they are involved in social relations through through the commune. economic exchanges or sharing the same The learning of new roles is also not a resources. one-sided affair. Roles are defined by rela- The NRMAC was designed with down- tionships. SANREM Project agents and their ward accountability in mind. Each village partners in the CRRA/Mopti, CARE/Djenné, chief designates members who are expected Groupe de Recherche Actions pour le to report back to village user groups on a Développement (GRAD) and other decon- routine basis. Indeed, as committee members centrated technical services have also been themselves have recognized, this reporting learning the new skills and roles necessary has posed serious difficulties. Village chiefs to empower local populations. Another full appear to receive reports, but broader dis- cycle of seasons must pass before these new semination throughout the village is often relationships can be consolidated into limited to informal and chance occasions, accepted behavioural practices where role with the exception of meetings held for representatives properly respond to their specific purposes, such as discussing and counterpart’s role. Understanding each negotiating a multi-village bourgoutière other’s new roles and building mutual trust management agreement. has been a central project task. At stake is For villagers, commune and regional the transformation away from the ‘welfare decision makers to think and act responsi- mentality’. This transformation is difficult, bly, they need knowledge and information because it depends not only on develop- about those outside their village terroir who ment agent behaviour, but also on the behav- share the same resource base and how these iour among and between segments of the resources can best be exploited. The bio- population. The new roles that the various physical, geographic and socio-economic groups in the population should play must modelling tools developed by SANREM be reinforced by the successes of their researchers have produced results that can actions and the manner in which their part- be extrapolated beyond the boundaries of ners respond to them in their activities. the commune. This is as true for technolo- What We Have Learned 251

gies adapted to the Madiama landscape connections and networking between all the types found throughout the Inland Delta as different groups. it is for intra-village dependencies within In recent years, two approaches to col- commune economies. lective NRM have been taken: one, targeting processes for establishing local agreements for co-management; and the other, targeting Multiple stakeholders/networking the development of human capital and social infrastructure to create the social capital nec- The SANREM pillar of involving multiple essary for successful co-management. While stakeholders invokes another dimension of one cannot be done without the other, participation, social networking. Encour- SANREM activities favoured the second aging participation of multiple stakeholders approach. Examples of the first approach are in community activities without precedent mentioned in the next section. models that the population and researchers We began with the assumption that have experienced is time consuming and social infrastructure can be developed to difficult. Scheduling and other obligations increase the chance that various ideas and hamper collaborative action, because it people can intensify open debate, increase takes time to coordinate the activity sched- the circulation of information, especially ules of different partners. Conflicting agen- technical information, and bring about das and obligations of the various partners more considered decisions (Flora, 1998). also inhibit the consensus building neces- However, skills in developing/negotiating sary for collaborative action. Behaviour pat- stakeholder relations and organizational net- terns of different partners need to coalesce works do not guarantee success in achieving for coherent programme implementation, inter-community management agreements. and their schedules need to synchronize so The SANREM experience clearly illustrates that the fundamental communication neces- the difficulties of achieving substantial NRM sary for collaborative participation can achievements in the short term, when the occur. Each aspect of the participatory early project focus is on developing the skills process requires negotiation among all par- to master the processes of decentralized and ticipants, and where possible structural participatory management. impediments, differential power relations The question becomes what social learn- and latent conflicts must be removed. ing proceeds in such a situation. The In the SANREM research and develop- SANREM experience validates observations ment programme, we experimented with of Ribot (1999) who noted that the processes various methods of building a new network of decentralization and participation can of relationships within rural civil society become perverted in the act of managing nat- between neighbouring villages, various ural resources. Instrumentalist approaches ethnic groups, production systems, the pop- to NRM inhibit empowerment. Although ulation, technical services, local govern- often characteristic of government and pro- ment and other partners. Providing a ject agents, the instrumentalist approach community with social infrastructure cre- may also be replicated by NGOs as they ates positive conditions for participation. To become project implementers. However, assure that all key stakeholders have the NGO facilitation has proven to be crucial to same level of information, training was first the success of civil society efforts. NGOs can provided to a wide range of development serve an important role by providing the agents and local officials before focusing ‘guide on the side’, instigating committee efforts on training commune level stake- members to step outside of their precon- holders. Communication of commune level ceived roles and become the community activities at the village level has been leaders and spokespersons for rural civil encouraged and has required vigilance in society, facilitating and monitoring commit- monitoring and follow-up efforts. The object tee activities, and coordinating human of this social infrastructure is to multiply capacity development among leaders and 252 K.M. Moore et al.

