Front Cover
 Front Matter
 Title Page
 Table of Contents
 Executive summary
 Conceptual framework
 Target and research area selec...
 Problem identification and development...
 Planning on-farm research
 On-farm research and analysis
 Extension of results
 Deciding on an FSR & D approac...
 Back Cover

Group Title: Westview special studies in agriculture/aquaculture science and policy
Title: Farming systems research and development
Full Citation
Permanent Link: http://ufdc.ufl.edu/UF00097355/00001
 Material Information
Title: Farming systems research and development guidelines for developing countries
Physical Description: 414 p. : ill. ; 29 cm.
Language: English
Creator: Shaner, W. W ( Willis W )
Philipp, P. F ( Perry Fred ), 1913-
Schmehl, W. R.
Consortium for International Development
Publisher: Westview Press
Place of Publication: Boulder, Colo.
Publication Date: 1982
Copyright Date: 1982
Subject: Agriculture -- Research -- Developing countries   ( lcsh )
Agricultural systems -- Research -- Developing countries   ( lcsh )
Agriculture -- Developing countries   ( lcsh )
Agricultural systems -- Developing countries   ( lcsh )
Genre: bibliography   ( marcgt )
non-fiction   ( marcgt )
Bibliography: Includes bibliographical references (p. 401-405) and index.
General Note: At head of title: A Consortium for International Development study.
General Note: Westview special studies in agriculture / aquaculture science and policy
Statement of Responsibility: W.W. Shaner, P.F. Philipp, W.R. Schmehl ; editorial assistance, Donald E. Zimmerman.
 Record Information
Bibliographic ID: UF00097355
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: oclc - 08613783
lccn - 82050365
isbn - 086531389X

Table of Contents
    Front Cover
        Front Cover
    Front Matter
        Page i
        Page ii
        Page iii
        Page iv
    Title Page
        Page v
        Page vi
    Table of Contents
        Page vii
        Page viii
        Page ix
        Page x
        Page xi
        Page xii
        Page xiii
        Page xiv
        Page xv
        Page xvi
        Page xvii
        Page xviii
    Executive summary
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
    Conceptual framework
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
        Page 38
        Page 39
        Page 40
    Target and research area selection
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
        Page 53
        Page 54
        Page 55
        Page 56
        Page 57
        Page 58
    Problem identification and development of a research base
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
        Page 67
        Page 68
        Page 69
        Page 70
        Page 71
        Page 72
        Page 73
        Page 74
        Page 75
        Page 76
        Page 77
        Page 78
        Page 79
        Page 80
        Page 81
        Page 82
        Page 83
        Page 84
    Planning on-farm research
        Page 85
        Page 86
        Page 87
        Page 88
        Page 89
        Page 90
        Page 91
        Page 92
        Page 93
        Page 94
        Page 95
        Page 96
        Page 97
        Page 98
        Page 99
        Page 100
        Page 101
        Page 102
        Page 103
        Page 104
        Page 105
        Page 106
        Page 107
        Page 108
    On-farm research and analysis
        Page 109
        Page 110
        Page 111
        Page 112
        Page 113
        Page 114
        Page 115
        Page 116
        Page 117
        Page 118
        Page 119
        Page 120
        Page 121
        Page 122
        Page 123
        Page 124
        Page 125
        Page 126
        Page 127
        Page 128
        Page 129
        Page 130
        Page 131
        Page 132
        Page 133
        Page 134
        Page 135
        Page 136
        Page 137
        Page 138
        Page 139
        Page 140
        Page 141
        Page 142
        Page 143
        Page 144
        Page 145
        Page 146
    Extension of results
        Page 147
        Page 148
        Page 149
        Page 150
        Page 151
        Page 152
        Page 153
        Page 154
        Page 155
        Page 156
        Page 157
        Page 158
        Page 159
        Page 160
    Deciding on an FSR & D approach
        Page 161
        Page 162
        Page 163
        Page 164
        Page 165
        Page 166
        Page 167
        Page 168
        Page 169
        Page 170
        Page 171
        Page 172
        Page 173
        Page 174
        Page 175
        Page 176
        Page 177
        Page 178
        Page 179
        Page 180
        Page 181
        Page 182
        Page 183
        Page 184
        Page 185
        Page 186
        Page 187
        Page 188
        Page 189
        Page 190
        Page 191
        Page 192
        Page 193
        Page 194
        Page 195
        Page 196
        Page 197
        Page 198
        Page 199
        Page 200
        Page 201
        Page 202
        Page 203
        Page 204
        Page 205
        Page 206
        Page 207
        Page 208
        Page 209
        Page 210
        Page 211
        Page 212
        Page 213
        Page 214
        Page 215
        Page 216
        Page 217
        Page 218
        Page 219
        Page 220
        Page 221
        Page 222
        Page 223
        Page 224
        Page 225
        Page 226
        Page 227
        Page 228
        Page 229
        Page 230
        Page 231
        Page 232
        Page 233
        Page 234
        Page 235
        Page 236
        Page 237
        Page 238
        Page 239
        Page 240
        Page 241
        Page 242
        Page 243
        Page 244
        Page 245
        Page 246
        Page 247
        Page 248
        Page 249
        Page 250
        Page 251
        Page 252
        Page 253
        Page 254
        Page 255
        Page 256
        Page 257
        Page 258
        Page 259
        Page 260
        Page 261
        Page 262
        Page 263
        Page 264
        Page 265
        Page 266
        Page 267
        Page 268
        Page 269
        Page 270
        Page 271
        Page 272
        Page 273
        Page 274
        Page 275
        Page 276
        Page 277
        Page 278
        Page 279
        Page 280
        Page 281
        Page 282
        Page 283
        Page 284
        Page 285
        Page 286
        Page 287
        Page 288
        Page 289
        Page 290
        Page 291
        Page 292
        Page 293
        Page 294
        Page 295
        Page 296
        Page 297
        Page 298
        Page 299
        Page 300
        Page 301
        Page 302
        Page 303
        Page 304
        Page 305
        Page 306
        Page 307
        Page 308
        Page 309
        Page 310
        Page 311
        Page 312
        Page 313
        Page 314
        Page 315
        Page 316
        Page 317
        Page 318
        Page 319
        Page 320
        Page 321
        Page 322
        Page 323
        Page 324
        Page 325
        Page 326
        Page 327
        Page 328
        Page 329
        Page 330
        Page 331
        Page 332
        Page 333
        Page 334
        Page 335
        Page 336
        Page 337
        Page 338
        Page 339
        Page 340
        Page 341
        Page 342
        Page 343
        Page 344
        Page 345
        Page 346
        Page 347
        Page 348
        Page 349
        Page 350
        Page 351
        Page 352
        Page 353
        Page 354
        Page 355
        Page 356
        Page 357
        Page 358
        Page 359
        Page 360
        Page 361
        Page 362
        Page 363
        Page 364
        Page 365
        Page 366
        Page 367
        Page 368
        Page 369
        Page 370
        Page 371
        Page 372
        Page 373
        Page 374
        Page 375
        Page 376
        Page 377
        Page 378
        Page 379
        Page 380
        Page 381
        Page 382
        Page 383
        Page 384
        Page 385
        Page 386
        Page 387
        Page 388
        Page 389
        Page 390
        Page 391
        Page 392
        Page 393
        Page 394
        Page 395
        Page 396
        Page 397
        Page 398
        Page 399
        Page 400
        Page 401
        Page 402
        Page 403
        Page 404
        Page 405
        Page 406
        Page 407
        Page 408
        Page 409
        Page 410
        Page 411
        Page 412
        Page 413
        Page 414
    Back Cover
        Page 415
        Page 416
Full Text
A Consortium for International Development Study
Farming Systems
Research and Development
Guidelines for Developing Countries

W. W. Shaner
P. F. Philipp
W. R. Schmehl

ti 4r~~ iA'

Westview Press


,f.A 1

Also of Interest

Small Farm Development: Understanding and Improving Farming Systems in the
Humid Tropics, Richard R. Harwood

Successful Seed Programs: A Planning and Management Guide, edited by Johnson E.

Food Security for Developing Countries, edited by Alberto Vald6s

Wheat in the Third World, Haldore Hanson, Norman E. Borlaug, and R. Glenn Anderson

Irrigated Rice Production Systems: Design Procedures, Jaw-Kai Wang and Ross E. Hagan

Managing Pastures and Cattle Under Coconuts, Donald L. Plucknett

Readings in Farming Systems Research and Development, edited by W. W. Shaner, P. F.
Philipp, and W. R. Schmehl

Small-Scale Processing and Storage of Tropical Root Crops, edited by Donald L.

Tomatoes in the Tropics, Ruben L. Villareal

I Managing Development in the Third World, Coralie Bryant and Louis G. White

Managing Renewable Natural Resources in Developing Countries, edited by Charles W.

Cassava, James Cock

Azolla as a Green Manure: Use and Management in Crop Production, Thomas A.
Lumpkin and Donald L. Plucknett

The Role of Centrosema, Desmodium, and Stylosanthes in Improving Tropical
Pastures, edited by Robert Burt, Peter Rotar, and James Walker

t Available in hardcover and paperback

Westview Special Studies in
Agriculture/Aquaculture Science and Policy

Farming Systems Research and Development:
Guidelines for Developing Countries
W. W. Shaner, P. F. Philipp, and W. R. Schmehl

Farming systems research and development (FSR&D) is an approach that is being
used increasingly to meet the need for greater food production and a better standard of
living for small-scale farmers in developing countries. This book synthesizes the
FSR&D procedures used by national governments and international research centers
around the world, emphasizing methodologies that have proved successful in practice.
The authors describe the characteristics and objectives of FSR&D, then present in-
formation on selecting target and research areas, problem identification and develop-
ment of a research base, research design, on-farm research, extending research results,
and implementation and training procedures. They emphasize that the FSR&D ap-
proach requires a clear understanding of farmers and their families, farmers' conditions,
and governmental staffing and organizational capabilities, and in one chapter discuss
how to determine whether an FSR&D approach is in a particular country's best in-
terests. Appendixes present detailed examples of procedures described in the text,
covering a variety of countries with different cropping and livestock systems, en-
vironmental conditions, and research and development capabilities.

Dr. Shaner, project director of the Farming Systems Research and Development
Methodology Project and associate professor of engineering at Colorado State Univer-
sity, has been chief economic advisor to the Technical Agency, Ministry of Planning,
Ethiopia, and associate director of the Consortium for International Development. He
has written Project Planning for Development (1979) and edited with Professors Philipp
and Schmehl Readings in Farming Systems Research and Development (Westview
1982). Dr. Philipp is professor emeritus in the Department of Agricultural and Resource
Economics at the University of Hawaii. Dr. Schmehl is professor and associate depart-
ment head of agronomy at Colorado State University.


Guidelines for Developing Countries

W. W. Shaner
P. F. Philipp
W. R. Schmehl
Editorial Assistance, Donald E. Zimmerman


/I j~k~J1 i ;ilki

Westview Press / Boulder, Colorado

Westview Special Studies in Agriculture/Aquaculture Science and Policy

This book was prepared for the United States Agency for International Development (Contract No.
AID/DSAN-C-0054), September 1981.

All rights reserved. No part of this publication may be reproduced or transmitted in any form or by
any means, electronic or mechanical, including photocopy, recording, or any information storage
and retrieval system, without permission in writing from the publisher.

Copyright 1982 by The Consortium for International Development
Suite 1500, 5151 E. Broadway
Tucson, Arizona 85711-3766

Published in 1982 in the United States of America by
Westview Press, Inc.
5500 Central Avenue
Boulder, Colorado 80301
Frederick A. Praeger, President and Publisher

Library of Congress Catalog Card Number: 82-050365
ISBN 0-86531-389-X
ISBN 0-86531-425-X(pb)

For further information about this document, please contact J.K. McDermott, USAID S&T/AGR,
Department of State, Washington, D.C. 20523; W.W. Shaner, College of Engineering, Colorado
State University, Fort Collins, Colorado 80523; or J.L. Fischer, Consortium for International
Development, Suite 1500, 5151 E. Broadway, Tucson, Arizona 85711-3766.

Printed and bound in the United States of America.

10 9 8 7 6 5 4 3 2


PREFA C E ....................................................... ...... xv
ACKNOW LEDGM ENTS................................................ xvii

1. EXECUTIVE SUMMARY. ............................................. 1
1.1. Background................................................... 3
1.2. Natureof FSR&D.... ............................... 3
1.3. FSR&D Activities.............................................. 4
1.3.1. Target and Research Area Selection ......................... 5
1.3.2. Problem Identification and Development
of a Research Base ..................................... 5
1.3.3. Planning On-Farm Research. .............................. 5
1.3.4. On-Farm Research and Analysis ............................ 6
1.3.5. Extension of Results.................................... 6
1.4. Issues of FSR&D Implementation................................. 6
1.4.1. Tim ing.............................................. 6
1.4.2. Organizational Flexibility. ................................ 6
1.4.3. Staffing Requirements ................................... 7
1.4.4. Training............................................. 7
1.4.5. FSR& D Costs .......................................... 8
1.4.6. Governmental Support ................................... 8
1.5. W hat This Book Offers .......................................... 8

2. INTRODUCTION ................................................ 11
2.1. Purposes ofFSR&D ............................................ 13
2.2. Definition of FSR&D .......................................... 13
2.3. Additional Definitions and Comments ............................ 16
2.3.1. Farming .... ...... ............. ......... 16
2.3.2. H households ......................................... 16
2.3.3. Small-Scale Farming. ................................... 16
2.3.4. Enterprises ...................................... ... 16
2.3.5. Farming Systems ....................................... 16
2.3.6. Cropping and Livestock Systems .......................... 16
2.3.7. Mixed Systems ........................................ 17
2.3.8. Cropping Systems Research. ............................. 17
2.3.9. Livestock and Mixed Systems Research ..................... 17
2.3.10. Commodity-Oriented Research. .......................... 17
2.4. Characteristics of FSR&D ....................................... 18
2.4.1. Farmer Based.......................................... 19

viii Contents

2.4.2. Problem Solving. .................................... 19
2.4.3. Comprehensive ...................................... 19
2.4.4. Interdisciplinary ....................................... 19
2.4.5. Complementary ....................................... 19
2.4.6. Iterative and Dynamic. ................................. 19
2.4.7. Responsible to Society .................................. 20
2.5. Objectives of These Guidelines. ................................. 20
2.6. Users of These Guidelines. ..................................... 21
2.7. Scope....................................................... 21
2.7.1. Applicability. ......................................... 22
2.7.2. Coverage.................. ........................... 22
2.7.3. Country Situations ..................................... 22
2.8. Approach to the Guidelines. .................................... 23
2.9. Contents of the Book ............................. ............ 23
R eferences ...................................... ............ 23

3. CONCEPTUAL FRAMEWORK. ...................................... 25
3.1. FSR&D Activities ............................................. 27
3.1.1. Target and Research Area Selection ........................ 27
3.1.2. Problem Identification and Development
of a Research Base.................................... 28
3.1.3. Planning On-Farm Research. ............................. 28
3.1.4. On-Farm Research and Analysis ........................... 28
3.1.5. Extension of Results .................................... 29
3.1.6. Collaboration. ......................................... 30
3.1.7. Feedback............................................. 31
3.2. Tim ing of Activities ........................................... 31
3.2.1. New FSR&D Programs for Predominantly
Cropping System s .................................... 31
3.2.2. Ongoing FSR&D Programs for Predominantly
Cropping Systems ................................... 32
3.2.3. New FSR&D Programs for Predominantly
Livestock Systems .................................... 32
3.2.4. Closing Comments on Timing. ........................... 33
3.3. Staffing.................................................... 33
3.3.1. Field Teams ........................................... 33
3.3.2. Regional Headquarters Team. ............................ 33
3.3.3. National Headquarters Team. ............................ 34
3.3.4. Others............................................... 34
3.3.5. Involvement of the Principal Groups
in FSR&D .......................................... 34
3.4. FSR&D Strategies ................... ......................... 35
3.4.1. How Much Change? .................................... 35
3.4.2. How Soon to Attempt Change?. .......................... 36
3.4.3. Where the Ideas for Change Originate?. .................... 37
3.4.4. What Type of Research? ................................. 37
3.4.5. A Partial Resolution of These Issues. ...................... 38
3.5. Summary .................................................. 38
References ................................................. 38

4. TARGET AND RESEARCH AREA SELECTION ......................... 41
4.1. Amount of Data to Collect. ..................................... 43
4.2. Selection of Target Area and Subareas ............................. 43
4.2.1. Definitions ................... ....................... 43
4.2.2. Criteria and Methods. .................................. 44
4.2.3. Information Needed .................................... 46
4.2.4. Actual Selection. ....................................... 49
4.3. Selection of the Research Area. .................................. 51
4.3.1. Definition........................................... 51

Contents ix

4.3.2. Criteria and Methods .................. ............. 51
4.3.3. Staffing ............ ................................. 53
4.3.4. Information Needed. ................ ................ 53
4.3.5. Collection of Information. ............................... 54
4.3.6. Selection by Stages. ................ ................. 55
4.3.7. Actual Selection. ................. ................... 55
4.3.8. Abandoning a Research Area .......................... 56
4.4. Early Identification of Opportunities for Action ................... 56
4.4.1. Research Opportunities. ................................ 57
4.4.2. Government Action. ................ ................ 57
4.5. Summary .................................................. 57
References...................................... ........... 57

OF ARESEARCH BASE .......................................... 59
Part 1: Identification of
Problems and Opportunities. ............................ 61
5.1. Definitions of Problems and Opportunities ......................... 61
5.2. General Approach ............. ................................ 61
5.3. Variations in Problem Focus. ................. ................. 62
5.4. Understanding the Farmers' Situation. ........................... 62
5.4.1. Describing the Farmers' Environment .................... 64
5.4.2. Description of the Farming System ........................ 64
5.4.3. Analyses of the Farming System ........................... 67
5.5. Analysis of Problems and Opportunities. .............. .......... 69
5.6. Setting Priorities for Problems and Opportunities .................. 70
5.6.1. Seriousness of a Problem. .............................. 70
5.6.2. Potential for Solving the Problem ......................... 71
5.6.3. Importance of the Problem in the
Research Strategy ................................... 71
5.7. Summary of Problem Identification ................. .......... 72
Part 2: Development
of a Research Base............................... 72
5.8. Collecting Data: General Comments. ........................... 72
5.9. Assembling Secondary Data ................ .................. 72
5.10. Collecting Primary Data. ................. ................... 73
5.10.1. Informal Methods ............... ................... 73
5.10.2. Formal Methods. ................. ................... 77
5.11. Combining Data Collection Methods .......................... 81
5.12. Data Management ........................................... 82
5.13. Summary of Research Base Development .......................... 82
References .................................................. 82

6. PLANNING ON-FARM RESEARCH. ................. ................ 85
6.1. Laying the Groundwork ...................................... 87
6.1.1. Reviewing Priority Problems and Opportunities .............. 87
6.1.2. Appraising the Organization's Capabilities
and Resources ................. ..................... 88
6.1.3. Appraising Present Technologies .......................... 88
6.1.4. Setting Assumptions About Near-Term
Conditions ................... .................... 88
6.1.5. Categorizing and Setting Research
Priorities ................... ....................... 88
6.1.6. Developing Hypotheses for Testing. ...................... 89
6.1.7. Establishing Research Collaboration ................... .... 89
6.2. Making Preliminary Analyses of On-Farm Experiments ............... 89
6.2.1. Alternative Solutions ................ ................ 89
6.2.2. Farmers' Conditions ............... .................. 89
6.2.3. Perspectives ............... ....................... 90

x Contents

6.2.4. Technically Viable Designs. ............................. 91
6.2.5. Estimating Values ................... .................. 91
6.2.6. Eventual Consequences ................................. 92
6.3. Considering Alternative Research Activities
and Methods ............................................... 92
6.3.1. Research Activities. .................................... 92
6.3.2. Research Methodologies ................................. 95
6.3.3. Applying Methods to Activities. .......................... 97
6.4. Finalizing Plans for On-Farm Experiments. ........................ 97
6.4.1. Deciding on the Design Conditions ................... ... .97
6.4.2. Searching for Improvements ............................ 98
6.4.3. Setting Design Standards. ............................... 99
6.4.4. Gathering Additional Data .............................. 102
6.5. Conducting Regional Planning Workshops ........................ 103
6.5.1. Nature and Purpose. ................................... 104
6.5.2. Attendance .......................................... 104
6.5.3. A activities ........................................... 104
6.6. Summary ................................................. 106
References..................................... .......... 106

7. ON-FARM RESEARCH AND ANALYSIS ............................... 109
Part 1: On-Farm Research. ...... ........... ......... .. 111
7.1. Researcher-Managed Trials ..................................... 111
7.1.1. Research on Crops ..................................... 111
7.1.2. Research on Livestock. ................................ 113
7.2. Superimposed Trials .......................................... 114
7.2.1. Research on Crops. .................................... 114
7.2.2. Research on Livestock. ................................ 115
7.3. Farmer-Managed Tests ........................................ 116
7.3.1. Research on Crops ................... .................. 116
7.3.2. Research on Livestock. ................................ 121
7.4. Team Organization. ........................................... 122
7.4.1. Team Leader's Activities ............................... 122
7.4.2. Assignment of Resources ............................... 123
7.4.3. Review Sessions ...................................... 123
7.4.4. Integration with Local Organizations ..................... 123
7.5. Summary ................... ..... ......... .. ............. 123
Part 2: Analysis ......................................... 123
7.6. Concepts of Analysis .......................................... 124
7.6.1. An Integrative Approach. ............................... 124
7.6.2. Prediction Versus Acceptance ........................... 124
7.6.3. Partial Budget Analysis Versus Whole
Farm Analysis ..................................... 125
7.7. Illustrative Designs and Analysis Procedures ....................... 126
7.7.1. Illustrative Designs .................................... 126
7.7.2. Biological Results. .................................... 126
7.7.3. Net Benefits ......................................... 128
7.7.4. Economic Criteria. .................................... 131
7.7.5. Financial Feasibility ................................... 139
7.8. Acceptability of New Technologies .............................. 139
7.8.1. Analysis of Farmer-Managed Tests ....................... 139
7.8.2. Acceptability Index. ................................... 141
7.9. Sociocultural Feasibility ....................................... 141
7.10. Other Analysis Procedures ..................................... 142
7.10.1. Further Data Analysis .................................. 142
7.10.2. Long-Term Investments ............................... 142
7.10.3. Analyses from Society's Point of View ..................... 143
7.11. Analysis Workshops .......................................... 143
7.12. Sum m ary .................................................. 144
R eferences ................................................. 144

Contents xi

8. EXTENSION OF RESULTS .......................................... 147
8.1. Integration of Extension into FSR&D ............................. 149
8.2. Staffing and Organizing for Extension's
Involvement.............................................. 149
8.2.1. Extension Specialist in Farming Systems ................... 149
8.2.2. Proposed Organization of Extension at
Three Levels................. ......... ...... .. 150
8.3. Details of Extension's Involvement in FSR&D ................... 151
8.3.1. Target Area Selection .................................. 151
8.3.2. Subarea and Research Area Selection ...................... 151
8.3.3. Problem Identification and Development
of a Research Base.................................. 151
8.3.4. Planning On-Farm Research. ............................ 151
8.3.5. On-Farm Research and Analysis .......................... 152
8.3.6. Extension of Results ................... ........ .... 152
8.4. Multi-locational Testing. ...................................... 153
8.4.1. Nature and Participation. .............................. 153
8.4.2. An Example from Southeast Asia. ........................ 154
8.5. Pilot Production Program. ..................................... 155
8.6. Problems in Extension ....................................... 156
8.6.1. Ties Between Extension and Research ..................... 156
8.6.2. Training............................................ 157
8.6.3. Orientation. .......................................... 157
8.6.4. Organization. ......................................... 158
8.6.5. Budgets ............................................ 158
8.7. Conclusions ............................................... 158
References.................................................. 158

9. DECIDING ON AN FSR&D APPROACH ............................. 161
9.1. Basic Issues in Research and Development ........................ 163
9.1.1. Are the Activities of the Research and Development
Process Consistent with the National
Development Goals? ................................ 163
9.1.2. Is the Research and Development Process Producing
Results that are Relevant to Small
Farmers' Needs? .................................... 163
9.1.3. Som eAnswers ................. ... ............ 163
9.2. Development Policy and the Role of FSR&D ..................... 164
9.3. Farmers' Needs and the Role of FSR&D. ........................... 165
9.3.1. Modifications to Existing Systems ................. ...... 165
9.3.2. Introductions of New Systems. .......................... 166
9.3.3. Decisions on How Much Change ......................... 166
9.4. Organizational Capacity and the Role of FSR&D .................... 166
9.5. FSR&D and Supporting Organizations ............................ 166
9.6. Adoption of an FSR&D Approach. ............................... 167
9.6.1. Emphasis on Applied Research ........................... 168
9.6.2. Adaptability to Ongoing Development
Programs .......................................... 168
9.6.3. Requirements for Skilled Personnel ....................... 169
9.7. Cost-Effectiveness of FSR&D ................................... 169
9.7.1. Comparison of Expenditures. ........................... 169
9.7.2. Comparison of Rates of Adoption ......................... 169
9.7.3. Comparison of Numbers of Farmers Affected ............... 169
9.7.4. Conclusions on Cost-Effectiveness ....................... 170
9.8. A Concluding Comment. ...................................... 170
References ................................................. 170

10. IMPLEMENTATION.............................................. 171
10.1. Deciding on an Approach. ..................................... 173
10.1.1. Project Approach. ..................................... 173

xii Contents

10.1.2. Program Approach. .................................... 173
10.1.3. Project Versus Program Approach. ....................... 173
10.1.4. Government Support. ................................. 174
10.2. Organizational Structure ....................................... 174
10.2.1. For the Project Approach. ............................... 175
10.2.2. For the Program Approach .............................. 175
10.2.3. A Generalized Organizational Diagram .................. 176
10.3. Staffing .......................................... .......... 178
10.3.1. FieldTeams......................................... 178
10.3.2. Research Specialists ................................... 178
10.3.3. Extension's Input ............... ...................... 179
10.3.4. Team Leadership ..................................... 179
10.3.5. Approach When Trained Staff are
Severely Limited. ................................... 180
10.4. Off-Site Management. ......................................... 180
10.4.1. General Considerations ................................ 180
10.4.2. Personnel M management .............. .................. 181
10.5. The Roles of the Field Team. ................................... 182
10.5.1. The Field Team and the Farmers. ........................ 182
10.5.2. The Field Team and the Research
Specialist .......................................... 182
10.5.3. The Field Team and the Extension
Service................................. ........ 182
10.5.4. Functional Assignments for the Field
Team ............................................. 182
10.5.5. Composition and Organization of the
Field Team......................................... 183
10.6. Interdisciplinary Teamwork. ................................... 184
10.6.1. A Model for Interdisciplinarity. .......................... 185
10.6.2. Application of Interdisciplinarity to
FSR&D ...........................................188
10.7. Getting Started.............................................. 190
10.8. Evaluation of Projects ......................................... 191
10.8.1. Types of Evaluation ................. .................. 191
10.8.2. Which Types of Evaluations to Use? ...................... 192
10.8.3. Developing Evaluation Procedures ........................ 192
10.8.4. A Caveat on Evaluations ................................ 192
10.9. Project and Program Management: A Two Country
Perspective ............................................... 193
10.9.1. An Example from Honduras ............................. 193
10.9.2. An Example from the Philippines ......................... 193
10.10. A Summary Perspective ....................................... 193
References ................................................. 194

11. TRAININ G ....................................................... 195
11.1. Initial Exposure to FSR&D Concepts ............................. 197
11.2. Development of National Training Programs for FSR&D ............. 197
11.2.1. Programs for Researchers ............................... 197
11.2.2. Programs for Extension. ................................ 199
11.2.3. Programs for Technicians ............................... 200
11.3. Complementing Activities ..................................... 201
11.3.1. Non-Degree Training. ................... .......... .201
11.3.2. Graduate Degree Training. ............................. 201
11.3.3. Short-Term Activities .............. .................... 201
11.4. International and Regional Centers. ............................. 202
11.5. University Programs in the United States ......................... 203
11.5.1. Objectives for University Programs ....................... 204
11.5.2. Approach for University Programs ........................ 204
11.6. Training Materials on FSR&D .................................. 204

Contents xiii

11.7. Summary................... ................. ........... . 204
References ................................................ 205

