MULTIPURPOSE TREES selection and testing for agroforestry IIv .., '..' . ... ........ . .,. kI ., *.L- Peter A. Huxley and Sidney B. Westley editors MULTIPURPOSE TREES: SELECTION AND TESTING FOR AGROFORESTRY Multipurpose trees and shrubs for specified agroforestry technologies and land-use systems-the ICRAF approach Peter A. Huxley and Sidney B, Westley, editors 2 The International Council for Research in Agroforestry (ICRAF) was established in 1978 with headquarters in Nairobi, Kenya. ICRAF is an autonomous, non-profit international research council governed by a Board of Trustees with equal representation from developed and developing countries. The mandate is to initiate, stimulate and support research leading to more sustainable and pro­ ductive land use through the integration or better management of trees in land-use systems. The Council derives its operational funds from voluntary contributions by several bilateral, multilateral and private organizations. In 1989, these included the World Bank (International Bank for Reconstruction and Development- IBRD), the African Development Bank, the International Fund for Agricultural Development (IFAD), the International Development Research Centre (IDRC), the Swedish Agency for Research Cooperation with Developing Countries (SAREC), the Ford Foundation and the Governments of Australia, Canada, Finland, France, the Federal Republic ofGermany (BMZIG TZ), The Netherlands, Norway, Sweden, Switzerland and the United States of America (USA). Published by the International Council for Research in Agroforestry ICRAF House, off the Limuru Road, P.O. Box 30677, Nairobi, Kenya. C International Council fr Research in Agroforestry 1989 All rights reserved. No part of this publication may be reproduced without written permission from the copyright owner. Printed by English Press P.O. Box 30127, Nairobi, Kenya. ISBN 92 9059 067 X FOREWORD Well over 20() woody species have been identified and listed as iseful 'multipurpose trees'- using this term in its broadest sense. A few are well knowi and have become widespread, others are so far only of local importance and their potential have yet to be fully explored. Early in ICRAF's development, initiatives were taken to collect information about these useful species and about the land-use systems and agroforestry technologies in which they were being used. But how do we match the potential of these species with the needs of the technologies, the land-use systems and the land users? At ICRAF we have focused on developing methods and 'tools' to achieve this match. This collection of papers was presented at a two-day technical seminar on the occasion of ICRAF's 10th Anniversary. The objective was to indic;ate the extent of work on multipurpose trees at ICRAF and the integration of this work within ICRAF's 'research­ for-dcvelopment' process. The contributions cover the steps required for the selection and testing of woody species for particular biophysical settings, agroforestry technologies and land-usc systems. If the outcome of the nIc:!-L'dologies described here do not provide all the answers, it is because we still do not have a great deal of information about multipurpose trees and their uses. The challenge for us all in the next decade of agroforestry research is to acquire and exchange present and new information, and to learn how to use multipurpose trees as effective y as possible :s components in tprropiatc agroforestry technologies. Wc would like to thank all those donors and organizations that funded participants in the technical seminar, and in particular the International Development Research Centre (IDRC), the Canadian International Development Agency (CIDA) and the Government of The Netherlands for their financial support. Bjoni 0. Lundgren Director-General ICRAF July 1989 CONTENTS Foreword 3 Contents 5 Introduction 7 Session 1: Analysing the Problems and Defining the Solutions 9 The diagnosis and design methodology, John B. Raintree 11 A case study from Zambia, DavidN. Ngu'i 19 Discussant's comments, K F. JWiersuin 25 Session 2: Preselecting Multipurpose Tree and Shrub Species for Particular 29 Agroforestry Technologies Matching trees with technologies using ICRAF's MPTS database, 31 PeterG. von Carlowitz Specifying tree characteristics: a case study from Burundi, DirkA. Hoekstra 37 Preliminary MPT selection for Burundi's central plateau region: 41 an example, PeterG. von Carlowitz Discussant's comments, Dominic E. Iyaniabo 47 Session 3: Formulating Agroforestry Research Programmes 51 Formulating agroforestry research programmes, PeterJ.Wood 53 Agroforestry research formulation: a case study of the initial stages 61 from Cameroon, SaraJ. Scherr Agroforestry research formulation: a case study of project implementation 75 from Cameroon, Bahint Dugwina Discussant's cemments, James L. Brewbaker 79 Session 4: Experimental Designs for Multipurpose-Tree Research 81 Experimental designs for multipurpose-tree research: concepts 83 and methods, Pete-A. Huxley Discussant's comments, F.J. Wangati 103 List of participants at ICRAF's Tenti. Anniversary Conference and 109 Technical Seminar 6 Semi I rid Lowlands LjHumid Lowlands [ZUriimodal Plateau SBimodal Highlands 0 '000 i(M G.At5, ICRAF in Africa: agroecological zones and countries where ICRAF.planning and/or Implementing agroforestrv research projects. As of mid-1989, projects are in progress in Burkina Faso, Burundi, Cameroon, Ethiopia, Ghana, Kenya, Malawi, Mall, Niger, Rwanda, Senegal, Tanzania, Uganda and Zambia. All projects are conducted in collaboration with national institutions through the Agroforestry Research Networks for Africa (AFRENA). INTRODUCTION This set of papers describes the position reached at iCRAF in 1987 after something less than 10 years of active scientific work related to a central part of the Council's programme-mullipurpose-trce research. It represents a statement of the progress made up to that time. From the start of research activities, it became clear that agroforestry research differed in some essential respects from research in agriculture or related disciplines. This volume helps to indicate what some [,fIlcse differences are. When work started, precise information about the woody components of agroforestry systems was scanty and/or mainly botanical or ecological in nature, with just a few notable exceptions such as the information available on Leucaena lettcocephala.At ICRAF, the priority was to obtain information about potential uses and appropriate management practices for species suited to particular agroforcstry technologies and ecological situ­ ations. This was undertaken primarily through two major projects--the 'Agroforestry Systems Inventory' and the 'Multipurpose Tree and Shrub Database'. Specialists in other areas also made important contributions. Establishing useful knowledge bases on multipurpose trees was clearly not enough. Another major problem for agroforcstry research was the limited understanding of the constraints affecting the land-use systems under study and of precisely in which ways, and to which extent, any agroforestry intervention could help remedy these. Thus, a major activity at ICRAF has been the development of a methodology for the rapid appraisal of land-use systems-ICRAF's 'diagnosis and design' methodology. Improving our know­ ledge of multipurpose trees and gaining an understanding of how they might be incorpor­ ated into land-use systems-these became two closely integrated activities. Accompanying all this was the development of a logical approach to agroforestry experimental research. How could the wide range of research tasks be structured and prioritized to promote relevant research programmes that could produce the required information in a cost-effective manner? In most cases, initiating this process has meant starting research on the introduction and testing of multipurpose trees. Finally, field experimentation in agroforestry inevitably encounters problems origina­ ting from the inherently different dimensions of space and time required for agricultural crops and for woody perennials. The complexities of managing trees and shrubs in order to provide different outputs and services also create problems. Important work at ICRAF has led to better understanding of appropriate field designs for agroforestry experimen­ tation and suitable methods for data analysis. To summarize work completed is often a useful way to establish where to go next. In planning the technical seminar, we soon realized that the completion of a holistic 're­ search-for-development' process involved several essential components. Some needed to be carried out in parallel (e.g. technology testing through 'prototype systems' design), others were part of the interactive cycle (e.g. research monitoring and evaluation proce­ dures). Summing up progress at the time of the 10th Anniversary was thus of considerable value for the future development of ICRAF's programmes. This exercise resulted in a 8 major emphasis on on-farm agroforestry research-again with its own set of essential components and procedures. Hopefully, the reader will discern not only an account of the past in this volume, but glimpses of what the future can bring. In most cases, we have attempted in the sec ions that follow to introduce the issues "nd then to show how research procedures have been developed, using practical fie!d examples from some of ICRAF's collaborative pro­ grammes. The material covered has bccn drawn from the work of many ICRAF staff, and is the result of a great deal of interdisciplinary interaction. The outc, ,me, we hope, is not just an account of the results of ICRAF's activities, but informative material for others involved in planning and implementing agroforestry research. Peter A. Huxley Filemon Torres, Seminar Organizers Session 1: Analysing the problems and defining the solutions ICRAF's 'diagnosis and design' methodology exposes the constraints of a land-use system and helps research planners choose agroforestry technologies to help overcome them. J.B. Raintree 11 THE DIAGNOSIS AND DESIGN METHODOLOGY John B. Raintree Research Developiment Division ICRAF The potential contribution of agroforestry to sustainable, integrated land use can only be realised if appropriate agroforestry technologies can be matched with specific land-use situations. ICRAF has developed adiagnosis anddesign (D&D) methodology to help achieve this goal. The D&D methodology can be used at the initial planning stage of an agroforestry project- for example by a multidisciplinary team charged with formulating research plans for national agroforestry programmes on the basis of rapid appraisal. The methodology can also be used iteratively throughout the implementation of agroforestry projects to refine the match between technology and land-use system. The author discusses the procedural aspects of the D&D methodology (asking the right questions) in relation to macro- and micro-level applications. Initial diagnosis leads to the identification of 'best-bet'protoype designs which are refined through testing and relevant on-farm and on-station research. The methodology is then used iteratively to help keep the research and development process on track towards the eventual optimization of the agroforestry design for the target land-use system. The substantive aspects of D&D (arriving at the right answers) depend on building up a knowledge base on agroforesty. Sources of information for agroforestry design are briefly discussed. BACKGROUND Agroforestry has great potential as an approach to rural development through integrated land use. Thcre arc many potentially useful ways to grow trees together with crops and/or livestock. However, there is a dearth of scientifically validated information on which to base the choice of suitable agroforcstry practices and systems for specific land-use situations or the selection of promising agroforestry technologies for further research. For any applied science, it is axiomatic that research and development efforts should focus primarily on technologies that hold pomisc for addressing important problcms. Following this principle, and in order to avokl squandering scarce resources on ad hoc, picccmcal rescaich projects, ICRAF set out in 1981 to develop a methodological tool to help agroforestry research and extension vw(rkcrs identify relevant research goals and formulate sound recommendations for agroforcstry development. This tool-thc diag­ nosis and design, or D&D, methodology- is nothing more, or less, than a systematic approach to agroforcstry planning based on the common-sense principle that 'diagnosis should precede trcatmcnt'. The D&D methodology was first developed in response to a need for a coherent interdisciplinary procedure, to be used by multidisciplinary teams on rapid appraisal