the community. However, differentiating cases, the NGO developed relations between facilitation for social empowerment from the local population and government service administration of a population is often diffi- agencies in support of co-management of cult. Government service providers have natural resources. However, each evolved a always had great difficulty maintaining an different social infrastructure. In Bankass, empowering stance as their formally defined the Projet d’Appui à la gestion de l’environ- role is as administrators of the population. nement (PAGE) formed no permanent inter- This raises the question of the proper village association; they simply facilitated roles and behaviours for NGO, administra- the linkage between government services tion and technical service agents. The and individual village associations. In the Malian government has taken an initial step zone of Kelka (near Douentza), the NEF by opening up the terrain for rural CSOs and coordinated with government services to the decentralized administration of services. support a traditional multi-village associa- It is important for local officials and techni- tion called Waldé Kelka. The CRRA/Sikasso cal service agents to cultivate relations with in collaboration with the forestry and exten- these CSOs. Following the conclusions of sion services helped develop a local forest Evans (1996), this synergy (the quality of the co-management agreement among three vil- interaction between state and civil society) lages. These cases confirm and reinforce the should develop in two dimensions. State following lessons learned from the NRMAC agents should focus on building relation- experience in Madiama. ships through providing goods and services complementary to those provided by the pri- Lesson 1: Including all stakeholders is a nec- vate sector. Further, state agents at the local essary but problematic task. All four pro- level need to become embedded in the com- jects experienced difficulty maintaining munity, building trust and the networks for sustained communications and assuring the increased social capital in rural areas. full participation of women and pastoral- ists (particularly non-resident pastoralists) in their activities and decision making Lessons Learned structures. Furthermore, participation is not limited to natural resource end users; cus- By learning new empowered roles, the tomary authorities, government officials, SANREM-trained local community leaders technical service agents and local NGOs have brought about behavioural changes in should also be included. Accountability and community relations leading to attitudinal sustained communication are necessary changes in the commune population as a reinforcements for stakeholder inclusion. whole. Future successes will be built on this cohort of local leaders, and their mobiliza- Lesson 2: There is no single model for build- tion of social capital. These new leaders have ing social capital and developing local man- learned to act on behalf of their communities agement agreements. The process must be in the development of sustainable and prof- iterative, allowing for the adaptive learning itable practices for decentralized NRM. that takes a considerable amount of time While SANREM has been working in (years) to bring all stakeholders together in Madiama, similar inter-village level experi- an informed and voluntary manner, and to ments in decentralized NRM have taken negotiate consensus on each component of place at the local level in Mali. In two cases, an accord. As we have documented, the they have been assisted by locally based process needs time to evolve and mature, NGOs: SOS Sahel in Bankass and the because building trust between groups is a Near East Foundation (NEF) in Douentza time-consuming process. (Bocoum et al., 2003). A third case, Siwaa, was supported by CRRA/Sikasso researchers Lesson 3: Project and partner personnel in collaboration with extension personnel need to be well trained to encourage open (Hilhorst and Aarnink, 1999). In all three debate, foster consensus building and guide What We Have Learned 253

without leading. Respect for and recogni- and applied in structuring the analysis and tion of local know-how is essential. This modelling in Chapter 11, build on farmer is something which we, as development insights noted by Earl et al. in Chapter 6 and agents can control, because respect and Crane and Traoré in Chapter 10. The teth- recognition depend on our own behaviour ered grazing research conducted by Abaye et regarding the population. This is as true for al. in Chapter 15 was shaped by interaction local leaders as it is for government service with the model of Holistic Management agents. Mistakes will be made and should be (HM) introduced in Chapter 9. The evalua- expected. They can be used as lessons. tion of technologies by Wyeth et al. in Training programmes in conflict resolution Chapter 12 relies on layers of interdiscipli- and consensus building have made major nary work from the field through several contributions in this regard. different laboratories. Interdisciplinarity assures dialogue Lesson 4: Power relations and stakeholder between multiple perspectives in the interests need to be carefully taken into process of holistic research. As researchers, account. If the interests of any stakeholder, we are obliged to choose an aspect, to iso- whether socially powerful or not, are late and examine it, and finally to apply ignored, full implementation of collective the findings to improve livelihoods. endeavours will be compromised. Training This reductionist approach is the way in conflict management can assist in bring- modern science advances, but it has also ing divisive issues into the open so that they produced monoculture cropping and the can be confronted and managed. weaknesses of an unbalanced landscape. Through interdisciplinarity, we have Lesson 5: Development agents must focus attempted to re-establish the balance in our interventions on developing synergy recommendations. However, it should be between the public and private sectors and noted that there are tensions between disci- on reinforcing the networks of local associ- plinary and holistic perspectives. It has ations, cultivating horizontal bonds across not always been an easy path to follow, communities, thereby facilitating the mobi- particularly across the divide between the lization of viable, locally defined initiatives. social and biophysical sciences. Some of This is the new agenda for research and these issues were highlighted in Bingham’s development of community life in West (Chapter 9) discussion of HM which advo- Africa. cated a unified perspective for SANREM that conceptually linked the biophysical and social sciences. The research/development process: Although it has provided a framework interdisciplinarity for discussion among researchers and development agents and with local nat- The final SANREM pillar, interdisciplinar- ural resource users, HM has been very diffi- ity, is a form of participation peculiar to cult to implement in practice. Scientists the philosophers and planners of develop- from divergent disciplinary perspectives ment. In this respect, the SANREM CRSP began to think about and understand some Landscape/Lifescape approach (amplified basic concepts across disciplines. However, by Holistic Management) has integrated these discussions often broke down into a local perspectives and allowed researchers multi-disciplinary parallelism, leaving to address real world problems in ways that researchers to integrate individual insights are meaningful to all natural resource users. into their own narrowly focused work. The Further, we have seen that listening to forming of joint projects did not really occur farmer perspectives improves the quality of until individual scientists had made disci- decision making tools (models) for higher- plinary advances of their own. level decision makers. The landscape types Ultimately, HM concepts did help Badini and Dioni introduced in Chapter 3 drive research directions, such as manure 254 K.M. Moore et al.