ACRON YM S ........................................................ 207
G LO SSARY ......................................................... 211
APPENDIXES......................................... ..........219
P-A. Project Contributors: Field Contacts, Participants
in the Workshops and Pretesting, and Reviewers................... 223
2-A. Mathematical Modeling. ........................................ 231
3-A. Review of FSR&D Activities ..................................... 237
4-A. Illustration of the Use of General Farm Data to Help
Establish Recommendation Domains. ......................... .. 243
4-B. Grouping Farmers into Homogeneous Populations ................... 243
4-C. Selection of FSR&D Areas. ....................................... 246
4-D. Climatic Zones in Southeast Asia. ................................ 248
4-E. Physiographic Regions in Southeast Asia ........................... 249
4-F. Research Area Selection in ICRISAT's Village
Level Studies............................................... 249
5-A. Physical Resources of the Research Area Affecting
Biological Production ...................................... .. 255
5-B. Land Types and Land Evaluation. ................................ 255
5-C. Marketing Factors Affecting Small Farmers ......................... 259
5-D. The Sociocultural Environment. ................................. 261
5-E. An Example of Sociological Research .............................. 262
5-F. Decision Making by Small Farm Families .......................... 264
5-G. Decision Trees: A Method for Learning About
Farmers' Decisions ..........................................265
5-H. Describing Existing Cropping Systems ............................. 267
5-I. Sample Forms for Describing On-Farm Resources
Used in Crop Production...................................... 269
5-J. Conceptual M odels ................................ ..... ..... 270
5-K. Mixed Cropping and Livestock Systems ............................ 275
5-L. Data Collection in a Rural Setting ................................. 278
5-M. Gathering Data About Women. .................................. 281
5-N. Assessment of Secondary Data. .................................. 283
5-0. Analysis of the Content of Informal Interviews ...................... 284
5-P. Illustrative Tables for Collecting Data During
Reconnaissance Surveys ...................................... 284
5-Q. Summary of the Sondeo Methodology Used by ICTA ................. 289
5-R. Guidelines for Pre-Survey Sequence. .............................. 293
5-S. Suggestions for Dealing With Farmers' Recollection
of Inform ation .............................................. 296
5-T. Validity From the Social Science Perspective ........................ 297
5-U. Questionnaire Design. .......................................... 299
5-V Sam pling................................................. 303
5-W. Selecting, Training, and Supervising Interviewers .................... 306
5-X. Farm Record Keeping. ......................................... 309
5-Y. Monitoring and Observational Activities ........................... 316
5-Z. Data M anagem ent ............................................. 316
6-A. Selection of Variables for Climatic Monitoring of a
Research Area............................................... 321
6-B. The Land Equivalent Ratio (LER) ................................. 323
6-C. A Guide for Locating On-Farm Experiments ........................ 324
6-D. Field Designs and Statistical Procedures for
On-Farm Experiments ....................................... 325
6-E. Example of a Procedure for Designing a Cropping
Pattern Experiment .......................................... 335
6-F. Field Assignments ............................................ 337

xiv Contents

7-A. Forms for Collecting Data for On-Farm Cropping
Experiments ................................................ 341
7-B. Forms for Collecting Data for On-Farm Livestock
Experiments............................................... 345
7-C. Data Collection Form for Farmer-Managed Tests .................... 348
7-D. Field Design for Farmer-Managed Cropping Test ..................... 349
7-E. Field Design of a Farmer-Managed Cropping Pattern
T est...................................................... 350
7-F. Example of a Mixed Crop-Livestock Research Project ................. 350
7-G. Estimating Net Benefits from Alternative Treatments ................ 352
7-H. Analysis of Cropping Pattern Research in Indonesia .................. 354
8-A. Memorandum of Agreement to Establish a Pilot
Production Program in the Philippines ........................... 359
8-B. Letter of Understanding between ICTA and DIGESA .................. 360
10-A. An Agricultural Research Project for the Senegalese
Institute for Agricultural Research (ISRA) ......................... 365
10-B. Alternative Organizational Diagrams for Farming
Systems Research and Development. ............................. 369
10-C. Team Building............................................... 374
10-D. Summarized Checklist for Successful
Interdisciplinarity. ........................................... 375
10-E. USAID's Logical Framework. .................................... 376
11-A. Honduran Training Program in FSR&D ............................ 383
11-B. Outline of an In-Service Training Program in FSR&D
Prepared by ICTA for DIGESA.................................. 388
11-C. Training in FSR&D at Selected International Centers
for Agricultural Research. ...................................... 389
11-D. A Cropping Systems Training Program at IRRI....................... 393
11-E. Six Principles for Technical Change ............................. 397
References Cited in the Appendixes. .............................. 399
IN D EX .................................. .............. ...... ..... 407


In recent years, policy makers have been paying
more attention to the problems of small farmers in
developing countries with the idea of increasing their pro-
duction and standard of living. The policy makers' objec-
tives are twofold: (1) to help those whose welfare is
materially below the rest of society, and (2) to help a coun-
try increase its agricultural production. With adequate
agricultural policies, these two objectives are mutually
reinforcing. For example, increased food production gives
farm households additional food for consumption and
surpluses for sale. Farmers can then use the money from
these sales to buy items they do not produce, and the
buyers of farm products benefit from the increased sup-
By focusing on these two objectives, leaders in
developing countries work toward other national objec-
tives. For example, increasing production on small farms
may (1) improve a nation's self-sufficiency in food produc-
tion, (2) supply more raw materials for industry, and (3)
improve the nation's foreign exchange position. Improving
small farmers' production should narrow the range of in-
comes among groups in the country, reduce political in-
stability, slow rural-to-urban migration, lessen the need
for food relief, and so on.
In developing countries, farmers with limited
resources often do not adopt new technologies because (1)
their conditions are not like those where the technologies
were developed, (2) they do not have resources to purchase
the required inputs, (3) the technologies do not apply to
the crops grown or the livestock raised on their farms or
the way they operate, or (4) they do not know about the
new technologies. For whatever reason, development of
new technologies sometimes leaves small farmers worse
off than before. This happens when large farmers adopt
new technologies and small farmers do not.
But conditions are changing. Recently, more na-
tional and international research organizations are direct-

ing their attention to the conditions and problems of
small farmers. As a result, small farmers and their en-
vironments are better understood; more research is applied
to solving small farmers' problems; and the extension ser-
vice works with better technologies. An approach now be-
ing applied more widely to make research relevant for
small farmers is called farming systems research and
development (FSR&D) or sometimes simply farming
systems research (FSR).
With more countries interested in FSR&D, the
United States Agency for International Development
(USAID) contracted with the Consortium for International
Development (CID) to write a set of guidelines on FSR&D
methodology. CID made Colorado State University (CSU)
the lead university, entered into agreements with other
CID schools, and subcontracted part of the work to the
University of Hawaii.
The result is this set of guidelines for those in the
developing countries who wish to learn about FSR&D and
who might apply the procedures. In keeping with this em-
phasis, the contract required the writers to synthesize

".. a set of integrated, multidisciplinary farming
systems R&D methodologies adapted to the person-
nel and financial constraints of the LDC's, packaged
for easy delivery in the form of a comprehensive
handbook or handbooks to LDC institutions."'

These methodologies refer to ways for gathering data
on farming systems and farmers' environments, identify-
ing problems and opportunities for improvement, develop-
ing research data for farmer and experimental plot condi-
tions, generating alternative farming systems, diffusing
improved technologies, and receiving feedback for further
To produce these guidelines, we synthesized data on
FSR&D and related practices from documents and

'USAID. 1978. Farming systems R&D methodology contract. USAID, Washington, D.C.


xvi Preface

meetings with individuals and representatives of organiza-
tions throughout the world. Because the published
literature includes few documents on farming systems
concepts and procedures, we relied heavily on visits to
organizations with agricultural research programs for
small farmers. Early in our study, we identified issues in
FSR&D and discussed them at a workshop that a group of
internationally recognized practitioners in FSR&D at-
The review process entailed critiques of three
preliminary drafts of the guidelines. Our first draft was re-
viewed during a two-day workshop in June 1980 attended
by a small group of FSR&D practitioners and the project
team. Our second draft was pretested at a two-week ses-
sion in August 1980 attended by a group from the develop-
ing countries, the project team, and observers. We
distributed the third draft worldwide in December 1980,
and received comments from more than 30 reviewers. In
Appendix P-A we provide further comments on these ac-

tivities and the names of contributing individuals and
To conclude, we stress that FSR&D concepts and
procedures are evolving rapidly. Moreover, the early con-
centration on crops is giving way somewhat to allow more
consideration of livestock-either as pure livestock
systems or in combination with crops. Furthermore, we
anticipate other topics will be integrated into the FSR&D
approach such as mathematical and ecological modeling,
agro-forestry, and soil and water conservation. Conse-
quently, this book of guidelines will require updating as
more is learned and better procedures become available.
With this in mind, we welcome your suggestions.

W. W. Shaner
P. F. Philipp
W. R. Schmehl

Fort Collins, Colorado


We, the three co-authors of these guidelines, are
grateful to the other members of the team; to consultants;
to advisers; to the governments of Guatemala and Hon-
duras, and New Mexico State University for hosting the
protesting sessions; to the participants in the protesting
and their parent organizations; to the reviewers of the
various drafts and outlines; to those who met with team
members during their travels and provided insight and in-
formation on agricultural research and extension; to the
national governments and international centers that
cooperated; to USAID's project monitor; and to the
"pioneers" in farming systems research and development
for their vision and courage in developing new approaches
to solve the problems of poor farmers around the world.
While we acknowledge the contributions of the above, we
do not hold them responsible for any of this book's short-
As co-authors, we had the responsibility for writing
the book of guidelines, but we could not have completed
the manuscript without help from others. This help came
in conceptualizing the approach, gathering and analyzing
data, drafting chapters and appendixes, and reviewing
others' work. Below, we list the project staff and their
principal inputs. By being an enthusiastic and effective
member of a team effort, each person contributed much
more than the contributions listed.
Gary Hansen drafted Chapters 9 and 10 on the merits
of the FSR&D approach and on organization. Tom Trail
drafted Chapters 8 and 11 on extension and training.
Richard Tinsley drafted sections on the cropping systems
research that went into Chapters 6 and 7. Helen Hender-
son drafted sections on sociocultural aspects for the whole
book. Howard Stonaker drafted sections on livestock
systems in Chapters 6 and 7. James Meiman drafted the
section on interdisciplinary teamwork and Tom Sheng
drafted the section on evaluation in Chapter 10.
Contributors to the appendixes were George Beal on
sociocultural research, Jen-hu Chang on agroclimatology,
Michael Read on experimental design and statistical pro-
cedures, Tom Sheng on mathematical modeling and other
topics, Don Zimmerman on analysis of data and research
results, and Ann Perry-Barnes and John Roecklein on a
variety of topics.

Don Zimmerman was project editor. Robert Dils
was responsible for the protesting. We obtained help in
gathering data on farming systems in Francophone West
Africa from Derrick Thom, in Central America from
Albert Ludwick, and in Indonesia from Martin Waananen.
Tom Sheng was research associate. Ann Perry-Barnes was
research assistant. Michael Read and John Roecklein were
graduate research assistants.
Other project staff included Dale Rosenbach for the
book's format and artwork. Carol Marander drafted the
figures. William Shaw created the sketches. Don Zimmer-
man coordinated the preparation of the photographs.
Miriam Palmer was responsible for the bibliography. Jan
Schweitzer prepared the index. Marilee Long helped with
the editing. Patty Sheng helped with the preparation of the
drafts. Veryl Meyers typed the final copy of the text.
Regular and temporary secretaries were Debbie Bartow,
Cheryl Buster, Lillian McKee, Lori Neubauer, Peggy Neff,
Vicky Lynn, Christine Stanley, and Imogene Wood.
The project's advisory committee included Gerald
Burke, Frank Conklin, Jack Keller, Shelley Mark, Gerald
Matlock, Martin Waananen, and James Meiman as chair-
Consultants, who contributed greatly to the project,
were Peter Hildebrand, David Norman, and Robert Waugh
on FSR&D in general; Ramiro Ortiz on the farming
systems program in Guatemala; George Beal and Edward
Knop on sociology; and Elmer Remmenga on experimental
design and statistics.
Kenneth McDermott-USAID's project mon-
itor-briefed the project team at the outset, assisted in
coordinating the interactions between the project and
USAID, and, most importantly, contributed substantially
to our understanding of farming systems concepts.
Finally, we wish to thank all those who provided us
with photographs to illustrate farming activities and con-
ditions in developing countries. Those photographs
selected for each of the chapters were supplied by the
following photographers and organizations: Chapter 1:
Food and Agriculture Organization of the United Nations
(FAO) photographs by A.E. Deutsch of the International
Plant Protection Center (IPPC), Corvallis, Oregon;
Douglas Horton of the International Potato Center (CIP),

xviii Acknowledgments

Peru; International Crops Research Institute for Semi-Arid
Tropics (ICRISAT), India; and Tom Sheng, FSR&D Proj-
ect, Colorado State University (CSU); Chapter 2: Tom
Sheng and W.W. Shaner, FSR&D Project; Donald
Sungusia, Tanzanian government; Wayne Freeman, Inter-
national Agricultural Development Service (IADS), Nepal;
and ICRISAT; Chapter 3: International Center for
Agricultural Research in Dry Areas (ICARDA); FAO; and
IPPC; Chapter 4: LADS, Nepal; Tom Sheng and P.F.
Philipp, FSR&D Project; and ICRISAT; Chapter 5: P.F.
Philipp and Tom Sheng, FSR&D Project; ICARDA;

Douglas Horton, CIP; and IADS, Nepal; Chapter 6: ICAR-
DA; IADS; Dan Lattimore, CSU Egyptian Project; and
Tom Sheng, FSR&D Project; Chapter 7: P.E. Hildebrand,
Agricultural Science and Technology Institute (ICTA),
Guatemala; ICARDA; ICRISAT; Tom Sheng and P.F.
Philipp, FSR&D Project; FAO; and IPPC; Chapter 8:
ICRISAT; ICARDA; and Douglas Horton, CIP; Chapter 9:
ICRISAT; FAO; IPPC; and ICARDA; Chapter 10: Douglas
Horton and Francis Tardieu, CIP; and Tom Sheng and
W.W. Shaner, FSR&D Project; and Chapter 11: Douglas

Chapter 1

This book provides guidelines for farming systems
research and development (FSR&D) as applied to condi-
tions in developing countries. The purpose of the
guidelines is to assist national governments interested in
helping poor farmers-primarily small-scale farmers with
limited resources. Therefore, the guidelines discuss the
nature of FSR&D, processes and methodologies ap-
propriate for various conditions, and alternative means for
implementation. Because most of the applied work in
FSR&D has been with cropping systems, this book of
guidelines emphasizes cropping systems research. By syn-
thesizing implemented and successful approaches, these
guidelines have a strongly applied orientation.
As a synopsis of the principal features of the
guidelines, this executive summary is intended for those
who wish a quick review of FSR&D's principal features.
This summary contains brief sections on the background
of FSR&D, its nature and activities, issues of implementa-
tion, and the contents of this book.

Considerable attention is currently being given to
improving the lot of small farmers in developing coun-
tries. An important way of helping them is through
agricultural research, extension, and related programs
specific to their needs. A better approach for such efforts
became necessary because farmers' conditions were not
improving adequately. Research and development pro-
grams had often been undertaken without having small
farmers in mind or without knowing much about them. In
contrast, the FSR&D approach starts and ends with small
farmers and thereby focuses specifically on their condi-
tions and aspirations.
While much of FSR&D has been directed toward
farmers with limited resources, the approach has
relevance for improving agricultural research and develop-
ment in general. Some argue FSR&D is simply a modified
version of farm management that has been widely prac-
ticed in the United States during the 20th century. While
this claim has merit, the general feeling among those ac-
tively engaged in FSR&D is that FSR&D is new-at least
as applied to the needs of small farmers in developing
countries. The accomplishments of some national and in-
ternational research organizations support the contention
that improved technologies can be designed for and will be
adopted by small farmers.

A common thread among alternative approaches to
FSR&D is the selection of relatively uniform sets of condi-
tions for conducting research and implementing change.
FSR&D allows researchers to (1) both intensively in-
vestigate the individual conditions of small farmers and
(2) make an impact on large numbers of farmers. This
result is accomplished by selecting reasonably uniform
physical, biological, and socioeconomic environments,
where farmers' cropping and livestock patterns and
management practices are similar. Improved technologies
developed for farmers in these research areas are expected
to be applicable to farmers operating elsewhere under
similar conditions.
The FSR&D approach typically uses inter-
disciplinary teams, whose composition varies according to
the task. Field teams conduct on-farm research and are aid-
ed by (1) disciplinary specialists in the physical,
biological, and social sciences who may operate out of
regional or national headquarters or experiment stations,
(2) extension specialists, and (3) others concerned with
agricultural production.
Together, they study

physical conditions such as rainfall, temperatures,
and land forms
biological factors such as production potential and
pest problems
socioeconomic conditions such as the size and
nature of landholdings, farmer and community
customs, markets, and local services
the farming system.

The farming system is the complex arrangement of
soils, water sources, crops, livestock, labor, and other
resources and characteristics within an environmental set-
ting that the farm family manages in accordance with its
preferences, capabilities, and available technologies.
Farmers manage the household's resources involved in the
production of crops, livestock, and nonagricultural com-
modities (e.g., handicrafts), and may also earn income off
the farm.
Farms are classified according to major character-
istics-e.g., grazing systems, permanent cultivation on
rain-fed land, or irrigated farming-and the environ-
ment-e.g., agroclimatic zone, soils, and terrain. Re-

4 Executive Summary

searchers classify farms according to the area, the needs of
the study, and the available information.
FSR&D focuses on the interdependencies among the
components under the farmers' control, and between these
components and the physical, biological, and socio-
economic environments. Also, FSR&D identifies and
generates improved technologies and adapts, tests, and
promotes them.
The various production activities are subsystems of
the whole farming system. For example, crop production
is a subsystem of the whole farm and is, in turn, made up
of individual cropping activities. The study of a cropping
system comprises everything required for the production
of one or more crops, including interactions between dif-
ferent crops. More specifically, research on cropping
systems concentrates on

crops and cropping patterns
alternative management practices in different en-
interactions between crops
interactions between crops and other enterprises
interactions between the household and en-
vironmental factors beyond the household's control.

A similar description could be given for livestock systems

Thus, FSR&D can be summarized as being farmer-
based, problem solving, comprehensive, interdisciplinary,
complementary, iterative, dynamic, and responsible to
society. The approach is

farmer-based because FSR&D teams pay attention to
farmers' conditions and integrate farmers into the
research and development process
problem solving in that FSR&D teams seek research-
able problems and opportunities to guide research
and to identify ways for making local services and
national policies more attuned to the farmers' needs
comprehensive in that FSR&D teams consider the
whole farming activity (consumption as well as pro-
duction) to learn how to improve the farmers' output
and welfare, to identify the flexibilities for change in
the environment, and to evaluate the results in
terms of both farmers' and society's interests
interdisciplinary in that researchers and extension
staff with different disciplinary backgrounds work
with farmers in identifying problems and oppor-
tunities, searching for solutions, and implementing
the results
complementary because it offers a means for using
the outputs of other research and development
organizations and for giving direction to others' work
iterative in that FSR&D teams use the results from
research to improve their understanding of the
system and to design subsequent research and im-
plementation approaches
dynamic in that oftentimes FSR&D teams introduce
relatively modest changes in the farmers' conditions
first and the favorable results encourage more sig-
nificant changes later
responsible to society in that FSR&D teams keep the
long-run interests of the general public-both pres-
ent and future-in mind, as well as those of the
farming groups immediately affected.

While much of the above is true of other forms of
agricultural research and development programs, the com-
bination of these factors distinguishes FSR&D from other
approaches. Even more, FSR&D is systems oriented in
that the researchers study the farmers' conditions at the
outset, keep these conditions in mind during research and
implementation, and use their knowledge of these condi-
tions in evaluating the results. In this sense, FSR&D
departs from reductionism, which is an approach that
breaks the whole into parts and studies them more or less
independently. Furthermore, FSR&D uses acceptance by
the whole family as its key measure of success, rather than
some abstract or narrowly defined criteria of effectiveness.

The approach to FSR&D varies according to the
organization's mandate, which may be for certain com-
modities or which may be localized, countrywide, or inter-
national. Approaches also vary by the physical, biological,
and socioeconomic characteristics of the target areas and

Executive Summary 5

groups, as well as by the preferences of FSR&D ad-
ministrators and researchers. Some approaches are com-
prehensive, taking many factors as variable, including
public policy; but more frequently, FSR&D works within
existing conditions or assumes only modest changes in the
existing conditions.
The basic FSR&D activities are target and research
area selection, problem identification and development of
a research base, planning on-farm research, on-farm
research and analysis, and extension of results. Each of
these is summarized below.


Using national and regional objectives, key decision
makers-including those from the FSR&D team-select
one or more target areas. Then, the FSR&D team divides
the target area into subareas with relatively uniform
characteristics and selects a research area representative of
the selected subareas. The team continues by choosing the
target group-farmers who have common environments
and common production patterns and farming practices.
This group of farmers might be those with a particular
cropping, livestock, or mixed (e.g., crops and livestock)
pattern; alternatively, the approach could be based more
on environmental conditions. Such classifications are
usually adequate for identifying problems and oppor-
tunities of sufficient magnitude to justify the research ef-
fort. Where practical, the FSR&D team tries to apply the
research results to farmers operating under similar condi-
tions beyond the target area.

The FSR&D team identifies and ranks problems and
opportunities according to such criteria as the short-run
and long-run significance to the farmers and society,
availability of suitable or potentially suitable tech-
nologies, and ease of implementation. Besides ideas arising
out of the previous activity, the team commonly iden-
tifies problems and opportunities through quick recon-
naissance surveys of the area. The study of livestock
systems tends to take longer and may involve aerial
photography, satellite imagery of rangelands, and monitor-
ing of development programs to learn how herding
societies function over time. A subject with considerable
and yet untapped potential is research on mixed farming
systems in which the researchers consider the influence of
crops and livestock on each other.
In the process of identifying problems and oppor-
tunities, the team gains considerable knowledge about the
area. This knowledge and the collected data form the ini-
tial research base for developing improved technologies for
the area's small farmers.

Once the FSR&D team has identified and ranked
problems and opportunities, gathered preliminary data,



and set out hypotheses, it plans the on-farm research ac-
tivities. Early in the process, the team needs to decide the
extent to which the farmers' environment can be changed.
For the most part, the team takes resource availability,
support services, and government policy about as they are.
But, an important part of FSR&D is to identify where and
how much change of this type is possible. Given an under-
standing of this, the team then considers opportunities for
improving farmers' conditions.
On-farm research emphasizes alternative cropping
and livestock patterns, management practices, and other
activities of the farm household. The team incorporates
the farmers' conditions into the design procedures by
working closely with farmers. The team meets with
farmers in their fields and learns farmers' terms such as
those for farmers' activities and units of measure. Re-
searchers also learn how the farm household divides its ac-
tivities, which members perform which activities, who
has responsibility for the different family decisions, who
controls which resources, how members tend the family's
crops and livestock, and how they market their surplus
production. Farmers, in turn, take part in the research ex-
periments and evaluate the results. This collaborative
style calls for integration of experiment station and other
research and development personnel who are specialists in
(1) disciplines such as entomology, economics, and soil
conservation; (2) commodity topics, such as plant
breeding and cattle production; and (3) extension.

6 Executive Summary

Furthermore, the team designs record keeping
systems, special studies, climatic monitoring, and surveys
to provide additional information about the farmers and
their environment. Often the team initiates recording of
farmers' activities early in the FSR&D process to develop a
continuing base of information on farmers' productive ac-
tivities throughout the cropping and livestock seasons.
The team uses special studies of selected topics, such as
cultivation practices, to help fill in gaps in its knowledge
about the area. The team needs information on the en-
vironment, including climatic data, to help design
research and interpret the results from crop and animal ex-
periments. Also, the team uses long-run studies of farm
households, local conditions, and related topics to provide
a sound basis for understanding the situation and im-
plementing change.
Before finalizing the research plan, the team
evaluates the proposed technological changes. It does this
to learn if the results are biologically feasible and in the in-
terests of the farmers and society. Finally, the team
assesses the extent to which local support systems and na-
tional policies will accommodate the new technologies.


Most national FSR&D programs emphasize applied
research by conducting much of the research on farmers'
fields. Three types of biological production experiments
are common: researcher-managed trials to experiment
under farmers' conditions where control of the experiment
is important; farmer-managed tests to learn how farmers
respond to the suggested improvements; and super-
imposed trials to apply relatively simple researcher-
managed experiments across a range of farmer-managed
The researchers initiate experiments, studies and
other activities, and gather data. Then, they analyze the
results in terms of the statistical meaning of biological per-
formance, actual resource requirements, economic and
financial feasibility, and sociocultural acceptability. They
estimate the overall impacts on both farmers and society.
Researchers study the acceptability of the experiments to
farmers through observations of farmers' actions, talking
with farmers, and in other ways. Finally, the researchers
examine the opportunities for improving support services
and government policies.

Throughout the research process, the FSR&D team
maintains contact with support organizations in the area.
Extension plays an especially important role in the pro-
cess. Inputs from extension should occur at all levels of
FSR&D-from initially identifying areas to the broad im-
plementation of results. FSR&D practitioners generally
recommend that the extension staff be trained in FSR&D
and become regular members of the field and regional
Extending the results involves multi-locational
testing-an activity that spreads the improved tech-

nologies more broadly than the previous on-farm trials and
tests. Multi-locational testing helps define the specific
conditions by applying the results on a broad scale. In this
process, extension agents learn the details of the technolo-
gies and how to apply them.
Another means of extending research results is
through pilot production programs-an activity that ap-
plies the improved technologies on a scale large enough to
effectively test the area's support systems. This activity
provides further insight into the needs for modifying the
technology, altering the support system, or both. How-
ever, the concept of FSR&D is that the derived technol-
ogies should fit the farmers' and environmental conditions
sufficiently well so that few adjustments are needed at this
Once these steps have been taken, the country can
broadly apply the new technologies among the groups for
which they have been designed.

Some of the issues concerning FSR&D implementa-
tion relate to the time required to obtain results, organiza-
tional flexibility, staffing requirements, training, FSR&D
costs, and governmental support.

1.4.1. TIMING
The general approach to FSR&D is rapid initiation of
on-farm experiments combined with adjustments in the
program's direction as results provide feedback. With ade-
quate planning, researchers often start experiments
without missing a cropping season. Sometimes they try
exploratory experiments to learn how farmers respond to
new opportunities; at other times, researchers conduct
trials to screen locally available technologies for their ap-
plicability to specific farmers' conditions. Under favorable
conditions, some research results may be ready for
widespread diffusion to farmers within a few seasons.
However, more fundamental changes in farmers' cropping
patterns and management practices normally take longer.
The approach being developed for livestock systems
is an exception. For larger animals such as cattle, the en-
vironment, livestock systems, and growth stages often re-
quire more careful study than most crops or small

FSR&D is primarily a modification of existing
research and extension methods; therefore, the approach is
adaptable to a variety of situations, as illustrated by the
following possibilities. A country can implement FSR&D
through a semiautonomous government corporation that
has more flexibility in operations, budgeting, and person-
nel management than ministerial research and develop-
ment organizations. A country can implement FSR&D
through a ministry of agriculture if the ministry is respon-
sible for research and extension. A country can apply
FSR&D to the activities of an experiment station in which

Executive Summary 7

one or more teams trained in FSR&D methods work close-
ly with experiment station staff. Or, a country can build
FSR&D into a project to increase production; in such a
case, FSR&D methods can improve the efficiency of the
overall project.
Each approach has its advantages and disadvantages,
so the approach selected depends on the situation. Here,
we emphasize that FSR&D, whether in whole or in part,
can be and has been implemented in a variety of ways.

FSR&D strongly emphasizes working with farmers
in their fields. To the extent that this emphasis is new,
those currently at research stations or at regional or na-
tional headquarters will require some reorientation. This
reorientation includes research methodology as applied to
field conditions and methods for working with the whole
farm family-male and female, young and old. Where ap-
propriate, females may need to be added to the research
and extension staff.

However, FSR&D does not replace existing research
or extension; rather, it builds on the existing base. Conse-
quently, experienced researchers and extension specialists
usually remain in their existing organizations and much of
the field staff consists of young professionals trained
specifically for FSR&D's purposes. Enough senior staff
members will be needed-whether nationals or expatri-
ates-to guide the younger members of the staff until they
gain adequate experience.
One approach is to begin FSR&D activities in one or
two regions and, after several years of experience, to
choose leaders from these teams when moving to new
areas. Heads of FSR&D programs must also train staff to
replace those who periodically leave the program.


An early activity when implementing an FSR&D ap-
proach is to train the staff about the objectives, processes,
and methodologies of FSR&D. Training materials will
need to be collected from ongoing programs elsewhere and

8 Executive Summary

augmented by new materials appropriate for the country.
During this early stage, the International Agricultural
Research Centers (IARCs) and organizations with similar
activities can be especially helpful.
The principal objectives of the training are to

acquaint team members with on-farm techniques
give them guidance and experience working as an in-
terdisciplinary team
instill in the team members an enlightened apprecia-
tion of small farmers as a useful source of informa-
tion and as valuable partners in the research and im-
plementation process.

Where members of the FSR&D team are recent graduates,
in-service field training under the guidance of experienced
staff is needed.
Initially, program leaders may want to take advan-
tage of production and farming systems training at one or
more of the IARCs and any regional center specializing in
applied agricultural research. With such training as a base,
in-country training programs for both research and exten-
sion personnel can then be developed and implemented.
Some staff members may be selected and sent abroad for
further academic training.
In training, as well as in other aspects of FSR&D pro-
grams, national governments may want to consider using
expatriate staff experienced in FSR&D. As the program
matures, the expatriates can be phased out gradually. In
one case, this occurred about six years after the program

1.4.5. FSR&D COSTS
A discussion on the relative costs of FSR&D centers
on expenditures, rates of adoption, and breadth of
coverage. This discussion must be general since carefully
quantified appraisal of FSR&D's costs, relative to other
research and development approaches, has not, to our
knowledge, been made. While firm estimates are not
available, those closely associated with FSR&D generally
feel that the approach is cost effective. The reasoning
The first of the three issues concerns expenditures
for facilities and operating costs. To the extent that
FSR&D reduces experiment station activity, costs of ex-
panded installations, operations, and the accompanying
staff will be lowered. In its place will be more work on
farmers' fields by generally less expensive staff. However,
the field work requires increased expenditures for vehicle
purchase and maintenance, field equipment, per diem, and
incentives. Overall, the combined initial and recurring
costs of FSR&D appear to be less than the costs of com-
parable levels of activities on experiment stations, when
administrators consider the costs of building, staffing, and
equipping the stations. However, such comparisons are of
limited value since FSR&D replaces only a portion of ex-
periment station activities.
The second issue concerns the generation of new

technologies acceptable to farmers. This too is not a
straightforward issue, because the target group for FSR&D
is sometimes different from that of general agricultural
research. Proponents of FSR&D, however, point to the
high levels of adoption of improved technologies by small
farmers targeted by the FSR&D process.
The third issue centers on the range of applicability
of research results. Opinion differs about how widely
FSR&D can be applied. Traditional research, by its nature,
often has general and wide applicability. FSR&D is design-
ed to be more specific, but it may also be applied broadly if
the team can identify environmental conditions sufficient-
ly wide ranging and target groups in sufficiently large
numbers. FSR&D practitioners expect work in categoriz-
ing research areas to eventually make it easier to locate
situations in which the new technologies generated by
FSR&D will have broad applicability. Eventually, the
study of environmental gradients will permit a better
understanding of the relationship between research results
and the conditions leading to these results, but this latter
possibility, especially when speaking of national pro-
grams, lies in the future.