exchange, through demonstration of field conflict resolution, leadership development stabling of livestock and the identification of and empowerment, and by using science- grazing management indicators for holistic based technical expertise in agricultural pasture management. Also, by providing a research and social infrastructure develop- framework for community discussion and ment to create technological options and decision-making concerning NRM alterna- build social capital for local decision tives, these concepts helped lay to rest agri- making. culture versus environment debates and Despite its short history, SANREM can highlight complementarities and integra- point to significant project impacts, the tion. most important of which have been social This work may be criticized for missing and institutional. Building capacity through or ignoring important issues or not going conflict resolution has been empowering, as deeply enough into specific themes. Indeed, has the networking of villagers created by we have recognized these shortfalls and the NRMAC. Together these two interven- expect criticism, because our particular tions have led to increased social capital, disciplinary and professional perspec- building greater trust and ultimately lower- tives often focus attention on certain issues ing transaction costs for community devel- and blind us to others. However, multi- opment activities. The NRMAC is now disciplinarity (the foundation for interdisci- widely perceived as an important actor in plinarity) is in itself difficult to achieve given environmental protection and economic the resources available to such research pro- development for the Commune. Its role in jects. Inter- or multi-disciplinarity poses sustaining and developing gains made serious financial constraints on a programme through the project is evidenced by their because of the necessity to fund individu- support from within the community (col- als representing the number of disciplines lection of membership fees). The mandate required for adequate problem identification of the current NRMAC was renewed with and sufficient cross fertilization of ideas and their re-election to a new term and the sup- perspectives. Furthermore, there is a need port of the mayor and village chiefs. The for greater flexibility in financial manage- NRMAC is building social capital within ment of projects working to empower local the community and region by bridging rela- populations. The iterative process of bring- tionships between villages within and out- ing all stakeholders to consensus does not side the Commune, and with government follow a set schedule. services and local authorities. In one of its first activities, the committee organized and implemented the Mayor’s campaign for pro- Conclusion tection of the Acacia albida,anitrogen- fixing tree native to the Sahel, thereby The SANREM CRSP–West Africa Research directly impacting soil fertility and produc- Project has been a participatory programme tivity. The NRMAC is also orchestrating the building science-based solutions while pro- establishment of inter-village accords for moting a holistic approach for sustainable management of wetlands (bourgoutières) agriculture and natural resource manage- and for open-range rotational grazing. ment. Our work in Madiama is replicable in Training in HM has built the NRMAC’s other communes of Mali and throughout capacity to act as the primary mechanism West Africa. Indeed, we have already found for channelling farmers’ and herders’ input considerable interest among local leaders in in assessing soil fertility and pasture man- neighbouring communes. This approach agement technologies being experimented has consisted of institution building activi- with by the project. ties supported by agronomic and livestock More importantly, the NRMAC has management interventions, databases, and proved to be an effective catalyst for decision-support systems. The process rein- improvements in social relations between forces decentralized NRM by applying villages and ethnic groups within the What We Have Learned 255

commune. The leadership and socio- equitable outcomes. The CROPSYST Model institutional surveys indicate that there has and the SAM are providing decision- been a marked decrease in the occurrence of making information to researchers as well NRM conflicts and in people’s perception of as planners at various administrative levels conflict as a major problem in the commune by enabling them to analyse the long-term since the committee has been established. implications of various technology and On at least two occasions the NRMAC has policy options. intervened to avert escalation of conflict by applying concepts and skills learned in the conflict management training workshops. References The Sous-Prefet, responsible for several communes in the Cercle, has testified that Bocoum, A., Cochrane, K., Diakité, M. and Kane, the project’s institutional innovations and O. (2003) L’inclusion sociale pour une ges- capacity building interventions have tion équitable et durable des ressources enabled the successful uptake of decentral- naturelles: deux expériences au Mali. ized NRM responsibilities by the Rural Securing the commons, No. 7. SOS-Sahel Commune of Madiama to a much greater International Institute for Environment and degree than elsewhere. Development, London, UK, pp. 1–45. Evans, P. (1996) Government action, social capital NRMAC members have been trained in and development: reviewing the evidence livestock and soil management decision on synergy. World Development 24(6), making through the introduction of Holistic 1119–1132. Management as a diagnostic and manage- Flora, J. (1998) Social capital and communities of ment tool for NRM interventions. On the place. Rural Sociology 63(4), 481–506. basis of this training, the NRMAC has Guèye, Bara (no date) Whither Participation? defined a common holistic vision for devel- Experience from francophone West Africa. oping their community and its members MARP Sahel Programme, Dakar, Sénégal. now function as trainers and facilitators Hilhorst, T. and Aarnink, N. (1999) Co-managing building local capacity and commitment to the commons: setting the stage in Mali and Zambia. Bulletin 346. Royal Tropical attain their collective goal. In addition to Institute, Amsterdam, The Netherlands, the documented advances in local institu- pp. 1–77. tional capacity, the database and modelling Ribot, J.C. (1999) Decentralization, participation efforts have improved resource managers and accountability in Sahelian forestry: and planners’ capacity to make NRM Legal instruments of political-administrative decisions that result in sustainable and control. Africa 69(1), 23–65.

Index

Page numbers in bold refer to illustrations and tables

Acacia albida protection 97 husbandry 37, 53–70, 87 acid detergent fibre (ADF) 201, 203, 204 impact 125 Active Delta 38 seasonal movement (transhumance) 25, actors 3, 7, 182–3, 231–2, 242 60–7, 182–3, 187, 188 see also stakeholders see also cattle; livestock adaptive learning 108, 113, 248–50 Annual Work Plan, SANREM 87 Agrarian Reform (1985) 9 associations 90–1, 102, 230–1, 234–5, 237–8 agreement development 98 assumptions 162–3 agricultural practices effects 143 autonomy 232, 234, 236, 239 agricultural production 81, 82, 86–7 see also production systems agricultural research weaknesses 130 Bani-Niger Midlands 38 agricultural systems 36–7 Bankassi 68 agroecological characteristics 38 barter system 182 agropastoral constraints and potentialities 48 Bas-Plateau Bobo 38 agropastoral development conflict 1–19 behaviours, new 113 agropastoralists 182, 183, 187, 188 Betail Huicoma (ABH) feed supplement 55 see also farmers biophysical modelling 15–16 alluvial plains 42–8 biophysical viability 148, 153–6 Alternative Conflict Management Approach biosystems modelling 159, 160 (ACM) 101, 102, 103–5, 113–14 black sand 136 Alternative Conflict/Dispute Management botanical composition assessments 213, literature 103 216–24 American soil classification 132–3 botanical response, controlled grazing 211–26 amoral individualism 238 bourgou grass analysis of variance (ANOVA) procedures 207 conflicts over 84 analytical methods 160–1 decline 126 animal manure see manure description 19, 51, 127 animal traction 25–6, 37, 81, 83, 124 regeneration 96–7, 98 animals bourgoutierès 62, 64, 70, 96–7 controlled grazing performance 211–26 Bracaria sp. 216 health 56 Brazilian soil classification system 133

257 258 Index

brittle environments 119–20 coping process 108–9 broilers, Cassia tora seed impact 197–8 natural resources 81–2, 83–4 process questions 109 resource use 8, 9 carbon/nitrogen (C/N) ratio 42, 46, 47 conflict management CARE/Djenné 95 classification system 103 CARE/Mali training workshops 14 strategies 82–3, 84 Cassia tora 96, 195–207, 216, 218 tools and capacities 249 see also feed training 14–15, 90, 95, 101–15 cattle conflict resolution 6, 87, 98–9, 102–3, 111–12, feed supplements 59 243 grazing 55, 64 Conflict Resolution Consensus Building (CRCB) movements 53–7 101, 111–12, 113–14, 122–3 numbers 54 confrontation situations 8 pasture management 54–7 consensus building semi-transhumant herd management 59–60 conflict resolution 101, 111–12 transhumant management 60, 62, 63 decision making responsibilities 239–40 see also animals; livestock; transhumance development 249 centralization 229 Natural Resource Management Advisory Centre Régional Recherche Agronomique Committee (NRMAC) role 13–14 (CRRA)/Mopti 91 training 113–14 Chakunda 198 workshop 122–3 Change Process diagram 111, 112 constraints chef d’arrondissement 29–30 agropastoral 48 chemical fertilizers see inorganic (chemical) food production 195 fertilization natural resource management (NRM) 13 chemical inputs see fertilization participatory landscape/lifescape appraisal chiefs 82, 83, 91–2, 231 (PLLA) list 123–5 civil society 27–30, 31–2, 91, 229–31 production 46, 47, 49, 50 civil society organizations (CSOs) 2, 101, social capital development 241–2 230–1 soil fertility 129–40 see also Natural Resource Management sustainability 78 Advisory Committee (NRMAC) contested terrain 27–9 classification system, conflict management 103 contextual adaptations, module tools 111 classification systems, soil 45, 74, 131–4, 135–6, cooperation 238–9 139–40 Cooperative for Assistance and Relief climate 39–40, 134, 145 Everywhere, Inc. (CARE) 14, 95 see also rainfall; weather data corralling 137, 139, 161–2, 167–70, 177 CLIMGEN 145 see also grazing, controlled; tethered Commission of Pedology and Mapping of Soils grazing effect 133 corvée labour 231 Communaute Economique de Betail et de la costs 57, 164, 165, 177 Viande (CEBV) countries 4–5 see also prices commune formation 126 cowpeas 171–6 commune-level authorities decision making 82, crop 83, 84–5 /livestock, production relationships communication 97, 99, 117, 160 25–7 Community-Based Natural Resource forage 219 Management (CBNRM) project 6 prices 164 compensation 55 production expansion 247 competition 7, 26–7, 81–2, 102, 126 rotation 147–50 complementary interaction 25 sales 186 computer-based remote sensing 35–51 see also Cropping Systems Simulation conferences 2, 78 Model (CROPSYST); intercropping conflict cropping pattern 36 characterization 102 see also intercropping cooperation 238–9 cropping system defined 144 Index 259