Because FSR&D concentrates on field activities, the
government will need to take steps to allow team
members to effectively carry out this work. Materials for
conducting experiments need to be available at appropriate
times, otherwise the experiments may not be completed.
Reliable transportation is essential, especially where the
terrain and weather conditions make travel difficult. The
FSR&D team needs adequate servicing and spare parts for
its vehicles. Finally, incentives are often required to at-
tract and hold qualified staff. Incentives such as the
recognition of team accomplishments will be needed to
overcome the uncertainties of working in a new and dif-
ferent program and the hardships of living and working in
remote areas.
FSR&D does not place great demands on the govern-
ment, but these demands must be met to create and main-
tain the momentum necessary to sustain an effective
FSR&D effort. Where the central or regional organization
cannot meet some of the above requirements, the orga-
nization should give the field teams adequate local auton-

To repeat, this book of guidelines describes an ap-
proach to agricultural research and development for
governments of developing countries interested in improv-
ing the output and welfare of small farmers. We present
the FSR&D activities, methods, and illustrations of
various approaches in the main body of this book and
elaborate on these points in the appendixes. We emphasize
cropping systems research because most experience lies
here; however, we include materials on livestock systems.
Systems concepts are included, but few analytical tools for

Executive Summary 9

systems analysis such as simulation or linear program-
ming are included because we found few examples of their
use in national FSR&D programs.
This book of guidelines is for those in the developing
countries who must decide whether to accept FSR&D and
bear the responsibility for its implementation. This book
is also for the expatriate who aids in this process.
In designing FSR&D activities, administrators must
decide on the approach, methods, organization, staffing,
training, and ways to secure technical assistance and
funds. The book should aid such individuals in making
reasoned decisions on these topics. Because of the diver-
sity of conditions and the wisdom of allowing those in a

country to make their own decisions, the book does not
prescribe how a country should implement FSR&D ac-
tivities. Instead, the book presents general concepts, offers
alternatives that have worked in different countries, and
provides the reader with sources of additional information.
In conclusion, undertaking an FSR&D approach that
modifies a more traditional approach to agricultural
research presents a considerable challenge to any country.
Existing institutions and individuals may feel threatened
by the change. False starts are possible. Still, if the en-
thusiasm of those who have been most active in the
FSR&D movement is any indication of its validity, the ef-
fort is justified.

Chapter 2




During the past decade, considerable attention has
been focused on the plight of the rural poor in the develop-
ing countries. One aspect of this emphasis has been to
direct agricultural research specifically to the needs and
aspirations of farmers with limited resources. Historically,
these have been small farmers who have not adequately
benefited from agricultural research because the research
was not specific enough for their needs. Instead, research
in the less developed countries has typically been under-
taken for farmers who have more resources and who often
produced for export.
Generally, technologies offered to the small farmers
have come from a top-down approach. By that, we mean
the research would be largely initiated and conducted on
experiment stations and then offered to small farmers to
accept or reject. As a result, farmers rejected many of the
proposed changes because the suggested improvements
were unprofitable or too risky, or the farmers lacked ade-
quate inputs or suitable markets. In short, the
technologies were not suitable because the researchers did
not know or consider the conditions of small farmers.
Therefore, research, extension, and other programs
are needed to correct these deficiencies, if small farmers in
developing countries are to be helped. One approach that
considers farmers' conditions specifically is called farming
systems research and development (FSR&D), or simply
farming systems research (FSR). In this book we use the
term FSR&D to emphasize the integration of research and
the development of technology for dissemination through
extension and by other means. The FSR&D approach pro-
vides a means for dealing with the close interaction of the
many on-farm activities that characterize subsistence
This book of guidelines was written for those with
national programs in developing countries who wish to
orient part of their research efforts toward benefiting small
farmers. These guidelines concern FSR&D processes and
procedures and include examples from specific situations.
This chapter covers the purposes and definition of FSR&D,
additional definitions and comments, characteristics of
FSR&D, objectives and users of these guidelines, the
guidelines' scope, and the approach and contents of the

As with other national approaches to agricultural
research and extension, FSR&D's purpose is to generate

more appropriate technologies for farmers and, where
possible, to improve policies and support services for farm
production, to raise farm families' welfare, and to enhance
society's goals. But more specifically, FSR&D aims at in-
creasing the productivity of farming systems by generating
technologies for particular groups of farmers and by
developing greater insight into which technologies fit
where and why. This latter purpose concerns using scien-
tific methods for generating hypotheses and then, by
deduction, determining which technologies to use in a par-
ticular farm setting. Such an approach contrasts with an
empirical approach that through trial and error arrives at
suitable technologies for the conditions of specific farmers
(Harwood, personal communication).
We include the farm family in the above description
because the collective interests of the family are impor-
tant, not just the interests of the head of the household.
Furthermore, we include agricultural production because
FSR&D concentrates on increasing crop and livestock
yields and overall farm output. And we include family
welfare because improved welfare is the ultimate goal of
individual families just as societal interests are the
ultimate concern of an enlightened government.

FSR&D is an approach to agricultural research and
development that

views the whole farm as a system
focuses on (1) the interdependencies between the
components under the control of members of the
farm household and (2) how these components
interact with the physical, biological, and
socioeconomic factors not under the household's

Farming systems are defined by their physical, biological,
and socioeconomic setting and by the farm families' goals
and other attributes, access to resources, choices of pro-
ductive activities (enterprises), and management prac-
The systems approach applied to on-farm research
considers farmers' systems as a whole, which means

1) studying the many facets of the farm household and
its setting through close and frequent contact with
household members on their farms

2) considering problems and opportunities as they in-
fluence the whole farm
3) setting priorities accordingly
4) recognizing the linkages of subsystems within the
farming system and considering them when dealing
with any part of the system
5) evaluating research and development results in
terms of the whole farming system and the interests
of society.

The FSR&D team implements the FSR&D process

1) selecting areas and groups of farmers with
reasonably similar characteristics as targets for
research and development
2) identifying and ranking problems and opportunities
and setting forth hypotheses for alternative solutions
3) planning experiments, studies, and procedures for
data collection

4) undertaking experiments on farmers' fields, in con-
junction with other research, to identify or generate
improved technologies suitable for farmers' condi-
5) coordinating the on-farm experiments and studies
with commodity and disciplinary-oriented research
6) evaluating the acceptability of the results of these
experiments to the targeted farmers and society
7) extending the results widely to farmers within and
outside the target area
8) focusing attention on ways to improve public policy
and support services to assist both the targeted
farmers and those operating under similar condi-

The distinction between FSR&D and "con-
ventional"1 research can be summarized in the follow-
ing way. FSR&D looks at the interactions taking place
within the whole farm setting and measures the results in
terms of farmers' and society's goals. Traditionally, con-

'The authors encountered differences of opinion as to the meaning of conventional research, consequently, the use of the quotation

Introduction 15

ventional research separates tasks into progressively nar-
rower subject areas to be studied more or less independent-
ly and then evaluates results by standards within the
discipline, not by their contribution to the whole (Dillon,
1976). Furthermore, FSR&D places relatively more impor-
tance, than in the past, on integrating the social sciences
into the research and development process. This is ac-
complished by considering such factors as farmers'
preferences, community norms, markets, public policies,
and support services (Norman, personal communication).
FSR&D's comprehensiveness can be illustrated by
discussing the differences between FSR&D's results and
those of a single disciplinary approach. For example, a
breeder may seek to obtain the highest physical yield for a
single crop through variety and fertilizer trials. In contrast,
an FSR&D approach integrates the breeder's work by con-
sidering more objectives and means of improvement. For

SAn earlier maturing variety might be sought that
allows time for planting a second crop, even though

the yield from such a variety is less than from other
Net profits from fertilizer application could be in-
creased by reducing the application rate to a lower
level than is needed to produce the maximum
biological yield.
Recognizing farmers' aversion to risk could suggest a
less profitable crop whose yields are more stable dur-
ing unfavorable growing conditions.
Social and cultural study could explain why some
farmers accept improvements and others do not, so
that the resulting technologies could be applied to
more farmers.
Integrating the extension service into the FSR&D
process could result in the extension staff suggesting
modifications to the technologies; these changes, in
turn, could help the extension service serve farmers
more effectively.

Not all aspects of a farming system must be ad-
dressed for the process to be considered FSR&D. Crop-

k'-. t- ,


L 7, v ** II
S- "- -- A

^ i."", "'r h --'- ,

. r"r

3L C '

16 Introduction

ping and livestock systems and even commodity research
may qualify. What is needed is for the research on sub-
systems-e.g., cropping systems-to be taken within
the context of the whole farm. Such an approach for crop-
ping systems requires a study of the farming system to
that research into cropping systems is justified
that the research on the subsystems and the
resulting recommendations fit within the overall
that the final evaluation is within the whole farm

Finally, some improvements in farmers' conditions
may not result from breakthroughs in agricultural
technologies; instead, the improvements may result from
identifying and implementing more suitable agricultural
policies and support services.

To help clarify the above concepts, we will now
define additional terms used in this book. They are farm-
ing, households, small-scale farming, enterprises, farming
systems, cropping and livestock systems, mixed systems,
cropping systems research, livestock and mixed systems
research, and commodity-oriented research.

2.3.1. FARMING
Farming is an activity carried out by households on
holdings that represent managerial units organized for the
economic production of crops and livestock (Ruthenberg,


The household is a social organization in which
members normally live and sleep in the same place and
share their meals. They may or may not be a joint family.
A joint family is one consisting of two or more lineally
related kinfolk, their spouses, and offspring.
Women may be heads of households in various
ways, as (1) recognized heads of households such as when
they are widowed or divorced, (2) acting heads such as
when their husbands are away for extended periods, or (3)
informal heads such as when they have command over
resources and make decisions on their initiative. Even
when they are not heads of households, women usually
have a recognized and important role through their con-
tribution of labor, management, marketing, and owner-
ship of resources. At times, individuality among males
and females leads to competition within the households as
when husbands sell firewood to their wives (Venema,
1978) or when wives and husbands lend each other money
with interest (Robertson, 1975-76).
Most farm households in developing countries strive
to produce a dependable and continuous food supply and
many of their other needs such as clothing and shelter, and

surpluses for sale. To do this, the members of the
household engage in several on-farm enterprises using
primarily their labor. Furthermore they are cautious about
adopting changes that threaten their ability to maintain a
reliable food supply. Members customarily have duties
within the household according to sex, age, and relation-
ship that are dictated by custom and practical considera-


In this book of guidelines, we emphasize small-scale
operations in which the farmers frequently have difficulty
obtaining sufficient inputs to allow them to adequately
use the available technology as would medium-scale and
large-scale commercial farmers. Small does not necessar-
ily refer to the area of land held because some farmers that
meet our definition of small-scale farming have access to
considerable amounts of land- as do pastoralists and shift-
ing cultivators. Such small-scale farmers are unable to
easily raise their levels of production because of limited
resources and technologies suitable for their needs.


Enterprises mean activities undertaken to produce
an output that contributes to total production or income of
the farm family. Enterprises in FSR&D typically concern
crops, livestock, processing or otherwise upgrading
agricultural commodities produced on the farm, produc-
tive nonagricultural activities carried out on the farm such
as handicrafts, and productive off-farm activities of the
household members.

For this book of guidelines, we consider a farming
system as a unique and reasonably stable arrangement of
farming enterprises that the household manages according
to well-defined practices in response to the physical,
biological, and socioeconomic environments and in accor-
dance with the household's goals, preferences, and
resources. These factors combine to influence output and
production methods. More commonality is found within
the system than between systems. The farming system is
part of larger systems-e.g., the local community-and
can be divided into subsystems-e.g., cropping systems.
Decisions as to classification depend on the needs
for analysis and decision-making. A system with a greater
cash income-e.g., when off-farm employment or sale of
handicrafts is possible-is different from one with lesser
cash income. Even though the same crops, patterns, and
management practices may be followed, farmers' reactions
to change will vary because of different capabilities, at-
titudes, and other factors (Harwood, 1979).

These are subsystems within the farming system. A
cropping system, a set of one or more crops, comprises all

Introduction 17

a~ r~?jjr

components required for production, including the in-
teractions between other household enterprises, and the
physical, biological, and socioeconomic environments.
Livestock systems can be defined similarly.


Cropping, livestock, and possibly other enterprises
are present within the farming system.


Research on cropping systems concentrates on crops
and cropping patterns, alternative management practices
in different environments and interactions between crops,
between crops and other enterprises, and between the
household and environmental factors beyond the
household's control. The procedures are similar to farming
systems research, but the breadth of cropping systems
research is generally less. However, when the initial

analysis considers the whole farm situation and then
focuses on cropping systems as the best area for research,
the differences between the approaches of cropping
systems research and of FSR&D are few.


The approach to livestock systems research and
mixed systems research follows a process similar to crop-
ping systems research except for the procedures that
reflect the inherent differences between cropping and
livestock systems-e.g., fewer numbers of animals than
plants. In mixed systems research, the team focuses
directly on the interactions between crops, livestock, and
possibly other enterprises.

Commodity-oriented research focuses on one or
more crops or animals by studying them in detail. Com-

18 Introduction

modities selected for emphasis should be the result of
prior investigation demonstrating their importance to the
farming system. While often conducted along disciplinary
lines such as breeding, physiology, and pathology,
commodity-oriented research supports FSR&D best when
it keeps the needs of the whole system in mind and takes
advantage of FSR&D results as a source of information for
making its programs more relevant to the farmers' cir-
cumstances. Commodity-oriented research organizations
such as the International Maize and Wheat Improvement
Center (CIMMYT), International Rice Research Institute
(IRRI), and International Center for Tropical Agriculture
(CIAT) work mainly with farmers and research organiza-
tions in those areas of their specialties that offer the best
potential for improving the farming system.

Further comments on the nature of FSR&D should
help in understanding the emphasis placed on this ap-

proach. FSR&D considers the farmers and their problems
in a comprehensive manner using an interdisciplinary ap-
proach that complements existing research and develop-
ment activities, and is iterative, dynamic, and responsive
to society. Many of these characteristics have their origin
in farm management that has been practiced in the United
States since the early 1900's. However, the emphasis on a
systems approach that considers the whole in terms of the
parts and evaluates results in terms of farmers' and
society's goals is relatively new in developing countries.
Moreover, the new emphasis in developing countries is on
research on farmers' fields using interdisciplinary teams.
In time, formal reference to FSR&D may fade away
as the process and procedures are absorbed into
agricultural research and development programs. But, in
the meantime, considerable opportunity remains for im-
proving FSR&D concepts, developing FSR&D procedures,
and expanding the FSR&D approach into areas such as
livestock and mixed-farming where experience and
literature are limited.

Introduction 19

FSR&D starts with farmers and learns about their
environments, resources, methods of production, pro-
blems and opportunities, aspirations, and how they react
to change. The FSR&D team designs experiments with
these factors in mind, carries out studies in farmers' fields,
and judges the results by farmers' standards. While other
factors enter into the process, FSR&D strongly emphasizes
obtaining a clear picture of farmers and their en-
vironments. Much can be learned by literally "walking in
the farmers' footsteps." Farmers may not have accurate
technical explanations of their problems nor know the
range of opportunities for improving their conditions, but
learning more about farmers helps the researchers produce
better technologies and extension workers promote
FSR&D's results more effectively.
Because the farm household both consumes and pro-
duces, the values of the marketplace and the farm
household are mixed. Farmers tend to be cautious about
change-especially involving their subsistence crops-but
they will change when suitable opportunities arise. They
have multiple goals and the community's norms influence
farmers in varying degrees. FSR&D practitioners consider
farmers rational according to the farmers' values and
perceptions of alternatives. But, individual farmers have
different values, perceptions, skills, and resources. Thus,
some farmers produce more and accept change more read-
ily than other farmers.


While some farming systems research applies broad-
ly and is long-run, a national FSR&D program tends to be
applied to specific, short-run objectives, as when adapting
available technologies. FSR&D identifies problems on
farms and introduces improvements that frequently re-
quire little governmental support. The approach identifies
farmers' constraints and distinguishes between those con-
straints that are within and those that are beyond their
control. For example, farmers can often implement
changes in varieties, planting distances, methods of apply-
ing fertilizers, and time of weeding, but may have difficul-
ty if a solution calls for more rapid plowing requiring oxen
that are either not available or too costly. In the first case,
we can look internally at ways to introduce change to the
farmers. In the second case, we must look externally to
changes that will make oxen available or else redesign the


FSR&D studies the whole farm setting to identify
problems and opportunities, notes their interrelation-
ships, sets research priorities responsive to farmer and
societal goals, carries out experiments, proposes changes
in light of this comprehensive perspective, measures
results in terms of impacts on the farmers and society,
observes farmer acceptance of change, and transfers accep-
table research results to implementing organizations. The

FSR&D team uses a whole-farm perspective to identify the
most relevant problems and to evaluate the acceptability
of results.
Even though FSR&D views the farming system and
its environment comprehensively, some aspects may not
be researched or considered for change-for example,
farmers' values and social customs or the level of support
services. As part of FSR&D's strategy, the FSR&D team
decides which areas offer the greatest potential for change
and are the most suitable for research and development
(Sec. 3.4. in Chapter 3).


Because of the comprehensive approach and interac-
tions of many technical and human factors, FSR&D teams
should be interdisciplinary. By interdisciplinary, we mean
frequent interactions among those from different
disciplines who work on common tasks and come up with
better results than had they worked independently. As a
minimum, both technical and social sciences should be
represented on a team with leadership strong enough to in-
tegrate the disciplines and direct their efforts toward team
objectives. Moreover, where cultures discourage com-
munication between unrelated men and women, field
teams may have to have members from both sexes if they
are to adequately communicate with male and female
farmers (Staudt, personal communication).


FSR&D replaces neither commodity nor disciplinary
research nor extension. On the contrary, FSR&D requires
a continuing inflow of improvements from such research
and close contact with farmers through extension. For ex-
ample, if a shorter season variety is needed to overcome a
problem associated with planting time, researchers will
have more reason to believe that success in identifying a
suitable variety will lead to its acceptance. Or, if extension
is having difficulty introducing change to its clientele,
FSR&D provides a means for bringing farmers' problems to
researchers. In view of these advantages, the direction that
FSR&D gives to commodity and disciplinary research,
coupled with FSR&D's influence on extension, is as im-
portant as improvements introduced to farmers par-
ticipating directly in FSR&D experiments.

The FSR&D approach calls for a conceptual
understanding of the farming system and its environment
from the very beginning. This framework provides the
basis for gathering data and directing the course of the
research and development effort. Initially, the system may
not be well understood, but the conceptualization im-
proves as the FSR&D team gathers data and gains ex-
FSR&D's iterative nature is shown by the process by
which the team begins by acting on partial information,
gains insight through studies and experimentation, and

20 Introduction

-' U

modifies its actions. This process continues until research
and extension staff are satisfied that changes can be
broadly implemented. Such an approach encourages the
FSR&D team to begin working within a whole farm
framework from the outset, rather than working
haphazardly or waiting for excessive precision before in-
itiating on-farm research. In this way, FSR&D seeks to
provide better solutions to farmers' conditions, but not
necessarily the best solutions.
Solutions to one set of problems usually generate op-
portunities for further research. FSR&D is dynamic in that
objectives and approaches for future work can be adjusted
in light of the accomplishments. For example, FSR&D
might initially work with only slight modifications in
farmers' existing cropping and livestock patterns. After the
farmers grow accustomed to change, greater modifications
to their farming systems could be tried.

As with other national programs that rely on private
initiative, FSR&D needs to produce results acceptable to

small farmers and society. Consequently, FSR&D operates
from the farmers' and society's viewpoints. The two can be
brought into accord by identifying issues of possible con-
flict and agreement, by measuring possible impacts of
alternative courses of action, and by devising appropriate
incentives and restrictions. For example, farmer groups in
Kenya receive government, technical, and financial
assistance in organizing and constructing terracing, in-
terceptor ditches, and other forms of soil and water conser-
vation. Without some government assistance, the farmers
would often not be inclined to do this work, which is in
their and society's long-run interests.

A few very capable persons have been able to apply
FSR&D concepts quickly and accurately. They succeed
because of their considerable experience and gift for pro-
posing practical solutions to complex problems. These
persons are extremely useful, but too few of them are
available to the developing countries. A number of
organizations and individuals are currently setting down

Introduction 21

their thoughts and findings on this subject in the attempt
to institutionalize the FSR&D approach. In this way,
FSR&D can be described, taught, learned, and applied on a
much broader scale. This book of guidelines is part of that
effort. These guidelines:

show the general process for FSR&D programs at the
national level
provide procedures that illustrate how to carry out
the FSR&D process under different conditions
supply an overview of the current state of FSR&D so
that readers will understand its nature, its relation-
ships to other agricultural research, and its potential
for contributing to increased production and the
welfare of small farmers.

This book of guidelines was written specifically for
those in the developing countries who are responsible for
deciding on and implementing an FSR&D approach and for
those with a general interest in FSR&D. This audience in-
cludes five groups.
The first group consists of top decision makers,
possibly at the ministerial or cabinet level, who decide
whether or not an FSR&D approach will be undertaken.
This group's members need to understand the advantages
and disadvantages of FSR&D, as well as its implications,
such as

effectively reaching the intended farmers
staffing and training requirements
length of time necessary to start
institutional and policy relationships with ongoing
programs in agriculture and other areas
relative costs
special governmental support.

We prepared Chapter 1, a summary of the principal
features of these guidelines, for this group.
The next group of potential readers includes those
who are responsible for administering FSR&D activities.
They might be technical directors and their seconds in
command. Although such individuals need to know about
FSR&D's broad implications, they will be more concerned
with understanding the entire process, how the parts fit
together, how others have organized programs and pro-
jects, and the strengths and weaknesses of alternative ap-
proaches for particular situations. They will need to know
enough about the procedures to select appropriate staff and
integrate their activities. We provide this information in
the remaining chapters.
The third group consists of regional directors, field
team leaders, and technical researchers who carry out the
various FSR&D tasks under the direction of the technical
directors and their assistants. We expect them to be in-
terested in individual chapters and the appendixes.
The fourth group includes expatriates who may be
technical advisers to each group. They could be involved
in all phases such as advising on whether to accept an

FSR&D approach, aiding in program and project design,
training staff, implementing FSR&D activities, and
assisting in evaluation of results. For them, all parts of the
book-the summary, main chapters, and appen-
dixes-should be useful.
The last group comprises all the rest who have an in-
terest in FSR&D, including others within the developing
countries, members of international agricultural research
and development organizations, agricultural researchers,
and others with similar interests outside the developing

2.7. SCOPE
This book of guidelines was written to convey the
approach to FSR&D generally being followed by or recom-
mended for national governments. By being general, this
book provides information on concepts applicable to a
variety of situations; however, this book alone will not
provide enough information for designing an FSR&D ap-
proach for a particular country. Because of the breadth of
FSR&D, many topics are included in this book. Some are
developed in considerable detail, while others are not. In
this section on the scope of FSR&D, we discuss the topics

22 Introduction

included in this book and why we emphasized them. At
times, we have said little about a topic because informa-
tion and experience are lacking.


These guidelines reflect the state of the art of
FSR&D at the national level. They encompass a broad
range of conditions, which can help the reader understand
FSR&D's general process. The illustrations provide the
reader with the opportunity to see how some of the prin-
ciples have been applied to specific situations. With these
concepts and illustrations, those in developing countries
should be in a position to design an FSR&D approach
specifically for their conditions. For example, the FSR&D
team needs to select an overall strategy responsive to the
country's goals, resources, stage of development, and types
of farming systems. Moreover, once the approach has been
designed, training is needed. These guidelines include the
basis for selecting training materials, but we do not pro-
vide detailed instructions on training in FSR&D.

The farming system consists of subsystems-e.g.,
cropping systems-and in turn is part of larger
systems-e.g., the local community. Fig. 2-1 depicts the
general categories of factors that influence small-farmer
production and welfare. The figure shows that the farming
system is made up of crops, livestock, and other on-farm
subsystems. The farming system is greatly influenced by

Figure 2-1. Farmers' setting and scope of the guidelines.

KEY. 0
Guideline's area of primary concern

the physical setting and the rural community where the
farm household members trade, socialize, and otherwise
take part in local affairs. The rural community, in turn, is
part of a regional and national setting. Government's na-
tional and regional policies and institutions reach the
household level to help the members through research, ex-
tension, and agricultural production programs, and by pro-
vision of infrastructure. Private and government
enterprises large-scale agriculture, manufacturing, trade,
etc.-also interact with the farm household by providing
services and supplies, facilitating the sale of farm output,
and offering off-farm employment.
Because FSR&D is comprehensive, these guidelines
are concerned directly and indirectly with all activities
that touch upon the farmers' lives. But, for practical
reasons, we have included FSR&D procedures primarily
on (1) agricultural research, (2) extension's part in transfer-
ring the technologies, (3) organization, and (4) training.
Also, many of the procedures relate to cropping systems
because most experience is there.
In this book, we use only general concepts of sys-
tems analysis and provide little discussion of the tech-
niques of systems analysis such as linear programming,
simulation, or other forms of mathematical modeling.
This omission should not be construed to mean that we
think that such modeling is not useful. We have excluded
mathematical modeling because our review of FSR&D
practices did not reveal at this time its significance among
national programs. In contrast, interesting work is occur-
ring at some of the international centers, such as the con-
ceptual and detailed livestock models of the International
Livestock Centre for Africa (ILCA), the simulation and op-
timization models applied by IRRI, and the ecology-based
models of the Tropical Agricultural Research and Training
Center (CATIE). We describe this work briefly in Appen-
dix 2-A. Included in some of the modeling work at IRRI
and other institutions are studies of relatively narrow
topics, such as soil-water-plant relationships.
Except for a few cases, we have not included, in this
book, procedures that are either commonly available in
the literature or are narrowly confined to a discipline and
are not central to the FSR&D process. For example, we
considered experiment station procedures as being general-
ly well-known and soil-water-plant relationships as nar-
rowly confined to a discipline and not central to FSR&D
methodology. For this same reason we have not provided
detailed procedures on topics such as marketing, transpor-
tation, credit, and extension at the agent's level.

As noted above, the guidelines in this book apply to
a broad range of situations encountered in different coun-
tries throughout the world. They apply to countries that
contemplate instituting FSR&D activities, as well as to
those with full-fledged FSR&D programs. They apply to
semiautonomous research corporations that conduct
FSR&D or ministries that have FSR&D as part of a
broader program. And they apply to production programs
in which upgrading agricultural research is important.

Introduction 23



In this book, we have

described alternative approaches to various situa-
suggested reasons for the differences in approaches
identified references where research procedures are
reasonably standard
included detailed descriptions of procedures that are
not standard or where we felt emphasis was needed.

We have included numerous references to the
FSR&D work of national and international groups
throughout the book. These references should prove useful
for details about specific concepts and procedures.
Especially helpful in preparing this book of guidelines
were the published and draft guidelines prepared by CIM-
MYT (Perrin et al., 1976; Byerlee et al., 1980), and the In-
ternational Rice Research Institute (Zandstra et al., 1981),
two state-of-the-art papers on farming systems (Gilbert et
al., 1980 and TAC, 1978), and a paper on cropping systems
research in Indonesia (McIntosh, 1980).


We have organized the remainder of this book of
guidelines so that readers begin with a review of the
overall approach to FSR&D, and then learn the details of
each of the principal activities. With that background,
readers should be able to decide whether or not FSR&D is
suitable for their situations. Assuming that some readers
will be interested, we then provide information on how to
implement the approach. The appendixes elaborate on the
More specifically, Chapter 3 contains the conceptual
framework of the principal features of FSR&D. The next
five chapters elaborate on FSR&D's major activities:
Chapter 4 explains target and research area selection,
Chapter 5 presents problem identification and develop-
ment of a research base, Chapter 6 discusses planning on-
farm research, Chapter 7 presents on-farm research and
analysis, and Chapter 8 discusses extension of results.
Then, Chapter 9 considers whether or not an FSR&D pro-
gram is in a country's best interests. The next two chapters
explain implementation: specifically, Chapter 10 presents
alternative organizational approaches for FSR&D, inter-
disciplinary teamwork, staffing, and other management
issues, and Chapter 11 reviews training as a way of im-
plementing new or improved FSR&D programs. The ap-
pendixes contain detailed information supporting the
various chapters, some general procedures, and worldwide
examples and illustrations. Most chapters end with
references and suggested readings.

Byerlee, D., M.P. Collinson, R.K. Perrin, D.L. Winkelmann, S.
Biggs, E.R. Moscardi, J.C. Martinez, L. Harrington, and A.
Benjamin. 1980. Planning technologies appropriate to
farmers: concepts and procedures. CIMMYT, El Batan,
Dillon, J.L. 1976. The economics of systems research. Agric.
Sys. 1:1:5-22.
Gilbert, E.H., D.W. Norman, and F.E. Winch. 1980. Farming sys-
tems research: a critical appraisal. MSU Rural Dev. Paper
No. 6. Dep. of Agric. Econ., Michigan State Univ., East
Lansing, Mich.
Harwood, R.R. 1979. Small farm development: understanding
and improving farming systems in the humid tropics.
Westview Press, Boulder, Colo.
McIntosh, J.L. 1980. Cropping systems and soil classification for
agrotechnology development and transfer. In Proc.
Agrotech. Transfer Workshop. 7-12 July 1980. Soils Res.
Inst., AARD, Bogor, Indonesia and Univ. of Hawaii,
Perrin, R.K., D.L. Winkelmann, E.R. Moscardi, and JR. Ander-
son. 1976. From agronomic data to farmer recommenda-
tions: an economics training manual. Inf. Bull. 27.
CIMMYT, El Batan, Mexico.
Robertson, C. 1975-1976. Women and change in marketing con-
ditions in the Accra area. In Rural Africana. Winter
1975-1976. Michigan State Univ., East Lansing, Mich.
Ruthenberg, H. 1971. Farming systems in the tropics. Clarendon
Press, Oxford, UK.
Technical Advisory Committee (TAC). Review Team of the Con-
sultative Group on International Agricultural Research.
1978. Farming systems research at the international
agricultural research centers. The World Bank,
Washington, D.C.
Venema, L.B. 1978. The Wolof of Saloum: social structure and
rural development in Senegal. Center for Agricultural
Publishing and Documentation, Wageningen, Neth-
Zandstra, H.G., E.C. Price, J.A. Litsinger, and R.A. Morris. 1981.
A methodology for on-farm cropping systems research.
IRRI, Los Banos, Philippines.