Cropping Systems Simulation Model (CROPSYST) Echinoloa stagnina 19, 127 15, 144–57, 255 ecological drift theory 196 crude protein (CP) 195–6, 201–2 ecological soil classification 133 cultivars, new 3 economic currency 70 activities diversification 87 customary actors 182–3, 232 association structures 230–1 evaluation, soil management technologies authority 27–30, 229 159–79 conflict resolution 102–3 growth 181, 187, 247–8 land tenure 229 interdependencies 185–7 law 24–5, 27–9, 32, 84 linkages, occupational groups 181–94 power 90 modelling 15–16 society confidence 236 efficiency measure 147, 148, 161 elections 91 embeddedness 231–3, 236, 239, 242 decentralization empowerment 106, 107–8, 113, 249, 252 dialogue opportunities creation 31 energy 124 implementation 30 ensiling Cassia tora 198–9, 201–7 inter-village level experiments 252 entrepreneurs 27 land tenure relationship 2 erosion 82, 147, 148, 153–4 natural resource management (NRM) see also soil, loss dynamics change 89 ethnic groups network creation opportunities 243–4 amoral familism 238 promotion mechanism 229 association membership 235 see also Natural Resource Management dominant 35 Advisory Committee (NRMAC); farmers 234 power; village chiefs herders 234 decision making land use conflict 7 analysis 238–41 networks autonomy 236 commune level authorities 82, 83, 84–5 participatory landscape/lifescape appraisal improvement 2 (PLLA) composition 80 responsibilities, consensus 239–40 relations 23–5 science-based information 15 resource use conflict 8, 9 tools 15 social capital formation 234 traditional 230 territories 25 training 255 trust in institutions 236 village chiefs 83 ethnopedology 131, 139 village-based 28 experiential process, imitation learning deconcentration 30 106 defoliation 211–12, 223, 225 extension activities 3 degradation 14–15, 83 Delta Vif 38 democratization 30–1 facilitation skills 113 development agents, interventions focus 253 facilitators 104, 105–6 development assistance strategies 3–6 fanga 136 differentiation 27 FAO-UNESCO soil classification 133 Dina 24–5 farm trials tables 165 Djoula ethnic group prosperity 83 farmers drought relief 3–4 diversification strategy 182 dry season ethnic groups 234 animal health 56 /herders conflicts 238–9 cattle and pasture management 54–6 income increase 187–8 feed 54–5, 195–207 knowledge system 129–40 herding 55 major activities 182 shared resource use 182 microenterprise activities 187 stabling 55–6, 58 survey data 183 watering 55 see also agropastoralists 260 Index

feed 54–5, 56–8, 59, 195–207, 218 overgrazing 118–20 see also Cassia tora; forage rights 119 fertilization greeting circle 107–8, 113 costs 177 grounding 107 microdosing 177, 178 prices 163, 164 recommended rates 177–9 Hamdallahi regime (1818–1862) 24 trials 96, 170–1 harima 29 yield probability 152, 153, 154–5 health, animal 56 see also inorganic (chemical) fertilization; herd management 53–70 manure; PNT application; soil herders fertility cost 57 fieldwork, participatory landscape/lifescape ethnic groups 234 appraisal (PLLA) 78, 79–81 /farmers conflicts 238–9 financial management training 14 payment 59 fire 124 remuneration 56 fisherfolk migration 25 sedentary 54, 82 flexibility 120 visiting 53 flood plains 42, 44 holistic management (HM) flooding 41 conflict resolution 111 folk knowledge 129–40 four pillars 10 see also local knowledge grazing 124–5 folk model, soils 135–7 natural resources degradation 14–15 food 6–7, 81, 82, 85, 195 SANREM 117–27 Food and Agriculture Association (FAO) training 95, 121–3, 255 102 honouring 108 forage horizontal conflicts 8 crops 219 household wealth 27 diversity 215–20 householders, panel surveys sample 233–4 grasses 211, 212 hydrology 41 mass assessments 213 plant height before grazing 212, 214 resources 96–7, 195 imitation learning 106 stocks 83 in vitro dry matter digestibility (IVDMD) 195–6, see also bourgou grass; feed 201, 203, 204 forestry services 97–8 income increases 187–8 French soil classification 133 indicators 147, 148, 212 funding 99 information dissemination 99–100 information exchange 97 information sharing 80 General Assembly 92, 94 infrastructure building 89–100 geographic information systems (GIS) 35–51 inorganic (chemical) fertilization geology 37–8 application rates 160–1 geomorphology 37–8 comparison with nógó farafin 137 gestion de terroir 1, 5–6, 11, 19, 30–2 dosing 170–1 goats 57 microdosing 138, 151–3, 160, 170, 177 see also small ruminants practices 81 governance training 14 rates 37 Granovetter, M. 231–3 superior alternative 157 grass species 211, 212, 215, 216, 218–20 see also fertilization; manure; soil fertility see also bourgou grass; vegetation inputs 164, 165 grazing see also fertilization; labour cattle 55, 64 Institut d’Economie Rurale (IER), research controlled 211–26 partner 78 holistic management (HM) 124–5 institutional environment 90–1 maximization 37 institutional impact capacity building 14–15 opportunistic 71 institutional reinforcement training 95 Index 261

institutions 23–32, 236–8, 240, 242 literature review 78 integration (embeddedness) 231–3, 236, 239, 242 livestock intensification 248 /agricultural production competition 7 Inter Tropical Convergence Zone (ITCZ) 40 concentration zones 60–7 inter-institutional collaboration 77, 78 corridor 70 inter-state relations, natural resource conflicts 8 /crop, production relationships 25–7 intercropping 138, 160, 161, 171–3 decision making training 255 see also crop; cropping pattern exports 186 interdisciplinarity 10, 77–8, 253–4 herding 55, 57, 58–9 Interstate Committee for the Fight against management practices 61 Desertification in Sahel (CILSS) 2, 3 market development 4 interventions 3–6, 161–2, 248, 253 mobility 26 see also fertilization; intercropping; movement 53–70 rotation passage calendar 84 Interview Process training module 105 production systems survey 53–4, 119 inundation basin landscape 42 watering 55, 56, 57, 58 invasive species 196–7 see also animals; cattle; feed; small ironstone plateau 49–50 ruminants; transhumance irrigation projects 4 living organisms habitat 125 loamy plains 46, 48–9 local control see decentralization jowro 24, 28 local institutions 30–2, 240, 242 local knowledge 15–16, 130, 131, 136, 139 local management 13–14, 31, 252 Keïta regime (1960–1968) 29 Looking back at the Landscape: Impacts and Lessons Learned from SANREM-CRSP 78 labour 26–7, 162–3, 164, 165, 231 low Bobo Plateau 38 land low hardpan plateau landscape 49–50 allocation practices 102–3 degradation 14–15, 83 management 25 Madiama Commune 53–75, 85, 117–29, 181–94 tenure 2, 24, 28, 84, 229 manure units 43 acquisition through corralling 137, 150 use 7, 71–5, 143 applications 137, 157, 161 land-based production systems 26 collection 55, 161–2 see also agricultural systems; crop; contracts 25 livestock microdosing 138 landscape 41–51, 77–88, 250–1 natural resource management (NRM) see also participatory landscape/lifescape practices 82 appraisal (PLLA) with PNT 177 lateral basin landscape 42 prices 163, 164 laterized indurated pediment landscape 49, 50 strength 136 laws 9, 24–5, 27–9, 32, 84 trials 165–7, 168, 177–9 learning 106, 108, 113, 248–50, 252 types 151 see also training viable alternative 157 legitimacy, Natural Resource Management see also corralling; fertilization; inorganic Advisory Committee (NRMAC) 92–4 (chemical) fertilization legumes 211, 212, 216, 220 map drawing 84 levees and alluvial terraces landscape 42–8 Marka ethnic group 83 leydi 24, 29 mayors 91 lifescape 23–32, 77–88, 250–1 medicinal plant, Cassia tora 197 see also participatory landscape/lifescape men appraisal (PLLA) association membership 234–5 linkage (autonomy) 232, 234, 236, 239 autonomy preference 239 listening skills 106 confidence in local institutions 242 literacy training 14, 89, 95 conflicts with women 9–10 262 Index

men continued roles 13–14 decision making responsibilities consensus soft technology 243 239 trainers 107 division of labour 26–7 training 14–15, 94–5 NRMAC activities participation 241 villagers networking 254–5 perceived problems 238–9 see also civil society organizations (CSOs); social capital characterization of villages decentralization; power 237 natural resource management (NRM) trust in local institutions 240 committees 92 Méthode Accélèrée de Recherche Participative conflict and decision making analysis (MARP) methodology 79 238–41 microdosing 139, 151–3, 177, 178, 248 constraints 13 see also fertilization; inorganic (chemical) improved 229–44 fertilization; manure practices 81, 82, 83 microenterprise activities 183, 186, 187 related conflicts 6–10 millet needs 11–13 animal manure treatment 165–7, 168 Nérékoro chemical fertilization dosing 170–1 animal performance 220–2, 225 intercropping trials 171–3 commune-level authorities decision making night corralling trials 169 83 profits 176, 177, 178 conflict management strategies 82–3 rotation trials 173–6 forage diversity 215–16, 217, 218, 219 yield trials 149–52, 156, 176, 177 soil tests results 226 models and modelling transhumant livestock 68–9 biosystems 159, 160 village production 82 natural resource baseline information and villagers, primary problems identification modelling system 143 85 social capital 232, 233 networks 232, 234, 238, 239, 251–2 soils folk model 135–7 neutral detergent fibre (NDF) 201, 203, 204 soils scientific models 132–3 nitrogen 48, 156–7, 196, 207 see also Cropping Systems Simulation nógó 136, 137 Model (CROPSYST); holistic non-governmental organizations (NGOs) 5, 31, management (HM); social matrix 230 model (SAM) non-tradable sector 186, 188 monitoring and evaluation (M&E) system 16 Nouna 69, 70 monitoring research trials 95–9 NRM see natural resource management (NRM) Moyen Bani-Niger 38 multiplier analysis 185, 187–8 occupational groups, economic linkages 181–94 old landscape formations 47–50 National Agriculture Research Services (NARSs) on-farm trials 160 6 organic (manure) fertilization see manure National Domain Law 9 organic matter (OM) national legal system 103 addition 134 national policy, institutional context 29–30 content 46, 47, 48 natural resource baseline information and levels 42 modelling system 143 nógó 136 Natural Resource Management Advisory organizations 6 Committee (NRMAC) see also associations communication mechanism 160 outcomes 109, 110, 111 downward accountability 250 overtrampling 119 establishment 87, 247 formation 234 innovation 89–100 panchromatic arial photograph, Madiama and legitimacy 92–4 vicinity 72 mission 94 participatory approach perceptions 240–1 commitment 10 262 Index