IRRI. 1977. Symposium on cropping systems research and
development for the Asian rice farmer. IRRI, Los Banos,
De Tray, D.N. 1977. Household studies workshop. In A/D/C
Sem. Rep. No. 13. The Agricultural Development Council,
Inc., New York.
Staudt, K.A. 1978. Agricultural productivity gaps: a case study of
male preference in government policy implementation.
Development and Change 9:439-457.

Chapter 3

FSR&D is a process that involves a set of interrelated
activities. While individual programs reflect specific con-
ditions within a country and the preferences of its leaders,
the various approaches to FSR&D have much in com-
mon. Thus, the conceptual framework described in this
chapter shows this commonality. And, this framework
illustrates the nature and interrelationships of each of
these activities and should help the reader understand the
material presented in subsequent chapters. The fol-
lowing sections briefly describe FSR&D activities, timing
of activities, staffing requirements, and strategies. The
chapter ends with a summary of the conceptual

A natural sequence for the FSR&D process, as out-
lined in Fig. 3-1, includes

target and research area selection
problem identification and development of a
research base
planning on-farm research
on-farm research and analysis
extension of results.

Experiment station collaboration, which interacts
primarily with the first four activities, is set off to the side
to emphasize its supporting role in on-farm research. Also,
extension is shown as collaborating with each of the five
principal FSR&D activities. Results from the last two
activities on-farm research and analysis, and extension
of results-feed back to the earlier activities.
While these activities are shown as being sequential
and discrete, in practice, an activity often overlaps with
other activities, some activities are not taken in the se-
quence indicated, and some are repeated. In fact, the
iterative nature of FSR&D usually calls for several repeti-
tions, which leads to progressively improved results.
While flexibility of the approach exists, omitting any one
of the activities may jeopardize the value of the final
The FSR&D approach typically uses interdis-
ciplinary teams and the team composition varies accord-
ing to the task and available staff. During FSR&D's early
stages, disciplinary representation should be broad and
flexible to allow adequate response to problems and oppor-

tunities concerning the targeted group of farmers
(Hildebrand, personal communication). Field teams, nor-
mally residing in or near the research area, conduct on-
farm research aided by specialists and others concerned
with agricultural production. The specialists come from
the physical, biological, and social sciences, and include
extension. These specialists may also reside in the
research areas, or operate out of regional or national head-
quarters or experiment stations. Section 3.3. provides ad-
ditional details on staffing requirements.

The FSR&D team aids national decision makers in
selecting target areas and target groups of farmers. The
team subsequently divides the target areas into subareas
and establishes the boundaries of the research area. In
FSR&D, the government chooses, implicitly or explicitly,
target areas or target groups of farmers to receive increased
attention. These choices are generally based on national
policies reflecting governmental objectives such as the
better use of resources, raising the income of poor farmers,
and greater domestic food production. A response to these
objectives is more appropriate agricultural technologies,
policies, and services.
The FSR&D team usually divides the target areas or
target groups of farmers into subareas according to com-
mon physical, biological, and socioeconomic char-
acteristics. Such a stratification separates environmental
conditions and farming systems into reasonably
homogeneous segments. By working with these
homogeneous segments, the FSR&D team is able to
develop improved technologies for farmers operating
under similar conditions throughout the target area. And,
sometimes the technologies are suitable for farmers out-
side the target area. Stated somewhat differently, ap-
propriate stratification according to environmental condi-
tions and farming systems enables the FSR&D team to
identify farmers who are expected to benefit from the same
recommendations (Byerlee et al., 1980).
The number and location of research areas
throughout a country depend on FSR&D goals and
resources and the characteristics of the areas. Because the
relationships between farmers' practices and their en-
vironments take time to understand, the research areas
should be broad enough to allow the team to adjust its ap-
proach as new information is obtained.

28 Conceptual Framework

Figure 3-1. The five basic activities of on-farm research in


After the research area has been selected, the FSR&D
team moves to more careful and detailed studies of farm-
ing systems and the area's characteristics. The team
studies, analyzes, and ranks farmers' problems and oppor-
tunities and either acts upon them immediately or plans
further studies and experiments.
The team gathers information about the farmers and
their environments through a review of secondary infor-
mation1 and through direct observation and discussions
with members of farm households. Many experienced
FSR&D practitioners strongly endorse rapid surveys at the
outset of FSR&D activities in an area. However, FSR&D
teams sometimes conduct experiments to help understand
the farming system and to identify problems. As part of the
next set of activities, the FSR&D teams plan more
elaborate surveys, studies on special topics, monitoring
activities, and experiments.
While starting research in the right direction is im-
portant, the nature of FSR&D allows the team to adjust its
approach as information is gained from experiments,
studies, and other forms of research. So, rather than delay
until a precise plan of action can be prepared, FSR&D
teams are advised to begin research early. In fact, FSR&D
practitioners urge that a cropping season not go by without
conducting some form of on-farm experiment. Livestock
systems, on the other hand, are more complex and may re-
quire more careful planning.


The research team begins the research design and
planning activity by

1) reviewing the priorities given to problems and op-
portunities and the hypotheses for solution
2) reviewing previous research findings
3) seeking whatever help is needed and available from
regional, national, and international sources
4) deciding whether to accept the current environmen-
tal conditions as they are, or to assume some degree
of change.

With its research agenda set, the team analyzes cur-
rent farming practices and environmental conditions as
they specifically relate to the proposed research program.
Team members investigate such topics as alternative crop-
ping and livestock patterns, management practices, the
number and types of experiments, and the level to set
nonexperimental variables.
The FSR&D team works collaboratively with
farmers in their fields. In this way, a solid base is
developed for the team to understand farmers' conditions
and to design and implement appropriate experiments.
From this base, FSR&D activities can also be more effec-
tively coordinated with the work of other research and
development organizations.
Moreover, the researchers design record keeping
systems, special studies, climatic monitoring, and
surveys. The FSR&D team needs these back-up activities

1) better understand the research area
2) implement the on-farm experiments
3) measure progress
4) evaluate the results.

Before finalizing the research plan, the team makes a
preliminary analysis of possible impacts of the proposed
technological changes on the farmers and the environ-
ment. This analysis provides advanced information on
likely biological performance, resource requirements,
economic and financial feasibility, and sociocultural ac-
ceptability. Included in economic and financial feasibility
are requirements placed on support systems and the need
for any changes in national and regional policies.


During this set of activities, the FSR&D team con-
ducts on-farm experiments, initiates studies, gathers data,
helps to coordinate supportive research, and analyzes the
The FSR&D team commonly conducts three types of
biological experiments: researcher-managed trials, farmer-
managed tests, and superimposed trials. Researchers use

'Secondary information is published or unpublished data collected for purposes other than the current activity.

the first of these experiments to develop new technologies
under farmers' conditions. Such experiments are used
where control is important. The second of these ap-
proaches seeks to learn how farmers respond to suggested
improvements by allowing the farmers to introduce the
new technologies in their fields. The third approach in-
volves relatively simple researcher-managed experiments
applied across a range of farmer-managed conditions.
In conducting these experiments, the FSR&D team
cooperates with extension in establishing contacts with
farmers and local groups, identifies the locations for on-
farm experiments, secures resources, and coordinates ac-
tivities with the experiment station. Agreements need to
be reached among farmers, researchers, and cooperating
groups concerning who supplies which inputs, how the
outputs will be distributed, and who bears the risk of loss.
In addition, the team initiates special studies and surveys,
keeps records of farmers' practices and experimental
results, monitors local conditions, samples yields, and
analyzes experimental results.
The team's analysis of experimental results involves
evaluation of biological performance, actual resource re-
quirements, economic and financial feasibility, and
socioeconomic acceptability. These analyses are similar to
those made during the planning of on-farm research. The
difference is that the analyses of on-farm experiments are
based on actual biological results, whereas the analyses
made during planning (Sec. 3.1.3.) are based on estimates
of biological performance. For farmer-managed tests, the
economic, financial, and sociocultural analyses incor-
porate farmer reactions directly.
The researchers also estimate the acceptability of
new technologies by noting farmers' reactions to the pro-
posed changes during and following the experiments.
Results from these analyses assist the team in evaluating

1) the readiness of the technologies for diffusion
2) the need for improved support services and govern-
ment policies.

The FSR&D team can promote technologies accept-
able to farmers and society at large2 in several ways. For
example, the technologies can be tested in multi-
locations, incorporated in pilot production programs,
turned directly over to the extension service, or promoted
through other organizations.
Through multi-locational testing, the FSR&D team
extrapolates the improved technologies to other locations
within the target area. Extrapolation is usually limited to
environments similar to those encountered in the research
area. The FSR&D team can facilitate this process if, at the
outset, it adequately describes the research area and

FSR&D team members continue to be actively in-
volved in multi-locational testing by assisting extension
staff in identifying the conditions under which the
technologies apply. In some cases, researchers may have to
modify the recommendations slightly to accommodate
moderate differences in farmers or environmental condi-
tions. In assisting with these tests, extension agents not
previously associated with the FSR&D team will learn the
details of the technologies and how to apply them.
Up to this point, the team's appraisal of the ability of
the support systems to accommodate the introduced
technologies is based on analytical studies. Before turning
the technologies over for widespread diffusion, the team
may want to initiate a pilot production program within the
target area. The purpose of the program is to test the im-
proved technologies under conditions similar to those
likely to be encountered when the technology is broadly
diffused. Based on results, the team makes whatever
changes in the technology that are necessary, or alter-
natively suggests changes in the support services or
government policies. These changes should be small,
however, since the FSR&D approach is intended to pro-
duce technologies that fit farmers' and environmental con-
Sometimes extension or production programs can
diffuse improved technologies without additional testing.

2Technologies acceptable to society at large are those that take into account not only the long-run interests of the farmers receiving
the technologies, but also the interests of present and future generations indirectly affected. Generally, decision makers with a regional
and national perspective are called upon to make this judgment.


30 Conceptual Framework

4, '~~~~~


This might occur when the new technologies do not call
for major changes or when the government prefers not to
spend the time and funds for such refinements. However,
we caution FSR&D teams that bypassing multi-locational
tests and pilot production programs increases the risk of
To summarize, researchers, extension staff, and
those from production programs have mutually reinforc-
ing roles during the diffusion process. Researchers can help
advise others about ways to identify conditions suitable
for the new technologies; they can also suggest alternate
recommendations in light of the slight variations that oc-
cur throughout the research and extrapolation areas. Ex-
tension and production staff who become familiar with the
improved technologies can then take over when
widespread diffusion begins. The close links established
throughout the FSR&D process aid in this transfer of
responsibility and implementation.

Effective collaboration with other organizations
closely associated with the needs of small farmers is
critical to successful FSR&D activities. The two most rele-
vant groups will usually be experiment station staff and
the extension service. We show the linkages between the
five FSR&D activities and these two organizations in Fig.

Experiment station staff and facilities may be in-
volved in some or all of the following ways. The target and
research areas may be selected because of available
knowledge of the agricultural potential generated by the
presence of a research station. Experiment station person-
nel usually aid the field teams in identifying farmers' prob-
lems. The FSR&D team frequently brings such personnel
directly into the FSR&D process because of their
knowledge and experience in the research process. Thus,
experiment station personnel will be able to help carry out
research on farmers' fields, to conduct other research in
support of FSR&D's needs, and to help analyze and inter-
pret the results. Sometimes, the stations provide the hous-
ing, office, storage, training, and related needs of the
FSR&D field teams.
Similarly, the extension staff can help the FSR&D
team understand the characteristics of the research areas
and particularly the target groups of farmers. Furthermore,
this staff can aid in problem identification by introducing
the FSR&D team to individual farmers and local leaders.
As members of the FSR&D team, extension specialists can
advise the researchers on farmers' conditions when the
team designs experiments. Then, when the team under-
takes research on farmers' fields, the extension specialists
can help select farmers and conduct the experiments.
Finally, extension specialists trained in FSR&D can
transfer the improved technologies to the extension ser-
vice and other promotional organizations.

Conceptual Framework 31


FSR&D does not require that all aspects of the farm-
ing system be studied or that a plan of research be com-
pleted before on-farm research begins. As long as the team
has prepared the general framework of analysis, on-farm
research can start and the results used to refine the ap-
proach. The framework of analysis needs to include a con-
ceptualization of the farming system and a preliminary
identification of relevant problems for research.
Moreover, for practical reasons, the approach to
FSR&D seeks to uncover improvements that are substan-
tially better than existing practices, but not necessarily
the best practices. Initially, FSR&D tends to focus on
readily identifiable and researchable problems for study
and experimentation. This pragmatic approach starts
quickly and uses feedback from experiments and other
sources to produce better results than from more complex
and lengthy approaches. As the FSR&D program matures,
more fundamental constraints to improved farmer produc-
tion may come to light. Then, the FSR&D team may work
on more complex problems.
As Fig. 3-1 illustrates, the feedback goes from ac-
tivities 4 and 5 to activities 1, 2, and 3. As the team
analyzes studies and experiments and evaluates and ex-
tends improved technologies, it gains new insights into
the target and research areas and about farmers' problems
and opportunities. This information, in turn, may be used

for selecting new target areas
for redefining subareas or the research area
for improving research designs and planning
for altering the approaches to on-farm research and

Such information eventually becomes part of the general
body of knowledge of the area.
By setting experiment station and extension off to
the side in Fig. 3-1, we stress the central role of on-farm
research. Such emphasis focuses attention on farmers'
problems and opportunities. At the same time, we do not
underestimate the importance of effective collaboration
with experiment station and extension staff. Effective in-
tegration of the efforts of experiment station, extension,
and field team staff is essential to FSR&D.
Finally, analysis is listed as part of activity 4 because
of its importance at this point in the FSR&D process (Fig.
3-11. The available information on biological performance,
resource requirements, economic and financial viability,
and farmer response gives the FSR&D team the first good
indication of how well farmers are accepting the new
technologies. Actually, the FSR&D team analyzes each ac-
tivity to evaluate results and plan the next activity.

The time required to generate and transfer improved
technologies varies because of several factors: types of
farm enterprises, research team's knowledge of the area,
the backlog of suitable agricultural technologies, strength

of research and extension programs, need for training staff,
perceived urgency for improvements, governmental sup-
port, success in identifying better technologies, and so on.
For example:

Some researchers working on irrigated lands will
want to obtain an understanding of the water
balance-i.e., accounting for water inflow, outflow,
uses, and losses-before initiating irrigation ex-
Research results can be obtained more quickly in
areas suitable for growing the same crop more than
once per year than for single season crops that are
part of a three-year rotation.
Livestock experiments involving pastoralists are
generally longer and more drawn out than cropping
experiments, and experiments with tree crops
generally take even longer.

Table 3-1 presents a representative schedule for
FSR&D activities for cropping systems in new areas. This
schedule allows on-farm experiments to begin 21 weeks
after initiating target area selection. Time could be re-
duced if the target area has been selected or if conditions are
sufficiently uniform that subareas do not have to be iden-
tified. Alternatively, the time could be increased if the
area's characteristics, potential, or problems are complex.
The six weeks for gathering and reviewing secondary data,
conducting reconnaissance surveys, and ranking of the
possibilities for improvement represent a fairly rapid, and
yet, a reasonable schedule according to FSR&D practi-
tioners. Note, however, that this schedule presumes that
researchers already know the FSR&D process.
The team uses the time for planning on-farm
research for a detailed analysis of the alternatives sug-
gested during the previous set of activities; deciding on the
types, locations, and numbers of experiments; planning
data gathering; establishing contacts with cooperating
farmers; acquiring materials; assigning tasks for team
members; and other activities.
The schedule in Table 3-1 assumes

good sequencing of activities
no serious problems
an available cadre of capable researchers
synchronization with the cropping seasons.

Delays in recruiting and training staff, securing equip-
ment, having to build facilities, obtaining permission to
proceed, securing operating funds, and the like would ex-
tend the schedule.
The following example illustrates the time the IRRI
staff needed to help the Government of the Philippines
initiate a cropping systems program in a new area:

IRRI helped launch a new program in the Philippines in 4-1/2
months between the time the research area was selected and

32 Conceptual Framework

Table 3-1. Representative schedule for Initially completing the
first three activities in the FSR&D process for cropping
systems in a new area.

1. Target and Research Area Selection
Selection of general target area
Division into subareas
Selection of research area
2. Problem Identification and Development of
a Research Base
Gathering and reviewing secondary data
Conducting reconnaissance surveys
Analyzing results and setting priorities
3. Planning On-Farm Research
Reviewing identified problems and
gathering additional information*
Designing experiments
Planning for research, including
identification of collaborating

)tal 6.0

tal 6.0

tal 9.0

Special studies and surveys, which are identified dur-
ing this review, may be designed and initiated after the nine
weeks indicated for this step.

experiments commenced in farmers' fields. The sequence
follows: (1) the decision to go to the field was made on the 15th
of November; (2) the research team looked at the farm setting
and farmers' practices in the area, including such factors as
soil classification and fertility, climate, major crops grown,
varieties, yields, types of farming enterprises, farm and family
size, markets, and credit; (3) the team began to identify soil
types and to plan the season's activities in terms of cropping
patterns, treatments, factors to hold constant, and so on by
January 8th; and (4) the team started the on-farm experiments
by April 1st. In the meantime, the staff was acquired, offices
were set up, equipment secured, and other tasks necessary for
field operations were completed. The approach was timed so
that experiments began with the first available cropping
season (Zandstra, personal communication).


Once the FSR&D program has been initiated and on-
farm experiments begun, the team usually repeats a series
of trials and tests until improved technologies are iden-
tified and passed on to the farmers. We have not included
the time for on-farm research and analysis and diffusion of
results in Table 3-1 because conditions vary widely. For
example, on-farm research and analysis might take 18
weeks for experiments with short season crops. If two or

three crops are sequenced, a year might be required. Ex-
periments with sugar cane could take two years and rota-
tions of single-season crops in temperate climates would
take longer.
The overall program may take several years to
develop technologies with broad applications. The
Agricultural Science and Technology Institute (ICTA) in
Guatemala often begins with researcher-managed trials,
follows with farmer-managed tests, and then evaluates
farmer acceptability.
As experiments are completed each season, ICTA
teams quickly analyze the results so that they can plan
next season's experiments. And, ICTA teams use the
time between seasons to conduct reconnaissance surveys
before moving into new areas. According to Hildebrand
(personal communication), the following activities
associated with problem identification and planning on-
farm research can be completed in 5 to 10 weeks at the rate
of 1 to 2 weeks per activity:

1) gathering and analyzing background information
2) conducting reconnaissance surveys, identifying
problems and opportunities, and setting priorities
3) locating collaborating farmers
4) designing experiments and identifying farmers'
5) obtaining experimental materials.

We consider such a rapid schedule as attainable once an
FSR&D team has gained experience in the region and has
mastered FSR&D methodologies.
On rice-based systems, IRRI follows another ap-
proach that takes about three years to adequately identify
new cropping patterns. Using both researcher-managed
trials and farmer-managed tests, IRRI's researchers begin
the first year with several cropping patterns with few
replications of experiments and by the third year they con-
centrate on the most promising patterns using more
replications. Improved technologies may then pass
through multi-locational testing and pilot production pro-
grams, before being handed over to promotional organiza-
tions (Zandstra et al., 1981).

The staff of the International Livestock Centre for
Africa (ILCA) is considering a substantially different ap-
proach to FSR&D because livestock systems are distinct
from cropping systems (deHaan, personal commu-
Researchers studying the productivity of herds and
flocks may want to classify substantial numbers of
animals (perhaps 2,000) by age and sex and to consider
such factors as livestock sales and cow-oxen ratios. Cow
histories can be obtained from interviews with farmers to
trace such factors as animal growth from birth through
calving and progeny history. These histories provide infor-

Conceptual Framework 33

mation on fertility and mortality rates, but little on the
causes for these rates. To learn about the causes for these
rates, record keeping and analysis should continue for at
least 18 months, or more when variable annual rainfall in-
fluences herd and flock conditions. Other ways to learn
about livestock systems are through modeling and
monitoring changes accompanying livestock development
programs. Modeling helps to identify important features of
the system for research. Monitoring seeks to learn, among
other things, about the sociocultural responses to the in-
duced change.
These characteristics of livestock systems, which
often depend heavily on grazing for animal nutrition, sug-
gest a longer period for data gathering than for cropping
systems. However, not all livestock or mixed systems
take so long to research before improved technologies can
be introduced. For example, research can be conducted
faster on mixed-system relationships and factors such as
manure and stubble, supplementary feeding from surplus
crops, livestock as traction and transportation, competing
and complementary land use, and opportunities for labor
specialization (Delgado, 1978).

Experienced researchers have opposing views about
the time required to identify improved technologies. Some
researchers caution against rapid initiation of FSR&D ac-
tivities. They argue that

The team requires time to gather base-line data for
use in measuring research accomplishments.
The team needs time to understand the farmers and
for farmers to become accustomed to the team.
The team must be cautious when introducing new
technologies to farmers; otherwise, farmers might be
harmed by the change, thereby causing farmers to
lose confidence in the team.

While recognizing the validity of these arguments,
we interpret the emphasis of FSR&D practitioners as
pressing for early initiation of on-farm experiments.
Rather than start off by collecting base-line data, farm
records can be used for measuring FSR&D impacts. Farm
records are accounts usually kept by one or more members
of a farm household of the inputs and outputs for a single
crop or animal type. Researcher-managed trials can ac-
quaint the researchers with farmers' conditions before sub-
jecting the farmers to the risks of farmer-managed tests.
Furthermore, FSR&D practitioners generally feel that ex-
perimentation provides a good way to learn about the far-
ming system. Some practitioners argue that early improve-
ment in farmers' conditions helps to interest the farmers.
In fact, frequent researcher contact with farmers without
producing tangible benefits usually dampens the farmers'
interests in further cooperation.
To conclude, FSR&D's approach needs sufficient
flexibility to adapt to local conditions and to take advan-
tage of new opportunities. Having said this, we recom-
mend initiating experiments early and relying on FSR&D's

checks and balances to guard against actions that might
harm the farmer.

This section provides information on those who
carry out the FSR&D process. The principal groups are
teams at the field, regional, and national levels. Whether
all of these teams will be actively involved depends on
how FSR&D activities are organized. The team make-up
will vary according to the types of FSR&D programs, the
availability of staff, and which FSR&D activity is being
undertaken. In this section we assume a national program
with teams at each of the three levels and sufficient per-
sonnel to implement the approach.

The field teams work with farmers in their fields.
Such teams often consist of agronomists (called ingenieros
agr6nomos in Latin America), economists, and technical
assistants. Where livestock is important, an animal science
specialist should be part of the team; where irrigation is
practiced, an irrigation engineer can be a key member of
the team; and where women are responsible for growing
important crops or performing critical operations, field
teams should include women.
These teams are assisted from time to time by
specialists in disciplines such as extension, sociology, en-
tomology, and pest management. Often, such field teams
range from two to five professionals supported by
technical assistants. As a minimum, FSR&D practitioners
generally recommend that the team has a representative
from the physical or biological sciences and another from
the social sciences. Based on his experience in two Central
American countries, Waugh (personal communication)
recommends staffing the field team with agronomists who
also specialize in another discipline useful to the field
team. For example, agronomists might take short courses
or classes at the master's level in economics, plant
physiology, diseases of one of the commodities, soil fertil-
ity, statistical procedures, and so on. Furthermore, for this
example, one of the members should specialize in general
agricultural production and serve as liaison between the
FSR&D team and the extension service.
These teams report to the regional headquarters and
may live there or in the area where they work. The
technical assistants often live in villages within or near
the research area. Experience has shown that one member
of the field team can often manage 15 to 20 researcher-
managed cropping trials and more farmer-managed tests.
In round figures, a team of five might manage about 100
experiments during a season.

The regional staff may include a regional director, the
field teams, commodity and disciplinary specialists
assigned to the area, and support staff. Commodity
specialists are generally assigned to those areas where

34 Conceptual Framework

... I-
1 Ijr
;~~' ~ " '~a~ I-f C~~ ~L I'
Lr ---- .I;*~r"
~rll Cli ~
3 CV "Y~,~lz~t~L~iaP;~u,. -U
r3 1C u
~5 .*
1- C~
,, -J
4 r, h~~~ Cw9 Ft

i7~L i.

2' "
~ J. - .,1

'.- .. s -

.. ,' -

crops and livestock of their specialty are most important
to the country. These specialists may work in other
regions as the need and opportunity arise. If a country does
not have enough specialists for each region, they might
receive short-term assignments in several regions. Nor-
mally, experiment station staff is part of the regional
FSR&D team.


The staff at national headquarters administers the
FSR&D program and may include senior management, a
technical director, commodity coordinators, heads of the
disciplines, the director of experiment stations if the ex-
periment stations are the responsibility of the FSR&D pro-
gram, commodity and disciplinary specialists, and other
technical staff. This team may be located at some central
headquarters or dispersed throughout the country.

3.3.4. OTHERS

Other groups and organizations become involved with
the FSR&D staff from time to time such as farmers'

*1..' -- -..-'

,/ ,

organizations, regional and local planning and ad-
ministrative organizations, production organizations,
educational and training organizations, key decision
makers at the national level, and international organiza-
tions in agricultural research.


The seven principal groups listed in Fig. 3-2 are in-
volved in the FSR&D process to varying degrees depending
on the activity and individual country situations. Others
are involved as well, but to a lesser degree.
Target area selection requires high-level decisions,
which explains why national level staff is involved.
Dividing the target area into subareas and selecting
research areas, a more technical matter, requires those
knowledgeable about the area. Because problem identifica-
tion brings the FSR&D team into direct contact with
farmers and the surrounding environment, the team seeks
support and suggestions from regional and local
authorities, local businessmen and leaders, and a cross
section of local farmers. Planning on-farm research calls

c~ i

Conceptual Framework 35

Figure 3-2. The sequence of principal groups involved in the FSR&D process.

FSR&D Activities

Groups and Timing

Target Area IV to VI
Selection31 1.5 Weeks
Delineation of
Subareas and II* to IV, VI
Selection ofI 4.5 Weeks
Research Areas
Identification and I to IV, VI
Development of a I6 Weeks
Research Base
Planning On-Farm II to IV
Research 9 Weeks
On-Farm Research I to III, VI
and Analysis Varies
Extension of I to III, VI, VII
Results Varies I
Evaluation of IV to VI
FSR&D Activities 4 Weeks
*Provided field teams have been selected

Key: I. Farmers and Farmers' Organizations; II Field Team; III Regional
Headquarters Team; IV. National Headquarters Team; V. Key National
Decision Makers; VI. Extension Service; VII. Production Oriented

for sound technical input from the field team, and com-
modity and disciplinary specialists.
On-farm research concentrates on activities on
farmers' fields and can profitably use help from the exten-
sion service when supervising farmer-managed tests. Ex-
tending the results moves the center of activities to the ex-
tension service and other promotional organizations, but
requires inputs from research to assist in the transfer pro-
cess. Finally, evaluation of FSR&D activities generally re-
quires involvement of national and regional level staff
from the major organizations concerned with FSR&D.

Earlier sections of this chapter on FSR&D activities
did not dwell on how much change to attempt, how soon
to attempt change, where the ideas for change originate,
nor what type of research to consider. The possibilities are
numerous and FSR&D practitioners have not settled on a
common approach. Consequently, we advise those begin-
ning FSR&D activities to be flexible so they can make
changes as they gain experience and new information
comes to light. In this section we discuss some of the
issues to consider in deciding on an FSR&D strategy.


Agricultural researchers frequently debate the issue of
how much change to introduce into farmers' systems such
as changes to the farmers' cropping and livestock patterns,

management practices, and the farming environment. At
one extreme, nearly everything is, initially, considered
subject to change. Changes could range from improved
seed and pest management to better farm-to-market roads
and price supports for grains. The resulting farming
systems could be quite different from those prevailing
before introducing such changes. Such an approach might
be followed if the FSR&D teams answer "yes" to these

Are farmers willing to consider major changes to
their systems?
Is the research team capable of dealing with the com-
plexities of the whole farm?
Is the government prepared to respond to suggested
changes across a broad range of possibilities?
Are few opportunities apparent for materially im-
proving the farmers' conditions without having to
initiate major change?

The other extreme takes the farmers' cropping or
livestock patterns about as they are and works on improv-
ing the efficiency of farmers' practices. The reasons for
working with existing patterns are

The above four conditions are not present.
Some form of quick results are needed to capture the
interests of the farmers, the researchers, and funding
Farmers with limited resources are evaluated as be-

36 Conceptual Framework


~ -

ing particularly cautious about accepting new crops
or livestock on a significant scale, especially when
the change reduces production of one of their tradi-
tional food sources.

Deciding on which approach depends on accurately
appraising the farmers' responsiveness to change, the
capabilities of the FSR&D team, the flexibilities within
the farming system and the environment, and the urgency
for change.
This discussion can be summarized by what has been
described as a comparison of "farming systems in the
large" and "farming systems in the small" (Harrington,
1980). The ratio of variables to parameters is high for the
former and low for the latter. CIMMYT emphasizes "farm-
ing systems in the small" because this approach focuses on
those maize or wheat systems where the chances for im-
provement are high. CIMMYT's researchers consider both
the requirements of maize or wheat and how these crops
fit into the farmers' system. In contrast, those following a
"farming systems in the large" approach would not settle
on any particular part of the system without first
establishing that the parts selected offer the best chances
for improvement.


How soon to attempt change frequently comes up
when discussing FSR&D. Because of the diversity of farm-
ing systems and the risks of doing something that will
cause irreversible damage, some researchers delay research
and implementation of results until a careful study of
farmers' conditions is completed. These researchers feel
that this approach allows them time to gain a good
understanding of farmers' conditions and the research
potential, to gather base-line data for evaluating the effec-
tiveness of their programs, and to design an integrated
research and development program. Pursuing this ap-
proach might take one or more years before on-farm
research begins.
Other researchers associated primarily with cropping
systems research propose to introduce change quickly us-
ing on-farm experiments to gain an understanding of how
the farmers and the environment respond to change. In the
process, the FSR&D team will uncover opportunities for
further research and improvement. Opportunities for
change by this process may be less than optimum, but
such opportunities often interest farmers. Moreover, in-
troducing change in this way sometimes yields greater

Conceptual Framework 37

returns from scarce research funds than more precise and
drawn out procedures. In addressing a meeting of
agricultural economists, Collinson (1979) argued that
with resources remaining limited, "low cost/rapid
coverage approaches seem to be an essential starting point
for a bread and butter contribution from the profession."
Most FSR&D practitioners in cropping systems
whom we have met prefer the quicker approach. They
argue that waiting reduces benefits to present-day farmers
and that such costs are greater than the costs of
misdirected action. Moreover, the FSR&D approach con-
tains adequate controls for keeping the research on track
through frequent experiments, close interaction with
farmers, and feedback from research and development
results. Finally, record keeping of farmers' activities in the
target area can serve as one of the bases for evaluating the
effectiveness of the FSR&D approach.

Much of the research in developing countries has been
conducted on experiment stations, which generated
technologies that were subsequently brought to farmers
for acceptance or rejection. In contrast, FSR&D em-
phasizes early contact with farm households to learn about
their conditions so that improvements can be developed in
response to their needs. A critical element of this approach
is understanding farm households through on-farm
research in which farmers collaborate. Experiment station
and extension staff are integrated into the process. The
proper mix of on-farm and experiment station research
depends on each situation, including the ability of the sta-
tion to respond to research needs identified through the
on-farm research.

Another aspect of the foregoing issue is the debate
over the desired mix of site specific and generally applica-
ble research. Several writers have described this issue in
terms of the integration of "upstream" and "downstream"
research. "Upstream" research is characterized as being
partly basic, broadly general, and supportive; whereas,
"downstream" research is characterized as being site
specific, primarily adaptive, and useful without long delay
for target groups of farmers (TAC, 1978; Gilbert et al.,
1980; Harrington, 1980). Gilbert et al. (1980) described
"upstream" research as finding

". out how to overcome major constraints com-
mon to a range of farming systems extended across
one or more geographic zones. The partial or total
removal of a constraint such as water availability in
arid areas and soil fertility in the humid tropics can
significantly expand the range of enterprises and
techniques which can be potentially utilized by
farmers. Such programs mainly contribute to the
'body of knowledge,' rather than develop practices
specifically tailored to a local situation. Prototype
solutions produced by 'upstream' FSR programs

must be further adapted by 'downstream' FSR pro-
grams to specific local conditions. Further,
'upstream' programs may provide inputs into the
establishment of research priorities for commodity
improvement programs, since the 'upstream'
perspective is broader in terms of commodities and
disciplines than commodity improvement pro-
grams. And their geographic perspective tends to be
broader than that of 'downstream' programs.
Ultimately 'upstream' programs should rely on feed-
back from 'downstream' programs to sharpen their
own research priorities or objectives. Extensive use
of experiment station trials often characterizes
'upstream' programs" (Gilbert et al., 1980).

On the other hand:

". 'downstream' FSR programs begin with an
understanding of existing farming systems and the
identification of key constraints. However, in con-
trast to 'upstream' programs, 'downstream' FSR does
not always seek to significantly alleviate key con-
straints. ., but instead identifies areas of flexibility
in the specific system through accommodating in-
novations to the reality of existing constraints. In so
doing 'downstream' FSR, as emphasized earlier,
depends primarily on existing research results for
testing and incorporation directly -or with relative-
ly minor modifications-into farming systems. On-
farm trials and direct or firsthand interaction with
farmers predominate while experiment station
research tends to be minimal and restricted to adap-
tive rather than basic research" (Gilbert et al., 1980).

Gilbert et al. (1980) summarized their discussion
about "upstream" and "downstream" research by saying
that an FSR&D program should strive for some mixture of
the two as determined by the availability of innovations
that can be easily and rapidly integrated into existing farm-
ing systems. Where the pool of technologies is large,
"downstream" programs can be effective in identifying and
adapting the most promising approaches. Conversely,
where basic or more general research is needed, an
"upstream" approach may provide an appropriate mode for
organizing research to cut across traditional disciplinary
and commodity lines. At the minimum, a two-way flow of
information is needed from farm level to research institu-
tion and back again in the form of appropriate
technologies. To date, the IARCs have had the relative ad-
vantage in conducting "upstream" research, while the na-
tional programs have the advantage in conducting
"downstream" research; however, a national program may
engage in both types of research.
Another aspect of the type of research to favor con-
cerns whether or not researchers should stress small
farmers' welfare or their production. Sometimes increases
in farmers' production are large enough to satisfy national
objectives for larger output. When they are not, the
FSR&D team should seek guidance from national policy

38 Conceptual Framework


While the above discussion might suggest an ex-
cessive number of factors to consider in planning an ap-
propriate strategy for a country's FSR&D program, some
resolution of the issues is possible. Consider the following
possibility. An FSR&D program might be initiated that
emphasizes quick and modest changes in farmers'
systems. Most small farmers will probably not want to
risk much at first. In time and with adequate success,
farmers may become interested in more substantive
changes. Likewise, some early success will be needed to
show FSR&D's effectiveness to those who fund the effort.
Then, as researchers and farmers become better
acquainted, more imaginative improvements can be
Thus, the researchers could begin working to improve
the yields of the farmers' subsistence crops. After these
have been improved, part of the farmers' resources can be
directed to higher valued crops or livestock with which the
farmer is familiar. Finally, new crops and patterns can be
introduced as farmers become more successful. This con-
cept reflects the dynamics of a staged FSR&D approach.
For example, such an approach is occurring in Ethiopia in
which ILCA (1980) is experimenting with improved sub-
sistence crops, with the idea of freeing land for forage crops
and livestock improvement. Better livestock helps to im-
prove traction and, in turn, crops.


We began this chapter by describing the five major ac-
tivities common to many FSR&D programs and the col-
laboration with experiment station and extension staff.
Research and development results are fed back to earlier
activities to improve research designs and add to the body
of knowledge. Timing of FSR&D activities depends on the
situation: for cropping systems in the tropics, rapid im-
plementation of on-farm experiments is often possible; for
livestock systems and other situations, more time is
needed. FSR&D stresses on-farm research conducted by
interdisciplinary field teams supported by disciplin-
ary and commodity specialists. We closed the chapter by
presenting alternative strategies concerning how much
change, how soon to attempt change, where the ideas for
change originate, and what type of change to consider.
In the course of gathering information for this book,
we visited several national and international centers con-
cerned with FSR&D. In Appendix 3-A, we provide a sum-
mary of some of these programs that reveals alternative
strategies for FSR&D.


Byerlee, D., M.P. Collinson, R.K. Perrin, D.L. Winkelmann, S.
Biggs, E.R. Moscardi, J.C. Martinez, L. Harrington, and A.
Benjamin. 1980. Planning technologies appropriate to
farmers: concepts and procedures. CIMMYT, El Batan,

Collinson, M.P. 1979. Theme: Agrarian change, the challenge for
agricultural economists. Micro-level accomplishment and
challenges for the less developed world. A paper presented
at the Int. Assoc. of Agri. Econ. 17th Conf. 3-12 Sept. 1979.
Banff, Canada. CIMMYT, Nairobi, Kenya.
Delgado, C.L. 1978. The southern Fulani farming system in Up-
per Volta: a new old model for the integration of crop and
livestock production in the West African savannah. Center
for Res. on Econ. Dev. Univ. of Michigan, Ann Arbor,
Gilbert, E.H., D.W. Norman, and F.E. Winch. 1980. Farming
systems research: a critical appraisal. MSU Rural Dev.
Paper No. 6. Dep. of Agric. Econ., Michigan State Univ.,
East Lansing, Mich.
Harrington, L. 1980. Methodology issues facing social scientists
in on-farm/farming systems research. A paper presented at
a CIMMYT workshop on Method. Iss. Facing Soc. Sci. in
On-Farm/Farming Sys. Res. 1-3 April 1980. El Batan, Mex-
ILCA. 1980. ILCA the first years. ILCA, Addis Ababa, Ethiopia.
Technical Advisory Committee (TACJ Review Team of the Con-
sultative Group on International Agricultural Research.
1978. Farming systems research at the international
research centers. The World Bank, Washington, D.C.
Zandstra, H.G., E.C. Price, J.A. Litsinger, and R.A. Morris. 1981.
A methodology for on-farm cropping systems research.
IRRI, Los Banos, Philippines.


Clyma, W., M.K. Lowdermilk, and G.L. Corey. 1977. A research-
development process for improvement of on-farm water
management. Water Mgmt. Tech. Rep. 47. Eng. Res.
Center, Colorado State Univ., Fort Collins, Colo.
CRIA Cropping Systems Working Group. 1979. [Draft] Network
methodology and cropping systems research in Indonesia,
CRIA, Bogor, Indonesia.
Dalton, G.E. 1975. Study of agricultural systems. Applied
Science Publishers Ltd., London.
Ford Foundation. 1977. Agenda papers, vol. 1, Middle East and
Africa agricultural seminar. 1-3 Feb. 1977. Tunis, Tunisia.
The Ford Foundation, New York.
Fumagalli, A., and R.K. Waugh. 1977. Agricultural research in
Guatemala. Presented at the Bellagio Conf., Oct., 1977.
Bellagio, Italy.
Hernandez X., E. (ed.) 1977. Agroecosistemas de M6xico: contri-
buciones a la ensefianza, investigaci6n y divulgaci6n agri-
cola. Colegio de Postgraduados, Chapingo, and CIMMYT, El
Batan, Mexico.
Hildebrand, P.E., and S. Ruano A. 1978. Integrating multidis-
ciplinary technology generation for small, traditional
farmers of Guatemala. A paper presented at the Ann.
Meeting of the Soc. for App. Anthro. 2-9 April 1978.
Merida, Mexico. ICTA, Guatemala.
IITA. [Undated]. The IITA farming systems program. IITA,
Ibadan, Nigeria.
IRRI. 1976. International bibliography on cropping systems,
1973-1974. The Library and Documentation Center, IRRI,
Los Banos, Philippines.
ISRA/GERDAT. 1977. Recherche et d6veloppement agricole: les
units exp6rimentales du S6n6gal. CNRA Sem. 16-21 May
1977. Bambey, Senegal.
Laird, R. 1977. Agricultural research for development of tradi-
tional agriculture. Postgraduate College, National
Agricultural School, Chapingo, Mexico.

Conceptual Framework 39

Lawani, S.M., F.M. Alluri, and E.N. Adimorah. 1979. Farming
systems in Africa, a working bibliography, 1930-1978.
G.K. Hall & Co., Boston, Mass.
Mathema, S.M., and M.G. Van der Veen. [Undated]. Socio-eco-
nomic research on farming systems in Nepal: a key infor-
mant survey in five cropping systems research sites. HMG,
Min. of Food, Agric., and In. USAID/IADS, Nepal.
McIntosh, J.L., and S. Effendi. 1978. Cropping system research
activities in Indonesia. A paper presented to the Cropping
Sys. Working Group. 2-5 Oct. 1978. IRRI, Los Banos,
PNIA. 1979. Guia metodol6gica para conducci6n de ensayos de
finca. Secretaria de Recursos Naturales. PNIA, Coma-
yagua, Honduras.

Salter, L.A., Jr. 1967. Scientific method and social science, p.
39-77. In A critical review of research in land economics.
The University of Wisconsin Press, Madison, Wis.
Turrent, F.A. 1977. El agrosistema, un concept fitil dentro de la
discipline de productividad. p. 291-319. In Hernandez X., E.
(ed.) Agroecosistemas de M6xico: contribuciones a la
ensefianza, investigaci6n y divulgaci6n agricola. Colegio
de Postgraduados, Chapingo, and CIMMYT, El Batan,
World Bank. 1979. Senegal agricultural research project. The
World Bank, Washington, D.C.
Zandstra, H.G., K. Swanberg, C. Zulberti, and B. Nestel. 1979.
Caqueza: living rural development. IDRC, Ottawa,

Chapter 4

. 1w10

Effective selection of target and research areas is one
of the critical activities of the FSR&D process. Selection
generally begins with high-level decision makers in a
government deciding on one or more target areas as the
focus of attention, and selection of research areas con-
tinues until improved technologies are diffused
throughout the target area and possibly extrapolated into
other areas.
The process of selecting areas for FSR&D activities
is roughly as follows:

1) The decision makers select the target area or areas.
2) The FSR&D team divides the target area into
subareas according to characteristics most important
for the FSR&D effort.
3) The FSR&D team selects the research area within
the target area.
4) The FSR&D team selects farms and farmers within
the research area for conducting on-farm research.
5) When research results are promising, the FSR&D
team and the extension service select multiple loca-
tions within the target area for validation of new
technologies on a broader scale.
6) If the results of the multi-locational tests are
satisfactory, the FSR&D team and governmental
agencies select areas for pilot production programs to
evaluate the new technologies on a more intensive
7) After resolving any problems arising from the pilot
production programs, the extension service and
other relevant agencies implement the new
technologies according to suitable subareas within
the target area.
8) Finally, these governmental agencies may ex-
trapolate relevant technologies to similar areas out-
side the target area.

We illustrate this sequence of activities in Fig. 4-1.
In this chapter, we discuss (1) the amount of data to
collect for the various FSR&D activities, (2) the selection
of a target area and subareas, (3) the selection of the
research area, and (4) the early identification of oppor-
tunities for action. We close the chapter with a summary.

The FSR&D team collects data during the different
activities on similar topics, but at different levels of detail.

The team begins with a broad, cursory overview of infor-
mation relevant for target area and subarea selection. The
overview will include physical, biological, and
socioeconomic information. We represent this activity in
Fig. 4-2 by a thin, continuous horizontal bar. While some
of the material collected during target area and subarea
selection is adequate for the research area selection, the
team needs greater detail and other kinds of data.
The above requirements can be compared with still
more intensive data requirements for problem identifica-
tion and the research process. In Fig. 4-2 we show this
need for increasingly greater detail on more specific topics
by successively longer and narrower rectangles as we pro-
ceed from one activity to the next. Data gathering for
FSR&D is an integrated and comprehensive process. We
discuss the detailed information needed for FSR&D ac-
tivities later in this book.

We now turn to a discussion of the first major ac-
tivities in the FSR&D process-the selection of a target
area and its division into subareas. We define our terms
and then discuss the criteria and methods for selecting the
areas, information needed, and selection of the areas.

An FSR&D target area, may be selected for two basic

1) to meet the needs of the people living there
2) to take advantage of its agricultural potential.

In the second case, people may or may not already be liv-
ing in the area.
When conditions in a target area are substantially
different, it may be divided into smaller areas with similar
physical, biological, socioeconomic, and farming systems
characteristics. These smaller areas are the subareas for
which improved technologies can then be developed. For
example, Fig. 4-3 illustrates a hypothetical target area
with four subareas. Subarea A consists of valleys and low
plains with stony soil subject to flooding. Subarea B has
the same type of land as A, but is settled by a different
ethnic population with a different farming system.
Subarea C is fertile plain. Subareas A, B, and C have good
accesses to markets. The farmers in these three subareas

44 Target and Research Area Selection

100 0

000 0 a0 0 0



the FSR&D process.

I I I. -
:_ r~~ ,.,r. j,. ,[ E ,-"

Sin Target Area
' and Extrapolation
to Similar Areas

have mostly crop-based farming systems. Subarea D is
erodible hill country with poor roads. The inhabitants
have a livestock-based farming system and are nomadic
during the dry season.
Some FSR&D practitioners use other means for
dividing a target area. For CIMMYT's work, Byerlee et al.
(1980) call the focus of the FSR&D effort the recommenda-
tion domain and define this term as "a group of roughly
homogeneous farmers with similar circumstances for
whom we can make more or less the same recommenda-
tion. Recommendation domains may be defined in terms
of both natural factors-e.g., rainfall-and economic fac-
tors-e.g., farm size." In Appendix 4-A we present a
tabulation of farming practices that CIMMYT used to aid
in establishing recommendation domains.
ICTA in Guatemala searches for groups of farmers
who are using similar methods for the same cropping pat-
terns. If farmers grow the same crops in approximately the
same manner, the researchers assume that the environ-
ment and other conditions are similar. ICTA's researchers
assume an improved technology should apply to all
members of the group, because the crops, means of pro-
duction, and expected responses are similar.
In establishing the boundaries of a subarea, the
FSR&D team does not look for an area with complete
uniformity of all farmer circumstances. Such an area rare-
ly, if ever, exists. Instead, the team seeks an area where
large numbers of farmers are relatively homogeneous in
their characteristics. Enough farmers should be involved
to make the research effort worthwhile. By relatively
homogeneous, we mean most farmers will respond to the
new technologies in a similar way.
The first boundaries of a subarea may have to be ten-

Figure 4-1. The sequence of locations of FSR&D work during

Data for Target Area and
Subarea Selection
(Broad Overview)

Data for Research Area
Selection (Some Detail
on Some Subjects)

Data for Initial Problem
Identification (More
Detail on Some Subjects)

Data Needed During the
Research Process (Very
Detailed on Specific

Target Area

Target Area
Divided into

Research Area

Farms in
Research Area

Test Areas

Pilot Production
Program Areas

tative. Then, as the FSR&D process continues and the
team learns more about the subarea, the boundaries may
be revised.
Subareas are not necessarily contiguous. For exam-
ple, in Fig. 4-3, three of the four subareas are divided. Fur-
thermore, the team need not consider all subareas
simultaneously. It might rank subareas according to some
priority for deciding which ones to work in.
In Appendix 4-B we show the method Collinson
(1979a) used for grouping farmers "into relatively
homogeneous populations on the basis of their present
farming system" in the Central Province of Zambia. "The
key step in interpretation [of the collected information] is
deciding the sources of variation which are critical in dic-
tating resource allocation in farming systems of the area."


The first step in selecting a target area is for the key
personnel in the FSR&D program to identify potential
criteria for the selection. Then those holding top policy
and planning positions in the government discuss, select
and order these criteria, and determine the information
needed. FSR&D researchers collect and analyze the infor-
mation and submit alternatives to the decision makers for
their choice.
Compatibility with national and regional policies
ranks high on the list of criteria for selecting target areas.
For example, a national development policy in a country
may favor assistance to poor farm families, providing food
for urban populations, or increasing the country's foreign
exchange balance. Table 4-1 based on an unpublished
paper by Collinson (1979b) illustrates how selected
characteristics can be used in comparing two potential
target areas. If the government favors helping poor

Figure 4-2. The amount of data detail collected during
FSR&D activities. The width of the rectangle indicates
breadth of coverage; height of rectangle illustrates amount
of detail.


a Ir

mama am


Physical Biological Socioeconomic

I-1 -IZ]

00 000

Target and Research Area Selection 45

Figure 4-3. A hypothetical target area divided into four sub- Table 4-1. Comparison of five characteristics of two potential
areas showing a research area in three parts. target areas (Based on an unpublished paper by Colllnson,

Target Target
. Characteristics area area
"". A B

Number of poor farm families 50,000 15,000
::::Number of poor farm people 300,000 100,000
:: +: -Production of exports (US $) --- 1,000,000
Production of urban food
supplies (US $) 50,000 200,000
Average family's cash income
'^:/': -- as percent of the national
average 40% 150%

m Subarea A
SSubarea B
|:J Subarea C
| Subarea D
r- Research Area

families, the decision makers would select target area A.
Alternatively, if the government favors increasing urban
food supplies or expanding exports, the decision makers
would select target area B.
As another example, the national decision makers in
one African country selected a sparsely populated region
with improvement potential as a target area. They based
their choice on the national policy to alleviate land
pressure in its fertile, but overcrowded highlands.
Selection of a particular target area may satisfy more
than one development goal. As an example, one Latin
American country has the twin goals of improving the
economic level of poor rural areas and of increasing
political stability there. Farmers in one region of the coun-
try had traditionally supplemented their meager farm in-
comes with seasonal work in a neighboring country. When
the neighboring country closed its borders to these
migratory workers and threatened annexation of the
region, the farmers' national government set up an FSR&D
project to raise farm incomes and to solidify the farmers'
As another alternative, the decision makers may
select a target area on the basis of specific physical limita-
tions or problems such as erodible slopes, flooding, soil

salinity, inadequate pasture, or animal disease. In
Sumatra, Imperata grass overgrowing large areas presents a
major problem to Indonesia's land settlement program. For
this reason, the decision makers chose a target area that
had large sections of Imperata infested land for an FSR&D
The following questions may help the FSR&D staff
provide the information the decision makers need for
determining broad policy options:

Is the target area or subarea large and relatively
similar in those environmental characteristics that
have the most bearing on potential research results?
If the answer is yes, the FSR&D team can apply the
research results broadly and meaningfully within the
target area.
Is the target area similar to other areas? If so, some of
the technology developed in the target area can be
extrapolated to other areas.
Does the area have the potential for rapid payoff
from FSR&D? Factors that often determine the
potential are (1) physical and biological conditions,
(2) markets and infrastructure, (3) available
technology, and (4) farmers' willingness to accept in-
Do the area's environmental conditions facilitate ap-
plication of technologies developed elsewhere? If so,
the FSR&D process could be greatly shortened.
What are some of the cost factors? For example, is
the target area now served by existing governmental
programs or private institutions? If so, cooperation
with such programs or institutions could enable
FSR&D to accomplish more, reduce costs, or save
time. One example would be the help that an ex-
isting experiment station could offer. However,
FSR&D may be needed more in areas where such
links are not well established, and the costs are
higher. Thus, high-level decision makers may have
to make this choice.

In Appendix 4-C, we illustrate how environmental

46 Target and Research Area Selection

.-, .- ai* r..* '''S 3 "..
*-:--* ^ *" ^
-. .
.2 .Jt -

suitability and other factors influenced the choice of target
areas in Indonesia.
In Mali, several factors favored the Mali Sud region
as the target area for a farming systems project. Regional
characteristics (1) favored rapid development of the area
and (2) offered a wide range of farms at different stages of
modernization and a potential for integrating livestock
and cropping enterprises. Rapid payoff was expected
because of favorable agroclimatic conditions and good
roads. The program costs were kept down because the
area's supportive institutions were functioning well and
because a research station had initiated farming systems
work in the area.


As indicated in Fig. 4-2, the team uses a broad range

of data for target area and subarea selection that includes
information on national policy, farmers' characteristics,
and the environment. Below, we (1) relate national policy
with farmer characteristics, (2) discuss four aspects of the
physical environment and provide an example, and (3)
describe secondary and primary data sources.' In discuss-
ing this third item, we provide suggestions on gathering
socioeconomic data, as well as other types of data.

National Policy and Farmer Characteristics
The criteria the decision makers and FSR&D team
use in selecting the target areas will determine general
data requirements. Consider national policy criteria. Sup-
pose the primary objective of national policy is "to in-
crease agricultural income on small farms." Then, for each
potential target area the FSR&D team needs data such as
the number of small farmers, the size of their farms,

'Secondary information means published or unpublished data collected for purposes other than the FSR&D project; primary infor-
mation refers to data collected specifically for the FSR&D project.

... :I rcs

Target and Research Area Selection 47

agricultural income, cropping and livestock patterns, and
management practices. In addition, the team should in-
vestigate the practices used by successful farmers and why
others do not use these practices.
The exact wording of policy objectives is important.
Suppose the word "agricultural" is left out before "income"
in the above policy objective. Now the objective is "to in-
crease the income on small farms." Thus, increases in in-
come are not limited to agricultural income. Data collec-
tion must now include nonagricultural income on the
farm such as spinning, weaving, or pottery making.

Physical Environment
In its work in Southeast Asia, IRRI developed an ap-
proach to classification of the physical resources of an area
according to climate, topography, and soils. In the discus-
sion that follows, we report on (1) IRRI's work on these
three topics, (2) water, and (3) an example in Southeast
Asia that integrates these factors.
Climate. In parts of Southeast Asia, IRRI researchers

delineated four major climatic zones (with subdivisions)
for rice, based entirely on rainfall patterns. See Fig. 4-4
(IRRI, 1974). These researchers explained "in general a fair
amount of data on rainfall is available for most South-
east Asian countries. However, other macroclimatic
parameters such as evaporation, temperature, relative
humidity, wind speed, and particularly solar radiation are
seldom recorded. Therefore, these data could not be used
to evaluate evapotranspiration." In Appendix 4-D, we pre-
sent IRRI's criteria for selecting the major climatic zones,
the description of these zones, and their subdivisions.
For other regions of the world, for different crops, or
for livestock, FSR&D researchers may need to consider
other climatic factors.
Topography. According to IRRI (1974), "a convenient
method for a broad-scale topographic classification of the
rice lands of Southeast Asia is to divide the countries of the
region into the physiographic units with which different
forms of rice cultivation are closely associated." IRRI
researchers used four subdivisions as a first approxima-

Figure 4-4. The rainfall zones for parts of Southeast Asia (Adapted from IRRI, 1974). Wet month means at least
200 mm rain. Pronounced dry season means at least 2-3 months with less than 100 mm rain.


I More Than 9 Consecutive
Wet Months
= II 5-9 Consecutive Wet Months
and No Pronounced Dry
SII2 5-9 Consecutive Wet
Months and Secondary
Rainy Season
I 113 5-9 Consecutive Wet Months
and Pronounced Dry Season
114 5-9 Consecutive Wet Months and /
Pronounced End to Rainy Season
III, 2-5 Consecutive Wet Months and
No Pronounced Dry Season
S11I2 2-5 Consecutive Wet Months With
Pronounced Dry Season
IV Less Than 2 Consecutive Wet Months

; OC3

Target and Research Area Selection 49

coastal plains of marine, deltaic, and fluvial origin
inland terraces and plains of ancient origin, usually
well dissected
inland terraces, fans, and valleys of recent origin
sloping lands.

We show these four physiographic units in Fig. 4-5 and
briefly describe them in Appendix 4-E.
Soils. IRRI (1974) lists the major soils in each
physiographic unit that are important for rice growing
(Fig. 4-5). In classifying these soils, IRRI relied on the soil
taxonomy of the U.S. Department of Agriculture (USDA
Soil Survey Staff, 1975). This taxonomy lists six soil
orders-Inceptisols, Vertisols, Entisols, Ultisols, Alfisols,
and Oxisols. Here we present IRRI's description of Ultisols
as an example of soil information useful in target area

"Ultisols. These are soils with a sandier topsoil, a
more clayey subsoil, a low base saturation2 and
mainly 1:1 clays.3 They occur in the older, but not
the oldest surface in the tropics. In Southeast Asia,
they are largely on the rolling and hilly lands. Prob-
ably much of the upland rice grown in shifting
cultivation is on these soils, particularly in Sumatra,
Kalimantan, and Thailand. Apart from this use for
shifting agriculture, Ultisols form major areas for
tree crops like rubber and oil palms, but the area
under permanent annual cropping may be small. Ma-
jor problems in annual cropping are erosion due to
profile instability and high acidity with aluminum
and manganese toxicity. Lime is needed to grow the
more sensitive crops, e.g., corn and sorghum."

Water. With regard to water, especially for irriga-
tion, such factors as the amount and period of availability,
quality, and source are important. We discuss irrigation
water in more detail in Sec. 5.4.1. and Appendix 6-A.

An Example in Southeast Asia. Member countries of the
Asian Cropping Systems Network are striving to increase
small farm income by searching for ways to help farmers grow
a second rice crop. Target areas are selected using data on the
number of consecutive months of rainfall of more than 200
millimeters. Within this target area, data on topography, soil,
and irrigation water guide the selection of subareas. In addi-
tion, researchers gather data on other topics, such as popula-
tion density, agricultural production, prices, and cropping ex-

Secondary Information
Data on the broad physical, institutional, social, and
economic factors are generally available from secondary
sources. For example, census data provide information on
population density, size of land holdings, land use, and
crop and livestock production. Other sources include
weather records, cadastral surveys, commodity studies,
and reports of other programs.

Statistics and reports may be available at all
levels-national, provincial, district, subdistrict, and
village. Data sources include the offices of national plan-
ning, statistics, agriculture, community development, ir-
rigation, research stations, and other agriculturally related
offices, as well as those of private and public marketing
organizations. For example, in an Egyptian project,
statistics on agricultural production guided the selection
of three target areas representing typical cropping patterns
along the Nile River.
An FSR&D team might use satellite imagery for a
general classification of potential target areas and aerial
photography for more detailed analysis. The latter is
especially useful in gathering and confirming information
on land use. Such information helps in identifying soil and
land types for cropping and livestock experiments. If
available, current aerial photos can be compared with
earlier aerial photos-preferably 20 years earlier-to iden-
tify trends (de Haan, personal communication).
Also, the team can prepare maps for FSR&D from
sources such as satellite imagery, aerial photos, or govern-
ment maps. Topographic maps on a 1:50,000 scale are
often available for many areas of a country. Maps might
show such items as population distribution, ethnic com-
position, land use, soil capability, vegetation, land tenure,
ecozones, and percent of land cultivated. Fig. 4-6 is an ex-
ample of one of the many types of maps that can be
prepared as part of a reconnaissance soil survey (Thom,

Primary Information
Once relevant secondary data have been gathered
and analyzed, the FSR&D team may need additional infor-
mation to guide its selection of a target area. Primary
sources may then be used. Formal and informal leaders,
such as farmers, educators, local merchants, mayors, and
religious leaders may be good informational sources. Also,
technical specialists such as agricultural officers,
agronomists, climatologists, and social scientists are
useful. The team may obtain information from these
sources through either informal or formal interviews. In
Peru, a research team used informal interviews of mer-
chants and others to select a highland region with easy ac-
cess to the Lima market as a target area. In Appendix 4-B,
we discuss how a research team gathered primary informa-
tion for a province in Zambia, surveyed extension agents,
and tabulated the data.