men continued roles 13–14 decision making responsibilities consensus soft technology 243 239 trainers 107 division of labour 26–7 training 14–15, 94–5 NRMAC activities participation 241 villagers networking 254–5 perceived problems 238–9 see also civil society organizations (CSOs); social capital characterization of villages decentralization; power 237 natural resource management (NRM) trust in local institutions 240 committees 92 Méthode Accélèrée de Recherche Participative conflict and decision making analysis (MARP) methodology 79 238–41 microdosing 139, 151–3, 177, 178, 248 constraints 13 see also fertilization; inorganic (chemical) improved 229–44 fertilization; manure practices 81, 82, 83 microenterprise activities 183, 186, 187 related conflicts 6–10 millet needs 11–13 animal manure treatment 165–7, 168 Nérékoro chemical fertilization dosing 170–1 animal performance 220–2, 225 intercropping trials 171–3 commune-level authorities decision making night corralling trials 169 83 profits 176, 177, 178 conflict management strategies 82–3 rotation trials 173–6 forage diversity 215–16, 217, 218, 219 yield trials 149–52, 156, 176, 177 soil tests results 226 models and modelling transhumant livestock 68–9 biosystems 159, 160 village production 82 natural resource baseline information and villagers, primary problems identification modelling system 143 85 social capital 232, 233 networks 232, 234, 238, 239, 251–2 soils folk model 135–7 neutral detergent fibre (NDF) 201, 203, 204 soils scientific models 132–3 nitrogen 48, 156–7, 196, 207 see also Cropping Systems Simulation nógó 136, 137 Model (CROPSYST); holistic non-governmental organizations (NGOs) 5, 31, management (HM); social matrix 230 model (SAM) non-tradable sector 186, 188 monitoring and evaluation (M&E) system 16 Nouna 69, 70 monitoring research trials 95–9 NRM see natural resource management (NRM) Moyen Bani-Niger 38 multiplier analysis 185, 187–8 occupational groups, economic linkages 181–94 old landscape formations 47–50 National Agriculture Research Services (NARSs) on-farm trials 160 6 organic (manure) fertilization see manure National Domain Law 9 organic matter (OM) national legal system 103 addition 134 national policy, institutional context 29–30 content 46, 47, 48 natural resource baseline information and levels 42 modelling system 143 nógó 136 Natural Resource Management Advisory organizations 6 Committee (NRMAC) see also associations communication mechanism 160 outcomes 109, 110, 111 downward accountability 250 overtrampling 119 establishment 87, 247 formation 234 innovation 89–100 panchromatic arial photograph, Madiama and legitimacy 92–4 vicinity 72 mission 94 participatory approach perceptions 240–1 commitment 10 264 Index

rainfall continued sedentary patterns 39–40 agriculture 26 poor 81 farming 24–5, 81, 82 see also climate; weather data herder 54, 82 rainy season management practices 61 feed 56–7, 58 pastoralists 182, 183, 187, 188 pastures 66, 69 semi-transhumant herd management 59–60 pastures lack 53 see also transhumance shepherding 57, 58–9 sheep 57, 225 transhumance 66, 182 see also small ruminants watering 57, 58 shepherding, rainy season 57, 58–9 ranches 5 silage 198–9, 206–7 Rapid Rural Appraisal techniques 13, 119 Siragourou 216, 218–26 recent alluvial terrace landscape 42–6 site selection protocol 79 recent (young) formations landscape 42–7 skills, new 113 reforestation 97–8 small group organization 106 remuneration, herdsmen 56 small ruminants repetition learning 106 local herds 60 research protocol outline established 79 pasture management 57–9 research trials monitoring 95–9 sedentary, management practices 61 research/development process 253–4 semi-transhumant herd management 60 resources transhumance 64, 65, 67 allocation shift 247 see also animals; livestock management practices 85 social accounting matrix outline 189–93 mobilization 99 social capital 2, 89, 229–44, 247, 252 shared 181–94 social matrix model (SAM) 15–16, 181, 184–5, tenure 28 188–94 use conflict 8, 9 socialist era 29 rest 119, 124 socioeconomic modelling 15–16 Rethinking Range Ecology: Implications for soil Rangeland Management in Africa 119 alternative management technologies rice growing associations 90–1 159–79 root morphology 212 characterization 35–51, 131, 139–40 rotation 147–50, 161, 173–6, 178 classification 45, 74, 131–4, 135–6, 139–40 rules administration 23–32 components 132 Rural Communes establishment 91 database, CROPSYST 145 Russian soil classification 133 development 38 folk model 135–7 formation time 134 Sahel 1–19 grazing effects 226 SAM ( social matrix model) 15–16, 181, 184–5, loss 154, 155 189–90 see also erosion sand banks 42–6 management 130–1, 255 sandy plains 48–9 material movement 134 SANREM CRSP see Sustainable Agriculture and resources, World Reference Base (WRB) Natural Resource Management West 133–5 Africa project (SANREM-WA) scientific models 132–3 Sao, President 98 survey investigations 41 Savory, Allan 118–20 tests results 226 scale considerations, Cropping Systems types 41–51 Simulation Model (CROPSYST) 144 units 43 scale issues 250–1 water retention 136 schematic rendition 120–1 soil fertility Science article 100 controlled grazing 213 science-based information, decision makers 15 folk knowledge 129–40 seasonal movement (transhumance) 25, 60–7, improvement 95–6 182–3, 187, 188 low 82 Index 265