The selection of specific target areas depends on the
key decision makers' values and on the country's decision-
making structure. These values reflect the weights deci-
sion makers place on national and regional policy, as well
as on socioeconomic and technical factors. Top-level deci-
sion makers such as the ministers of agriculture, planning,
and finance are often involved. At other times, those

2Low base saturation refers to a soil with a small percentage of exchangeable metallic cations (i.e., Ca +, Mg + +, K NH4+).
31:1 clays consist primarily of kaolinite minerals. Kaolinite is a non-swelling mineral of low cation exchange capacity.

Figure 4-5. The physiography of rice growing areas in parts of Southeast Asia (Adapted from IRRI, 1974).


l Coastal Plains of Marine,
Deltaic and Fluvial Origins
* Inland Terraces, Fans and
Valleys of Recent Origin
[ Inland Terraces and Terraces
of Ancient Origin < S
Ultisols 1
Sloping Lands
D Rice unimportant crop > "(

Target and Research Area Selection 51

Figure 4-6. The percentage of land cultivated in an area of Kenya (Thom, 1978).

Cleared and Cultivated
Cleared and Cultivated

Cleared and Cultivated
Cleared and Cultivated

37 15'







--J / ,\\\\\\\\ \-.*..u'-'-'.

S.*.". \\\\Ndooa\\\\ .. ....
- cha doslu Ndoa ** ,
"" \\\\\\\\\\ .*.
S \ \\\\\\\ \.'.'.'.'

.".'. M \ \\\ .
'.,uto \\\ \Ndoo\
''\\\\\\\\ "

".acao ...... .
S ...
\ \\\\ .:.:.:::

;. \y,,.: .... /:.:..:.

below the ministerial level make the decision. For exam-
ple, the regional agricultural directors in one Central
American country have considerable influence in such
decisions, provided their decisions fall within broad policy
guidelines. In another country, the director general of
agriculture selects the target area.
Those responsible for deciding on the target area nor-
mally rely on their advisers to assist them in clarifying the
alternatives and in providing background data. In this
regard, FSR&D staff should be particularly helpful.
In summary, selecting a target area and subareas re-
quires choosing criteria and then collecting and analyzing
data pertinent to the criteria. Large amounts of detailed
data should not be amassed. Generally, less than one
month is enough time to select the target area.


Once the target area and the subareas have been
determined, the FSR&D team is ready to select the
research area. In describing this process, we will cover (1)
the definition of a research area, (2) criteria and methods
for selecting the area, (3) staffing, (4) information needed,
(5) means for collecting information, (6) selecting the area
in stages when the target area and subareas are large, (7) ac-
tual selection, and (8) abandoning an area.


A research area, where the FSR&D team develops
improved technologies, may represent the whole target

area or only some of the subareas. When the team properly
selects a research area, the research results are transferable
throughout the target area according to the subareas for
which the technologies were designed.
The research area may, in turn, be a single area or
comprise scattered areas. A single research area, reflecting
the variations within the target area, usually provides ad-
vantages such as better logistics and team organization.
If the team cannot find a suitable single area, it
might select several areas that when combined constitute
the research area. Within this area, the team locates the
trials and tests in a way that broadly samples the condi-
tions within each subarea (Fig. 4-7). In Fig. 4-7, the
research area is divided into three parts to cover minor
variations within subareas.
A convenient location for the team's headquarters
would be within or near the research area. Where the
research area is fragmented and the parts separated by poor
roads and communications, team operations could be dif-
ficult and costly. Where such conditions exist, the FSR&D
team may need to establish two or more independent
research areas.


In this section, we provide illustrative suggestions
that have helped others select research areas. Selection
criteria relate to (1) the area's representativeness, (2) size,
(3) accessibility, (4) closeness to an experiment station,
and (5) cooperation with farmer contact agencies and
leader support.



Z> 50%
] 25-50%
< 10%
L- l- *


52 Target and Research Area Selection

Figure 4-7. Hypothetical research area consisting of three parts. This figure is an enlargement of the 8,000 ha area shown in Figure
4-3. Selection of the location for the trials and tests is discussed in Chapter 7.

SResearcher-managed Trials in Corn
O Farmer-managed Tests in Corn
Researcher-managed Trials in Wheat
O Farmer-managed Tests in Wheat
,&Researcher-managed Trials in Livestock
& Farmer-managed Tests in Livestock
SUrban Areas

Representativeness of an Area
The overriding criterion for research area selection is
its representativeness of the target area or subareas. This
representativeness generally hinges on both environmen-
tal conditions-i.e., physical, biological, and socio-
economic-and the farming systems.
We suggest the FSR&D team determine the
representativeness of a potential research area in this way:

1) Decide which of the target area's or subarea's
characteristics such as rainfall, land type, distance to
market, or ethnic composition of the farm popula-
tion are the most important.
2) Take a sample of farm communities with these
desired characteristics.
3) Collect data on these characteristics for each farm
community in the sample.
4) Calculate the average value for the sample for each
characteristic. These averages, taken together,
describe the most representative farm community of
the target area or subarea.
5) Label the farm community that deviates the least
from these average values as the most representative
of the area.

We illustrate how this method is being used in Indonesia
in Appendix 4-C.

The size of research areas varies. For example, the
initial size of one of the research areas in Guatemala was

L | Subarea A
[ Subarea B
: Subarea C
| Subarea D

8,000 hectares. The area could be smaller in a densely set-
tled region, such as in parts of India where considerable
variability in farms and families occurs. The area could be
much larger in a sparsely populated area when transporta-
tion is adequate. Most often, FSR&D teams use research
areas large enough to include some variability in farmers'
conditions. This variability allows the research team to
test the adaptability of new technologies under different
conditions and to make appropriate adjustments.

Easy access to all parts of the research area enhances
cooperation between the research team and farmers, per-
mits timely provision of inputs and marketing of outputs,
and reduces the team's operating costs. Accessibility of the
research area enables the support staff- scientists from the
experiment station, government officers, and FSR&D
researchers stationed in the nation's capital-to reach the
research area quickly. Living facilities for the regional
headquarters and field teams should be available nearby.
Easy access is especially important during the early stages
of FSR&D because meaningful results are often needed
quickly. Later, when FSR&D becomes established, re-
search in areas more difficult to reach may be undertaken.

Existence of a Nearby Agricultural Research Station
When research stations are close to the research
area, the FSR&D team is more likely to receive effective
cooperation from the experiment station staff. Experiment
station staff can help the team with on-farm experiments
and by conducting experiments on the experiment station.

Target and Research Area Selection 53


a S "





Cooperation of Farmer Contact Agencies
and Leader Support
Support from farmer-contact agencies such as exten-
sion, credit, and production groups is helpful to the
FSR&D effort. Their provision of services such as informa-
tion, credit, purchasable inputs, or marketing of outputs
can be essential for the widespread adoption of FSR&D's
Help from formal and informal leaders at the na-
tional and local levels can also be critical to FSR&D's suc-

In forming the FSR&D team to select the research
area, the FSR&D leaders should select the organization's
most experienced staff. These leaders may need help from
specialists in agronomy, livestock production, agricultural
economics, and farmer and social organizations.
Since livestock or crops are usually the focus of at-
tention, at least one specialist in these fields is essential.
An agricultural economist is needed to translate technical
possibilities into economic practicality. A sociologist or

L._, t-. ';.

.* "'.. a -~ .. .

... ; " '. ... *." .'." ... .

someone with similar training may be needed for
evaluating the sociocultural characteristics of farmers and
the organizations affecting farmers. Other experienced
staff may also be required, such as an agricultural engineer
where irrigation or farm equipment is important, com-
modity specialists according to the most relevant crops,
and specialists in pest management.
This team should integrate its efforts by operating in
an interdisciplinary manner. For continuity, research and
extension members, who will be working in the area,
should be involved. Finally, women may need to be added
if one or more would not otherwise be part of the group.
Such involvement of female personnel is particularly rele-
vant when women farmers manage many or all of the
farming activities.

Keeping the criteria from Sec. 4.3.2. in mind, the
FSR&D team now gathers information on alternative
research areas. In Table 4-2, we list data categories for
research area selection. In view of the limited budgets in
most developing countries, the team needs to be selective
in how much data to collect.

54 Target and Research Area Selection

Table 4-2. Data categories for research area selection.

Physical environment
Climate: e.g., rainfall, temperature, wind, sunny days
Soil: e.g., physical, chemical, hydrological conditions
Topography: e.g., slope, flood plain
Irrigation: e.g., water source and quality, means and
frequency of delivery, on-farm practices
Biological environment
e.g., weeds, insects, diseases, birds, rodents, crop yields
Socioeconomic environment
Resource availability: e.g., land, labor, cash, means of
Infrastructure: e.g., supply of farm inputs, markets for
farm output, transportation, electricity
Market data: e.g., prices of farm inputs and commodities,
Sociocultural characteristics: e.g., land tenure and
inheritance systems, sexual division of labor in
agriculture, religious beliefs concerning agriculture,
openness to change
Political and economic structures: e.g., national regulations,
community groups, caste or clan systems, patron-client
relationships, cooperatives
Production systems and land use
e.g., major crops and livestock, cropping patterns,
livestock characteristics, management practices

In Table 4-3, we show the type of data the Indone-
sian Cropping System Program uses for research area


Whenever possible, the FSR&D team uses secondary
data for research area selection. Team members can usu-
ally obtain such data quickly, inexpensively, and simply.
The team uses independent checks-other secondary
sources, discussions with those knowledgeable in the
field, spot surveys, and observations-to verify the
reliability and accuracy of these data. In Table 4-4, we list
the typical sources of data that researchers in Indonesia
collect. Other countries may have different sources of
secondary data such as tax offices or regional experiment
stations. Often the latter are especially useful for soil
surveys, rainfall records, and research results. Alterna-
tively, the team may have to depend on data from national
research centers and agricultural offices. Too frequently,
data collected locally are sent to the national offices
without copies being retained in the local offices; hence,
local data resources may be limited.
When the FSR&D team finds secondary data insuffi-
cient for selecting a research area, direct observations and
contacts with farmers may be necessary. In such cases, the
team conducts reconnaissance surveys (sometimes called
exploratory surveys).

Reconnaissance Surveys
Reconnaissance surveys of the research area include
both aerial and ground surveys. For aerial surveys, the

Table 4-3. Data required for systematic selection of rural com-
munities as research areas in Indonesia (Adapted from CRIA,

Data Purpose

Distance from main road (km)

Area in each land use class

Relative area in each slope
class (percent)
Relative area in each soil
texture (percent)
Area planted to each crop,
by month (percent)
Population, by economic
activity (number)
Rainfall by month for past
10 years (mm)

Participants (number) in the
rice production program of
the government

Months during which irriga-
tion water is available (per-
cent of area with less than
5, 6-7, 8-9, and 10-12
months of irrigation per
Draft animal population
Tractor population (number)

To guarantee that the rural
community is easily
To permit the selection of
rural communities with the
largest area in the desired
land use class
To avoid rural communities
with atypical topography
To avoid rural communities
with atypical soils
To identify current production
To determine importance of
agricultural employment
To determine number of
months 100 mm or more of
rain and probability of less
than 100 mm at beginning
and end of cropping
To determine the availability
of credit and level of
technology in the rural
To identify areas according
to irrigation regimes

To determine the availability
of draft power
To determine the availability
of mechanical power

FSR&D team may use existing aerial photographs to ob-
tain a quick and general impression of the obvious features
such as terrain type, land use, transportation networks,
and population centers. In addition, team members often
fly over the area to observe conditions. For ground
surveys, the team gathers information by observations and
interviews. Besides the items listed under aerial surveys,
the team observes such features as crops and livestock,
field activities, power sources, soil types, and irrigation
The team often interviews government administra-
tors, agricultural researchers, personnel of farmer-contact
agencies, businessmen, community leaders, and farmers
and their families. Besides discussing and verifying their
observations, the team members may ask economic and

Target and Research Area Selection 55

Table 4-4. Typical sources of data for research area selection in Indonesia (Adapted from CRIA, 1979).

Extension office-district
Land use by type
Variety trials
Fertilizer trials
Pest surveillance
Demonstration plots

Extension office- subdistrict
Hectares in each crop
Planting and harvesting dates
Yields by crop

Extension office- field
Land ownership
Support services
Yield constraints
Constraints to crop intensification
Input subsidies

Village office
Land ownership
Tenant-landlord relations
Landless labor
Support services
Irrigation system
Employment profile

Farmer-group interview
Current cropping pattern
Historical cropping pattern
Landless labor
Input availability
Input prices
Off-farm employment
Migration of agricultural labor
Input use
Yield constraints
Planting decisions
Input levels
Constraints to cropping intensification

Input dealers
Input sales
Input availability
Input prices

Support services

Bank- field office

Irrigation office
Irrigation system

Statistician for district
Output prices
Agricultural production

social questions about farming systems, markets, and
other unobservable factors. While farmer interviews may
range from informal contacts to formal surveys, FSR&D
practitioners generally favor quick, informal surveys for
research area selection.

At times, the reconnaissance survey and the
available secondary data may still leave some doubt in the
minds of the team members about the suitability of a pro-
posed research area. If so, the FSR&D team may conduct
exploratory experiments to learn more about the research


When the target area or subarea is large, the FSR&D
team may select a research area in stages. First, the target
area or subarea may be broken down into subdivisions.
Where feasible, the team may use districts, subdistricts or
other administrative units because secondary data are

usually available for such entities. Then the team iden-
tifies the subdivision best suited for the research area.
Within this subdivision, the team selects the research
The methods a team uses in research area selection
and the emphasis on types of data vary from project to proj-
ect and depend on FSR&D's mandate. For example, in
selecting a subdistrict in Indonesia, the primary considera-
tion was the number of hectares of a particular land-use
type. In addition to the example in Appendix 4-C, we
discuss a different method in Appendix 4-F-the one
ICRISAT (Jodha et al., 1977) used in India to select the
most representative subdistrict and research area.


Based on analyses of the foregoing data, the FSR&D
team recommends a research area. The final selection,
however, may involve governmental representatives at the
national, regional, and district levels, and local leaders.
Remember, target area selection is largely a policy deci-

56 Target and Research Area Selection

1~, ~-~F~s

sion that matches national objectives with practical
possibilities, while research area selection is largely a
technical decision that concerns how best to accomplish
FSR&D objectives.
Some FSR&D practitioners consider one or two
weeks sufficient to select a research area, if (1) the target
area is fairly homogeneous in its natural and so-
cioeconomic setting, (2) the selection is based primarily
on physical and biological information, and (3) good
secondary data are available. However, six weeks is more
common. Even longer time is required if (1) the target area
varies greatly such as in mountainous terrain with dif-
ferent climates and is in various stages of socioeconomic
development, or (2) the team needs much socioeconomic,
farming systems, or other primary information.

Occasionally, after a research area has been selected
and research started, national interests would be served

better if the research area were abandoned. This may hap-
pen because of overlooked and unfavorable conditions
such as farmers' unwillingness to cooperate, the ap-
pearance of a new obstacle such as a plant virus, or the
discovery of a serious soil condition. In such a situation,
the FSR&D leaders should have enough flexibility to shift
to a new research area. The reasons should be explained to
the affected government officials, local leaders, and
farmers. Such actions do not preclude returning to the area

In the process of collecting and analyzing data on the
area, the FSR&D team may uncover opportunities for
research or government action that can be started im-
mediately. Solutions may appear so obvious and the con-
sequences of failure so slight that waiting for subsequent
FSR&D activities cannot be justified. These opportunities

Target and Research Area Selection 57

could relate to on-farm experiments, improvements to the
infrastructure, or changes in governmental policy.
However, we caution the team to be careful with such
early actions, since more often than not the solutions are
not as simple as they appear.


A situation may dictate that the team initiate
research as soon as practicable. For example, data col-
lected during research area selection may indicate that
drought during maize flowering regularly reduces yields. If
early maturing maize varieties are locally available, they
might be tested during the first available season. Test
planting would not be costly and, if successful, could solve
the problem quickly.


Occasionally, by systematically reviewing the
characteristics of an area, the team will be able to identify
situations that call for quick government action. Such
might be the case if government policy is not appropriate
for an area, such as regulations for trucking or grain
storage. Or from its experiences elsewhere, the team
might know how to quickly improve a facility essential for
the area's farmers. For example, the area may lack an all-
weather access road, but may have all other production
and marketing facilities for growing fresh vegetables to
supply a nearby city during the rainy period. Starting a
public dialogue on the advantages and costs of building the
road may be in both the farmers' and the nation's interests.
While the team will not have fully clarified the oppor-
tunities and problems of building the road, existing infor-
mation might justify initiating the dialogue.


The selection of the target area, subareas, and
research area is normally the first activity confronting the
FSR&D team. Proper choices at this stage can be crucial,
since subsequent FSR&D activities and possibly successes
are influenced by the suitability of the areas. Target area
selection involves top-level decision makers in the govern-
ment, FSR&D staff, and others who can help in matching
national and regional goals with the potential of alter-
native areas. Identification of subareas and choosing the
research area are largely technical matters that call on the
expertise of the FSR&D team. Even so, advice and accep-
tance by national and local leaders are often needed before
choosing the research area.
In choosing these areas, the FSR&D team gathers
data on (1) national objectives and policies, (2) physical,
biological, and socioeconomic environments, (3) farming
systems, and (4) farmers' characteristics. Generally, the
team begins this work by reviewing secondary sources of
information such as published reports and records by local
organizations, national ministries and agencies, and
agricultural research stations. The team fills in missing

data by direct means, such as reconnaissance surveys of
the area. The total time for this activity may be as short as
one or two weeks, but more often takes about six
weeks- sometimes longer.
In addition, we covered other points in the chapter,
such as the need for interdisciplinary teamwork, selecting
research areas in stages when the target area or a subarea is
large, justification for abandoning an area, and early iden-
tification of opportunities for action.

Byerlee, D., M.P. Collinson, R.K. Perrin, D.L. Winkelmann, S.
Biggs, E.R. Moscardi, J.C. Martinez, L. Harrington, and A.
Benjamin. 1980. Planning technologies appropriate to
farmers: concepts and procedures. CIMMYT, El Batan,
Collinson, M.P. 1979a. Understanding small farmers. A paper
presented at a conf. on Rapid Rural Appraisal. 4-7 Dec.
1979. IDS, Univ. of Sussex, Brighton, UK.
__ 1979b. Farming systems research (FSR). CIMMYT,
Nairobi, Kenya. (Unpublished).
CRIA Cropping Systems Working Group. 1979. [Draft] Network
methodology and cropping systems research in Indonesia.
CRIA, Bogor, Indonesia.
IRRI. 1974. An agro-climatic classification for evaluating crop-
ping systems potentials in Southeast Asian rice growing
regions. IRRI, Los Banos, Philippines.
Jodha, N.S., M. Asokan, and J.G. Ryan. 1977. Village study meth-
odology and research endowments of the selected villages
in ICRISAT's village level studies. Occ. Paper 16. Econ.
Prog., ICRISAT, Hyderabad, India.
Thom, D.J. 1978. Human resources and social characteristics. In
Kenya marginal/semi-arid lands pre-investment inven-
tory. Rep. No. 6. Consortium for Int. Dev., Tucson, Ariz.
USDA Soil Survey Staff. 1975. Soil taxonomy: a basic system of
soil classification for making and interpreting soil surveys.
Agric. Handb. 436. USDA Soil Conser. Ser. U. S. Govern-
ment Printing Office, Washington, D.C.


Benchmark Soils Project. 1979. Development of the transfer
model and soil taxonomic interpretations on a network of
three soil families. In Benchmark Soils Project Progress
Report 2, Jan. 1978-June 1979. Dep. of Agron. and Soil Sci.,
Univ. of Hawaii, Honolulu and Dep. of Agron. and Soils,
Univ. of Puerto Rico.
Buol, S.W., F.D. Hole, and R.J. McCracken. 1973. Soil genesis
and classification. Iowa State University Press, Ames,
Calkins, P.H. 1978. Why farmers plant what they do: a study of
vegetable production technology in Taiwan. p. 74-78. In
AVRDC Tech. Bull. 8. Shanhua, Taiwan, Republic of
FAO. 1976. A framework for land evaluation. Soils Bull. 32. FAO,
1974. Soil map of the world, Vol. 1. UNESCO, Paris.
Swindale, L.D. (ed.) 1978. Soil resource data for agricultural de-
velopment. Hawaii Agric. Exp. Sta., College of Tropical
Agric., Univ. of Hawaii, Honolulu.

Chapter 5

- .

After the target and research areas are selected, the
FSR&D team identifies the specific problems farmers face
and develops a research base. A clear identification of prob-
lems and opportunities and the development of a research
base provide the team with information for subsequent
research and development of improved technologies.
This chapter is divided into two parts. The first part
deals with the identification of problems and oppor-
tunities. We refer to opportunities as well as to problems
because they have different connotations. The second part
concerns development of a research base and how to col-
lect and manage data useful not only in problem identifica-
tion but throughout the FSR&D process.


Identification of problems and opportunities is an
iterative and dynamic process that continues throughout
all FSR&D activities. The process is iterative because
research and development results feed back to earlier ac-
tivities to help improve subsequent research activities; the
process is dynamic because accomplishments from one set
of actions can be used to set new or revised targets and
strategies for the next set of actions.
In the past, the problems of small farmers in the
developing countries were often not clear to researchers,
development specialists, or policy makers. These profes-
sionals failed to appreciate that existing farming systems
best met farmers' needs by using farmers' resources and
knowledge. This failure often led researchers to develop
and others to attempt to extend technologies that were in-
appropriate for small farmers.
Farming systems research is more apt to design
technologies that are appropriate and acceptable to small
farmers because the FSR&D approach stresses an under-
standing of the farming system and the farmers' environ-
ment. The FSR&D approach encourages researchers to be
open-minded and to revise their conceptualization of prob-
lems and opportunities as new information is gathered.
In the first part of this chapter, we will define prob-
lems and opportunities, and then discuss the general ap-
proach to problem identification, variations in problem
focus, understanding the farmers' situation, analysis of
problems and opportunities, and setting priorities for prob-
lems and opportunities. We close Part I with a summary.


Problems and opportunities are sometimes like two
sides of a coin. Problems could be the result of constraints
that prevent farmers from reaching their goals, and oppor-
tunities could be the potential for relieving the con-
straints. For example, farmers may have a problem in not
being able to obtain yields comparable to other farmers in
similar circumstances. Researchers cooperating with
farmers might try to learn what prevents the farmers from
doing as well as the other farmers. Once the problem's
cause is identified, the opportunity becomes the potential
for bringing the farmers' yields closer to those of the other
Sometimes farmers have problems they do not
recognize and yet the opportunities for alleviating prob-
lems exist. These opportunities could be in the form of
better use of underutilized or misused resources or the in-
troduction of new technologies.

The FSR&D team follows three basic steps as part of
problem identification. The team:

identifies existing farming systems and seeks to
understand them and the environment
identifies problems and opportunities for improving
the system, the environment, or both
sets priorities for research and implementation.

Fig. 5-1 sets out the various parts to the problem
identification process. The first preliminary analysis (2.a.
in Fig. 5-1) is based on information obtained during the
selection of the target and research areas. Sometimes, a
target area may be selected because of an overriding prob-
lem or opportunity. Such a situation was described in Sec.
4.2.2. where an area in Sumatra was selected because of a
serious problem with Imperata grass. At times enough will
be known about the problem to go directly to planning
research activities. More often, the team will need to learn
more about the farming systems and the environment of
the area. The team usually begins with the collection (2.b.
in Fig. 5-1) and analysis (2.c. in Fig. 5-1) of secondary data.
Then, the team may know enough about the problems and
opportunities to proceed directly to planning research ac-

62 Problem Identification and Development of a Research Base

Figure 5-1. A flow chart for the problem identification and de-
velopment of a research base. See Figure 3-1 for more details
on how the activities in this figure relate to other FSR&D ac-
tivities, the experiment station, and extension.
(From Previous Activity:
Target and Research Area Selection)

(To Next Activity: Planning On-Farm Research)

tivities; but, more likely, the team will find it necessary to
collect primary data (2.d. in Fig. 5-1).
We offer a word of caution at this point. While
bypassing part of the problem identification process is con-
ceptually possible and, at times, can be justified, more
often the team is ill-advised to do so. More likely, prob-
lems and opportunities that appear to be adequately
understood will not be understood and the benefits of the
FSR&D approach will be sharply reduced when a team
bypasses the problem identification process.
A reconnaissance survey is commonly an initial step
in collecting primary data. Some diagnostic analysis, such
as soil pH, insect identification, or fertility tests, may be
sufficient to identify the causes of some of the more ap-
parent problems. When problems are complex, the team
may initiate some experiments quickly to provide better
data for identifying problems and opportunities. During
the reconnaissance, the team seeks breadth of coverage so
that major problems and opportunities will not be
By analyzing the secondary and primary data (2.e. in
Fig. 5-1) the team divides the problems and opportunities
into those that are (1) well-defined, (2) poorly-defined, and
(3) not presently worth pursuing. Well-defined problems
and opportunities are those sufficiently clear and with
benefits potentially great enough that they can serve as the

basis for the next research activity. Poorly-defined prob-
lems and opportunities require further study and defini-
tion. Those not worth pursuing are problems with ap-
parently no practical solution under present conditions or
whose payoffs are substantially lower than those realized
from other possibilities.
The first pass through the problem identification ac-
tivity culminates with a set of well-defined problems
ranked by criteria as discussed in Sec. 5.6. (2.f. in Fig. 5-1).
During the first pass through the problem identification
process, information from sources such as formal surveys,
detailed studies, farm records, and on-farm experiments is
usually not available. Data from these sources come later
and materially add to a thorough understanding and
penetrating analysis of the fundamental problems farmers
encounter. In fact, problem identification, as an activity,
continues throughout the FSR&D process.
In this chapter, we concentrate on this initial iden-
tification of problems and opportunities because the team
often establishes the general approach to the research ef-
fort at this point. The methods for identifying problems
and opportunities remain essentially the same throughout
the FSR&D process, except for a narrowing of focus and a
greater emphasis on detail and accuracy.

While the approach to problem identification is
often similar, the details differ. The FSR&D team may
focus on any part of the whole farming system or it may
concentrate on a predetermined subsystem. However,
both approaches require some study of the whole farming
system. The first approach requires an understanding of all
subsystems and their interactions, while the second ap-
proach demands detailed knowledge only of the selected
enterprise and its relationship with the rest of the system.
The approach an FSR&D team uses depends partly
on its mandate. If the mandate is to increase productivity
anywhere in the farming system, the first approach will
probably be used. The FSR&D projects in Honduras and
Guatemala follow such an approach. Where the mandate
requires concentration on specific crops or livestock, the
second approach will probably be used. Examples are
FSR&D projects undertaken under the auspices of the In-
ternational Maize and Wheat Improvement Center (CIM-
MYT) and the International Rice Research Institute (IRRI),
in which the researchers focus on maize and wheat, and
rice, respectively.
A more specific FSR&D approach is used in coun-
tries that have research organizations separated according
to commodities. In Indonesia, for example, crop research
and livestock research are administered by different
governmental divisions. Thus, researchers would gain ad-
ministrative approval more easily for separate crop or
livestock projects than for combined projects.

Effective FSR&D requires the selection of well-
defined problems and opportunities that are applicable to a

Problem Identification and Development of a Research Base 63

significant number of farmers who operate in about the
same way in response to similar environmental condi-
tions. When such groups of farmers are identified, FSR&D
results can be broadly applied. The key to identifying this
farmer group is a good description of the farmers' environ-
ment and the farming system. For this, the FSR&D team
distinguishes between

environmental factors over which the farm
household has little control
farm management factors over which the farm
household has considerable control.

While FSR&D implements change primarily through the
latter, the team considers environmental factors too.
Therefore, the team examines both to learn

how the farming system interacts with the environ-
what flexibilities for change exist within the en-

-" ."ot"

what farming strategies are likely to succeed
how to influence local and national decision makers
concerning support services and policies.

In learning about these factors, the team will need to
use considerable judgment about the type and amount of
data to collect. FSR&D budgets are normally limited and
time and personnel are usually scarce. Furthermore,
misguided approaches to collecting data frequently cause
research programs to flounder. Thus, before collecting
data, the team should have a reasonably clear understand-
ing of how the data will be used; otherwise, the team
should wait until the need manifests itself. An exception
is when additional data can be obtained at little extra cost.
Then, the researchers should remember that such data
need not be analyzed just because they have been col-
We now turn to a description of the farmers' environ-
ment and the farming system and then to analyses of the
farming system.

64 Problem Identification and Development of a Research Base

The farmers' environment can be divided into the
physical, biological, economic, and social settings. While
many facets of these environments can be used in FSR&D,
some are more critical than others. The sections that
follow cover what we have found to be the more important
facets. Of course, individual situations and the experience
of the FSR&D team will dictate which factors to em-

Physical Setting
The more important physical factors are climate,
water, and land. Under climate, the research team
analyzes primarily temperature and rainfall as recorded at
the nearest weather station. For temperature, the average
monthly values, and the lows and highs during critical
periods in the growing season are important. Rainfall
analysis includes the study of monthly means and the
time of the beginning and ending of the rainy season.
Where irrigation is practiced, stream flow records should
be analyzed. For these types of data, the longer the records,
the better. The study of land takes location, slope, soil
type, whether rain-fed or irrigated, and other factors into
account. For additional information on physical factors,
see Appendix 5-A on the physical resources of the research
area affecting biological production, Appendix 5-B on land
types and land evaluation, Appendix 6-A on climatic
monitoring of a research area, Appendix 5-P for collecting
data during reconnaissance surveys, and Appendixes 7-A
and 7-B for data forms for characterizing and monitoring
on-farm cropping and livestock experiments.

Biological Setting
The biological setting includes those factors that in-
fluence the health and vitality of plants and animals and
the quality of harvested products. Most commonly the
team identifies prevailing diseases and insects, according
to the plants or animals being damaged or harmed. Other
biological factors are considered too. Insects can be a
serious destroyer of harvested grains. Weeds are particu-
larly important in humid, tropical areas primarily as they
interfere with plant growth; but, in some cases, weeds are
allowed to grow and are used as animal fodder. Rodents
and birds are other biological factors to consider. In some
cases, birds-e.g., the African weaverbird-can devastate
crops, especially if the crops are grown out of the normal

Economic Setting
Several aspects of the economic setting influence the
farming system. Access to markets during critical periods
is particularly important for increasing farmers' produc-
tion. These critical periods correspond with the need to
obtain credit, purchase inputs for production, and the
marketing of crops. In marketing, pricing, storage, and
reliable transportation are especially important when the
farmers grow perishable crops. When investigating these
factors, the FSR&D team should obtain information on
any services that might be needed to support new

technologies-for example, the roles of agricultural
chemicals in small amounts or repair services for farmers'
equipment. Other information that the team may need to
gather concerns such items as processing facilities for farm
products; channels of information; the performance of the
extension service and cooperatives; seasonal wage rates;
labor supplies; opportunities for off-farm employment;
sources and costs of traction; and government regulations.
In Appendix 5-C, we provide more information on
economic factors as they relate to markets.

Social Setting
The specific nature of the social factors that the
FSR&D team needs to consider is less subject to
generalization than the above considerations because con-
ditions vary so widely from location to location. Never-
theless, the team needs to keep a number of social factors
in mind-particularly those social factors that influence
farmers' acceptance of new technologies. These factors
concern societal norms and customs related to land owner-
ship and use, division of labor within society and the fam-
ily, rights and obligations according to sex and age groups,
descent and inheritance systems, and other community
norms and customs as they support or restrict individual
and cooperative efforts. We discuss the sociocultural en-
vironment further in Appendix 5-D.


As noted in Sec. 2.3.5., the farming system is a
unique and reasonably stable arrangement of farming
enterprises that the farm household manages according to
well-defined practices in response to the physical,
biological, and socioeconomic environments and in accor-
dance with the household's goals, preferences, and
resources. These factors combine to influence farm out-
puts and methods of production. Having touched on en-
vironmental issues, we now turn to (1) a description of the
household as an integrating factor, (2) the household's
resources, and (3) farming enterprises. The FSR&D team
needs to understand these aspects of the farming system
before attempting to develop appropriate technologies for
specific groups of farmers.
The team normally relies on secondary sources and
reconnaissance surveys for initially understanding the
characteristics of representative farms. These farms
should be of sufficient number to justify the FSR&D effort.
Usually the team bases its descriptions on the three
general categories mentioned above. After problems and
opportunities are identified and selected for further study,
the team can then seek more detailed information using
farm records, surveys, special studies, and other methods.
The general types of data within these three
categories are described below.

The Household as an Integrating Unit
The farm household is a key element in FSR&D.
Much of what FSR&D attempts to accomplish is the in-
tegration of the many factors influencing the farmers'
choices of enterprises and methods of production, given

the environment and the household's resources. In a sense,
the farm family is the ultimate integrator; the FSR&D
team is a helper in the process; and FSR&D methodology
provides the framework for bringing the farmers and the
team together. Consequently, the team needs to gain as
much understanding as is practical about the farm family
and the way it operates.
A starting point is for the team to distinguish be-
tween those families that are "nuclear" and those that are
"extended." A nuclear family consists of the parents and
their children. The extended family is a grouping together
of two or more nuclear families. Types of extended
families include not only parents and their married
children, but also one man or woman with more than one
spouse, members of several generations, or children of
only one spouse when married siblings live in the same
household. The extended family is common where
cultivation practices require a labor force larger than the
nuclear family. The extended family's solution to labor
problems relies on a large permanent labor force, which
must be fed, housed, and clothed over a long time. Such a
unit is especially productive if the extended family has
sufficient land and skills to maintain itself (Wolf, 1966).
With the family structure in mind, the team could
then turn to a study of family decision making. Beal and
Sibley (1967) propose a way of viewing decision making by
farmers that may aid the FSR&D team in understanding
why farmers do what they do. These authors distinguish
among farmer characteristics, knowledge, beliefs, at-
titudes, behavior, and goals.
Characteristics. These are facts about farmers, such
as their sex, age, education, literacy, and ethnic back-
Knowledge. This relates to what farmers know; for
FSR&D, the team focuses on such items as the farmers'
knowledge of alternative management practices, cropping
patterns, sources of inputs, information, and markets.
Beliefs. These concern what farmers think is true,
whether correct or not, based on the farmers' experiences
and common knowledge. Beliefs, in turn, may influence
the farmers' attitudes, behaviors, and goals. For example,
even though farmers may be treated fairly when seeking
credit, the anticipation of being treated unfairly may keep
farmers from seeking credit.
Attitudes. These relate to farmers' feelings, emo-
tions, and sentiments and may have a strong influence
over farmers' decisions to accept or reject new tech-
Behavior. Often farmers' past behavior may predict
how farmers will react in the future. Behavior relevant to
FSR&D concerns past actions such as farmers' work on
and off the farm and visits to communities outside the
region, marketing practices, use of credit, and adoption of
new technologies.
Goals. These reflect what the farm family desires,
i.e., what it is seeking. Goals are conditioned by the fam-
ily's beliefs about what is attainable, as well as its sen-
timents, and are based, in part, on the norms of the com-
munity and farmers' general level of welfare.
Beal and Sibley (1967) used the above concepts for a

Problem Identification and Development of a Research Base 65

sociological study of farmers in Guatemala to gain insight
concerning which factors would be most important when
introducing new technologies to farmers. This is one of
the types of special studies that the FSR&D team might
undertake itself or commission others to do. Alterna-
tively, the FSR&D team might simply use the foregoing
concepts as a checklist when studying the farm family.
We supplement the above description by providing
additional appendixes on family decision making. These
include (1) Appendix 5-E with additional details on the
Beal and Sibley (1967) study, (2) Appendix 5-F with addi-
tional thoughts on decision making by small farm
families, and (3) Appendix 5-G with a description of "deci-
sion trees" used as a method for learning about farmer
decision making. This last appendix refers to the work of
Gladwin (1979), which is an alternative approach to the
statistical methods Beal and Sibley (1967) described.

The Household's Resources
Conceptually, the household's resources can be
described as comprising land, labor, capital, and manage-
ment. Which aspects of these four factors an FSR&D team
considers depend on the situation. Below, we list some
factors that FSR&D practitioners have found useful.

Land. Characteristics ascribed to the farmer's land,
in the broad sense of the term, include

size of holdings
fragmentation of the holdings-e.g., whether the
holdings are a single unit or are broken into pieces
ownership- e.g., sole owner, joint husband and wife
ownership, communal owner, long-standing ten-
ancy, and short-term tenancy
permanency of use-e.g., permanent, shifting, or
landlord-tenant relationships-e.g., share of crops
retained by the farmers, access to milk by the one
who tends the herds, and division of inputs as be-
tween tenant and owner
land quality-e.g., soil depth, texture, and presence
of toxic substances
terrain-e.g., slope, terraced or not, and concave or
convex cross sections when on hillsides
water availability-e.g., nearness of ponds or
streams for livestock, irrigated or rain-fed farming,
and dependability of supply
location-e.g., access to markets and other services.

Labor. The family's labor includes members of the
household who are capable of working and also the
family's participation in cooperative efforts. Some relevant
characteristics are

number, age, and sex of the family members
division of effort among the members according to
preferences, individual profits, and customs
general level of productivity and health
division of time between farming and other ac-
tivities-on the farm and off the farm
extent and nature of cooperative efforts in terms of

66 Problem Identification and Development of a Research Base

obligations to others as well as help from others
other responsibilities and factors that influence the
way farm households allocate their time and effort.

Capital. This factor refers to physical and financial
assets that include

tools and equipment
buildings and improvements to the land
livestock and other assets capable of being sold to
meet the farmers' needs or wishes
cash from sale of crops, animal products, handi-
crafts, and from other sources
access to credit.

Management. The skills of the household in organiz-
ing and carrying out the many farming tasks is a valuable
resource. Management represents a considerable asset to
the family and determines the household's efficiency in
using its land, labor, and capital.

Farming Enterprises
The household integrates the foregoing resources in

conducting various enterprises that relate to crops,
livestock, and other activities such as processing of farm
products, handicrafts, and off-farm employment. Although
small farmers frequently engage in many enterprises to
meet their varied needs, the FSR&D team usually concen-
trates on a limited number of enterprises-perhaps four or
less. Even though this book emphasizes agricultural ac-
tivities, those responsible for FSR&D may decide to focus
on nonfarm activities. Some characteristics of enterprises
to consider relate to

general agricultural practices-e.g., methods of land
clearing and preparation, means of traction, and pest
principal crops and cropping practices-e.g., rota-
tions and combinations, varieties, yields, and
schedules for such activities as seedbed preparation,
planting, weeding, cultivation, and harvesting
livestock and livestock practices-e.g., breeds,
numbers, age and health, feed, yields, uses, and how
tended, where, and by whom
interactions between crops and livestock-e.g.,
complementarities such as crop residues used for

cattle feed, dairy cows used for traction, and com-
petition among enterprises for land, labor, and
overall cash, labor, and power requirements when
the above enterprises are combined into the farmer's
system with emphasis on periods of high re-
quirements and restricted inputs
purchased inputs-e.g., agricultural chemicals by
types, amounts, uses, prices, sources, and depend-
ability of supply
disposal of production through family consumption,
farm use, barter and sales, and marketing factors
such as amounts sold, seasonal and annual price
fluctuations, locations of sale, time spent selling,
and means of transportation.

The Asian Cropping Systems Working Group
developed a format for describing the existing cropping
systems (Zandstra et al., 1981). The Group begins by
listing the major crops and varieties produced on each land
type within the research area, the time the crops are
grown, and estimated yields. Next, the Group lists the ma-
jor cropping patterns. Finally, the Group enumerates the
principal cropping systems and the percentage of farms in
the research area that follow each system. With such infor-
mation, the Group is in a good position to understand the
significance of cropping patterns and practices in the area
and has the raw material with which to identify oppor-
tunities for improvement. We present further details on
this approach in Appendix 5-H.
Furthermore, we include forms for recording and
summarizing some of the information on the char-
acteristics of farming systems in Appendix 5-I. These
forms facilitate

recording of labor required at various periods during
the year, available family and non-family labor, and
the costs of hired labor
determining periods of cash scarcity and sources and
costs of credit (Note: Careful attention needs to be
paid to cash flows, since inopportune cash shortages
may force farmers to sell their products at low prices
at harvest time.)
summarizing power and machinery available on the
farm according to whether they are owned or rented,
or the services are hired
gaining an impression of the farmers' technical
knowledge and the technical history of the area
(Note: Technical knowledge and experience depend
greatly on the area's stage of agricultural develop-

Following initial data gathering, the FSR&D team
will need to analyze the data it has collected. Possibilities
for study of the farming system are to develop conceptual
models of the system and hypothesize about ways to im-
prove the system.

Problem Identification and Development of a Research Base 67

Conceptual Models
Study of the interrelationships among the com-
ponents of a farming system are useful in two ways. One is
to help classify the system, which is part of the selection
of relatively homogeneous groups of farmers who are to be
the focus of the team's efforts. The other is to help under-
stand the system so that the team's efforts will be produc-
tive. Conceptual models in their simpler form identify the
major components of the system and the links both among
the components within the system and between the
system and its environment. Later, additional data can be
gathered to quantify these relationships and eventually to
search for improvements in the system's functioning.
Below is a model in its simpler form developed by
McDowell and Hildebrand (1980). We include a similar
model with additional details in Appendix 5-J, along with
other examples of conceptual models.
Fig. 5-2 shows the McDowell and Hildebrand model
for a humid-upland farming system in Asia. This is one of
eleven models they have developed for small farming
systems in Asia, Latin America, and Africa (McDowell and
Hildebrand, 1980). The authors' explanation follows:

"The box identified as 'Market' represents all off-
farm activities and resources (except land); hence it
includes products sold or labor going off the farm as
well as purchased inputs and household items. The
'Household' is the core of the farm unit. In preparing
the models of the systems, labor use, sources of
human food, household income, animal feed, and
the roles of animals were the main focus. The solid
arrows (--- ) depict strong flows or linkages (e.g.,
more than 20 percent of total income arises from the
sale of crops, animals, or household-processed prod-
ucts). Broken arrows (- - -) are used when sales
of crops or animals contributed less than 20 percent
of household income, the interchange among func-
tions was intermittent, or there was no routine pat-
tern identifiable ... .Family labor applied on the
farm was identified, but off-farm employment or the
amount of hired labor was not quantified except
generally and is indicated by broken or solid arrows.
"For most products there is a direct relation to
market, absent ... when the household changes the
characteristics of the product before sale (e.g., . .
milk to cheese . .). Household modification is
shown by solid arrows from crop or animal products
to household to market."

In Appendix 5-K, we present a further discussion of
this model of the humid-upland farming system in Asia. In
this same appendix, we also include a discussion on the in-
teractions between cropping and livestock systems that
might help in designing research activities for these two
While we believe conceptual modeling, as above,
improves the understanding of the farming system, just
viewing the farming system from a broad perspective helps

68 Problem Identification and Development of a Research Base

Figure 5-2. Humid-upland farming system in Asia, permanent cropping, moderate integration of crops and
animals (animals tethered or herded) [----depicts strong flows or linkages among the parts, while -- --- de-
picts weaker flows or linkages] (McDowell and Hildebrand, 1980).

the FSR&D team. The following are examples of how a
comprehensive view of the farming system led to the iden-
tification of interactions within the system and
possibilities for improvement:

Farmers may combine two or more enterprises to make
better use of scarce resources. For example, in a dry region of
Northern Mexico, farmers are growing maize and castor beans
together. Maize plants, with a shallow rooting system, use the
moisture in the upper part of the soil more effectively while
castor bean plants, with a deeper rooting system, use the
moisture in the lower part of the soil more effectively.
Farmers in New Zealand often put both sheep and cattle
into the same pasture, because the two species mostly use dif-
ferent parts of the pasture's growth. In addition, the cattle
trample and break, and thereby help control the young fern, a
weed, that sheep cannot adequately control.
In many cases, two or more farm enterprises use each
other's products. For example, Egyptian farmers use the lower
leaves of maize plants as feed for their cattle, and use cattle
manure to help fertilize the maize.
Certain farmers in Central and South America grow
maize and beans together. The maize stalks support the climb-

ing beans and the bacteria on the bean's roots fix atmospheric
nitrogen and make the nitrogen available to the maize plant.
Sometimes interrelationships between enterprises are
quite intricate. For example, in Taiwan, farmers grow sweet
potatoes for food or starch processing. The sweet potato
peels are fed to swine, the swine manure is fed to fish, and the
muck from the fish ponds is used to help fertilize the sweet

Hypotheses for Improving the System
By working with conceptual models of the foregoing
type, the FSR&D team should be in a better position to
develop meaningful hypotheses for understanding and im-
proving the farming system. These hypotheses will aid the
team in sharpening its focus on the more important
aspects of the farming system. The following example is
hypothetical, but representative of situations described by
McIntosh (1980):

An FSR&D team observed great variation in the number
of rice crops farmers planted each year, the times and
methods of planting, and the varieties used. The team tested

several hypotheses regarding the relation of these variations
to (1) availability of soil moisture according to rainfall, ir-
rigated conditions, and type of soil, (2) variation in prices of
rice sold and seed bought, and (3) the previous crop in the
field. After analyzing available data, the team found that price
fluctuations did not appear to influence farmers' rice growing
methods. The other hypothesized relationships, however,
proved meaningful. As illustrated in Fig. 5-3, in rain-fed fields
with rather light textured soils, farmers raised only one crop of
a transplanted, traditional rice variety. In rain-fed fields with
clay soil, farmers planted two crops of early maturing varieties
of rice. The first of these crops was dry seeded and the second
transplanted. Where year-round irrigation was available,
farmers produced three early maturing crops of transplanted

Figure 5-3. The relation of rainfall distribution, availability of
irrigation water, and soil type to the variety of rice and timing,
method, and number of plantings (Adapted from Mclntosh,



Rainfed, Light Soil

Months of Year
Transplanted Rice
S Traditional Variety

Rainfed, lay Soil Dry Seeded Rice Transplanted Rice
Early Maturing Varieties

Months Irrigation Transplanted Rice Transplanted Rice Transplanted Rice
12 Months Irrigation I r I |
Early Maturing Varieties

OPPORTUNITIES (Activity 2.e., Fig. 5-1)

With data gathered, the farming system concep-
tualized, and hypotheses tentatively identified, the
FSR&D team is ready to begin a detailed analysis of the
identified problems and opportunities. A starting point is
for the team to select farms representative of the prevalent
farming systems and see how well they compare with
other farming systems. The team will have identified a
problem if results from representative farmers are poorer
than other farmers. These others may be the better farmers
in the research area, good farmers in other areas, or even
managers of experiment stations. The latter comparison is
valid if the team allows for the better production environ-
ment such as better soils, water availability, information,
and resources.
The team may identify a problem in the farmers'
resource utilization. The farmers may be putting too many
resources into one enterprise and too few into another.
Suppose they are picking green beans daily, at a return of
25 cents per hour of labor, but are picking tomatoes only
once a week, with a labor return of $5 per hour of labor. An
opportunity for increased profits might be realized by

Problem Identification and Development of a Research Base 69

harvesting tomatoes more often and green beans less
A low monetary return to the farmers' labor in an
enterprise may, however, not be a problem at all; it may be
the result of a purposeful decision by the farmers to
mitigate a more serious problem. For example, farmers
may be growing a low value, drought-resistant crop of
cassava as insurance against the loss of their other,
drought-susceptible food crops. The farmers' low return on
labor from raising cassava should then be regarded as a cost
of overcoming a serious problem-an excessive risk of
drought loss. Alternatively, female farmers in areas with a
high out-migration of males may plant cassava that,
because of its long time to maturity, helps to spread out
labor demands (Cloud, personal communication).
In identifying cropping and livestock problems, the
team may use performance criteria, such as

yield per unit of area, animal labor, or human labor
value of product per farm or per unit area or animal
net returns above variable cost
net income to farm resources.

Among the other criteria that the team can use to
measure the efficiency of specific resources are

returns per unit of labor or cash
amounts of carbohydrates or protein per millimeter
of rain.

The team may measure intensity of land use by the
indexes for land use intensity or multiple cropping; we
show these calculations in Table 5-H-2 (Appendix 5-H).
The team may use still other criteria to measure stability
of returns over time or across locations. For an additional
discussion of performance criteria, see Sec. 6.4.2. in
Chapter 6, and Part 2 of Chapter 7.
In trying to identify farmers' problems, the team first
looks for the factors that are limiting plant or animal
growth and productivity. These factors may be part of the
physical environment, such as frequent droughts or floods
or poor soil fertility. Problems could also be of a biological
nature such as weeds, insects, or diseases, or could be due
to management factors. For example, farmers may be
planting a wheat variety that tends to lodge when heavily
fertilized; or they may be using a cropping pattern that re-
quires more labor at critical periods than they have
Next, the team investigates why farmers have not
solved these problems. It may find that farmers have good
reasons for not using technology that, at first glance, ap-
pears to be better than what they are using now. For exam-
ple, farmers may not use an insect spray that successfully
controls a rice stem borer elsewhere in the country.
Through investigation, the team might discover that the
reasons for not using the spray include one or more of the

The cost of the insecticide is greater for the farmers
than the value of the rice they save by using it.

70 Problem Identification and Development of a Research Base

Farmers do not know how to use the insecticide cor-
rectly or may be concerned about its effects on their
Farmers cannot get sprayers when they need them.
The insecticide is not available in the area or is sold
in too large a package.
The effectiveness of the insecticide has not been
proven to the farmers' satisfaction.

Other reasons may explain why farmers cannot over-
come their problems. For example, farmers may stay with
a strain of chickens that produces just a few white-shelled
eggs, instead of shifting to a much more productive strain
that lays brown-shelled eggs because they think they have
no market for eggs with brown shells.
Sometimes, an existing remedy for a problem is
against farmers' beliefs or sentiments. For example,
Hildebrand (1977) reported:

"Among the indigenous farmers, young
[maize] plants are treated as a child . ., so they are
almost never knowingly destroyed until they can
provide a useful product. Hence, the farmers plant
only a few seeds and then reseed if the number of
plants drops too low in any hill. The net result is a
less than optimum productive population. The usual
technical solution is to plant a higher than necessary
number of seeds and thin after germination to the
desired number of plants per hill. But, for obvious
reasons, this meets a tremendous cultural resistance
on the part of these farmers, and will probably not be
adopted on any large scale in this area."

A solution to one problem may create problems for
another part of the farming system. For example, plowing
a harvested maize field in the fall would allow earlier
seeding in the spring and better utilization of early rains.
However, in some places these harvested maize fields
serve as cattle pastures in the fall, for which farmers have
no substitute.
In another instance, Hildebrand (1979) reported,

"In an irrigation project in a very dry area of
eastern Guatemala, weeds are a severe problem and a
definite limit to yields of the vegetables that are
raised as the priority crop. However, this area was
historically a dual-purpose cattle zone and the
farmers still maintain their herds for production of
meat and milk. During most of the year, forage is
very scarce, so the farmers use the vegetable fields
immediately following harvest as a source of feed.
Hence, they tend to let the weeds grow to increase
the feed supply, even though it knowingly reduces
the yield of their principal crop and is counter to the
recommendation made by agricultural technicians."

Farmers' abilities to see problems and opportunities
in their farming system are often limited by the narrow
range of their experiences. This is particularly so if they
have little communication outside of their immediate

area. In contrast, the FSR&D team tends to have ex-
perience in other regions of the country and may be
familiar with literature on new agricultural technologies
and research. Thus, the team may see opportunities for
improving the farmers' system that the farmers do not. For
example, farmers may only know the traditional way of
farming in their area. They may produce one crop of rice
per year, using slow methods of field preparation and a
long-season variety of rice. The FSR&D team may be able
to introduce new quick maturing rice varieties and her-
bicides that reduce the required time for field preparation.
Together, these new technologies may enable farmers to
produce two rice crops per year instead of one. The team
may also know of other crops, such as soybeans or
sorghum, that could be grown before or after the rice crop.
Thus, the team may see opportunities for doubling or tri-
pling the farmers' agricultural output, but perceiving such
opportunities is beyond the limited experience of the
Winkelmann (personal communication) sums it up
when he says that farmers have misconceptions. For ex-
ample, in West Africa they believe that leaf mosaic on
cassava is the natural way for the plant to grow. On the
other hand farmers are "street wise," meaning they know
how to survive in a hostile environment. In contrast the
researchers, because of their education and training, can
draw from a larger set of possibilities in solving problems.
Integrating the capabilities of farmers and researchers is
one of the major opportunities for FSR&D.

OPPORTUNITIES (Activity 2.f., Fig. 5-1)
Setting priorities is the final activity in the initial
pass through the identification of problems and oppor-
tunities. The FSR&D team might set its priorities on the
following considerations:

the seriousness of the problem as viewed by both the
farmers and society
the potential for solving the problem and gaining ac-
ceptance of the solution, and the ease of implement-
ing the results
the importance of the problem in some overall
research strategy.


The seriousness of a problem should be evaluated
from the viewpoints of both the farmer and society. The
farmer will usually be interested in the severity and fre-
quency of the problem as related to the farmer's household
and immediate associates. Decision makers who are
responsible to society at large (farmers and others) will be
concerned with the extent of the problem throughout the
target area, and the long-run interests of the present and
future generations.
Where a problem is important to both the farmers
and to society at large, then both groups would agree that
it should be included as part of the FSR&D team's research

agenda. Where the interests of the farmer and society
diverge, two possibilities exist. If the problem is in the in-
terests of the farmers but not of society at large, the
FSR&D team, being a representative of the government, is
advised not to work on the problem. Alternatively, if the
problem is in the interests of society and not of farmers,
the government has the choice of whether to alter the
farmers' environment, including incentives, to bring the
farmers' interests in line with those of society or to leave
conditions as they are.
These examples should clarify the above.

Commonality of Interests
If the government has concern for a specific group of
farmers and the FSR&D team identifies a problem among
them, such as damage from flooding or a disease affecting
one of their major crops, then a commonality of interests ex-
ists. The FSR&D team could include such a problem in the
priority list, provided the severity and frequency of the problem
are great enough.

In Farmers' Interests but Not Society's Interests
Farmers may be pursuing a slash and burn pattern of
farming on erodible hillsides and may be seeking help from the
FSR&D team in ways to expand this activity. If the government
sees this method as depleting the natural resource, the in-
terests of society would be better served by not responding to
the farmers' stated interests. A better solution would be to try
to meet the farmers' needs in some other way.

In Society's Interests but Not in Farmers' Interests
The government may view unsettled lands as represent-
ing a major opportunity to increase agricultural production
and to alleviate land fragmentation in overcrowded areas. If
the farmers are unwilling to resettle, the government can
upgrade the support services in the new area and offer incen-
tives to individuals or groups of farmers in the hopes of per-
suading them to resettle and thereby bring the two viewpoints
into agreement.


The FSR&D team can evaluate the potential for solv-
ing a problem, possible farmer acceptance of the solution,
and ease of implementation according to these criteria:

biological potential
resource availability
economic and financial feasibility
sociocultural acceptability.

At this stage of the FSR&D process, the team tries to
gain only a preliminary impression for these criteria. The
team will make a more careful and detailed investigation
of these factors when planning on-farm research (Sec.
6.2.4.) and during analysis of results (Part 2 of Chapter 7).
Under the heading of biological potential, the team
investigates such questions as

Do the physical and biological conditions in the
research area provide opportunities to solve the
What information on potential solutions is available

Problem Identification and Development of a Research Base 71

from experiment stations, farmers in the research
area and in other areas, and the technical literature?
Do the proposed technologies fit into the farmers'
existing system?

Under the heading of resource availability, the team

whether available resources are adequate to meet the
resource requirements
whether potential solutions reduce the employment
of scarce resources
whether the employment of under-utilized resources
is increased
whether farmers are able to apply the new

Under economic and financial feasibility, the team
tries to determine

whether the benefits of potential improvements in
the farmers' system offer sufficient incentives to in-
terest family members (Note: incentives include
coverage of the costs of purchases and any additional
labor by the family and provide some crop or
livestock surplus to offset the risk and effort of
whether the potential solutions increase or decrease
the stability of the farmers' production and earnings
whether the farmer has sufficient cash or credit to
pay for any increase in purchases and whether
lenders differentiate between males and females in
granting credit
whether the potential solutions change the farmers'
perception of risk through changes in the stability of
production and requirements to obtain credit.

Under sociocultural acceptability, the team tries to

whether the community's social and cultural values,
norms, and customs help or hinder the acceptance of
the proposed solutions.
whether the farmers' perceptions, beliefs,
knowledge, and attitudes facilitate or make more
difficult the acceptance of the proposed solutions
whether field team members have social or cultural
values that hamper their working with certain
groups or types of farmers
whether farm family goals are served or altered if the
proposed solutions are successful.

After problem identification, the FSR&D team may
think through a preliminary research strategy, which we
call a development path. For example, ILCA (1980)
planned a development path for small ruminant produc-
tion in the Forest Zone of West Africa as shown in Fig. 5-4.

72 Problem Identification and Development of a Research Base

Figure 5-4. Generalized development path for small ruminant production in the forest zone [of West Africa] following completion of
initial problem analysis and determination of market potential (ILCA, 1980).

Researchty Mgte a ase Forag Breed Testing New
Research Mortality/ Mgmt x Disease Selection Mgmt Units
Morbidity x Nutrition

Genetic Increased
Existing System / Improvement Anirmal nce Zero
Performance Grazing
Changes Small Flocks/ Reduced Increased mproed Increased Limit of Intensified
in High Mortality Numbers/ Numbers/ Available Production
System and Morbidity Mortality Interest Vigour Growth I Forage Systems
Improved Pastwe
Forage F- Fallow/ Systems
Production Pasture

Improved New Plant/pi
Improvements Health Management/ Animal Capital/

This development path calls for research to reduce disease
incidence among the animals, followed by improvements
in management, marketing, forage production and
breeding stock, in that order.
The relative importance of the identified problems
and opportunities in this development sequence is a useful
criterion for establishing priorities and establishing the
initial direction of the research program. While this activ-
ity is important, the FSR&D team will probably want to
limit the time spent on such planning because the develop-
ment path will change, possibly drastically, over time.


The identification of problems and opportunities
serves an important purpose, coming as it does after the
preliminary collection of secondary data and the comple-
tion of the reconnaissance survey. Such identification
guides the FSR&D team in what research to initiate, and
possibly what steps to take toward implementation. Since
problem and opportunity identification is, at this point,
based on preliminary and largely qualitative information,
the team may not always start in exactly the right way,
but hopefully in the right general direction. As more and
better data are collected, the team, keeping an open mind
and remaining flexible, should have little trouble in mak-
ing the necessary adjustments in its research and im-
plementation activities.


The purpose of Part 2 on the development of a
research base is to present methodologies for collecting in-
formation needed in describing and analyzing farming
systems and in developing new technologies. While these
methods are first used in identifying problems and oppor-
tunities, they are also employed during the rest of the
FSR&D process. Our emphasis in this part is on data col-
lected through surveys, observations, and record keeping,

rather than from experiments. The latter will be discussed
more fully in Chapters 6 and 7. The sections in this part
cover general comments on collecting data, assembling
secondary data, collecting primary data, combining data
collection methods, data management, and a summary.


Data collection in a rural setting requires careful
preparation and a special perspective. Often, farmers and
their associates have never encountered researchers
collecting data. Surveys may arouse both curiosity and
suspicion-in some cultures researchers asking questions
of and about women will arouse suspicion. We describe
some of the more important points for collecting data in a
rural setting in Appendix 5-L and for collecting data about
women in Appendix 5-M.
The methods of data collection to be used in FSR&D
are determined by

the character of the FSR&D activities
the stage in the FSR&D process at which data collec-
tion takes place
the amount of detail and degree of accuracy con-
sidered necessary
the size of the data base needed
the resources-e.g., money, time, personnel, and
equipment-available for data collection and

Fig. 5-1)

The FSR&D team can gather secondary data from
several sources such as

the national census-if the data are specific enough
national archives
reports of the village administration, local

marketing or credit offices, and the local extension
service, irrigation or production program office
reports on locally conducted experimental or exten-
sion demonstration plots
agricultural publications and reports of university
research and consultant studies -such reports often
include excellent detail and good summaries of
secondary data.