management 134–5, 146, 148–54, 156 total digestible nutrient (TDN) 201, 205 non-tradable sector 187 Toucouleur invasion 24 technologies 143–57 Toumadiama, transhumant livestock 67–8 see also fertilization; inorganic (chemical) traction 25–6, 37, 81, 83, 124 fertilization; manure tradable sector 185–7, 188 soil water erosion rates (SWER) 148 training sorghum 165–7, 169, 170 conflict management 14–15, 90, 95, 101–15 Sorghum Line 8 conflict resolution 95, 243 space, competition 26 consensus building 113–14 specialization 27 facilitators 104 stability 161 financial management 14 stabling, dry season 55–6, 58 governance 14 stakeholders holistic management (HM) 95, 121–3, 255 demographic distribution 183 individual capabilities development 89–90 inclusion 252 institutional reinforcement 95 interests consideration 253 Interview Process training module 105 involvement 11 literacy 14, 95 key 101 Natural Resource Management Advisory networking 251 Committee (NRMAC) 14, 94–5 see also actors organizational and leadership skills 95 State 1, 27–30, 90–1, 230–1 participatory approach 105 Striga hermonthica 160 requirement, project and partnership surveys 41, 53–4, 183, 233–4 personnel 252–3 sustainability tools, repetition and imitation 106 assessment modelling 161 training of trainers (TOT) 104–5 constraints 78 workshops 14, 104–6, 111–12, 122–3 development 89–100 transhumance 25, 60–9, 182–3, 187, 188 growth crisis 247–8 Traoré regime (1968–1991) 29 indicators 147, 148 treatments 162, 165–7, 176–177,201, 202–5 Natural Resource Management Advisory see also fertilization Committee (NRMAC) 13–14 trust 234, 236–8, 240 soil fertility management practices 153–4 Sustainable Agriculture and Natural Resource Management West Africa Project United States Agency for International (SANREM-WA) Development (USAID) support level Annual Work Plan 87 decision 88 approach 10–19 United States Geological Survey (USGS) Level I holistic management (HM) 117–29 classification scheme 74 lessons learned 78, 247–55 use rights 28 objectives 143 user groups development 92 primary hypothesis 234 programme description ix–x Working Paper 81 vegetation 38, 42, 44, 46–9 verbal territory establishment 107 vertical conflicts 8 tax collection 231 viability, biophysical 148, 153–6 taxes 83 village chiefs 82, 83, 91–2, 231 teaching techniques, Alternative Conflict village elders 29–30 Management Approach (ACM) 105 village-based, decision-making 28 Téguégné, transhumant livestock 69 villagers, primary problems identification 85–6 tethered grazing effect 225 violent conflict 98–9 see also corralling Voisin, André (1903–1964) 118–19 Thomas-Willey Mill 201, 207 Tombonka 68, 83–6 tonba 234 water 85, 86, 87, 136 tools 15, 107–8 watering livestock 55, 56, 57, 58 Torokoro, transhumant livestock 68 wealth 27, 83 266 Index

weather data 145, 213, 215 perceived problems 238–9 see also climate; rainfall resource management practices discussion weathering 133–4 85 weeds see Cassia tora; Striga hermonthica tenure security 9 Western European soil classification 133 trust in local institutions 240 women work plan process 87–8 association membership 234–5 workshops 14, 104–6, 111–12, 122–3 associations 91 World Reference Base (WRB) soil resources confidence in local institutions 242 133–5 conflicts with men 9–10 decision making responsibilities consensus 239 yield 148–57, 165, 176 division of labour 26–7 young landscape formations 42–7 ensiling and sampling procedures 200–1 food production constraints 195 land inheritance rarity 28 Zai (anti-erosion, moisture retaining rock land tenure insecurity 84 structures) 82 participation in NRMAC activities 241 Zornia sp. 216