For some types of information, using secondary sources is
the most efficient method of collecting data, and the team
should explore this possibility before considering other
The availability and quality of secondary data for
FSR&D varies by subject matter and country. Secondary
data may provide good information on physical en-
vironmental factors such as rainfall patterns, soil types,
etc., but secondary data are often inadequate on the
socioeconomic and biological aspects of the farming
The FSR&D team can determine the usefulness of
secondary data by four criteria:

The relevancy and specificity of the data for the par-
ticular study. For example, if the team needs weekly
rainfall data, monthly rainfall records will not suf-
fice; and biological data for a region may not apply to
the research area.
The clarity of defined terms. For example, a category
of data labeled "returns" without further clarifica-
tion is unsatisfactory. Do "returns" mean gross
returns without any deductions or have some or all
costs been subtracted?
The accuracy and reliability of the data. Specifically,
the team should check these data by comparing
secondary data from different sources, by in-
vestigating the methods used in getting these data,
and by conducting spot surveys.
The recency of the data. As a general rule, the team
should verify secondary data on socioeconomic
topics that are more than 10 years old. This need for
recency is much less critical for physical factors, for
example, soil classification and rainfall patterns.

We present additional guidelines for judging secon-
dary data in Appendix 5-N.

5.10. COLLECTING PRIMARY DATA (Activity 2.d.,
Fig. 5-1)
FSR&D teams commonly use several methods of
primary data collection. These methods fall into two ma-
jor groups-informal and formal. The informal methods
discussed here include reconnaissance surveys, informal
follow-up observations and interviews, and participant
observation. Formal data collection includes single and
frequent interview surveys, farm record keeping, monitor-
ing, and case studies. The reconnaissance survey is the
principal method for obtaining primary data during the

Problem Identification and Development of a Research Base 73

initial identification of problems and opportunities. The
other methods tend to follow the reconnaissance survey.

Informal methods refer to surveys undertaken
without questionnaires. This section begins by giving
some of the advantages and disadvantages of informal
methods, discusses approaches for conducting informal
surveys, and closes with descriptions of reconnaissance
surveys and participant observations. Reconnaissance
surveys occupy a strategic position in FSR&D and include
observations, discussions, and sometimes collection of
physical data. Participant observations provide a means for
gathering data over time, but, to date, have not been
widely integrated into FSR&D methodology.

Advantages and Disadvantages of Informal Methods
During initial problem identification, the FSR&D
team customarily relies heavily on informal methods to
quickly gather broad-gauged data about farmers and local
conditions. For example, informal interviews allow
farmers and others to express their experiences without
excessive structuring by the interviewer. This approach
allows both the interviewer and the interviewee to pursue
topics of interest freely and in depth. When interviews are
conducted in a relaxed and friendly manner, the research-
ers and farmers will have a chance to become better ac-
Informal interviewing also gives the FSR&D team an
opportunity to become acquainted with farmers' words,
concepts, and ideas. This should lead to a much deeper
understanding of farm families, their farming systems and
environment, how they reason, and their decision-making
The team can use informal observations to check
farmers' answers to questions administered in formal
surveys. By verifying the accuracy of observable facts,
such as pests and planted crops, the researcher can judge if
the farmers understand the questions being asked and are
accurately answering the questions.
Data gathered informally have some limitations
because rigorous methodologies are not followed. For ex-
ample, the farmers interviewed may have been selected
purposively and not randomly. Thus, these data should
not be subjected to statistical tests. Furthermore, research-
ers must be cautious in generalizing from informally col-
lected data. Without a written questionnaire to work
from, interviewers may not ask the same questions of all
farmers, nor are they likely to ask questions in the same
way. Thus, quantification, coding, computer analysis, and
summarization become more difficult and the reliability
of conclusions is more subject to question.
One way to overcome some of the disadvantages of
informal surveys is to combine informal investigations
with formal ones. For example, when administering a for-
mal questionnaire, the interviewers may spend a few
minutes on informal conversation and questioning about
matters connected with the questionnaire. In general, the

74 Problem Identification and Development of a Research Base

team usually sets the quantitative data gathered from for-
mal surveys within the context of largely qualitative infor-
mation obtained from informal surveys and observations.
Another way is to carry out a content analysis of the
results of informal interviews. This approach measures
the substance of the interviews in a quantitative manner,
which allows researchers to draw more meaningful infor-
mation from the results. In Appendix 5-0, we include fur-
ther details on content analysis.
To summarize, while informal methods have their
disadvantages, they also have an important role for
FSR&D. They aid the team in (1) quickly learning about
farmers and farmers' conditions and (2) obtaining an early
appraisal of researchable problems and opportunities. At
times these problems and opportunities are so apparent
that formal methods of data gathering are unnecessary and
thus research can begin immediately and results can be ob-
tained quickly.

Approaches for Informal Methods
Information gathered by informal observations and
interviews will differ in both topic and detail, depending
on the capabilities of those who do the gathering. Thus, in-
formal data are best collected as a team effort, involving
male and female staff from the physical, biological, and
social sciences.
Some FSR&D practitioners believe that hypotheses
and general guidelines should be developed before inter-
viewing begins. They argue that this requirement forces
interviewers to conceptualize the farming system and to
take a systematic approach to farmers' problems. Con-
sistency is thereby given to the research process, which
prevents different teams from coming up with substan-
tively different results for similar conditions. Others
believe that a framework prepared before meeting farmers
will predispose team members toward their own ideas,
thereby blocking out opportunities to gain new insights.
Such FSR&D practitioners prefer to go with a "blank
mind." Which approach to take is a judgmental factor and
depends on the nature of the team and the situation.
A difference of opinion also exists regarding the
desirability of recording information in front of farmers.
Some interviewers prefer writing down the information
during the interview, if acceptable to farmers. Others
believe that this practice restricts the spontaneity of
farmers' reactions. Except for recording such items as
quantities, names of products, and varieties, these inter-
viewers prefer waiting to record answers and observations
until they are out of the farmers' sight.

The Reconnaissance Survey
The reconnaissance survey, also generally known as
a quick, informal, or exploratory survey, is a method of
data collection that usually follows secondary data collec-
The reconnaissance survey has several distinguish-
ing characteristics:

It is conducted by the FSR&D team, assisted by com-

modity and disciplinary specialists, and extension
It emphasizes the collection of qualitative data.
It uses informal, largely unstructured interviews
combined with observation.
It is carried out quickly. Sometimes farmer inter-
views are completed and the results written up in
two weeks or less.

Objectives and Approach. The primary objectives of
the reconnaissance survey are to provide orientation for
the research, and to educate and develop teamwork among
the FSR&D personnel. During the survey, the team will
become aware of what additional data are needed to better
understand present farming practices and to design useful
and acceptable changes. The survey involves following up
on the largely quantitative data gathered from secondary
sources by obtaining information firsthand from farmers.
This is the first time in the FSR&D process that the main
attention is focused directly on what farmers do and why
they do it.
The team members attempt to gain impressions of
what enters into the farmers' decision-making with regard
to their farming systems such as their knowledge and
beliefs, their obligations, their goals, and their perceptions
of risk. At this stage, the team formulates hypotheses to
explain present farming practices.
The FSR&D team also undertakes reconnaissance
surveys for other reasons. A survey early in the FSR&D
process helps the team understand farmers' technologies,
terms, units of measure, and explanations of why things
happen. Besides being essential to on-farm experiments,
such information helps the team develop material for
subsequent surveys. Thus, the team will be better able to
phrase questions in formal surveys so farmers can under-
stand them.
Other objectives of the survey include helping the
team to

establish appropriate research objectives and
decide on the type and size of the sample for formal
define more precisely the characteristics of
homogeneous farmers for whom the research will be
publicize forthcoming FSR&D activities
locate collaborating farmers.

Interviewees. The team should consider interview-
ing a cross section of farmers such as

Farmers who hold leadership positions. (They
sometimes have useful perceptions about the
reasons behind traditional practices and how these
practices have changed over time.)
Farmers identified by the extension service who will
often have tried recommended technologies. (They
will have information and opinions about problems

Problem Identification and Development of a Research Base 75

-^ u

stq \ i
S,- *.- ^ jfll
'* . ^^^Br'rf.. J

and potentials for these and other technologies.)
Innovative farmers who have successfully developed
improved technologies. (They will be valuable
sources of information on potential technologies for
other farmers in the area.)
Women farmers who are both members and heads of
households. (They can provide information on fam-
ily decisions and resource allocation in areas where
they have major responsibilities such as the care of
small animals, garden crops, food processing,
storage, and trading (Cloud, personal communica-
Above all, farmers who are representative of major
farming systems in the area.

The team may also want to continue interviews
with people contacted previously during research area
selection such as (1) extension agents, (2) bankers, (3)
buyers of agricultural products, and (4) suppliers of inputs.
Observations. The two basic aspects of the recon-
naissance survey method are (1) observation of farms and
farm families, and (2) interviewing family members.
Team members, often working in pairs, observe
farmers' fields and animals and notice such things as inter-
cropping, spacing between rows, and condition of plants,
animals, tools and equipment, and buildings. They may
take plant and weed counts, study the rate of disease in-
festation, and grade pest damage. Nygaard (personal com-

munication) reports that these types of observations and
measurements by researchers and farmers were important
to the FSR&D surveys undertaken in Syria by the Interna-
tional Center for Agricultural Research in the Dry Areas
Tables 5-1 and 5-2 and Appendix 5-P illustrate the
types of information that field teams have collected and
found useful when observing and recording crop and
livestock conditions during reconnaissance surveys. With
suitable adjustments to meet local conditions, FSR&D
researchers can use tables such as these as aids in quickly
noting important characteristics about crops and animals.
Team members should also take advantage of their
trips to the field to observe other aspects of the farming
system. Possibilities include observations of household
members and the type and quality of their food. In addi-
tion to observations, team members might take soil
samples to measure depth, texture, and pH.
Team members' observations can be used for check-
ing the validity of information obtained through farmer in-
terviews. In this regard, the expertise of experienced
researchers is particularly useful in identifying the
underlying causes of observable and stated problems. Such
information will often suggest follow-up questions to
reconcile any inconsistencies between farmers' statements
and the team's observations.
Interviews. At the beginning of the visit with
members of the farm family, the FSR&D team may ask

76 Problem Identification and Development of a Research Base

Table 5-1. Summary of crop and soil observations from the
reconnaissance survey.*

A Crop
B. Cropping pattern
C. Stage of growth
D. Surface soil texture
E. Dates of observations
F. Observer
G. Top growth characteristics indicative of possible problems
1. Soil moisture relationships: moisture stress_
excessive moisture salinity stress
2. Nutritional relationships: nutrient deficiency
soil pH soil salinity toxicity
3. Pests (e.g., weeds, diseases, insects)
4. Cultural practices (e.g., cultivation methods, weed control)

5. Other (e.g., varietal and field uniformity, plant population)

H. Root growth characteristics indicative of possible problems
1. Soil moisture relationships (e.g., color, distribution, depth)

2. Nutritional relationships (e.g., toxicities, nodulation)

3. Pests (e.g., cutworms, parasites)
4. Other
I. Soil characteristics indicative of possible problems
1. Surface (e.g., crusting, cracking, salts, structure)

2. Subsoil (e.g., compact, layer, mottling, structure)

3. Other
J. Other

*A table should be prepared for each crop in each crop-
ping pattern.

wide-ranging questions such as

How much of the family's harvest will be marketed
and when?
How much labor, credit, and other inputs are
available to the family? Do women have difficulty
obtaining these inputs? How are the inputs used?
What are the family's food needs and preferences?
What are the family's obligations to neighbors,
relatives, friends, and religious institutions?

The team may then use the responses to decide what
specific topics to emphasize during the remainder of the
After each day's work, the team members gather and
discuss such questions as

What have they learned?
Have they observed similar phenomena?
Do they agree on their interpretations of what they
have seen?
What do they need to explore further?

During the reconnaissance survey, the team pays at-
tention to the general practices of farmers in the area and

Table 5-2. Summary of animal observations from the recon-
naissance survey.*

A. Kind of animal
B. Animal use
C. Animal feed
D. Dates of observations
E. Observer
F. Appearance of the animal
1. General condition (e.g., size, amount of fat, skin condi-
tion, liveliness)
2. Symptoms of problems (e.g., swellings, growths, dis-
3. Other
G. Nutritional problems
1. Deficiencies (e.g., protein, minerals, vitamins)
2. Excesses
3. Toxicities
4. Control measures
5. Other
H. Disease and parasite problems
1. Diseases or parasites
2. Control measures
3. Other
I. Sanitation problems
1. Unsanitary conditions
2. Control measures
3. Other
J. Other

*A table should be prepared for each kind of animal.

to the variations in practices among the farmers. In each
case, trying to understand why variations occur helps the
team understand why farmers use certain practices. Varia-
tions among farmers should help identify environmental
changes across the research area. Identification of
trends- e.g., what traditional practices are being discarded
and what new practices are becoming common-sheds
light on farmer reactions to change such as increasing
population or different market conditions.
Toward the end of the reconnaissance survey, the
team might estimate the approximate frequencies-e.g.,
0-10 percent, 11-25 percent, 26-50 percent, 51-75 percent,
and 76-100 percent of farmers who use a particular practice
(Byerlee et al., 1980). Finally, if all goes well, the team
should be able to identify at least some of the more impor-
tant constraints and opportunities in the farming system.
We include a description of the sondeo, the Spanish
term for reconnaissance survey used by the Agricultural
Science and Technology Institute (ICTA) in Guatemala, in
Appendix 5-Q and we provide further guidance on the
kinds of data to collect during reconnaissance surveys in
Appendix 5-R.

Participant Observation
One type of informal observation and interview
method is participant observation-a method from the
social sciences. To use this method, a researcher lives
with a farm family for several months, observing and

recording what is going on and participating in the family's
daily life to the extent possible. The process involves con-
siderable informal interaction with the family.
Participant observation could be used in FSR&D for
collecting information on farmer decision making and the
factors that influence farmers' decisions such as social
obligations, food preferences, and beliefs about plants.
Participant observation would also identify interactions
between different parts of the farming system and between
the farming system and the environment-particularly
socioeconomic aspects. By providing an understanding of
the context into which changes are introduced, informa-
tion from participant observation could serve as a basis for
hypothesizing the possible effects of alternative actions by
the farm family.
Participant observation could be particularly useful
in situations where little is known beforehand about the
farmers' culture. These observations could be a starting
point for recording labor activities according to type, tim-
ing, and worker. Furthermore, these observations might
provide the background for working with more complex
Two disadvantages of participant observation are the
time the method requires and its high cost. However,
since participant observation provides an insight into farm
families that is not easily obtained in other ways, the team
may want to include an element of participant observation
in its data gathering activities. Alternatively, a literature
search of previous participant observations may provide
information about farmers' past practices that helps to ex-
plain their present practices. Studies that relate farmer
changes with environmental pressures could, in turn, aid
the team in predicting farmers' future reactions to changes
in the environment and to opportunities for adopting new

In FSR&D, formal surveys are generally undertaken
to test and otherwise clarify the FSR&D team's recon-
naissance and other findings and to follow up on important
topics. Verification comes primarily through statistical
procedures, but also through additional insights gained by
experienced researchers. This section briefly covers some
of the characteristics and issues of single and frequent in-
terval surveys, questionnaire design, sampling, implemen-
tation of formal surveys, and three variations of formal
surveys-farm record keeping, monitoring, and case
studies. We provide additional details on data collection in
Sec. 6.4.4.

Single Interview Surveys
The single interview survey often follows soon after
the reconnaissance phase. A questionnaire is administered
to farmers usually selected according to formal sampling
procedures. The questionnaire may be used to

Verify and quantify findings of the reconnaissance

Problem Identification and Development of a Research Base 77

collect information on topics of specific interest to
the FSR&D team.

For example, in Syria an FSR&D team followed a general
crop survey with a single interview survey designed
specifically to learn about the cost of harvesting lentils.
The single interview survey is the least costly of the
formal methods per unit of usable information and is best
used for gathering data on phenomena that

change slowly-e.g., land tenure or size of farm
are one-time or infrequent occurrences-e.g., fer-
tilizer purchases, date of planting, and similar
farmer practices
deal with such information as farmer knowledge,
beliefs, attitudes, and goals.

Farmers can easily remember and describe this informa-
tion. Also errors in farmer's statements for such informa-
tion, called measurement errors, tend to be small.
On the other hand, the team is likely to obtain poor
results when using the single survey method to collect
data such as the amount of labor used daily for various
farm operations during the growing season. Farmers do not
remember these kinds of data, called continuous data,
very well. In the single survey approach, measurement er-
rors tend to be large for continuous data because the lack
of a follow-up survey or observation does not permit direct
verification of results.
Another error, called sampling error, occurs when
the results of a survey are not representative of the group.
Much variation may exist among farmers for any
characteristic such as farm size. For a given situation, the
larger the sample size, the less the sampling error. Thus,
to improve the statistical reliability of results, researchers
should take large samples when conducting single inter-
views. We discuss how to deal with measurement and
sampling errors in Appendix 5-S.
Moreover, a single interview is a poor instrument for
gathering sensitive information such as farm income,
because the interviewers have little chance to build rap-
port with the farmers. Also, the FSR&D team must be
cautious in making cause and effect statements based on
data from single interviews. Often, several explanations
can be offered for the information obtained from a single
interview survey. Finally, in societies in which members
do not pool their incomes, income data obtained from
heads of households can be incomplete (Simmons, 1976).
Husbands and wives may not know the income of the
other, nor wish to divulge their own income (Staudt, per-
sonal communication).
In the single interview method, the team may take
an hour and a half, or even more, of a farmer's time.
However, over the life of the project, the single interview
approach takes less of a farmer's time than any other for-
mal data collection method. This is an advantage of the
method, because team members need to be sensitive to
their use of the farmer's time, both in gaining the farmer's
cooperation and in not being a burden during those times
of the year when the farmer is particularly busy.

78 Problem Identification and Development of a Research Base

Frequent Interview Surveys
The frequent interview method involves collecting
data from a limited number of farms on a repetitive basis.
This may continue for a year or more. For example,
periodic visits to farmers by the National Agricultural
Technology Center (CENTA) in El Salvador continued for
more than a year. Other frequent interview surveys may
cover shorter periods such as the growth period of an an-
nual crop.
The frequency of visits depends on the topics, the
degree of accuracy required, and the funds available. For
example, a researcher with experience in farm manage-
ment surveys in Africa prefers twice-weekly interviews
even though they are costly and require much time to
summarize. He found that daily interviews pestered
farmers too much while more than three days between
visits resulted in loss of accuracy (Friedrich, 1976).
The frequent interview survey can be used to show
progress, trends, and fluctuations over time, and to gather
information on specific aspects of the farming system at
specific times or places. This type of survey is well-suited
for collecting continuous data such as labor and cash flows
and food consumption. Because data are continuously
recorded, errors due to faulty farmer recall are minimized,
as are errors in measurement and observation on the part
of the interviewer. This method, because of the frequent
contact, has the potential for establishing close rapport
with the farmer. Furthermore, when carried out for any
length of time, the method provides a mechanism for
transmitting information from the researcher to the farmer
and back again.
Frequent interviewing does, however, have disad-
vantages. The more serious ones concern the time to ob-
tain results, staff requirements, and validity of the data.
Interviewers have to be in the field continually for long
periods, ranging from several months to a year. In addi-
tion, data collected by means of frequent interviewing are
open to sampling error, since the high cost of the method
permits only a small number of farmers to be included in
the sample. Since the frequent interview method requires
much time to collect the data, it cannot be used when the
information is needed quickly.
The team must be aware of the threats to the validity
of the data during interviews. For example, farmers may
unknowingly try to help the team by responding with the
socially correct answers. Or farmers may feel obliged to
answer questions about topics for which they have little
feeling or experience. In such cases, farmers' responses
may not reflect the real conditions on the farm or in the
community. Therefore, the FSR&D team should consider
the validity issue when designing surveys and in inter-
preting the results. In Appendix 5-T, we provide some sug-
gestions on how to do this.

Questionnaire Design
A good questionnaire is important to the FSR&D
team when it makes a formal survey. The questionnaire
links the FSR&D team with the farmers. A sequence for
the development of multidisciplinary questionnaires is

shown in Fig. 5-5. Each stage is discussed below.
The team begins by reviewing what is known from
secondary sources and the reconnaissance survey, and
develops a list of additional data needs. Keeping in mind
the characteristics of both farmers and interviewers, the
team next decides on the type of questions, for example,
multiple choice or open ended-i.e., respondents are not
forced to select from a set of predetermined answers. All
questions need to be carefully and clearly worded, so as to
communicate the intended meaning of the inquiry. This is
helped when the team uses appropriate local terminology
and units of measure.
The questions should be arranged in a logical pro-
gression from the farmers' standpoint, starting with sim-
ple, more general questions and proceeding to the more
specific, difficult, and sensitive areas. Sometimes, break-
ing the logic or sequence of questioning may be desirable
to keep from leading farmers to what they believe are the
expected answers. A way to check on the validity of the
farmers' responses is to ask the same question in more
than one way.
Next the team has to decide on the appropriate
layout and length of the questionnaire and to take time to
train interviewers. These interviewers then assist in

Figure 5-5. The steps in designing a ques-
tionnaire (Adapted from Marketing Re-
search by T.C. Kinnear and J.R. Taylor).
Copyright c 1979 McGraw-Hill Book Com-
pany. Used with the permission of McGraw-
Hill Book Company.

Determine Specific Information Needed

Determine the Type of Questions

Decide on Wording Questions

Decide on Question Sequence

Determine Layout and Length

Pre-Test and Revise

protesting the questionnaire. Pretesting is a trial run dur-
ing which the questionnaire is administered to a limited
number of farmers who are then asked if they had prob-
lems understanding the questions. During this trial run
the team also has an opportunity to test its data analysis
techniques. The team then revises the questionnaire, in
cooperation with the interviewers and perhaps the help of
a farmer advisory group. We provide additional details on
questionnaire design and examples of poorly versus clearly
worded questions in Appendix 5-U.
Designing a questionnaire should not only take into
account the best way in which to ask questions, but
should also consider ways to increase the efficiency of data
collection and processing. In Sec. 5.12., we discuss data
management further.

If the essential characteristics of farm families and
their farming systems in an area were the same, the
FSR&D team would not have to be concerned with sam-
pling. The team would only need to select one farm family
to find out about the rest of the families. Farm families,
however, vary; no one is exactly like any other. Thus, by

Problem Identification and Development of a Research Base 79

sampling, the team selects a small group of farm families
to represent a larger group. The team uses information
from this small group to generalize its findings to all farm
families of a particular type in the research area and,
therefore, does not have to interview all of the farm
families. Sampling saves considerable staff time and
money as compared with taking a census and usually pro-
duces results adequate for FSR&D purposes. In Appendix
5-V, we discuss sampling procedures at greater length.

Implementation of Formal Surveys
With a questionnaire developed and pretested and a
sample drawn, the formal survey is ready to be im-
plemented. Successful completion of the interviews calls
for a cadre of trained interviewers and close supervision by
the FSR&D team. We elaborate on these requirements and
others in Appendix 5-W.

Variations of Formal Surveys
Researchers may normally think of sample selec-
tion, questionnaires, and frequency of contact when hear-
ing about formal surveys, but other approaches to formal
surveys are possible. One of these, farm record keeping,

". ".

, ,
"4" ,/

'y' g ;:: *^ ^--If
kJ ^ ff 7 Y


80 Problem Identification and Development of a Research Base

has played a key role in some FSR&D projects. Two other
approaches are monitoring and case studies.
Farm Record Keeping. A variation of the frequent in-
terview method involves farmers keeping regular, often
daily, records of specific farming activities. The FSR&D
team initiates farm record keeping once it knows the types
of farmers and situations it will be studying. These records
generally focus on inputs and outputs associated with
specific crops and livestock activities. To keep these
records simple enough for farmers and technicians to
manage, data on household items such as family income
and personal expenditures are generally left out. Market
prices and technical descriptions of inputs can usually be
obtained as part of special studies or other data gathering
procedures-not from farm records.
When records are kept by farmers participating in
the on-farm experiments as well as by other farmers in the
research area, the records can form the base for evaluating
FSR&D's effectiveness. Such information, supported by
studies and surveys, also assists the team in interpreting
its research results and in planning research for subse-
quent seasons. If farm records are started early in the pro-
ject's life, they can be substituted for baseline surveys.
When kept for several years, these records reveal the ef-
fects of technological changes in the farming systems.
Farm records are especially useful for recording the
kinds of data that farmers and their families soon forget.
Examples of data recorded include how much work the
farm family performs daily on each crop or animal type,
when and how long the family hand weeds, and when it ir-
rigates certain fields. In Appendix 5-X, we provide addi-
tional details on the types of data to collect and the ap-
proach taken to record keeping by ICTA in Guatemala.
If well managed, farm record keeping is less expen-
sive and more accurate than the other methods of collect-
ing data over a long period of time. However, obtaining
reliable data from farm records requires the FSR&D team,
usually the technical assistants, to regularly and carefully
check these records. These assistants generally live in the
research area during the life of the project. While making
their visits, the assistants should observe the farming
system and report their findings to the rest of the team. In
Guatemala, technical assistants with ICTA have been able
to oversee from 40 to 50 farm records per season.
Farm record keeping normally requires some
literacy. However, two methods have been devised to in-
clude illiterate farmers in this process. One is to have
literate members of the farm household, usually young
children, help in keeping the records. The other method
involves forms with pictures and symbols. Hatch (1980)
designed such a form which we also include as part of Ap-
pendix 5-X.
Monitoring. Another variation of formal surveys is
monitoring. In this section we will report on three types of
monitoring useful to FSR&D. These three relate to obtain-
ing data on climate, recording data from on-farm ex-
periments, and gathering information on livestock
Climatic data need to be monitored as part of plan-
ning on-farm experiments and interpreting the results.

Such data are used to locate experiments according to
different climatic conditions. Also, such data are needed
for judging whether the climatic conditions prevalent
during the experiments represent typical or atypical con-
The FSR&D team generally collects data on rainfall,
temperature, wind, and sometimes solar radiation. This
information should be summarized weekly and monthly
by the team's technical assistants. Then senior members
of the field team should review these data periodically. If
the assistants are having difficulty preparing the weekly
summaries, the team might consider omitting some of the
items, or reducing the number of locations where
measurements are taken. In Appendixes 5-A and 6-A, we
provide more information on climatic monitoring.
Monitoring of cropping and livestock experiments
includes two types of observations-those about crop and
livestock performance and those about farmers' manage-
ment practices. This monitoring requires field team
members to visit the experiments, particularly farmer-
managed tests, to record farmers' activities and plant
growth or animal performance during the experiments.
Such monitoring is particularly important for farmer-
managed tests, because the team needs to keep track of
what the farmers do and what results they obtain. For
researcher-managed and superimposed trials, the team ob-
tains the necessary data as a natural part of its experimen-
tal activities.
Observations of crops and animals, for example, in-
clude those that indicate symptoms of stress, deficiencies,
toxicities, pest and disease infestations, plant stands, and
changes in animal health. The researcher records farmers'
activities as they relate to such factors as planting dates
and animal feeding rates. The results from such monitor-
ing help the team to (1) identify production problems and
(2) understand farmers' management practices.
ILCA uses monitoring techniques to observe how
livestock production systems change with the introduc-
tion of production-oriented projects. This method of data
collection complements ILCA's other work such as identi-
fying constraints and carrying out detailed analyses. ILCA
(1978) comments on the purpose of its monitoring ac-
tivities as follows:

"It will also be important to examine the
response of traditional systems to development pro-
cesses. Indeed, the monitoring of ongoing develop-
ment programmes needs to receive a high priority,
since these programmes represent unique ex-
periments which can never be reproduced in the con-
fines of a research station. If not given early atten-
tion, a great volume of information crucial to future
livestock development will be lost. At first these
studies are likely to be mainly in eastern Africa,
where existing development programmes already af-
fect a wide range of pastoral societies, though they
would be selected also for their wider relevance to
Africa as a whole."

In Appendix 5-Y we provide additional information

Problem Identification and Development of a Research Base 81

on ILCA's approach to the monitoring of livestock
Case Studies. The case study approach, a special
form of the frequent interview survey, involves an in-
depth analysis of a small number of farms selected because
of their representativeness of farming systems in the
research area. Formal interviews, combined with observa-
tion and informal discussions with the farm household are
repeated regularly, sometimes for an entire year.
The case study method is particularly suited for in-
vestigating the whole farming system and the interrela-
tionships among its parts and between the system and its
environment. One output of such a study is an integrated
model of the farming system, showing, in quantitative
terms, flows of money, materials, energy, and information
(Hart, 1979). Such information is useful in predicting
probable repercussions of changes in any part of the sys-
The primary advantage of the case study is the ac-
curacy and detail of the data. This in-depth view can be
most helpful in interpreting the data gathered on many
farms during reconnaissance and single interview surveys.
However, the method is costly and time consuming. In ad-
dition, as with the frequent interview survey, sampling er-
rors can occur, since few farms are usually studied. In Ap-
pendixes 2-A and 5-J we discuss Hart's (1980) case study
work in Honduras.

FSR&D teams use several methods to gather data.
Each method is limited in the quality and type of informa-
tion it can provide. Achieving the right combination of
methods involves a careful study of research needs and
resources. As different kinds of data are often needed at dif-
ferent stages of FSR&D, the methodologies used should be
reevaluated periodically. A carefully developed combina-
tion of data collection methods can take advantage of the
best points and minimize the drawbacks of each.
Following are some possibilities:

A reconnaissance survey is almost always followed
by one or more of the other data collection methods.
Informal observation and casual interviewing should
accompany formal data collection methods.
The single interview survey is often combined to
good advantage with frequent interview surveys,
farm record keeping, or case studies.

Concerning this last point, a single interview survey at the
outset minimizes the delay in moving from the descriptive
phase to the problem solving stages of FSR&D. Much of
the information, however, may not be detailed enough for
later research. Following the single interview survey, one
or more of the other methods can provide the required

University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs