Journal of farming systems research-extension

Material Information

Journal of farming systems research-extension
Running title:
Journal for farming systems research-extension
Abbreviated Title:
J. farming syst. res.-ext.
Association of Farming Systems Research-Extension
Place of Publication:
Tucson Ariz. USA
Association of Farming Systems Research-Extension
Publication Date:
Physical Description:
v. : ill. ; 23 cm.


Subjects / Keywords:
Agricultural systems -- Periodicals -- Developing countries ( lcsh )
Agricultural extension work -- Research -- Periodicals ( lcsh )
Sustainable agriculture -- Periodicals -- Developing countries ( lcsh )
serial ( sobekcm )
periodical ( marcgt )


Dates or Sequential Designation:
Vol. 1, no. 1-
General Note:
Title varies slightly.
General Note:
Title from cover.
General Note:
Latest issue consulted: Vol. 1, no. 2, published in 1990.
Electronic resources created as part of a prototype UF Institutional Repository and Faculty Papers project by the University of Florida.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
The University of Florida George A. Smathers Libraries respect the intellectual property rights of others and do not claim any copyright interest in this item. This item may be protected by copyright but is made available here under a claim of fair use (17 U.S.C. §107) for non-profit research and educational purposes. Users of this work have responsibility for determining copyright status prior to reusing, publishing or reproducing this item for purposes other than what is allowed by fair use or other copyright exemptions. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder. The Smathers Libraries would like to learn more about this item and invite individuals or organizations to contact Digital Services ( with any additional information they can provide.
Resource Identifier:
22044949 ( OCLC )
sn 90001812 ( LCCN )
1051-6786 ( ISSN )

Full Text
Vol e Number 2
o urn al for Farming Systems
Research- Extension
I V. Message Flom file President of AFSRE IX. From the Editor's Desk I Resource Poor Farmers N% ith Complex Technical Kiio%% Icdgc ill a High
Risk SNsfein ill 17 fhiopia (';ill Rescarch lIcIp"
ujisak(l. ( 71(11-1c.s [F)IIIIIIIIII. (111(i HablaIIIII. 1(hunassu 1- IicIcrogcnciI\ and Coinplcml-N ill Farming S.\slcni ToNN:irds all
FA011111ollar.\ PClSpCCh\C
Reductionisni. SNsIcnis Approaches. and Farmer Parlicipalion Conflicts
and Connihilions ill Ilic Norlh American Land (irant SNstc[li
lolM .')', ( Ihhl c// wl& I I.( her /I ( 711,1'sti(III
45 Inicgialing I-SR iwo file Mitional I'xiciision S.\stem A C;Ise of'
Bangladesh III(InIIII Rm
.5.5 Appl.\ ing Fainiing and Regional S\sIcins Approaches for Sustainable
Wood Fnerp Rcsom-cc De\c1opinciii A Case of'Northern Thailand
. h-elacch'(1111kill 1110 Ff, I)oPpIcr
61 Contribulions from Social SNsferrisThinking to Farin Systems Research
and Extension
X ') Book Re\ ic \ s

for Farming Systems Research-Extension
Volume 6, Number 2, 1996
Published by
the Association for Farming Systems Research-Extension

Journal for Farming Systems Research-Extension
ISBN: 1051-6786
George H. Axinn
Department of Resource Development
Michigan State University
East Lansing, Michigan,U.S.A.
Nancy W. Axinn, Editorial Assistant
The Journal for Farming Systems Research-Extension is published by the Association for Farming Systems Research-Extension (AFSRE), an international society organized to promote the development and dissemination of methods and results of participatory on-farm systems research and extension. The objectives of such research are the development and adoption through participation by farm household members of improved and appropriate technologies and management strategies to meet the socioeconomic and nutritional needs of farm families; to foster the efficient and sustainable use of natural resources; and to contribute toward meeting global requirements for food, feed, and fiber.
The purpose of the Journal is to present multidisciplinary reports of onfarm research-extension work completed in the field, and discussions on methodology and other issues of interest to farming systems practitioners, administrators, and trainers. The Journal also serves as a proceedings for the international Farming Systems Symposia from which selected and refereed papers are included. It also welcomes contributed articles from members of AFSRE who are unable to attend the symposia. Contributed articles will be judged by the same review process as invited articles.
The AFSRE President is: The AFSRE Secretary/Treasurer is:
Dr. P. Anandajayasekeram Dr. Virginia Cardenas
SACCAR Education and Rural Studies
P/bag 00108 College of Agriculture
Gabarone University of the Philippines at
BOTSWANA Los Bafios, College, Laguna
Tel: (+267) 328769 PHILIPPINES
Fax +267 328806 Fax +63-2-813-5697 or +63-2-942-914
e-mail: 100075.251 e-mail:
Correspondence regarding articles for this journal should be addressed to: George H. Axinn, Editor, JFSRE, E-Mail -- FAX 517-353-8994 Postal: 313 Natural Resources Building, Michigan State University, East Lansing, Michigan 48824-1222, U. S. A.
Journalfor Farming Systems Research-Extension

Journal for Farming Systems Research-Extension
Volume 6, Number 2, 1996
IV. Message from the President of AFSRE
IX. From the Editor's Desk
1 Resource Poor Farmers with Complex Technical Knowledge in a High
Risk System in Ethiopia: Can Research Help?
Sam Fujisaka, Charles Wortmann, and Habtamu Adamassu
15 Heterogeneity and Complexity in Farming Systems: Towards an
Evolutionary Perspective
G. Weber
33 Reductionism, Systems Approaches, and Farmer Participation: Conflicts
and Contributions in the North American Land Grant System
John S. Caldwell andArcher H. Christian
45 Integrating FSR into the National Extension System: A Case of
Bangladesh Indrajit Roy
55 Applying Fanning and Regional Systems Approaches for Sustainable
Wood Energy Resource Development: A Case of northernn Thailand
S. Praneetvatakul and W. Doppler
61 Contributions from Social Systems Thinking to Farm Systems Research
and Extension
T. E. Kleynhans
83 Book Reviews
VoL 6, No. 2, 1996

Dear Colleague,
Greetings from the Southern Cone of Africa. First of all, let me thank you all sincerely for the responsibility that you have entrusted on my shoulders to lead this Association on your behalf. I will assure you that I will try to live up to your expectations. I also would like to congratulate the outgoing president and the board for the excellent job they did during the last two years to continue to maintain the image and credibility of this Association. Without the clear vision of the board and the commitment and dedication of the members, this could not have been possible.
The major objective of the Faming Systems Approach (FSA) is to help improve the relevance and efficiency of research, extension, training, institutional, and policy support systems in accelerating the process of agricultural development, with a view to improve the livelihood and well-being of farm families. The renewed interest in Farming Systems ResearchExtension (FSR-E) in the seventies arose as a result of the failure of the traditional approach (in much of the world), and the so-called "Seed-fertilizer" revolution to create the necessary impacts in the less favourable environments where the majority of the resource-poor farmers live.
Traditional production systems in many parts of the world reflect an enormous degree of ecological awareness. The underlying principle in the evolution of the wide variety of farming systems was the selection of a sustainable agriculture so that people could survive. The guiding principle in many African fanning systems is that of risk minimization. The traditional agriculture was, to a high degree, integrated with the ecosystems; but unable to meet the demands of the growing human population and economic challenges. Modern agriculture is highly productive, but also often unstable, especially in rainfed farming conditions, and considered to be environmentally un-friendly. The challenge for FSR-E practitioners is to develop, with farming people, new and improved sustainable agricultural production systems by integrating what is good in the different systems in a compatable, economically and ecologically sound manner -- a Win-Win situation; increased productivity coupled with ecological/environmental sustainability. This challenge is further complicated by recent economic developments. Market liberalization as well as structural adjustment programmes are currently being implemented in many countries in Sub-Saharan Africa, Asia, and Latin America. These recent changes often call for fundamental changes in existing production systems -- a reality to be addressed in the current context.
Journalfor Farming Systems Research-Extension

Over the years, FSR-E procedures have undergone various modifications giving rise to an incredible range of methods, projects, vocabularies and technologies. FSR-E methods have grown and matured sufficiently that there is likelihood that they will continue to be a major approach among those concerned with technology development and transfer (TDT). At the beginning the focus of FSR-E activities was to better understand the various components of the fanning system; the linkages among those components, and the linkages between farming systems and the larger social/political/economic systems of which they were a part. (It was an application of systems science to farming). The goal was often technology generation and dissemination. Now there is growing consensus among practitioners that the approach could be used effectively to address issues related to policy and institutional support systems affecting the farming community. While significant attention has been given in the past to developing FSA methods, provision for fully integrating this approach within the research and extension process has been inadequate and in some instances the institutional challenges are being undermined. This aspect deserves more attention in the future.
FSA as a tool for successful technology development and transfer has been accepted by the majority of decision makers and development practitioners within agricultural and natural resources sectors. The TDT is a dynamic and evolutionary process and will continue to evolve as new challenges emerge. In a dynamic world such as ours, the challenges confronting the National Agricultural Research Systems (NARS) are continuously changing. As a reflection of these changes, the current FSA methods and procedures are also being challenged to respond to these changing needs. The contemporary challenges confronting the FSA users include, among others, increased farmers' participation, cost effectiveness of FSR-E, need for sustainability (ecological, economical, as well as institutional) considerations in FSA, need for increased focus on policy research (including micro-macro linkages), improved research-extension linkages, impact of FSA on professional agricultural research and extension personnel as well as on the ultimate farming community (the so-called "people level" impact), effective use of indigenous technical knowledge (ITK) in FSA, and empowering the farming community. Currently research and development practitioners are working on these issues and developing innovative methods and techniques to address them. I would like to see a greater participation and contribution of our members in these debates.
In my opinion, the emerging methods and techniques are not substitutes for FSA, but can complement the methods that are already available for clientoriented research and development. FSA practitioners are thinking innovatively in addressing the practical problems encountered in the field. In coming up
VoL 6, No. 2, 1996

with new ideas, one needs to be very careful about the limitations and inherent dangers in stretching these ideas beyond their limits. This is especially true when ideas/techniques/methods are developed in an external environment. Therefore, we need to critically evaluate these emerging ideas and methods. I trust that all of you will participate in these evaluations, so that the appropriate modifications are grafted on to the existing procedures. FSA procedures are flexible enough so that the new technologies and methods can be effectively combined in addressing the new challenges.
Now, turning to the challenges confronting the Association itself, the FSA family is continuously growing. Today, in addition to the global mother association, the Association for Farming Systems Research-Extension (AFSRE), there are several others. There is one in Asia, several in Latin America, one in North America, and four in Sub-Saharan Africa (SSA). Three of the African associations are regional (Eastern, Western, and Southern) and one is continental in terms of geographical coverage. In SSA we are still struggling to work out the linkages and responsibilities between the regional and continental associations. Given this expansion and growth, there is an urgent need to redefine the roles and responsibilities of AFSR-E, so that it complements and reinforces the activities of the regional associations. At the Sri Lanka Symposium a committee was constituted by the General Assembly to review the current constitution of the Association. The terms of reference of this committee were to review the goals and mission of the Association, including the philosophy, strategy, and future of AFSR-E. This issue should be addressed in relation to the goals, objectives, and activities of the regional organizations. Currently I am working very closely with this committee to get a preliminary report on this matter. This report will be circulated widely to all members of the Association for review and comments. I would urge all of you to participate actively in this process of chartering the future direction and priorities of this Association.
Each of the sub-regional associations are also planning to publish their own journals. AFSR-E is still facing problems in getting articles for publication because a good number of FSA practitioners from the developing countries are not members of the Association. There is a need to revisit the editorial policies of the AFSR-E journal, and the regional journals, to ensure that the various issues of the journals are released at regular intervals without undue delay at the same time that the quality of this Journal is also maintained. The review procedures followed by these journals need to be harmonized. Several policy questions regarding the journals should be addressed and resolved in the near future. At present the AFSR-E Board is in the process of constituting an Editorial Conmmittee. The editor and the chairperson of the Editorial Committee are currently addressing these policy issues confronting this
Journalfor Farming Systems Research-Extension

Journal, and seeking views of the committee members. Once again, this information will be circulated to you all in due course. Please take time to study these materials when you receive them and give us your views and expert opinions. I earnestly request all of you to submit your contributions (articles, book reviews, notes, etc.) to these journals. It is the responsibility of each and every member to ensure that the journal is published regularly while maintaining its quality.
Another issue of concern is the declining membership. The only way to increase the membership is to raise the image and credibility of the Association and to provide services to the FSA community, so that they can be proud to be members. One of the reasons for this declining membership of the AFSR-E is the formation of regional associations. Some members are unable to invest in more than one association. This issue should also be addressed while reviewing the constitution of AFSR-E and its linkages with regional organizations. If you have any suggestions to address these issues and challenges please feel free to communicate with me or any of the AFSR-E Board members.
One of the distressing problems confronting the broader TDT activities in many developing countries is the declining support for agricultural research. In the recent past, the high payoffs to investments in agricultural research and extension are well documented. Despite these, in times of budgetary tightness research seems to be a prime target for cuts. This is largely due to the long gestation period to generate results. Even within research and extension, the on-farmn research component gets affected very much because of the need for a higher level of recurrent expenditures to continue these activities. Although the reintroduction of a systems approach to research and extension has celebrated its silver jubilee, the FSA community is not consciously documenting the accomplishments and success stories. Research managers at the national level and the donor community are looking for the impact of FSA to justify their continued investments. Therefore, in order to continue to maintain, as well as to mobilize additional resources to support FSA activities, there is need to document and disseminate the success stories and accomplishments. I would urge the FSA community to use our journal(s) as well as other forms of publication to widely disseminate this information.
As all of you may be aware, preparation for the 15th International Symposium of AFSR-E to be held in Pretoria, Republic of South Africa, from 30 November through 4 December 1998, is proceeding well. This is the first time the global symposium is going to be held in Sub-Saharan Africa, and the Rainbow Nation is proud to host this symposium, The major theme of this symposium is "Farming Systems Approach (FSA) in Rural Livelihoods, Empowerment, and the Environment: Going beyond the farm boundary." The
VoL 6, No. 2, 1996

sub-themes include ecologically sustainable development and farming systems; short term farmer survival versus long term sustainability; capacity building; empowerment through building capacity; the institutional environment and farming systems; and methodological issues and challenges. Please mark these dates in your calendar. In addition to providing food for your thought, you can explore the unexploited natural potential as well as the warm hospitality of the democratic South Africa. We expect a large participation in this symposium from all parts of the world. I am looking forward to seeing you all in Pretoria!
Last but not least, remember this is your Association. The strength of this Association depends on your active participation in its activities, dedication and commitment to its cause. The membership should take a pro-active role in deciding the future direction and activities of this Association. Let us all work together to strengthen this Association, make it more relevant and productive so that we can be proud that our Association is second to none and we are members of a successful, credible, professional association serving the resource poor farmers around the world.
Dr. P. Anandajayasekeram (Ananda)
President, AFSR-E
May, 1997
Journalfor Farming Systems Research-Extension

The world is always changing, and a new look is evolving for the global Association for Fanning Systems Research-Extension. Readers will find evidence of that in the pages of this issue of the Association's Journal. If you started reading this issue from the first page, you have already seen the message from our new president, Dr. P. Anandajayasekeram. For members of AFSRE, Dr. Ananda has called our attention to many of the opportunities lying ahead of us as regional associations in most parts of the world gain strength, and the leaders of the global AFSRE seek to rationalize our evolving relationships.
This Journal is also evolving. As mentioned in the President's letter, we now have a global editorial committee, under the leadership of former AESRE President, John Caldwell. As they struggle to work out appropriate relations among the regional or continental journals and this one, the Journal for Farming Systems Research-Extension, opportunities abound for those involved in farming systems approaches to agricultural research and extension to reach an ever-widening audience through their publications. Personally, I look forward to some kind of an arrangement through which, as each one of us joins her/his regional organization we automatically become members of the international AFSRE, and thus receive both information from the place where we work and from the rest of the Globe.
Readers who are interested are urged to communicate with President Ananda, Chairperson Caldwell, or myself to share your views in these matters. The next issue of this Journal will feature a Letters to the Editor section in which your suggestions for our associations, as well as other ideas for the future of systems approaches to the study and development of fanning may be shared with others.
A new feature of this issue of your Journal is the Book Review Section. This time we include reviews of some of the newest and the oldest of relevant books. Robert Chambers' 1997 book, Whose Reality Counts? Putting the first last should bring back warm memories and deep understanding to readers who themselves work very closely with farming people. U. P. Hedrick's 1948 book, The Land of the Crooked Tree, may cause some readers to raise their eyebrows, but it was only this year that your editors read it for the first time, and we were amazed by the accuracy and detail of the descriptions of fanning systems in this book which were made at the beginning of this Century.
Different from the first two, the next two reviewed books take a broad sweep of the global situation with respect to an evolving situation most relevant to those concerned with systems approaches to farming, to agriculture, and to the broader ecological, social, economic, and political systems of which they are a part. One is Lester Brown et. all's State of the World 1997; the other is Idriss
VoL 6, No. 2, 1996

Jazairy et. al.'s 1992 IFAD book, The State of World Poverty -An Inquiry into its Causes and Consequences.
Finally, because it is so personal from a dedicated field worker who has a lifetime of experience learning from rural people, Akhtar Hameed Khan's 1996 book, Orangi Pilot Project Reminiscences and Reflections is included. That is followed by a brief notice of our loss of one of the best known authors in this field, Willard R. (Bill) Schmehl. And then, you will find notice of a new book just being released by AFSRE, summarizing much of what members of this Association have written in the last two decades about farming systems.
Before you get to the Book Reviews, however, I hope you will read the analysis of resource poor farmers with complex technical knowledge in a high risk system in Ethiopia by Sam Fujisaka, Charles Wortmann, and Habtamu Adamassu. It is followed by an evolutionary perspective on heterogeneity and complexity in farming systems by Georg Weber.
The next three articles deal with the challenge of actually implementing farming systems approaches in a national system of agricultural or forestry research and extension. The first, by John Caldwell and Archer H. Christian focuses on the North American Land Grant System; the second, by Indrajit Roy deals with integrating the farming systems approach into the national extension system of Bangladesh; and the third analyzes the application of farming and regional systems approaches in Northern Thailand, and was written by S. Praneetvatakul and W. Doppler. The final paper, by T. E. Kleynhans, includes an overview of the evolution of systems thinking and systems applications. It provides those who practice farming systems approaches with a useful conceptual framework.
Here are two other important thoughts from your Editor. There is an urgent need for more papers from readers like you which meet the criteria for this Journal. (See p. XII) It is great that many farming systems articles can now be published in regional and other publications. But this Journal cannot serve your needs unless you send me your manuscripts. The other thought regards the importance of your participation in the AFSRE Symposium scheduled for South Africa in November-December 1998. Please see the notice of it which follows, and begin to make your plans.
George H. Axinn
Editor, JFSRE
Journalfor Farming Systems Research-Extension

Date: 30 November 1998 to 4 December 1998 Venue: Pretoria, South Africa
Theme: Farming Systems Approaches in rural livelihoods, empowerment,
and the environment: Going beyond the farm boundary Sub-Themes:
1.1 The changing role of researchers and extensionists in
development, and in the dissemination of knowledge and
1.2 Integration of micro (specific) strategies with macro
economic/social/political and ecological factors.
2.1 Reciprocity and exchange; farmers responses and initiatives.
2.2 Heterogeneity and multiple realities.
3.1 Training in FSRE and other participatory techniques: tips, tricks,
techniques, and materials.
3.2 Empowerment of personnel and organizations to meet the challenges of changing strategies for agricultural research and
4.1 Demand-driven research and development.
4.2 Government regulations: incentives or constraints?
4.3 Contributions of gender studies to increased efficiency and equity
in research and knowledge generation.
5.1 Building on local rural knowledge.
5.2 Working with farmer groups experiences, benefits, problems.
5.3 Cost effectiveness of FSA.
5.4 Production of knowledge and exchange.
FOR MORE INFORMAL TIONABOUT THE 15T INTERNATIONAL SYMPOSIUM, contact AFSR-E Symposium '98, PO Box 411177, Craighall 2024, SOUTH AFRICA. FAX: +27 (0)11 422 5927 E-mail:
VoL 6, No. 2, 1996

Criteria for Selection of Future Articles for this Journal
a. Articles will address farms as whole systems, including their production and consumption, and involving the plants, livestock, and humans in the farming system, or the institutional support services for fanning systems, including the policy structure.
b. If possible, field research would involve partnerships among professional agricultural/forestry scientists, extension personnel, and farmers (men, women, and children). Ideally, that includes collaboration in deciding what to study, designing the research, collecting the data, analyzing and evaluating the findings, etc.
c. Human data will be disaggregated by gender.
d. Where papers involve several layers of a systems hierarchy -- from soil microenvironments (for example), to whole farms, to communities of farms, to larger agro-ecological zones, to national policy -- they will deal with these phenomena from a systemic perspective.
e. Papers which deal with FSRE as a concept, and its relationship to other approaches to rural development, etc., will assume that FSRE refers to work like that described in (a) through (d) above.
f. Papers should demonstrate the relationship between new material being presented and the current literature in FSRE, as published in earlier issues of this journal, proceedings of AFSRE symposia, etc.
In addition to Regular Articles, the Journal will occasionally include a section of Field Notes. These are more brief, need not have literature citations, but are expected to share relevant field experience which may be useful to other practitioners and scholars.
Journalfor Farming Systems Research-Extension

Sam Fulisaka, Charles Wortmann and Habtamu Adamassu1
Dryland farmers in the Nazret area of the Rift Valley SE of Addis Ababa are resource poor and must deal with low and uncertain rainfall. They have fine-tuned a set of alternatives if rains arrive late, if rains stop after initial rains and sowing, and/or if any crop fails to establish; and they sow a range of crops and cultivars suited to different land, soil, and moisture conditions as the growing season unfolds. Farmers have changed crop and varietal preferences over time in response to problems and opportunities. Their complex technical knowledge and sound adaptation to circumstances suggest, first, that there may be limited opportunities for additional research-based problem solving; but that, second, any system improvements will require substantial farmer involvement.
Farmers and researchers were able to develop a set of participatory research activities directed towards specific problem areas, largely to intensify use of their more productive lands. Research activities include evaluation of varieties of several crops, improved low-level input use and
improved tillage.
S. Fujisaka is agricultural anthropologist, Centro Internacional de Agricultura Tropical (CIAT), A.A. 6713, Cali, Colombia; C. Wortmann is agronomist, CIAT, P.O. Box 6247, Kampala, Uganda; and H. Adamassu is agricultural researcher, Nazret Research Center, Institute of Agricultural Research (IAR), P.O. Box 436, Nazret, Ethiopia.
VoL 6, No.3, 1996

The agricultural area northeast of Nazret (and SE of Addis Ababa), Ethiopia, in the Rift Valley receives 800 nun rainfall per year with uncertain initiation and cessation and a peak in July and August (CIAT Agroclimatic Database). Rains are also uncertain in September during a critical stage for most crops. Soils are sandy to clay loams with little organic matter (Mulatu et al 1992). Major crops are tef, maize, beans, wheat, barley, and sorghum. Farmers kept small numbers of cattle and oxen, as well as a few goats and donkeys. Farmers matched crops and cultivars to different lands and soils and to the uncertain moisture conditions; used a range of techniques to control weeds; applied small amounts of farmyard manure (FYM) and inorganic fertilizers; but had few solutions to problems such as malaria, livestock diseases, lack of cattle feed during part of the year, and insect pests in peas and lentils.
Participatory diagnosis and planning suggested that although farmers were relatively "rich" in technical knowledge and solved many problems to the extent that limited resources allowed, there were also opportunities for farmer participatory research to make systems improvements. A previous study in Ethiopia in which farmers discussed varietal characteristics, problems, and technology evaluation; and participated in the formulation of recommendations and policies also highlighted the importance of integrating farmers' and researchers' knowledge (Franzel 1992). Agricultural research conducted in the Nazret area, however, has mainly consisted of on-station or researcher-managed on-farm trials on tie-ridging, row planting, seed rates, sowing dates, rates and timing of fertilizer application, herbicide use, and varietal trials of early-maturing maize, beans, sorghum, and tef (Regassa et al 1992).
Our work builds upon the increasing incorporation of indigenous technical knowledge and of farmer participation in agricultural problem solving that has taken place world wide over the past more than 15 years. Agricultural research approaches have evolved from somewhat tentative uses of ethnoscience in agricultural development (Brokensha et al 1980) to the emergence of partnerships among biophysical and social scientists in practical problem solving (Rhoades 1984, Richards 1985, Green 1986), and to currently relatively "mainstream" applications of indigenous knowledge in development (Warren et al 1995). At the same time, learning about farmers' technical knowledge has moved from ethnoscience to "rapid rural appraisal" (Carruthers and Chambers 1981, KKU 1987), "diagnostic" approaches (Raintree 1987, Fujisaka 1991), and more recently to "participatory rural appraisal" (IED Participatory Learning and Action Notes 1995). Similarly, farmer participatory problem-solving has become substantially more commonplace as "farmer-back-to-farmer" (Rhoades 1982) and "farmer first" (Chambers et al 1989) have matured into the facilitation of farmer experiments (Haverkort et al
Journalfor Farming Systems Research-Extension

1991, Fujisaka et al 1994) and even "beyond farmer first" which placed greater emphasis on "rural people's knowledge" (Scoones and Thompson 1994).
Our work necessarily incorporates farmers' technical knowledge in the design of participatory research. As will be discussed further,. "necessarily" because of the inherent riskiness and poverty of the agroecosystem combined with the relative wealth of farmers' problem-solving approaches.
Some 10 national and two CIAT researchers first conducted informal structured interviews in four communities in the area northeast of Nazret in late 1995. Farmers provided descriptions of their lands and soils, respective crop and cultivar choices, crop management practices (including allocations of resources), crop-related problems, farmer solutions, and farmer decisions in the face of risk and uncertainty.
Based largely on farmer interest, researchers next agreed upon a single community, Merebe Mermessa, for farmer participatory research, and then worked with small groups from the village using participatory diagnostic techniques. Among others, farmers provided a bar graph of rainfall by month; assessed the changing importance of their crops by assigning different numbers of counters (beans) from a total of 100 to each of eight crops using 1960, 1980 and 1995 as reference years; and estimated relative labor demand by assigning counters to crops by month giving the proportion of total management labor required per crop by month. Farmers also assigned counters by month to indicate incidence of malaria, availability of food, and availability of animal feed and fodder over the course of a year. Finally, farmers and researchers worked together to develop a decision model for planting, and to identify and discuss systems problems, farmers' solutions, and needed research.
Although the CIAT researchers emphasized the need to have women researchers and to interview women farmers, no women researchers participated and few women farmers were interviewed. Male farmers controlled participation in the appraisal activities such that women farmers were not represented.
Soils and lands
Farmers described their soils in terms of colors and texture and land types largely in terms of landscape position. Colors included black (tekur afar), and less fertile red (kyafar or gombore) and white (necht). Soils, mostly loams,
Vol 6, No.3, 1996

included shakite with a coarse pumice layer which occurred at varying depths and thicknesses, and served to conserve underlying soil moisture; black clay loam koticha; and gombore with low moisture holding capacity and most subject to soil moisture deficits.
More productive soils were the lower-lying (golba) shakite or koticha soils where farmers planted maize and sorghum. Dark soils in upper areas were also relatively productive and were sown to tef, beans, barley, and wheat. Lower (golba bota) areas received rainfall runoff and were planted earlier than upper (aoretti, goba or kafyale bota). Some low areas with black soils were frequently waterlogged and could only be used for grazing.
Farmers applied FYM largely to maize fields near the homestead. FYM was not applied to tef to avoid lodging from higher fertility and introduction of weeds. Many farmers used inorganic fertilizers, normally up to 50 kg DAP and and 50 kg urea per ha, for maize and occasionally for tef.
Crops and livestock
Farmers managed a range of crops, the most important of which are discussed in the farmers' order of importance (Table 1).
Table 1. Farmer ranking of crop importance over time, Nazreth, Ethiopia, 1995
Crop '60 '80 '95 Reason for changes
Tef 10 34 33 i Use of fertilizer & herbicides
Maize 26 23 24 (dieSome reduction due to more tef
Beans 19 17 22 ()
Wheat 0 10 11 Use of fertilizer & herbicides
Barley 15 10 5 fd.High seed rate, low milling recovery ,
Sorghtl 12 2 1 ( Birds, < area means greater damage
Peas 10 2 1 < Aphids
sLuentils 8 2 1 e Insect pests
c Farmers assigned a total of 100 counters (beans) to different crops for each ofthe three years
Tef, the main crop, has increased in importance over time due to high market demand and increased productivity due to use of herbicides and inorganic fertilizers. Tef was sown in late June through early August and harvested from early October through November. The crop was produced for consumption (as injera, a fermented bread-pancake) and sale. Straw was used as cattle fodder. Tef was sown on upper rather than lower areas with heavier, more fertile soils to avoid lodging and to allocate such soils for maize. Tef suffered from few insects or diseases, but required substantial labor for weed control. Farmers tilled fields up to six times and handweeded as many as three times (with Setaria sp a major pest). Farmers also used high seed rates to
Journalfor Farming Systems Research-Elxtension

obtain weed-competitive plant stands. Cultivars included longer duration white seeded types, shorter-duration and more drought tolerant reds, and mixtures of red and white. White tef was preferred for consumption. Shattering was a reported problem (confimed by flocks of pigeons gleaning grain from the ground at harvest). Reported tef yields were 0.4-1.4 t/ha.
Maize followed tef closely in importance and was also consumed (also as injera) or sold. Cultivars included the short duration Katumani sown in June, medium duration Lemat sown in early to mid-May, and the long-duration Israel sown in mid-April. Each was harvested in late November, with longer duration cultivars having higher yields if the season was without substantial soil moisture deficits or other problems. Farmers reported typical yields of 0.8 t/ha for short, 1.0 t/ha for medium, and 1.6 t/ha for long-duration cultivars. If plant stands of long or medium duration maize cultivars were poor, farmers re-plowed and re-sowed a short-duration maize or other short-duration crops, or undersowed short-duration crops such as bean leading to what then emerged as intercropped fields. Farmers cultivated maize fields with animal drawn plows (ards or maresha) at about 30 days after sowing (DAS) as a weed control measure in a practice called shilshalo. Besides water stress and weeds, other problems in maize were low soil fertility and poor emergence due to soil crusting, armyworm, stalkborer and cutworm.
Beans (largely small white seed cultivars such as Mexican 142) were the third most important crop, produced largely for market with the residues stored and used for fodder. Beans were sown in July and harvested in late September and October. Weeds were well managed without weeding through sowing in a clean seedbed at high seed rates on fields previously sown to tef. Typical yields appeared to be in the 0.5-0.9 t/ha range. Farmers recognized low soil fertility, leaf feeding beetles and bruchids as crop problems.
Farmers ranked wheat fourth, with importance increasing since the 1960s due to higher productivity associated with more use of inorganic fertilizers and herbicides. Farmers estimated wheat yields to be 0.7 to perhaps 1.5 t/ha.
Although Ethiopia is a center of barley diversity, farmers reported decreases in the importance of the cereal due to the high seed rate needed and low milling recovery. Yields were approximately 1.0 t/ha.
Sorghum, peas, and lentils each decreased in importance over the past 30 years. Sorghum was important for grain, fodder, building material, and fuel, but suffered from substantial bird losses. Farmers sowed bitter varieties as a way to reduce losses, but decreasing areas have placed more bird pressure on remaining fields. The crop occupied roughly the same agricultural niche as maize. Peas were a cash crop with high demand, but declined due to serious pest problems, especially aphids. Lentils were sown after failure of long-duration crops, but suffered losses from insect pests.
Farmers' assessment of crop importance for 1960, 1980, and 1995 (Table 1) indicates the changes mentioned above. Overall, maize shifted from first to
VoL 6, No.3, 1996

* 6 FuIisAXA, WoRTmANN, AND ADAmAssu
second in order of importance; tef shifted from a minor crop to the most important; and beans moved from second in importance to about equal importance as maize. Both tef and wheat increased in importance with increased productivity due to greater use of herbicides and inorganic fertilizers. Barley, sorghum, peas, and lentils have declined in importance because of unsolved production problems.
Farmers had small herds of 4-5 cattle, including draft oxen, plus a few donkeys and goats. Livestock were grazed in communal grazing lands, field margins, and in harvested fields; and were fed crop residues harvested and stored for the purpose. Farmers have reduced numbers of chickens due to diseases. Donkeys were important for transport, especially of domestic water supplies.
Farmers' perception of rainfall, labor demand, malaria incidence, and food and livestock feed availability
Farmers provided various types of data related to yearly cycles. Farmers' reported rainfall approximated the 10-year recorded pattern with the exception that September and October were perceived to be much drier than recorded data indicate (Figure 1). Farmers' rainfall estimates predicted their reported practices in that they would be expected to start land preparation in late February and March, sow lower areas as soil moisture became sufficient in April through June; and sow any remaining areas in July-August. Farmers' underestimation of September rains may reflect their perception of the riskiness and high negative consequences of drought in that month.
A farmer group-estimate of labor required per crop per month indicated low demands from January through June, a moderate high in July and peak requirements October through December, with November the busiest (Table 2). Another group estimated labor demand per month (independent of crop): early demand periods were April-May and June-July, and the heaviest requirement again in November (Table 2). Both estimates may be biased in that womens' weeding labor from late May through mid-August appeared to be under-estimated.
Malaria was identified by farmers as a major problem that reportedly increased in May as the rains, started and agricultural labor demands increased, peaked in September, and fortunately decreased substantially by November when labor was required for crop harvest (Figure 2). The same figure shows that farmers reported that their food stocks declined from January through May, remained very low until August or September when the first crops were harvested and increased to highest levels in December after harvest and processing.
Journalfor Farming Systems Research-Extension

Table 2. Farmers' labor allocations to different crops over the year
Tef,* 1 5 64 3 1 1 0 0 1 1 1 25
Maize 0 0 4 500 1 1 1 0 3 1 16
Beans 1 13 00 1 1 1 1 1 2 0 10
Wheat 1 14 20 1 0 1 1 1 3 1 14
Barley 1 1 500 1 1 1 1 1 3 1 14
Sorghum 0342 0 1 1 1 1 1 3 2 17
Peas 00300 1 0 0 0 1 1 0 5
TOTAL 4 1 2 1 3 5 3 3 4 4 1 7 100
iiiii{ 1 i 8 4. 4 .iiiiii
TOTAL 29 1 g 5 2 1 1 14 3 116 100
i.3 1 6
* Farmers assigned zero to 100 beans to each cell. Values were then converted
such that labor totals equalled 100%
** Another group of farmers drew a bar graph of labor demand over the year,
values were converted to percent per month
The decline of malaria coincided with both restoration of family food stocks and the end of the wet season.
Farmers indicated that animal feed stocks followed a pattern very similar to human food stocks, except that relatively more animal feed became available in September with harvest of the earliest crops; and stocks already started to decline in December when human food stocks were highest. Farmers used crop residues of tef, maize, barley, bean and sorghum as fodder; purchased oil cakes if cash was available; and practiced open grazing after September to maintain animals. Unfortunately, animal feed was least available at the time of peak demand for animal power for land preparation in March-July.
Farmers' solutions
Investments in any future research should provide innovations which can be adopted given farmers' circumstances and which provide systems improvements via problem solving. As such, farmers' solutions must be well understood if research-based innovations are to progress beyond what farmers already do.
Farmers dealt with their major constraint, erratic and often inadequate rainfall, by having responses to whatever conditions emerged over the course of each crop season. Farmers began tillage with the earliest rains, and practiced a stale seedbed technique when early rains were intermittent. Farmers could sow a long duration maize in lower areas with early April rains If and as rains
VoL 6, No.3, 1996

Figure 1. Farmers' v measured rainfall, Nazret, Ethiopia
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
U Avg monthly rainfall Farmersfeported
Figure 2. Farmers' reported malaria incidence, food and animal feed availability and labor demand over the year
20 . .
* -- ....
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Journal for Farming Systems Research-Extension

continued, farmers planted intermediate aid then shorter duration crops and cultivars while expanding area sown from lower to upper fields. If rains were delayed, farmers did not sow the longer duration maize or sorghum cultivars, waiting instead to sow shorter-duration crops. Farmers could resow or undersow shorter duration crops if rains stopped after planting. Observed maize plus bean, barley, tef, or wheat intercrops, for example, were results of sowing into a maize crop which had not established well.
The farmer decision systems were related to a weak bimodal rainfall pattern which resulted in a long but unreliable growing season for cereals which presented farmers with the many different scenarios to which they had to respond. The sophistication of such farmer decision systems made it difficult to identify research aimed at developing innovations that would increase system productivity and enhance sustainability. Likewise, however, an opportunity was available to involve farmers who were knowledgeable about their situation to participate in systems-improvement research.
Farmer-identified problems and solutions Farmers identified problems in addition to those named above. Although they started with "shopping lists" and did not reach a consensus regarding prioritization, farmers eventually developed a "short list" of low soil moisture, birds, aphids, low soil fertility, weeds, lack of animal feed, and a range of pests. Farmers and researchers then worked together to propose several participatory research activities. Although also developed, numbers of farmers conducting each experiment, plot layout and size, and data to be collected are not presented here, although open-ended farmer evaluations will be necessary in all cases as the experiments are established and as results become available.
1. Livestock feed. Crop establishment was often poor because of poor land preparation due to insufficient draft power. Lack of draft was in part due to lack of livestock feed in May through July. Farmers were aware that feeding oxen small amounts of high quality supplements such as cotton seed cake enabled animals to be in better shape for plowing. After rejecting forage undersowing, oversowing and "stock exclusion" as ways of improving livestock nutrition, farmers decided to test several herbaceous forages grown as sole crops and fodder trees grown in their household compounds and as hedges around compounds. Herbaceous forages are expected to contribute to improved tillage operations, while fodder trees may help to improve tillage from April to July.
2. Low soil moisture. Low soil moisture was a cause of poor crop establishment and low productivity. As discussed, farmers adjusted crops and sowing dates according to soil moisture. Farmers reported that shilshalo improved infiltration. Farmers will test a three tined animal-drawn tillage implement developed by researchers at Melkassa for seedbed preparation. The
VoL 6, No.3, 1996

implement is expected to require less draft power, result in less evaporation from disturbed soil, and improve tillage for all rainfall scenarios.
3. Weeds. Farmers identified weeds as a major cause of yield loss and/or of high labor requirements for crop production. Farmers repeatedly tilled soils prior to sowing thereby practicing a stale seedbed technique, hand weeded, used shilshalo in maize and sorghum, sowed beans in fields previously in tef (where investments in hand weeding were high), did not use FYM in tef, and used high seed rates in response to weeds. Farmers will test herbicide (2,4-D) application after shilshalo in sorghum. Improved weed control would also improve moisture use by the crop.
4. Cutworms. Maize was the most important crop in the area; and cutworms were a cause of poor maize establishment. Farmers dealt with the problem by resowing or undersowing beans and lentils in poor maize stands. Farmers will conduct trials to verify the use of carbosulfon wettable powder as a maize seed dressing for cutworm control.
5. Maize cultivars. Farmers and researchers thought that some of the recently bred maize varieties may be superior to local cultivars against some of the production constraints. Researchers and farmers decided to test medium-maturity maize varieties under farmers' conditions. A few promising varieties are to be identified for testing to improve maize production.
6. Bean cultivars. New varieties were identified as a possible solution to some of the problems affecting beans. Researchers and farmers thought that varieties exist that would perform better under farmers' conditions than those currently sown by farmers. Farmers will test a few varieties so identified.
7. Soil nutrients. Maize yields were constrained by low available N and P. Farmers recognized the need for improved nutrient supply, but needed to get high returns to their investments in fertilizers; and they suspected that rapid vegetative growth was related to susceptibility to drought. On shakite and koticha soils, farmers will test 0 fertilizer, 60 kg/ha DAP applied at sowing, 60 kg/ha DAP applied at shilshalo (about 30 DAS), 30 t/ha FYM applied at final land preparation, 1.5 T/ha FYM applied at planting time and 30 kg/ha DAP applied at shilshalo.
8. Birds in sorghum. Farmers have reduced area sown to sorghum due to bird damage. Farmers now grow some bitter varieties less favored by birds (and humans) and practiced bird scaring. Researchers and farmers thought that varieties exist that would suffer less bird damage than those being sown. Farmers would test a few early-maturity varieties.
Journalfor Farming Systems Research-Extension

9. Field pea declined in importance as farmers failed to adequately manage aphids. Farmers wanted to test new genotypes that might be resistant or tolerant to aphid.
Most of the proposed research topics were aimed at intensification of production on the less drought-prone and more productive shakite and golba soils. Apparently farmers recognize that they are likely to get better returns with these soils to the additional capital or family labor investment associated with some of the proposed innovations.
Although the priority problems, except for the insect pests, were very similar, the research needs identified through this participatory approach differed from those identified earlier by researchers using more conventional farming systems research methods (Regassa et al 1992). Proposed experiments which were similar for the two approaches include bird resistant sorghum variety testing, evaluation of fertilizer use on maize, and tests of bean varieties. Fourteen other areas of experimentation were identified using the FSR approach, while the participatory approach identified six areas not previously identified. IAR researchers agreed that the participatory research approach should first be implemented on a limited scale to determine its relative effectiveness in comparison to a more conventional farming systems research approach.
The question underlying this paper, however, is whether or not research could provide improvements over current farmer practices given, on the one hand, an agroecological environment featuring low ard uncertain rainfall, relatively poor soils, and low levels of farmer resources, and, on the other, complex technical knowledge employed by farmers to solve problems. Similar difficult environments combined with often elegant farmer solutions have been encountered in, among others, rainfed lowland rice in India, Indonesia, and Burma (Fujisaka, forthcoming, Fujisaka 1990, Fujisaka et al 1993); in upland agriculture in Mexico and Central America (Wilken 1987); in the famine-prone Horn of Africa (Walker 1995); and with subsistance Morn farmers in southern Sudan (Sharland 1995). In each of these cases farmers' own advances in problem solving: a) made understanding farmer knowledge and practices a pre-requisite to any further problem-solving; b) made it difficult to identify potential innovations that either were not superior to, or did not, in various ways, run counter to farmers' solutions; and c) made farmer participation very necessary in achieving any additional improvements.
In our case example, farmers broadcast seed, but still used animal cultivation after emergence in maize and sorghum; and therefore probably
VoL 6, No.3, 1996

12 FunisAKA, WoRTmANN, AND ADAmAssu
would gain little from additional investment in row planting--an area in which substantial on-station research investments have been made. Fanners used higher seed rates than has been recommended by researchers; although the benefits given possibilities of low germination and emergence combined with benefits in weed competitiveness makes this another "rational" fanner practice (although research recommendations were modified for beans after farmers' practice was "validated"). Farmers adjusted sowing dates in response to soil moisture availability. Although researchers have conducted many fertilizer and herbicide trials, farmers had substantial experience in the use of their limited amounts of organic and inorganic fertilizers and herbicides. Finally, farmers managed numerous crops and cultivars according to specific requirements and could probably benefit little from cropping pattern trials.
Will the participatory research program developed jointly by researchers and farmers provide new problem-solving answers that farmers do not currently possess? For the three-tined cultivator to be an improvement over farmers' implements it must function better at no increase in the draft power required. Experiments with herbicides versus handweeding following shilshalo of sorghum will provide new information, although farmers' current weed management practices for sorghum appear quite efficient. Given women's (observed) high labor inputs into weeding, however, trials with herbicides and haudweeding would require their evaluations.
Improving livestock feed supplies may require the introduction of new plant species and associated management practices. Field-sown forages require that farmers reduce area of food crop production to produce fodder to feed their animals, a scenario that appears unlikely in an area where people complain of land scarcity. Fodder trees may grow in the area, but would probably require community action to protect seedlings if plantings are to be extended beyond the household compounds. Seed treatments for maize are easy to test and may provide significant benefits.
It may be difficult to improve systems through improved fertilizer use. Farmers were aware of the benefits of fertilizer use and appeared experienced in the application of affordable amounts. Even the researchers' standby of new varieties may be problematic given that farmers already employed a range of crops and cultivars. Farmers, in addition to planting less sorghum because of birds, had already turned to more bitter sorghum varieties to reduce losses to birds in spite of these varieties also being less palatable.
We hope that research will provide problem-solving innovations, albeit these may represent incremental rather than quantum leaps forward. Where a quantum leap could be made, however, is in effectiveness of systems research as partnerships formed among farmers and teams of researchers becomes more effective in addressing farmers' problems. As both fanners and researchers improve their skills, participatory approaches to fanning systems research should become more effective and efficient. New problems and opportunities
Journalfor Farming Systems Research-Extension

will also arise as demographic pressures push for greater intensification (Netting 1993). Such a research partnership will enhance the development of more intensive, but sustainable systems; and that the modest investments needed to conduct participatory research (which may be substantially less than on-station research) may be especially appropriate for cases similar to what we have described for Ethiopia.
Brokensha, D, DM Warren, and O Werner, eds. 1980. Indigenous Knowledge Systems and
Development. Washington, DC: University Press of America.
Carruthers, I and R Chambers. 1981. Special issue on rapid rural appraisal. Agricultural
Administration 8:6.
Chambers, R, A Pacey, and LA Thrupp, eds. 1989. Farmer First: Farmer Innovation and Agricultural
Rersearch. London: Intermmediate Technnology Publications.
Franzel, S. 1992. Impact, institutionalization and methodology: research with farmers in Ethiopia. In:
Research with Farmers: Lessons from Ethiopia. S Franzel and H van Houton, eds. CAB International for the Institute of Agricultural Research, Ethiopia. Melksham: Redwood Press. pp
Fujisaka, S. forthcoming. Research: help or hindrance to good farmers in high risk systems?
Agricultural Systems.
Fujisaka, S. 1991. A set of farmer-based methods for setting post 'green revolution' rice research
priorities. Agricultural Systems 36:191-206.
Fujisaka, S. 1990. Rainfed lowland rice: building research on farmer practice and technical knowledge.
AgricultureEcosystems, and Environment 33: 57-74.
Fujisaka, S, E Jayson, and A Dapusala. 1994. Trees, grasses, and weeds: species choices in
farmer-developed contour hedgerows. Agricultural Systems 25: 13-22.
Fujisaka, S, K Moody, and K Ingram. 1993. A descriptive study of farming practices for dry seeded
rainfed lowland rice in India, Indonesia, and Myanmar. Agriculture, Ecosystems, and
Environment 45: 115-128.
Green, EC, ed. 1986. Practicing Development Anthropology. Boulder: Westview. Haverkort, B, J van der Kamp, and A Waters-Bayer, eds. 1991. Joining Farmers' Experiments:
Experiences in Participatory Technology Development London: Intermediate Technology
IIED. 1995. Notes on participatory learning and action. Sustainable Agricultural Programme,
International Institute for Environment and Development
Khon Kaen University (KKU). 1987. Proceedings of the 1985 International Conference on Rapid
RuralAppriasal. Rural Systems and Research and Farming Systems Research Projects, Khon
Kaen, Thailand.
Mulatu, Tilahun, Teshome Regassa, and Aleligne Kefyalew. 1992. Farming systems of the Nazret area.
In: Research with Farmers: Lessons from Ethiopia. S Franzel and H van Houton, eds. CAB International for the Institute of Agricultural Research, Ethiopia. Melksham: Redwood Press. pp
Netting, R McC. 1993. Smallholders, Householders: Farm Families and the Ecology of Intensive, Sustainable Agriculture. Stanford: Stanford University Press.
Raintree, J. 1987. The state of the art of agroforestry diagnosis and design. Agroforestry Systems
Regassa, Teshome, Tilahun Mulatu, and Roger Kirkby. 1992. Developing technologies for small-scale
farmers: on-farm research in the Nazret area. In: Research with Farmers: Lessons from Ethiopia.
S Franzel and H van Houton, eds. CAB International for the Institute of Agricultural Research,
Ethiopia. Melksham: Redwood Press. pp 126-142.
VoL 6, No.3, 1996

Richards, P. 1985. Indigenous Agricultural Revolution: Ecology and Food Production in West
Africa. Boulder: Westview.
Rhoades, R. 1984. Breaking New Ground: Agricultural Anthropology. Lima: International Potato
Center (CIP).
Rhoades, R, and R Booth. 1982. Farmer-back-to-farmer: a model for generating acceptable agricultural
technology. Working paper. Lima: CIP.
Scoones, I and J Thompson. 1994. Beyond Farmer First: Rural People's Knowledge, Agricultural
Research, and Extension Practice. London: Intermediate Technology Publications.
Sharland, RW. 1995. Using indigenous knowledge in a subsistence society in Sudan. In: The Cultural
Dimension ofDevelopment: Indigenous Knowledge Systems. Warren, DM, LJ Slikkerveer, and D
Brokensha, eds. London: Intermediate Technology Publications.
Walker, P. 1995. Indigenous knowledge and famine relief in the horn of Africa. In: The Cultural
Dimension ofDevelopment: Indigenous Knowledge Systems. Warren, DM, LJ Slikkerveer, and D
Brokensha, eds. London: Intermediate Technology Publications.
Warren, DM, LJ Slikkerveer, and D Brokensha, eds. 1995. The Cultural Dimension ofDevelopment:
Indigenous Knowledge Systems. London: Intermediate Technology Publications.
Wilken, GC. 1987. Good Farmers: Traditional Resource Management in Mexico and Central
America. Berkeley: University of California Press.
Journalfor Farming Systems Research-Extension

G. Weber'
Attending the diversity of farming systems in a given mandate area is a common challenge to technology
generation, testing and dissemination. Frequently, predominating systems are classified into domains using a wide range of structural and functional criteria. The utility of these classifications are mostly related to the purpose for which they were designed and have limited applicability beyond the original study area and discipline. Additionally, they may hide more than what they reveal. It is argued that the adoption of an evolutionary perspective permits a better insight for sustainable agricultural development and may provide a much needed general framework for systemsoriented research and development. Changes in demand for agricultural products are determined by market access and population increase, both of which are ultimate forces driving the evolution of agricultural system. These in turn impact at the supply side on the evolution of agricultural productivity and production practices. Resulting individual systems can accordingly be identified as stages along general evolutionary pathways.- Sucdh an evolutionary perspective offers the opportunity to integrate site-specific systems research and development with strategic eco-regional or discipline-oriented
research and provides a common framework for action.
PASOLAC, AP' 6024, Managua, Nicaragua. Former address: International Institute of Tropical Agriculture, IITA, PMB 5320, Ibadan, Nigeria.
VoL 6, No.2Z 1996

The biophysical, socioeconomic and institutional conditions of agricultural production vary widely from place to place and over periods of time. In the process of adapting cropping patterns, livestock management and farming practices to the conditions of the location and the aims of the farmer, distinct types of farm organization have developed (Ruthenberg, 1971).
A large body of literature has accumulated where the social, environmental and technological dimensions of farming have been described. Depending on the discipline and the research objectives, different approaches, modes of analysis, problem foci and geographical scales have been utilized. Thus, the term 'farming system research' encompasses a wide range of studies, experiences and views. Most studies are based on comparative system analysis with the objective to classify systems into a few representative units. The criteria for comparative system analysis and classification vary widely and may be based on structural, functional or evolutionary phenomena or any mix of these. Production-oriented agronomists and economists, for example, have provided location- or enterprise-specific studies with the major objective to define research and recommendation domains (Byerlee and Collinson, 1980; Norman et al., 1982). Livestock-scientists have described livestock systems and have analyzed changes in carrying capacity of the natural resource base for livestock production over time and as a function of technology (Jahnike, 1982). Geographers have focused attention on the geographical distribution of agricultural systems as part of the overall economy (Morgan and Munton, 1971), while historians have described the stages of past agricultural development (Andreae, 1981). These different approaches have contributed to a more holistic view of farming systems although a general, integrating framework is lacking.
A generalized concept of comparative analysis and classification of farming systems should be applicable at different geographical scales with the practical aim at enhancing the integration of strategic, applied and adaptive research with development efforts across disciplinary and, as far as possible, across geographical boundaries. It should be compatible with past approaches and should be simple to apply. In this paper, a comparison is attempted between an evolutionary approach to farming systems characterization with a number of widely used concepts which rely mostly on structural and functional criteria of classification. Implications for research and development are shortly discussed.
Journal for Farning Systems Research-Extension

Past approaches
Systems research aims at a holistic view, encompassing most or all parts of the total. The boundaries of the total may be delineated geographically at the farm, zonal, regional or supraregional. level or they may be set conceptually at a more or less specifically defined enterprise or management practice, e.g. maize-based systems or mechanized farming. The complexity of the system increases exponentially and only a few major components can be analyzed as the geographical scale increases. Therefore, a geographical or horizontal view of all major components of a system at the farm or local level, e.g. all cropping patterns and practices used in a cereals-producing area, may at the larger scale be switched to the analysis of a vertical or conceptual slice of the system, such as an assessment of the integration of maize production into the overall farmeconomy of production, marketing and consumption.
The vertical description is often depicted as a structural hierarchy of subsystems with functional relationships in material and information flow among them. The horizontal analysis concentrates on a comparative analysis of systems at a defined level of aggregation and a grouping of these into higher levels of aggregation. Historically, mainly structural criteria such as land use, predomiinant crops or farm size have been used for comparative analysis and classification. More recently, functional criteria such as market orientation, land use intensity or crop-livestock integration have increasingly been included to better describe the interaction among components. Further analysis of farmers' decision criteria has contributed during the past two decades to a better understanding of socioeconomic system dynamics while research on physical resource processes, ecological succession and pest dynamics have deepened the agronomic understanding of interdependencies.
Towards an evolutionary perspective
Functional analyses go beyond structural descriptions. They question not only 'What' and 'How' a system is organized but expand the question of 'How' systems function and ask 'Why' they have a given structure. This functional analysis, however, was initially limited to the proximate causes of interrelationships. Further understanding of the ultimate causes and driving forces of system dynamics over time requires an evolutionary perspective. For instance, the structural description has led to a differentiation of shifting
VoL 6, No. 2, 1996

cultivation, semi-permanent, stationary and permanent cultivation as organizational units of land use. A functional analysis reveals the transition from one stage to the next as land use intensifies and may identify associated processes such as fertility decline and changes in natural vegetation. An evolutionary approach expands this analysis and indicates that the demand for agricultural products from an increasing population or from better market access are the ultimate forces which drive the system along a pathway of intensification and that the supply of agricultural products and associated production practices change along the evolutionary pathway. Land-use intensifies along this pathway and the farming system shifts gradually from one organizational stage to the next. Depending on the demand forces, however, different functional relationships may develop. The evolutionary analysis of farming systems builds on accumulated knowledge about the structure and function of fanning systems while the latter two find a general framework for explanation in system evolution.
Examples for different classification systems are summarized in Table 1. Land use intensity and land use type are the most commonly applied agronomic criteria for classification while the utilization of modem production technologies and the degree of market-orientation are widely used economic criteria. A wide range of additional criteria have been suggested for multiplecriteria classifications. Underlying forces of long-term system change have been recognized by many authors since Malthus indicated a close relationship between population increase and the limited capacity of the resource base to satisfy demand. The Malthusian deterministic view has in the meantime been replaced by a more dynamic view of the supply side after the flexibility and elasticity of production have been revealed. Boserup (1965), for instance, indicated that population growth is an evolutionary demand force in initiating a transition from one land use system to another, more intensive one, while Pingali et al. (1987) added market access. These authors identified population growth and access to markets as the main determinants of agricultural intensification in Sub-saharan Africa, both leading to a transition from forest fallow and bush fallow systems of cultivation to annual and multiple-crop systems. In a similar approach, Turner and Brush (1987) use a trajectory of production types from exclusive household consumption or subsistence to full market-oriented commodity production as a criterium of classification. They add land use intensity and the degree of technological advancement as a second and third trajectory and locate 12 different farming systems along these trajectories. Boserup (1965), Pingali et al. (1987), and Turner and Brush (1987) have most clearly moved towards an evolutionary analysis and classification of farming systems. More recently, an evolutionary classification has been proposed by which systems are first classified into agricultural systems according to the underlying forces of demand, i.e. market access and population increase.
Journalfor Fanning Systems Research-Extension

Table 1: Classification of fannrming systems into structurally or functionally similar groups.
Classification system and criteria used Groups formed
S Land use intensity (R-value) Shifting cultivation
(after Nye and Greenland, 1961) Semi-permanent cultivation
Stationary cultivation Permanent cultivation
2 Land use intensity, crops rotated Shifting cultivation
(after Ruthenberg and Andreae, 1982) Semipermanent fields
Ley system
Permanent cultivation Perennial crops Grazing system
3 Land use intensity Gathering
(R-value), cropping system Forest fallow
(after Pingali et al., 1987) Bush fallow
Short fallow Annual cultivation Multiple cropping
4 Stationariness of herds Nomadism
(after Dittmar, 1954, in Ruthenberg, 1971) Semi-nomadism Transhumance Partial nomadism Stationary husbandry
5 Degree of commercialization Subsistence
(after Ruthenberg, 1971) Partly commercialized
Semi-commercialized Highly commercialized
6 Degree of market orientation, factor Subsistence with
productivity shifting cultivation
(after Doppler, 1991) stationary cropping
Subsistence/market-oriented with
- livestock integration
- water management
- intensive labour use Market-oriented with
- high intensity
- extensive
7 Degree of mechanization Pre-technical
(after Ruthenberg, 1971) Manual hoe farming
Semi-mechanized Mechanized
8 Land use intensity, market-orientation, Paleotechnic and consumption-oriented
technological complexity systems
(after Turner and Brush, 1987) Mixed-technic and production systems
Neotechnic and commodity-oriented systems
VoL 6, No.2, 1996

Subsequently, further divisi6ns into resource domains and farming domains are suggested based on the concomitant evolutionary changes in resource characteristics and the utilization of resource management and fanning practices (Smith and Weber, 1994; Weber et al., 1996). Different from earlier classifications, Turner and Brush (1987) and Smith and Weber (1994) suggest that the underlying forces of market access and population increase may be independent and thus may lead on different evolutionary pathways to distinct farming systems. These can roughly be assigned to four basic evolutionary phases: market-driven expansion or population-driven expansion at an early stage and market-driven intensification or population-driven intensification at a later stage of land-use intensity (Fig. 1).
Figure 1. Schematic representation of the evolution of agricultural systems from a market-driven (ME) or population-driven (PE) expansion phase to a market-driven (MI) or population-driven (P) intensification phase (after Weber et aL, 1996).
With increase in market access:
+ market demand for farm products
" availability of cash
" availability of purchased Inputs
" labour market
Market-driven < _,, ,>_ Populabon-driven
Intensification phase
w1 P1
Agricultural systems
With increase in population density: land use Intensity
" utiIation of marginal land
Early MI Early P + labour availability
- land availability
" cropping intensity
" crop husbandry efforts (weeding,.. livestock confinement
xp- phJournalfor Farming Systems Research-Extension

Past approaches in an evolutionary perspective
A comparison of classification systems summarized in Table 1 with an evolutionary approach of classification indicates that most commonly used classifications can approximately be located along the basic pattern of evolutionary change indicated in Figure 1. The descriptive differentiation of land use systems by Nye and Greenland (1961), Ruthenberg and Andreae (1982) and Pingali et al. (1987) are stages along the pathway of increasing land use intensity (Fig. 2). These studies, however, do not differentiate clearly between systems under population-driven intensification from those under market-driven intensification, although Pingali et al. (1987) have elaborated on the effect of these underlying evolutionary forces on agricultural mechanization. Additional criteria such as cropping system (Fig. 2-2), used by Ruthenberg and Andreae (1982) in order to separate the peculiarities of grazing systems and plantations from annual cropping systems,and multiple cropping (Fig. 2-3) used by Pingali et al. (1987) cannot be located on the twodimensional evolutionary graph as it relates to management rather than land use.
A classification exclusively along a transect of market-orientation has been mentioned by Ruthenberg (1971), describing systems according to their degree of commercialization. This approach, however, does not differentiate degrees or types of land use (Fig. 2-5). Doppler (1991) analyzed the functional relationship between the degree of commercialization, land use and farmers' decision criteria and suggests a first differentiation according to market orientation into subsistence, market-oriented and mixed subsistence/marketoriented agriculture which is followed by a second differentiation according to intensity and specialization. The first level of classification includes evolutionary and functional criteria while the second level is mainly based on functional and structural criteria. The evolution of systems at the second level is therefore difficult to analyze (Fig. 2-7). The classification of farming systems by Turner and Brush (1987) along the three trajectories of 'output intensity', 'production type and consumption' and 'technology type' coincides for the first two criteria with a differentiation according to land use intensity and market access. The third criterium relates to the degree of technological advancement and the replacement of labor by new resource and crop management practices. The three basic systems and the 12 farming systems described by Turner and Brush (1987) can approximately be located according to market-orientation and land-use intensity on the evolutionary graph (Fig. 2-8) indicating that systems under population-driven intensification are not clearly differentiated by their concept.
The comparison of these commonly used classification systems indicates that some criteria do not coincide with the first level of evolutionary
VoL 6, No.2, 1996

Figure 2. Approximate location of six different classification systems (see Table 1) in an evolutionary space of agricultural systems (see Figure 1).
Classification I (after Nye and Greenland, 1961) Classification 2 (after Ruthenberg and Andrea, 1982) Market-driven < > Population-driven Market-driven < > Population-driven
Permanent cultivation Intensilication Permanent cultivation Intensification
P systems
Stationary cultivation Ley farming systems
... .. .. . .. ... .. .. .. .. .. .. .. .. .. .. .. ................. .. .. ...... . . . .. /
Semi-permanent Semi-permanent field ,
cultivation systems e T
... ... ... ... .. ... ... ...... ............ ... ............. .. .=
Shifting hifling cutivation
cultivation fallow systems
Expansion Expansion
Classlication 3 (after Pingall et al., 1987) Classilloation 5 (after Ruthenberg, 1971)
Market-driven < > Population-driven Market-driven<( ) Population-driven
Annual cultivation intensification
Intensification Intnsiicaio
........................................................ Int ns-ic--on---In-ns--c-o
Short fallow systems
Bush-fallow .
Forest faillowE
Forest Expansion Expansion
Classification 7 (after Doppler, 191) Classiiaton 8 (alter Turner and Brush, 1987)
Market-driven < > Population-driven Market-driven < > Population-driven
I ntensification E intensification
CL -6
shdb'15~ E Z
- shifting cult. loehi
Espension onsumpton
Expansion nmpton- Expansion
Journalfor Farming Systems Research-Extension

differentiation into agricultural systems according to market-access and population increase, i.e.evolutionary changes at the demand side of agricultural production. An additional dimension has to be added defining evolutionary changes at the supply of agricultural production along the transition from extensive to intensified systems as a result of increasingly scarce land. Changes in supply and demand both determine together the evolution of farming systems.
The evolution of supply from agricultural production
Supply of agricultural products depends on two steps: first, the recognition, accessibility and management of internal and external resources and second, the efficient conversion of a given resource endowment into useful products. Both steps are clearly interrelated. As land use intensifies, resource and crop management technologies play an increasingly important role in determining ecosystem processes through human intervention.
Evolution of resource management.
The classification of the physical resource base has mostly followed descriptive criteria based on climate, soil or vegetational characteristics (e.g. K6ppen, 1931; FAO-UNESCO, 1974; Walter, 1973). These criteria have been used to describe potential biomass production of natural ecosystems and are widely used in agroecological studies (FAO, 1978). The analysis of major disturbances, in particular in tropical climates has revealed the limitations of such structural classifications without a thorough understanding of functional and evolutionary phenomena. For instance, in most areas of the humid tropics, biomass production is maintained through efficient nutrient cycling rather than through a large pool of plant-available nutrients in the soil. Major disturbances may therefore disrupt the nutrient cycling leading to major shifts in floral composition and long resilient times to regain the original climax. Human intervention, such as large-scale crop-production projects in the humid and subhunid tropics, have often failed where structural classifications were used without a thorough understanding of the function and evolution of the ecosystem. This has stimulated a shift in resource and crop management research towards a focus on system processes, such as soil nutrient dynamics, with the aim to develop more sustainable land use systems for major agroecological zones.
The dynamic interrelationship between the physical resource base and crop yield can, in a simplified way, be described as a function of plant-available water and nutrients within broadly defined climatic zones (Fig. 3). For instance, in the semi-arid tropics, cropping on newly cleared land without external inputs results
VoL 6, No.2, 1996

Figure 3. Schematic representation of changes in land productivity as a function ofplant-available water and nutrient supply in sub-humid savannas.
Pote per ha
Potential output
................ .......
Plant-available water
Path 1: Newly cleared land (A) degrades to (B) as soil organic matter is mineralized; the process is reversable through natural or improved fallow.
Path 2: The organic matter content is increased in compound fields (C) through the application
of manure and crop residues.
Path 3: Application of anorganic fertilizers increase the supply of plant-available nutrients (D),
plant-available water becomes limiting.
Path 4: Application of nutrients and water-harvesting technologies such as tied ridges increase
plant-available water and nutrients to (E).
A Pathwaz
6 Potential output
-- G
Plant-available water
Path 5: Newly cleared land (A) evolves to (F) as anorganic fertilizers are applied; soil organic
matter is mineralized causing a decline in plant-available water.
Path 6: The application of acidifying fertilizers and the decline in organic matter content can
cause soil acidification (G) if no lime is applied.
Path 7: Application of anorganic fertilizers and crop irrigation increase the supply of plant-available nutrients and water to a high yield potential (H).
Journalfor Farming Systems Research-Extension

in a decline of organic matter content through mineralization with a concurrent reduction in plant-available nutrients and water. This shift from (A) to (B) in Figure 3 results in a decline of potential output. The decline can be recovered from (B) to (A) through land fallowing within the resilient capacity of the ecosystem. The introduction of moderate amounts of fertilizer may increase nutrient availability, leading to an increase in potential output from (B) to (D). Further increases in output to (E) will, however, depend on the concurrent use of two resource management technologies: moderate fertilization and the introduction of water-harvesting technologies such as tied ridging. Alternatively, farmers may increase in (A) the pooi of organic matter in the soil through the application of manure and crop residues which in turn can contribute to higher output through efficient nutrient cycling in the system (C). Examples of such processes have, for example, been described by Prudencio (1993), Rugalema et al. (1994), Koudokpon et al. (1994) and have been summarized for the West African Savannas by Pieri (1989).
In areas with good access to external inputs, farmers may use inorganic fertilizers at an earlier stage. The system shifts from (A) to (F) with a certain reduction of organic matter content. Continuous use of inorganic fertilizers, however, may induce soil degradations, eventually soil compaction, erosion, acidification, if no additional resource management technologies will be introduced to sustain the system around (F). The introduction of water management technologies, in particular irrigation, may increase potential output to (H). The stages (A) to (H) describe distinct physical resource conditions and associated potential output along evolutionary pathways of the physical resource base. Pathways 1,5 and 6 lead to a decline of potential output unless appropriate resource management technologies, such as long fallow periods, manuring or application of lime, will be used. Pathways 2, 3, 4 and 7 may increase potential output. However, the adoption of any specific resource management technology will depend to a large degree on resource endowments as described for agricultural systems in Figure 1. Pathway 1 and 2 do not require purchased inputs although pathway 2 requires considerable local material input in terms of manure and crop residues and it is highly labordemanding. Stages (A), (B) and (C) can therefore be described as moderately productive, degraded and highly productive 'resource domains', respectively, in farming systems liely to evolve in areas with poor market access. Pathways 3, 5 and 7 depend on purchased inputs and (F), (G) and (H) are moderately productive, degraded and highly productive resource domains, respectively, in farming systems more likely to evolve in areas with good market access. Pathways 3 and 4 describe transitional stages to the moderately productive resource domains (D) and (E) likely to evolve in farming systems with moderate market access.
Figure 3 greatly oversimplifies the functional interrelationship between physical resource conditions and potential agricultural output. However, an
VoL 6, No. 2, 1996

analysis of the effect of human intervention on the physical resource base along basic evolutionary pathways can still be achieved although the specific technologies used by farmers in different environments may greatly differ. Dynamic simulation models may in the future be able to approximately quantify such processes including the potential sustainable output from a physical resource base. The most important challenge for farming systems research, however, is to relate these process level studies and the differentiation of 'resource domains' to the evolution of the overall farming system.
Such differences in the resource base have, for example, led to previous classifications of farming systems or land use units into structural or organizational groups -of farming according to resource status or resource management. Soil fertility status or management and water management were the most frequently used criteria (Table 2). Groups formed describe phases along an evolutionary process of increasingly intense resource management and concurrent differentiation of the resource base. These changes may evolve under the use of 'modem' as well as 'traditional' technologies, both of which can reach considerable advancement in complexity and in the degree of environmental control (see Klee, 1980 and Turner and Brush, 1987)). Although this general process is driven by the demand for agricultural products, the type of technologies adopted, thus the resulting differentiation of the resource base and the formation of different farming systems, is determined by the degree of resource endowment and access to resource management technologies at the farm level. Both of these in turn will be vastly different in areas under population-driven or market-driven expansion or intensification.
Conversion to agricultural products.
Farmers use production technologies in order to efficiently convert a given resource endowment into useful products for household use or marketing. As the endowment with socioeconomic and biophysical resources change along the evolutionary pathway of agricultural systems and resource domains, preferences for technologies change and different farming systems emerge. Many studies in farming systems research describe and compare the structure of such specific systems, many recent studies reveal also functional interrelationships and analyze farmers' production strategies and decision criteria. In most cases, however, the analysis is limited to the proximate causes of functional interrelationships and is rarely linked to the overall dynamics of the agricultural system. Recent summaries have, for example, been compiled by
Journalfor Farming Systems Research-Ddension

Table 2: Classification of land use in farming systems (examples).
Study area AS' Differentiating resource Groups formed
characteristics and management
Benin: P/M-I Fertility and organic matter 1 Forested zone
Subhumid management 2 Home gardens
savanna oil palm fallow 3 Oil palm
(Koudokpon et manure from small ruminants fallow
al., 1994) residue composting zone
- fertilizer to cash crop (cotton) 4 Cotton/maize rotation zone
Burkina Faso: P/M-I Fertility and organic matter 1 First ring fields
Semiarid management: 2 Second ring
savanna Manure field
(Prudencio, food legume rotations 3 Third ring field
1993) low amounts of fertilizer
- natural fallow
Water conservation:
- ditches
- earth or stone dikes
- mulch
Nigeria: P/M-I Fertility management: 1 Compound
Humid forest bush fallow fields
(Lagemann, mulch 2 Village fields
1977) household refuse 3 Bush fields
- composted residues
Tanzania: M/P-I Fertility management: 1 Homegarden
Humid forest banana residues (kibanja)
(Rugalema et tree litter 2 Fallow plots
al., 1994a, b) animal manure (omusiri)
- fallow
Nigeria: Sub- MI and Fertility management: 1 Fertile uplands
humid savanna PI fertilizer 2 Moderately
(Weber et al., food legume rotation fertile
1996) animal manure uplands
- residue mulching 3 Degraded
- compost uplands
- fallow
'AS Agricultural System under market- (MI) or population-driven (PI) intensification
VoL 6, No.2, 1996

Klee (1980), Turner and Brush (1987), Doppler (1991) and McIntire et al. (1992). An analysis of changes in farming systems along the evolutionary pathways of agricultural systems and resource domains with different market access, land use intensity and land productivity has most recently been attempted by Manyong et al. (1996) and Weber et al. (1996) for West Africa. The resulting classification of farming systems according to predominant crops, for example, into maize-based, sorghum-based, cotton-based, cassava-based and yam-based systems coincides broadly with common classification systems. The evolutionary perspective, however, broadens the view and emphasizes the heterogenous socioeconomic and agroecological settings under which these crop-based systems have evolved and how they might evolve in the near future.
Constraints to production as an evolutionary phenomena.
The previous chapters indicate how changes in productivity can be described within the context of the evolution of farming systems along the transect of increasing market-driven or population-driven expansion or intensification. Further evaluation of system stability and sustainability over time requires additional analysis as functional relationships in particular in biotic systems become increasingly complex as land use intensifies. For instance, many pests depend in their severity on the frequency of encounter between pest and host. Thus, an expansion of a particular crop may promote the increase of frequency-dependent biotic constraints. Others such as many leaf diseases are closely related to the overall physical (e.g. climate) determinants of the environment. Thus, their severity may hardly change as land use intensifies. Details of such functional relationships have been analyzed for many agricultural pests under studies on insect population dynamics, pathogen epidemiology and weed community research (Ruesink, 1976; Kranz and Hau, 1980; Teng and Savary, 1992). Extensive research has been done during the last two decades on developing dynamic simulation models of pests, pathogens and, more recently, their interaction with host plants. An important element of such research has been the identification of factors which drive individual pests to increase or decrease in severity. Much less research has been devoted to analyze the dynamics of biotic communities in agroecosystems and to relate them to the evolution of the overall farming systems. Differences in terminology and in conceptual approaches have additionally impeded the practical integration of pest dynamics into farming systems research.
Journal for Farming Systems Research-Exdension

A widely accepted, generalized approach to the classification of farming systems has not emerged, despite its potential contribution to agricultural research and development. Attempts to classify agricultural or farming systems commonly represent a potpourri of formal and informal criteria and many must be considered typologies which do not systematically apply to every system (Turner and Brush, 1987). The utility of such classifications are closely related to the purpose for which they were designed and it has been claimed that maps based on such classifications may hide more than they reveal about the overall farming system (Morgan and Munton, 1971). In spite of such criticism, classifications have been a useful tool for practical agricultural research and development.
The task of promoting agricultural development has become increasingly complex over the past decades. The initial focus on productivity has been widened to include stability, sustainability and equity (Conway, 1987). Additionally, efforts in research and development have shifted from favorable to more marginal environments and from external-input systems to low external-input systems. Research on the supply side of agricultural production has shown, that the overdeterministic interpretation of the population-land relationship underestimates the flexibility and elasticity of agricultural production through resource and crop management (Boserup, 1965; Pingali et al., 1987). Interest in comparative farming systems analysis has therefore shifted during the past three decades from describing the structure and organization of farming systems to an analysis of functional interrelationships towards evolutionary phenomena. This has recently been strengthened through an increasing emphasis on sustainability which requires the analysis of system dynamics and resilience over time. Similar trends in most disciplines concerned with agricultural research towards the analysis of functional interrelationships have resulted in a better understanding of soil processes, crop growth, pest dynamics and vegetation change. A common framework for system analysis, however, is lacking.
Much can be learned from previous experiences in ecology and biology where a shift from structural and functional comparative analysis and classification towards evolutionary biology, population analysis and community ecology provided a much needed common framework (Mayr, 1984). An evolutionary approach to comparative system analysis and classification may offer a much needed, generalized framework in sustainable agriculture. Structurally and functionally different farming systems, as previously for example described by Nye and Greenland (1961), Ruthenberg (1971), Pingali et al. (1987) and Turner and Brush (1987), can thus be interpreted as being at different phases in an evolutionary space. A considerable proportion of the
VoL 6, No.2, 1996

previously described and classified variability in farming systems can, in a first step, approximately be represented in a simple two-dimensional evolutionary space of agricultural systems, determined by demand for agricultural products according to market-access and population-increase. Additional variability can in a second step be determined by concomitant evolutionary changes at the supply side associated with the utilization of appropriate resource management and conversion-efficient production technologies. Previously used structural and functional classifications based on criteria such as water management, fertility status, land-use or predominant crops can therefore be represented as phases in a multi-dimensional space of the evolution of farming systems. Linking an evolutionary analysis of farming systems with recent advances in understanding insect, pathogen and weed systems facilitates an assessment of system sustainability over time (Smith and Weber, 1994; Weber et al., 1996).
The adoption of an evolutionary approach in sustainable agricultural development will require a common terminology and a general framework for comparative analysis and classification based more on thematic clustering rather than on geographic boundaries. Geo-referenced information systems, which presently attract considerable attention, may contribute to a delineation in the distribution of some major resource characteristics, but they might contribute little to a clustering of resource and fanning domains along evolutionary pathways.
It remains an open question for basic research if and to what extent complex systems evolve gradually or if system change is linked to rather sudden phase changes. Recent research in system modelling and in ecology indicates that there is a considerable degree of evolutionary self-organization in complex systems. Additionally there are indications that, although the dynamics are driven by local processes the final configuration is a reflection of the global structure (Lewin, 1992; Maxwell and Costanza, 1993; Judson, 1994). Thus, phase changes may better explain the evolution of complex biotic systems than gradual selection processes (Lewin, 1992). This has considerable implications for research on farming system dynamics as commonly used deterministic models may be far from explaining the complexity of systems in the real world. In particular the human dimension of change in agricultural systems needs considerably more research before it can be integrated into an evolutionary concept. It is an essential component for evaluating the implications of agricultural change on equity.
Although basic and strategic research at the regional or international level on analyzing system dynamics and the underlying forces will be necessary, there are immediate practical implications for site-specific technology generation, testing and dissemination. The identification of major opportunities and risks in predominating farming systems in a mandate area can be better identified under an evolutionary perspective and can contribute to priority setting in research and development projects (Manyong et al., 1996).
Journalfor Farming Systems Research-Extension

Andreae, B. 1981. Farming, Development and Space. Berlin: De Gruyter. Byerlee, D., and M. Collinson. 1980. Planning Technologies Appropriate to Farmers. Concepts and
Procedures. Mexico, CIMMYT.
Boserup, E. 1965. The Conditions ofAgricultural Growth. Chicago: Aldine. Conway, G.R. 1987. The properties ofagroecosystems. AgriculturalAdministration 24: 95-117. Doppler, W. 1991. Landwirtschaflliche Betriebssysteme in den Tropen und Subtropen. Stuttgart:
Verlag Eugen Ulmer.
FAO-UNESCO. 1974. Soil Map of the World. Vol.1. Legend. Paris, UNESCO. FAO. 1978. Report on the Agro-Ecological Zones Project. Rome, FAO. Jahnke, H.E. 1982. Livestock Production Systems and Development in Tropical Africa. Kiel:
Judson, O.P. 1994. The rise of the individual-based model in ecology. TREE 9: 9-14. Klee, G.A. 1980. World Systems of Traditional Resource Management. London: Winston and Sons. K6ppen, W. 1931. Grundri3 derKlimakunde. Berlin: De Gruyter. Koudokpon, V., J. Brouwers, M.N. Versteeg and A. Budelman. 1994. Priority setting in research for
sustainable land, use: the case of the Adja Plateau, Benin. Agricultural Systems 26: 101-122.
Kranz, J. and B. Hau. 1980. Systems analysis in epidemiology. AnnualReviewPhytopathology 18: 6783.
Lagemann, J. 1977. Traditional African Farming Systems in Eastern Nigeria. Munich: WeltforumVerlag.
Lewin, R. 1992. Complexity. Live at the Edge ofChaos. New York: MacMillan Publ. Comp. Manyong, M.V., J. Smith, G. Weber, S.S. Jagtap and B. Oyewole. 1996. Macro-Characterization of
Agricultural Systems in West-Africa: An overview. Ibadan: IITA, RCMD-Monograph No. 21.
Maxwell, T. and R. Costanza. 1993). An approach to modelling the dynamics of evolutionary selforganization. Ecological Modelling 69: 149-161.
Mayr, E. 1984. Die Entwicklung der biologischen Gedankenwelt. Berlin: Springer-Verlag. McIntire, J., D. Bourzat and P. Pingali. 1992. Crop-Livestock Interaction in Sub-Saharan Africa.
Washington: The World Bank.
Morgan, W.B. 1978. Agriculture in the Third World. A spatial analysis. London: Bell and Hyman. Morgan, W.B. and R.J.C. Munton. 1971. Agricultural Geography. London: Methuen and Co. Norman, D.W., E.B. Simmons and H.H. Hays. 1982. Farming Systems in the Nigerian Savannas.
Boulder: Westview Press.
Nye, P.H. and D.J. Greenland, D.J. 1961. The Soil under Shifting Cultivation. Reading:
Commonwealth Bureau of Soils. Technical Communication No. 51. Pieri, C. 1989. La Fertilit6 des Terres des Savannes. Paris: CIRAD. Pingali, P., Y. Bigot and H.P. Binswanger. 1987. Agricultural Mechanization and the Evolution of
Fanning Systems in Sub-Saharan Africa. Baltimore: John-Hopkins University Press.
Prudencio, C.Y. 1993. Ring management of soils and crops in the West African semi-arid tropics: the
case of the mossi farming system in Burkina Faso. Agriculture, Ecosystems and Environment 47:
Ruesink, W.G. 1976. Status of the systems approach in pest management. Annual Review Entomology
21: 27-44.
Rugalema, G.H,, F.H. Johnsen and J. Rugambisa. 1994. The homegarden agroforestry system of
Bukoba district, North-Western Tanzania. Agroforestry Systems 26: 53-64, 205-214. Ruthenberg, H. 1971. Farming Systems in the Tropics. Oxford: Clarendon Press. Ruthenberg, H. and B. Andreae. 1982. Landwirtschaftliche Betriebssysteme in den Tropen. Pages 125173 in von Blanckenburg, P., ed., Sozialakonomie der lindlichen Entwicklung. Stuttgart: Verlag
Eugen Ulmer.
VoL 6, No.2, 1996

Smith, J. and G. Weber. 1994. Strategic research in heterogenous mandate areas: an example from the
West African savanna. Pages 540-560 in Anderson, J., ed., Agricultural Technology: Policy
Issues for the International Community. Wallington: CAB International.
Teng, P.S. and S. Savary. 1992. Implementing the systems approach in pest management. Agricultural
Systems 40: 237-264.
Turner, B.L. and S.B. Brush. 1987. Comparative Farming Systems. New York: The Guilford Press. Walter, H. 1973. Die Vegetation der Erde in 6kophysiologischer Betrachtung. Stuttgart: Fischer. Weber, G., J. Smith and M.V. Manyong. 1996. System dynamics and the definition of research domains
for the northern Guinea savanna of West Africa. Agriculture, Ecosystems and Environment 57:
Journalfor Farming Systems Research-Extension

John S. Caldwell2 and Archer H. Christian3
A program in vegetable crop sustainability has been
developed from 1993 by a team of researchers in two universities and members of a farmer organization, applying the principles of FSR/E within the U.S. land-grant university context. Selection of cucurbit integrated pest management as the research focus was based on farmer focus group meetings and a farmer survey. Dialog with funding sources has revealed that importance is given to component research methods that conflict with farmer constraints of land area and time, and that farmer monitoring time and extensive farmerresearcher interaction are considered to be too expensive.
The research program has sought to introduce greater biological diversity into American cucurbit production systems. Exploratory trials in 1994 and 1995 have used the university experiment station and one on-farm site. The research program has been divided into three components (strip crop management techniques find effects on ground beetles; effects of flowering plants at the head of crop rows on parasitoids; and insect movement between natural vegetation and the crop field). The research has broadened beyond pestpredator/parasitoid insect interactions, increasing the number of biological interactions examined and adding labor and
This paper was presented at the AFSRE Symposium in Sri Lanka in November of 1996.
2 Associate Professor, Department of Horticulture, Virginia Polytechnic Institute & State University, Blacksburg, VA 24061-0327 USA.
3 Chair, Research Connittee, Virginia Association for Biological Farming, POB 10721, Blacksburg, VA 24062-0721 USA.
VoL 6, No.2, 1996

management interactions, while examining these interactions in smaller subsets. An approach is evolving in which systems interactions identify reductionist research as a basis for farmer assessment of the fit and functioning of new
components in their farming systems.
Farmer participation and system approaches have been goals of FSR/E since its inception in the late 1970s. Concurrently FSR/E has sought to apply the power of scientific methods and knowledge to the solution of agricultural problems of farmers, especially those in less favored environments. FSR/E has also recognized the importance of both public institutions and non-governmental organizations. FSR/E has sought to integrate farmer participation and systems approaches into agricultural research and extension in public institutions, and infuse their agendas with farmer-defined needs for agricultural research and extension.
Tension has existed in the pursuit of these goals. One source of tension has been methodological. Scientific methods in agricultural research have been based on reductionism: the study of components of systems individually (Dillon, 1976). Several assumptions, sometimes explicit and other times implicit, have underlain the acceptance of reductionism. One is tle mechanistic paradigm: the assumption that a system can be explained by aggregating separately-obtained sets of knowledge of the nature and working of its individual parts. A second assumption has been that it is not possible to study in a scientific way the workings of an entire system, because of complexity and confounding of variables.
A second tension has been institutional. Publicly-supported agricultural research and extension institutions have developed within the reductionist paradigm. Organizational structures have usually been established around components, either in discipline-based departments or in commodity-based programs. Both institutional reward systems and professional recognition have been primarily based on work in these component-centered departments and programs. In many countries, research and extension are also institutionally separated. Interdisciplinary teams are best suited to pursue a systems approach, but such teams cross component-based boundaries both inside and outside research and extension institutions. Efforts to resolve this institutional tension have been the focus of numerous FSR/E papers under the rubric of "institutionalization."
There has also been tension between science and farmer participation. This has been seen throughout the history of the International Symposia that AFSRE and its predecessors have sponsored. Most recently this tension was evident in differences in perspectives on FSR/E accomplishments and future directions
Journalfor Farming Systems Research-Extension

presented in several presentations at the 1994 symposium. From very different starting points, both Sebillotte (1994) and Olivier de Sardan (1994) questioned the relationship between farmer participation and science. While Sebillotte acknowledged the place of farmer participation in FSR/E, he also cautioned against loss of scientific objectivity. Olivier de Sardan saw an even more fundamental conflict between the individual and practical goals of farmers, versus the public and knowledge-generating goals of scientists. Jiggins (1994), in contrast, reaffirmed the centrality of farmer participation to achieve the goal of making science more relevant to the problems of farmers in less-favored environments. Antecedents to this debate can be found in papers and workshops in mid-1980s FSR/E symposia that sought to develop on-farm trial methodologies that simultaneously provided scientifically-valid assessment and real farmer participation (Barker and Lightfoot, 1986; Caldwell and Lightfoot, 1988), and in discussion following Chambers (1993) keynote address at the 1992 symposium.
The land-grant system in North America has often been held up as a model for linking research and extension. Researchers based at a central university and extension field staff based in local administrative units (counties) are both under one organizational structure. Some of the researchers at the central university are given explicit extension responsibilities in their job descriptions that require and reward them for work in the field.
Although the land-grant system links research and extension, it has not been based on systems approaches, and farmer participation has been mixed. Much pre-war agricultural research was implicitly systems in nature, reflecting the farm background of the first generations of agricultural researchers. Systems approaches were also evident in early farm management research and in a number of extension programs from the 1940s to the 1960s, including the Farm and Home Development Program, the Rapid Rural Adjustment Program, and the Balanced Farming Program (Johnson, 1982; Hagan, .1984). The latter programs included work that could be called on-farm research, even if the programs did not conceive of themselves as research programs. But even in these programs, on-farm research was valued more as a real-world means to "educate" farmers, and farmer participation tended to be more responsive to researcher initiative. Subsequently, as methods for reductionist study became more sophisticated, scientific knowledge became more compartmentalized. Industrialization favored reliance on off-farm inputs, resulting in less on-farm recycling, a more industrial approach to agricultural production on a commodity-by-commodity basis, and a reduction in the systems nature of agriculture. This in turn meant that knowledge produced by reductionist approaches became more easily applicable to production agriculture. Both systems approaches and farmer participation in research seemed to become less relevant in the 1960s and 1970s.
VoL 6, No.2, 1996

Into this context, since the early 1980s, has come concerns about the sustainability of agriculture in North America. These concerns have been based on the recognition that North American agriculture cannot continue indefinitely to "mine" the soil, even the rich soils of the U.S. Midwest; that offfarm inputs are largely non-renewable; and that industrial agriculture has negatively affected public health and the environment (Phillips and Shabman, 1991; NWA Foundation, 1994; Feath, 1993). By its very nature, sustainable agriculture is a multi-goal endeavor and process, seeking to achieve a balance among the often conflicting desired outcomes of environmental protection, high-yield production, economic gain, and enhanced quality of life. Sustainable agriculture is inadequately served by a strictly component approach (Hanson et al., 1995). Further, sustainable agriculture research is rarely amenable to the direct, shorter-term measurement that is characteristic of the reductionist evalup,'on that takes place in production systems focused almost exclusively on yit and return. Flora (1992) argues that farmer participation on multi-discipli, my teams and in on-farm research is essential for effective FSR/E application in sustainable agriculture, and that adaptations to all phases of this methodology are necessary, from diagnosis to extension, to accommodate the differences presented by sustainable agriculture. Yet this conception has not been put widely into practice, and the number of university associated on-farm research projects underway across the United States is still very small.
In this paper we report on the development of a program applying the principles of FSR/E in vegetable crop sustainability within the above institutional and historical context of the North American land-grant system. As a faculty member in the Department of Horticulture of Virginia Polytechnic Institute and State University (Virginia Tech), one of Virginia's two land-grant universities, the first author (Caldwell) began developing the program in August 1993 in close collaboration with the Virginia Association for Biological Farming (VABF). VABF is a non-profit/non-governmental farmer
organization whose mission is to promote and provide education in biological (sustainable) agriculture. We first describe how farmer needs and input guided the choice of Integrated Pest Management (IPM) as the initial focus of the program and how farmers have participated in the development of the research program. We then describe the technical difficulties encountered in developing a biologically-based vegetable IPM production system. Finally, we conclude by presenting ways in which the three methods, reductionism, systems approaches, and farmer participation, can be better integrated for each to make complementary contributions to research and extension in sustainable agriculture.
Journalfor Farming Systems Research-Extension

Selection of Cucurbit Integrated Pest Management as the Research Focus.
The process of setting a research agenda for the vegetable sustainability program utilized farmer input in both secondary information and formal survey methods. Secondary information was gathered from a 1993 farmer focus group meeting sponsored by the Southern Sustainable Agriculture Working Group (SSAWG), a network of non-profitlnon-governmental organizations created to promote more sustainable methods in agricultural production in the Southern region of the United States. The focus group meeting, conducted by SSAWG member VABF, sought to identify problems for which research and information are needed.
Among the four key production constraints identified in the meeting, two involved insect control. Farmers indicated a need for better information on which cover crops and living or applied dead mulches could best serve as habitats for beneficial insects predatory to six vegetable insect pests: stink bugs, Mexican bean beetle, flea beetles, cucumber beetles, Colorado potato beetle, and squash vine borer. In fall 1993, when the two authors came in contact, a first decision was thus made to focus the new vegetable sustainability program on integrated pest management (IPM).
Although simultaneously conducting research on all six pests and all their primary vegetable crop hosts might best reflect the diversity of growers' needs, the knowledge base and resources of the new program would be exceeded. Therefore, program focus was narrowed using a farmer mail survey on pest incidence and control measures for vegetable crops and available mulch and cover crop materials. The survey was designed by a team of two VABF members and six faculty (two horticulturalists, an entomologist, a plant pathologist, an agricultural economist, and a sociologist). It was sent to the 314 members of VABF and 147 other non-VABF participants at the 1993 Virginia Sustainable Agriculture Conference. A 31% response rate was obtained, but only 57% (82 respondents) of respondents grew vegetables. Among these 82 vegetable growers, cucurbits were most frequently cited (31% of all crop responses) as having insect pest problems. Cucumber beetles and squash vine borer comprised 74% of the pests cited on cucumber and squash (Caldwell et al., 1995). Thus, cucurbit IPM was selected as the initial focus for research.
VoL 6, No.2, 1996

Research program design,
In 1993, 1994, and 1995, research plans were developed by joint VABFuniversity teams. The overall goal of the research program has been to build on general agroecosystems concepts in order to develop methods to introduce greater biological diversity into cucurbit production systems in ways that are amenable to farmer implementation. A review of 150 studies comparing more and less diversified agroecosystems has shown that 53% of 198 herbivore species were less abundant in the more diversified systems (Risch et al., 1.982). Achievement of greater diversity would introduce into North American agriculture more mixed cropping and natural vegetation management approaches similar to those used by farmers in less favored environments of many developing countries.
In the current U.S. land-grant system, internal resources are primarily available for on-station research. Developing a consensus between farmers and researchers on how to carry out joint on-farm research and to develop farmers' monitoring and evaluation skills requires training, travel, and personnel costs that exceed internal resources. A major focus of team planning, then, has been the design of proposals for outside funding for on-farm research linked to onstation research.
Proposals have been reviewed by anonymous teams composed primarily of researchers, but in some cases including farmers. The resulting dialog between the planning team and the research community has revealed the tension among reductionism, farmer participation, and systems approaches.
In 1993, the joint VABF-university team developed a two-year plan for initial exploratory research, placing a strong emphasis on farmer participation. The team proposed testing four habitats for beneficial insects on 25 farms in three regions of Virginia, using farms as replications within regions in a nested design structure (Stroup et al., 1991; Caldwell, 1989; Milliken and Johnson, 1986). One farmer in each region would be trained and hired to serve as regional trial coordinator. This model conibines the concepts of farmer paraprofessionals of the Rapid Adjustment Program (Rich, 1982) and clustered trials of Heinrich and Masikara (1993) in Botswana. A 26th site would be managed by researchers, have on-site replication, and include supplemental researcher control treatments. All sites would be assessed in an incomplete block design analysis of variance model developed in collaboration with the Department of Statistics, Virginia Tech (Caldwell and Walecka, 1987; P. Palettas, Virginia Tech, personal communication, 1993).
Proposal rejections by two funding sources related to insect monitoring methods and farmer compensation. The suggested larger plot sizes and more
Journalfor Farming Systems Research-Extension

rigorous monitoring conflicted with farmer constraints of land area and time. Partial support proposed for farmer monitoring activities and travel costs for extensive farmer-researcher interaction were seen as excessive.
In 1994, the planning team was expanded to include additional VABF members, an off-campus research station entomologist, and an entomologist and an extension specialist from Virginia's second land-grant university, Virginia State University. This collaboration produced a three-year research plan, drawing from habitat research in 1994 (based on the 1993 research plan), and proposing on-farm and on-station technical research in the first two years on five crop agro-ecology modifications. Concurrent economic research was included to develop a "menu" of marketing techniques to improve the economic viability of biological pest management. These two research efforts would then converge in the third year in farmer-managed trials of improved agro-ecologies. Compared with the 1993 plan, this plan expanded the systems aspect of the research program, but delayed widespread farmer participation in on-farm research until the last year. It retained the overall structure of on-farm trials clustered in regions, but with a more gradual expansion of farm numbers, beginning with first year single-site trials in two regions, to 3-4 farms in one pilot region in year 2, and 10-15 farms in 3-4 regions in year 3.
Concern about the scientific validity of on-farm research methods was the primary factor in decisions by two funding sources to reject this proposal. The size of the budget for a three-year program was a second factor.
In 1995, a smaller team developed a less ambitious plan for two years of research, building on 1995 on-station research in three of the five agroecological modifications. This proposal was also rejected, despite the gradual phasing in of on-farm research, a more reductionist approach, and a considerably reduced budget compared with 1994. Concern about on-farm trial research methods was again the primary factor.
The above results of three years of efforts to obtain outside support for onfarm agro-ecology research indicate that it is difficult to combine technical research with extensive farmer participation in the design and conduct of the research on their farms in a way that is viewed by the research community as scientifically valid. Criteria derived from reductionist research remain the critical measure of the validity of on-farm research.
The research plans developed for outside funding in 1993 and 1994 have served as the basis for reduced programs of exploratory research at the university experiment station and one on-farm site in 1994 and 1995. These trials, supported by internal resources, have revealed that development of improved
VoL 6, No.2, 1996

systems is based both on understanding and skill in managing components, and on an understanding of interactions within a biological system.
Previous research in Georgia by Phatak (1992) and Bugg (1991) has shown that cover crops and living mulches can increase populations of generalist predators of pests of several vegetable crops. Their approach was based on the assumption that an overall increase in the types and numbers of predators results in stability at a lower level of pest pressure on the economic crop. We initially adopted this same assumption.
In the 1994 habitat trials, we attempted to assess both above-ground and below-ground predators in the economic crop and in natural vegetation. Due to the complexity, and thus time and monetary expense required, of each of these components, we were forced to abandon the natural vegetation assessment and reduce above-ground insect assessment.
We also found that other interactions affected the viability of strip cropping as an insect pest management approach. Poor crop establishment and competition with the strip crop was not offset by positive pest management effects in the economic crop. Improvement of those interactions is a precondition for assessing economic benefits of strip cropping for pest management. Grossman (1993) in California had similar problems in developing a living mulch system for crucifer production.
The research program was divided into three components in 1995, based on the 1994 findings. Objectives for the first component, a continuation of the strip cropping trial, were a) the development of strip crop management techniques supporting economic crop yields equal to or greater than those with sole cropping; and b) assessment of the effect of the different habitats on ground beetles. The second component examined the effect of flowering plants at the head of crop rows on above-ground predators and two types of parasitoids. The third component characterized natural vegetation adjacent to the crop field, and assessed insect movement on transects bisecting the natural vegetation and the crop field.
Each of these trials revealed both new needs for additional component understanding, and the importance of other interactions. Overall weed management problems and increased labor demands for strip management in the strip cropping trial led to investigation of equipment modification for no-till seeding of crop strips between habitats. Difficulties in assessing parasitism in the flowering border trial led us to restrict our focus to only one of the parasitoids, while adding a second sampling method in 1996. In the natural vegetation trial, difficulties in locating marked insects led to research into rearing methods.
These technical difficulties have, on the one hand broadened the scope of the research beyond pest-predator/parasitoid insect interactions, increasing the number of biological interactions examined and adding labor and management interactions. On the other hand, they have necessitated examining these
Journal for Farming Systems Research-Extension

interactions in smaller subsets, in order to deepen the knowledge and ultimate mastery of each component playing a role in a given interaction.
The results of two years of technical research suggest that reductionist research and systems research are complementary rather than conflicting alternatives. The testing of biological systems incorporating greater diversity has been necessary to reveal which reductionist research is most needed. This has conceptual parallels with the approach used by Atta-Krah and Francis (1986) to develop agro-forestry systems in Nigeria. In that program, farmers weie left free to try different combinations of agro-forestry components in their own systems, which were then monitored to identify problems needing reductionist component research.
Biological vegetable farmers typically produce a range of crops in small parcels with high diversity using machinery, animal traction in some cases, and considerable hand labor. They are thus more similar to farmers in the developing world than large-scale North American farmers who use industrial methods to produce monocrops on large pieces of land. It would be ideal to study each biological farm as a whole, examining interactions among crops, pests, beneficial insects, and natural vegetation, to develop a classification of types (research domains).
After making such a classification, however, attempting to improve systems would require specialized knowledge of a large number of pests and beneficial insects. Yet, a given researcher often spends many years in acquiring specialized knowledge, for example of a particular order of insects (Diptera), or" even of a particular family within the order (Tachinadae). Thus, while systems interactions can be used for classification and identification of research problems, only a few research problems can be addressed in depth at a given time. A complete specialized knowledge of all or most of the components of a system not only exceeds the capability of the individual researcher, but it is likely to exceed the capabilities of the institution as a whole.
Participatory approaches place high value on farmer knowledge. They are based on the expectation that farmers have developed an intuitive understanding of how complex, diverse systems function, and how they would respond to change, that is superior to the understanding that models of their systems built by scientists can provide. Because farmers are assumed to have acquired valid empirical knowledge over the course of time and this knowledge is not uniformly distributed among farmers, participatory approaches also incorporate the position that solutions to farmers' problems are often found
VoL 6, No.2, 1996

from group analysis and sharing of individual knowledge. Bentley (1989) has shown, however, that the validity of farmers' knowledge varies considerably depending on the type of phenomenon described. Reductionist research is needed to complement farmer knowledge precisely in these areas where empirical approaches are inadequate.
A 'more synergistic approach to combining reductionism, systems approaches, and farmer participation recognizes the different roles of each. Improved systems depend first on improved components. Reductionism is needed to understand components in ways that go beyond farmer empirical knowledge. At the same time, reductionism can only be useful if it fits into farmers' systems. Systems analysis and farmer participation are needed at the start of any program to select topics for reductionist research, and systems validation and tarmer participation are needed during and after reductionist research to insure component fit and functioning. Farmer participation is needed to develop and choose management components, and to assess how all components function in the management system. Adoption of this perspective acknowledges that farmers have an advantage in understanding how whole systems work, and that this understanding is integral to an effective research process.
In the North American institutional setting, the first criteria for the evaluation of improved systems'has been the validity of research on their components. The work reported here indicates that such research is a necessary but not sufficient criteria for the development of biological systems. The performance of improved components needs to be assessed by farmers in terms of the functioning of their entire system. Including on-farm research as the sufficient step parallel with reductionist research can harness farmers' assessment and integrating abilities to the application of the results of reductionist research.
Journal for Farming Systems Research-Extension

Atta-Krah, AN., and P.A. Francis. 1986. The role of on-farm trials in the evaluation of alley cropping.
Paper presented for International Workshop on Alley Farming for Humid and Subbumid Regions
of Tropical Africa, March, 1986. International Institute of Tropical Agriculture, Ibadan, Nigeria. Barker, R., and C. Lightfoot. 1986. Farm experiments on trial. In: C.B. Flora and M. Tomacek (eds.),
Farming systems research & extension: management and methodology. Paper No. 11, Fifth Annual Fannrming Systems Research and Extension Symposium, 13-16 October 1985. Kansas State
University, Manhattan, Kansas.
Bentley, J.W. 1989. What farmers don't know cant help them: the strengths and weaknesses of
indigenous technical knowledge in Honduras. Agriculture and Human Values 6(3,Summer):2531.
Bugg, R.L, F.L. Wackers, K.E. Brunson, J.D. Dutcher, and S.C. Phatak. 1991. Cool-season cover
crops relay intercropped with cantaloupe: influence on a generalist predator, Geocoris punctipes
(Hemiptera: Lygaeidae). J. Econ. Entomol. 84(2):408-416.
Caldwell, J.S. 1989. M6thodologie de I'analyse statistique des tests agro-zootechniques: facteurs de
groupement des UP, interactions entre types de paysans et les traitements, et contrastes orthogonaux. In: Volet O.H.V., Commissions techniques sp6cialishes sur les systhmes de
production rurale: r6sultats campagnes 1988-1989. Institut dEconomie Rurale, Bamako, Mali. Caldwell, J.S., J-P. Amirault, and A.H. Christian. 1995. Insect pests, beneficial insects, and cover crops
ofbiological farmers. HortScience 30(4):806.
Caldwell, J.S., and C. Lightfoot. 1987. A network for methods of farmer-led systems experimentation.
FSSP Newsletter 5(4):18-24.
Caldwell, J.S. (technical editor) and L. Walecka (coordinating editor). 1987. Design techniques for
on-farm experimentation, FSR/E Training Units: Volume II. 2nd ed. Fanning Systems Support
Project. Gainesville, Florida.
Chambers, R. 1993. Methods for analysis by farmers: the professional challenge. Journal for Farming
Systems Research-Extension. 4(1):87-102.
Dillon, J. L. 1976. The economics of systems research. Agricultural Systems 1:5-22. Feath, P. 1993. Evaluating agricultural policy and the sustainability of production systems: an economic
framework. J. Soil & Water Conserv. 48(2):94-99.
Flora, C. 1992. Building sustainable agriculture: a new application of farming systems research and
extension. p.37-49. In R. K. Olson (ed.), Integrating Sustainable Agriculture, Ecology, and
Environmental Policy. The Haworth Press, Inc., New York.
Grossman, J. 1993. Fighting insects with living mulches. The IPM Practitioner XV(10):1-8. Hagan, A 1984. Balanced farming in Missouri: A farming systems approach to assisting farm
families. University of Missouri, Agricultural Economics Department Paper No. 1984-31,
Columbia, Missouri.
Hanson, J. C., C. S. Kauffman, A Schauer. 1995. Attitudes and practices of sustainable farmers, with
applications to designing a sustainable agriculture extension program. J. Sust. Ag. 6(2/3):135-156. Heinrich, G.M., and S. Masikara. 1993. Trial designs and logistics for farmer-implemented technology
assessments with large numbers of farmers: some approaches used in Botswana. J. for Farming
Systems Research-Extension 3(2):131-145.
Jiggins, J. 1994. Prelude to conclusion. Closing address presented at the 13th International Symposium
on Systems-Oriented Research in Agriculture and Rural Development, Montpellier, France, 21-25
November 1994.
Johnson, G. L. 1982. Small farms in a changing world. Pp. 7-28. In W. J. Sheppard (ed.). Proceedings
of Kansas State University's 1981 Farming Systems Research Symposium Small Farms in a Changing World: Prospects for the Eighties. Paper No. 2, Kansas State University, Manhattan,
Milliken, G.A., and D.E. Johnson. 1984. Analysis of messy data. Volume 1: designed experiments.
Van Nostrand Reinhold Company, New York. 473 p.
Northwest Area Foundation. 1994. A better row to how: the economic, environmental and social impact
of sustainable agriculture. NWA Foundation, St. Paul, MN. 38 p.
Olivier de Sardan, J-P. 1994. La participation des acteurs sociaux: la grande illusion. Symposium
Letter N'2, 13th International Symposium on Systems-Oriented Research in Agriculture and Rural
VoL 6, No.2, 1996

SDevelopment, Montpellier, France (summary of paper presented at the Symposium, 21-25
November 1994).
Phatak, S. C. 1992. An integrated vegetable production system. HortScience 27(7):738-741. Phillips, S. R., and L. Shabman. 1991. Agricultural pesticide use and risk in Virginia's Chesapeake Bay
Region. Pub.448-03/REAP R004, Virginia Cooperative Extension, Virginia Tech, Blacksburg,
Rich, N. 1982. A model for providing assistance to limited resource farms in Smyth County Virginia.
Southwest Extension District, Abingdon, Virginia (mimeo).
Risch, S.J., D. Andow, and M.A. Altieri. 1982. Agroecosystem diversity and pest control: data,
tentative conclusions, and new research directions. Environ. Entomol. 12(3):625-629.
Sebillotte, M. 1994. The systems approach and action. Paper presented at the 13th International
Symposium on Systems-Oriented Research in Agriculture and Rural Development, Montpellier,
France, 21-25 November 1994.
Stroup, W.W., PE. Hildebrand, and C.A. Francis. 1991. Farmer participation for more effective
research in sustainable agriculture. Staff Paper SP91-32. University of Florida, Gainesville,
Journalfor Farming Systems Research-Extension

Indrajit Roy'
Accelerating transfer of technology to farmers is one of the key elements in the package of policies designed to achieving higher growth rates in agriculturalproduction in Bangladesh. In this context, the important parameter to judge the relevance of farming systems research (FSR) would be the efficiency with which it serves as an approach to technology transfer. The traditional transferof-technology approach is biased against resource-poor farm families in the sense that their technology needs are not only neglected but also they are frequently overlooked as a group for receiving messages about technological innovations. Besides, the existing approach is inadequate to take account of diversity of land, soil, and climatic resources. This often results in technological recommendations. that are overly general and farmers often find them inappropriate for solving their problems. Recognition of locationspecificity and greater relevance to farmers' socio-economic context are thus two important features that make FSR unique as an approach to ensure that technologies developed are demand-driven and client-oriented.
The proposition that FSR be incorporated into the country's mainstream extension services is no longer debatable. But the practical approaches and innovations in management required to foster a truly FSR perspective are yet to be developed. The purpose of this paper is to identify the opportunities and constraints in integrating the FSR approach into the national extension system.
Characteristics of FSR in Bangladesh
The FSR perspective in Bangladesh evolved through a decade-long experience of working with a cropping systems research (CSR) approach. The Bangladesh Agricultural Research Council (BARC) brought the scattered
1 Revised version of the paper presented at the 12'h Annual Symposium of the Association for Fanning Systems Research-Extension. Mchigan State University, East Lansing, USA, September 13-18, 1992
2 Principal Scientific Officer, Bangladesh Agricultural Research Council, Farnugate, Dhaka 1215, Bangladesh
VoL 6, No.2, 1996

46 Roy
efforts on FSR under one project titled the National Coordinated Farming Systems Research Program which started in 1985-86. The program involved six crop and non-crop research organizations of the national agricultural research system (NARS) and the Bangladesh Agricultural University. In 1989 the program was renamed as the National Coordinated Farming Systems Research and Development Program (NCFSRDP). Each component of the program opened one or more FSR sites in specific agro-ecological zones of the country. The site model was chosen to achieve a greater coherence in efforts to conduct research in crop and non-crop components of the farming system. The other purpose was to make use of the site to forge a close working relationship between research and extension staff at the local level.
The Bangladesh Agricultural Research Institute's FSR program built on its limited experience with on-farm client-oriented research (OFCOR). BARI's OFCOR program had limitations it was strong in service function and applied function which concerned mainly broad-scale on-farm testing, screening, and evaluation of technologies developed on-station. But its adaptive, feedback, and support functions were limited (Merrill-Sands 1989). The adaptive research trials network of the Bangladesh Rice Research Institute (BRRI) provided a basis for embracing the FSRD approach. Other research institutes incorporated FSR into the strong focus on their mandated commodities. The outcome of these efforts was development of technologies that had exclusive focus on specific farm components and tended to ignore the inter-relationship among other components of the whole farm. But as the need to link FSR with extension grows stronger, the research organizations mount efforts to move toward a FSR program that pursues a holistic approach of integrating commodities and searches for inter-dependencies.
Characteristics of extension in Bangladesh
The country's agricultural extension system involves both public and private sector organizations. The Department of Agricultural Extension (DAE) is the main public sector institution that provides extension services in crop agriculture. There are several parastatals which provide extension services for non-crop commodities, for example, fisheries, livestock, and forestry. Private sector's involvement in technology transfer is mostly limited to rural extension work conducted by non-government organizations (NGO).
The NGOs in their work focus mostly on marginal farmers, landless laborers, and destitute women groups largely by-passed by the public sector extension system. Such NGOs as the Bangladesh Rural Advancement Committee, Proshika, Rangpur Dinajpur Rural Service (RDRS), and Caritas have action programs that promoted the use of minor irrigation equipment (Biggs and Farrington 1991). The activities of another NGO the Mennonite Central Committee encompass a fanning system perspective that is reflected in farmer participatory research for developing cropping systems for small and
Journalfor Farming Systems Research-Extension

marginal farmers (Graham and Buckland 1991). In the forestry sector, the Village and Farm Forestry Project (VFFP) is the most notable non-government institution in the country that promotes tree planting on privately owned farmland and homesteads. The project supported by the Swiss Agency for Development and Cooperation operates through 19 NGO partners who assist in developing village-based and commercially run private nurseries (Hocking and Roy 1996).
The T and V model was the dominant extension approach till the early 1990s. The system has been modified and the grass roots level extension staff are now required to make contact with farmer groups rather than with individual farmers. Apart from the formalized extension system DAE and other extension parastatals conduct a large number of demonstrations with planning and funding support from research institutes and several international funding agencies.
The approaches to integrate FSR with the mainstream extension system can be broadly divided in two categories. The first one groups isolated and scattered efforts on the part of research institutions to build a collaborative relationship with the public sector extension organizations. The second one groups modifications which extension organizations themselves bring about in their working methods that allow an FSR perspective in conducting extension activities. The former approach is an adhoc and project-tied activity through which research institutes attempt to disseminate specific product of their FS research by large-scale demonstrations and training making use of institutional capacity of the public extension organizations. This approach is an adhoc one because it exclusively depends on external funding and extension organizations tend to treat them as activities not internal to them. The latter approach relies on changes in philosophy and methods of organizing extension work, due in part to the impact of the former approach, that are friendlier to internalizing FSR in routine activities of mainstream extension organizations. Below we present some case examples that illustrate the experience of the first approach:
Case 1. Homestead Vegetable Production Technology
The FSR team of the Bangladesh Agricultural Research Institute (BARI) developed this technology that envisages growing 14 vegetable crops in five patterns round the year in a 6m x 6m plot at farmers' homesteads. The technology was designed to achieve three-fold objectives: (1) to encourage farm families, particularly marginal and small ones, to grow vegetables on their homesteads for consumption and thus to cope with nutritional deficiency in
VoL 6, No.2, 1996

48 Roy
vitamin A, C, calcium,, and iron; (2) to sell the marketable surplus for additional cash income; and (3) to involve women and children in active homestead labor.
The success of this technology package during four years of field testing prompted BARC to initiate the homestead vegetable production program in the winter season of 1989 in collaboration with the Department of the Agricultural Extension System (DAE) and BARI. The program started with establishing 1000 homestead vegetable gardens and by 1993 the number of demonstrations soared to 57,500 spread in 135 of the 465 Thana (sub-district) of Bangladesh. It became possible to disseminate this technology to an estimated 2.1 million farmers through demonstrations, field days and training. Because seed availability was a major concern for the sustainability of the program, 2000 farmers were trained to produce and preserve vegetable seeds. The program had a tremendous impact on farmers' intake of vegetables 66.4% of the vegetables produced were consumed by the farmers, their neighbors and relatives and 33.6% were sold (Ahmed et al. 1991). Another estimate puts additional income across the country through production of vegetables on homesteads at around Taka 43 million (Stem 1993).
Case 2. Spaced Transplanting Technique (STP) of sugarcane
This technology envisages planting of pre-germinated cane seedlings raised from single node stem material in place of the traditional "end-to-end" stem cuttings (containing a minimum of three nodes). The STP has a comparative edge over traditional methods of planting: it ensures optimum plant stand that causes about two-fold increase in yield; it is cost-effective because it requires less planting material and holds potential to create spin-off effect in rural agribusiness.
The Bangladesh Sugarcane Research Institute (BSRI) included STP in 1985 as trials under their FSR program. Since 1988-89, BARC started funding demonstration and farmers' training program on STP in cooperation with DAE and BSRI. In 1988-89, DAE set up six demonstration plots each on 0.40 ha plots. In 1990-91 this figure rose to 335 covering 42 districts of the country. The mean yields of sugarcane in demonstration plots were consistently high ranging from 74.6 to 83.4 t/ha as against 50 t/ha obtained using traditional planting method. BSRI conducted farmers' training on STP. During the period from 1988/89 to 1990/91, they trained a total of 6000 cane growers.
The spread of this technology was quite rapid. From 251 hectares of land in 1988/89, this technology was expanded to cover 7626 ha of land in 1990/91 and about 10 thousand hectares by the end 1993 (Huda et al. 1992 and Stem 1993). For small-scale farmers, this technology offered the opportunity of making extra money from selling cane seedlings raised in polybags. Most important, however, is the generation of employment for women. In sugar millJournal for Farming Systems Research-Extension

zone areas under STP, women could earn US$ 0.79 a day for filling 1000 polybags with soil (BARC 1990).
Case 3. Fish Production Technology in ponds
Eight FSR sites incorporated fisheries as a component in their activities. In, 1990-91, BARC conducted a demonstration program on four fish production technologies tested at these sites. Based on the initial results, BARC in 1992 initiated a large-scale technology transfer program through the Department of Fisheries (the extension wing of the Fisheries sector) and NGOs. The technologies included were: (a) fish culture in seasonal ponds, (b) polyculture in perennial ponds, and (c) nursery pond management. Eight NGOs working in 14 districts were selected who established 165 demonstrations in perennial ponds, 127 in seasonal ponds and 7 in nursery ponds. They trained a total of 1227 farmers. This training and demonstration program created a positive impact on farmers' adoption of technologies. The trained farmers, on average, motivated five persons who have adopted fish culture technologies in their ponds using their own resources. This caused the level of their income to increase by about 33% (Dewan et al. 1994).
The major success of the first approach is perhaps the adoption of technologies by farmers at a pace much faster than what is usually observed when innovations are transmitted through the traditional extension system. Additionally it provided an opportunity to draw lessons on the type of institutional arrangements required to integrate inputs from FSR into the mainstream extension services. First, innovations in building strong horizontal linkages with different interest groups would be required. This is because lines of responsibility are vertical within departments, and the authority of one department does not usually extend to other departments and good linkage would always depend on how the interested parties perceive it as being in their interest to interact.
One good example is the institutional arrangement which BARC made for transfer of HVP technology. These were (i) the technical advisory and monitoring committee headed by the BARC's Member-Director (Crops), (ii) the national task force convened by the Additional Director of DAE's Field Services Division, (iii) the District coordination committee led by the Deputy Director of DAE, and (iv) the Thana coordination committee led by the Thana Agricultural Officer. These committees were formed with members drawn from the various organizations: BARC, DAE, BARI, Bangladesh Agricultural Development Cooperation, private agribusiness entrepreneurs, and NGOs.
VoL 6, No.2, 1996

50 Roy
The involvement of NGOs in transfer of technology proved to have an additional benefit creation of employment opportunity for resource-poor farm families. For example, a local NGO the Bangladesh Rural Advancement Committee (BRAG) conducted demonstration and training on three aquaculture technologies: polyculture of carps, culture of nilotica fish, and culture of an introduced fish species, Thai Sharpunti. Many marginal farmers and destitute women started using these technologies in their backyard ponds and seasonal ditches which were suitable for culture of those fast-growing fish species. This enabled them to earn a substantial cash income.
An important learning was the effectiveness of the District Technical Committee (DTC) in facilitating transfer of technologies. The DTC was conceived as an institutional framework to link agriculture research and extension at the district level. This committee meets once a month and among other responsibilities formulates extension recommendations and provides feedback on farmers' problems. Much of the success in examples cited above can be traced to good cooperation achieved at the DTC level. It has been observed that where there was a BARI FSR team, DAE's Subject Matter Specialists had a better appreciation of technological opportunities available for small-scale farmers (Phillips et al. 1991).
Some important changes took place in the Department of Agricultural Extension with launching of the Agricultural Support Services Project (ASSP) in the early 1990s. These changes enabled DAE to institutionalize many activities which were conducted on adhoc basis as mentioned in case examples. The T and V model of extension which puts emphasis on training field-level extension staff and their contact with individual farmers ceased to be DAE's dominant method of organizing extension activities. The ASSP restructured the extension approach based on five core principles: decentralization, targeting, responsiveness to farmers' needs and working with farmers' groups rather than with individual farmers. The homestead agricultural production component of ASSP became an important institution for promoting FSR technologies, notably homestead vegetable production through the NGOs.
Other non-crop extension organizations are also switching over to NGO-led extension for promoting FSR technologies. For example, the Department of Fisheries (DOF) under a European Union-supported project is now working with 37 NGOs in 21 districts of the country to disseminate pond fish culture technologies including those mentioned in case 3 above. Additionally, the project focuses on organizing marginal farmers, group action and gender dimension in creating income-earning opportunities. The new thinking on the need to reorganize extension in line with basic principles of FSR was reflected
Journalfor Farming Systems Research-Extension

in the New Agricultural Extension Policy (NAEP) which the Government of Bangladesh approved in 1996.
The key components of NAEP are: extension support to all categories of farmers; efficient extension services; decentralization; demand-led extension; working with groups of all kinds, strengthened extension-research linkage, training of extension personnel, appropriate extension methodology; integrated extension support to farmers; coordinated extension activities; and integrated environmental support. The NAEP emphasizes the use of participatory methods and techniques such as RRA, PRA, and problem census to identify farmers' problems. One extension project titled Thana Cereal Technology Transfer and Identification Project which DAE is currently implementing with support from FAO has employed PRA. In order to forge a whole farm approach in conducting extension work, the District Technical Committee has been reorganized into the Agricultural Technical Committee (ATC) comprising district-level representatives of not only crop-oriented but also non crop-related extension organizations.
The successes and failures of the attempts to integrate FSR with the national extension system should be seen in the context of institutional constraints in accommodating basic principles of FSR. FSR promotes a wide range of technological options in an understanding that the clientele would decide are useful. It thus allows the farmer to make a choice among the alternatives on the basis of criteria that most suit his or her circumstances, for example, net extra cash income, profitability, access to markets, and access to credits, etc. Traditionally, DAE and other public extension agencies attempt to make decisions for farmers by providing them with prescriptions which should have guaranteed returns rather than options which might be worth trying. While doing so, they promote only those innovations which are approved, declared as 'mature', and deemed to be 'transferable' by a national-level Technical Coordination Committee.
The process, therefore, misses an important link: it restricts farmers' access to technical information, limits their ability to innovate with technical knowledge at different levels of development, and, above all, does not recognize their ability to make decision -for themselves. The preference of public extension organizations for standard prescriptions was amply reflected in DAE's acceptance of only 5 technologies for dissemination to farmers out of 100 presented in a national workshop on FSR which BARC organized in 1989. These technologies were generated through FSR and on-farm research but most of them were not approved as 'mature' by the Technical Coordination Committee.
VoL 6, No.2, 1996

52 RoY
More recently, there are, however, signs of change in public extension agencies that came along with institutional reforms and the Government's commitment to implementing the New Agricultural Extension Policy. These changes are shaping the contours of an institutional framework conducive to working with farmers of different resource endowments, with technical information of varying levels of 'maturity' and more autonomy for local bodies to pick up research results and suggest location-specific recommendations.
The opportunities and constraints of integrating FSR into the national extension system have been explored. Earlier attempts to transfer FSR technologies were isolated and scattered and relied more on building linkages and working mechanisms of an adhoc nature. This mechanism worked and indeed led to many valuable learning experiences. But it was not sustainable since these activities were not looked upon as internal by the public extension agencies. Despite this limitation, the experiences gained through the technology transfer programs set the stage for institutional changes that favor diffusion of FSR technologies. One example of an institutional change that reflects the moving of the concept of FSR from commodity level to policy level is the replacement of the District Technical Committee by the Agricultural Technical Committee. The more recent experiences of transferring FSR technologies build on the elements of sustainability that stem from the institutional reforms in public extension organizations which allow them to internalize the philosophy and activities related to the transfer of FSR technologies.
Ahmed, M.,M. Hossain, and A.K.M.A. Gaffar. 1991. Evaluation report of the project on demonstration
of imported technologies of sugarcane cultivation implemented by the Department of Agricultural
Extension, Bangladesh Agricultural Research Council.
BARC. 1990. Transplant techniques promises sweet future for Bangladeshi sugarcane growers.
Farming Systems Agribusiness Newsletter, 1(3): 1-3.
Biggs, S. and Farrington J. 1991. Agricultural research and the rural poor: A review of social science
analysis. Ottawa: IDRC, 139p.
Dewan, S., Alam F. and A. Awal. 1994. Evaluation report on transfer of fish culture technologies
through demonstration by NGOs, Bangladesh Agricultural Research Council.
Graham, P and J. Brickland. 1991. Challenges and opportunities of govemment-NGO collaboration in
agricultural research and extension: The experience of Mennonite Central Committee in
Bangladesh, 1973-1990. Bangladesh Journal of Extension Education 6:118-141.
Huda, M.S., M.A. Quddus, and M.S. Ahmed. 1992. Evaluation report on impact of demonstration and
on STP training of sugarcane. Bangladesh Agricultural Research Council, Dhaka, Bangladesh.
Hocking, D. and Indrajit Roy. 1996. Village and Farm Forestry Project, Mission Report: Action
Research Planning for VFFP phase 5. Swiss Agency for Development and Cooperation, Dhaka and
Journalfor Farming Systems Research-Extension

Merrill-Sands and J. McAllister. 1988. Strengthening the integration of on-farm client-oriented research
and experiment station research in national agricultural research systems (NARS): Management lessons from nine country case studies. International Service for National Agricultural Research.
Phillips, J.C., Miah, H., Yeshewalul, A. and Bartlett, A.P. 1991. Report of the ex-post evaluation
mission: Strengthening the Agricultural Extension Services (BGD/79/034), February-March,
Stem, C. 1993. Final Report ARP II (S), Checchi and Company Consulting Inc., BACC/USAID/Dhaka.
Vol 6, No.2, 1996

S. Praneetvataku2 and W. Doppler3
Wood is an important source of rural energy in many developing countries. The rapidly declining forests have reduced wood energy supply, consequently, the negative environmental and economic impacts occurred. This problem requires a wider approach and solutions which cope with the specific needs of the region, country, community, village and people involved. Therefore, farming and regional systems approaches are applied. In conclusion, for the families to avoid wood energy shortages, using a more efficient woodstove and growing trees on farmland could help to reduce demand and to increase supply of wood. For the region, in order to avoid over-harvesting of wood from natural forests and to make wood a viable and sustainable energy resource, the policy makers should focus activities on (1) subsidising local stove industries in the initial phase to make more efficient woodstoves more available and cheaper for the family; (2) enhancing rural job opportunities to increase family income making the substitution of modem energy possible; (3) encouraging multipurpose-tree plantations on farm lands and woodlot plantations for rural industries; (4) promoting community forests by providing the
This paper was presented at the AFSRE Symposium in Sri Lanka in November of 1996.
2 Lecturer, Department of Agricultural and Resource Economics, Faculty of Economics, Kasetsart University, Bangkok 10900, Thailand.
3 Professor, Department of Agricultural Economics and Social Sciences in the Tropics and Subtropics, University of Hohenheim, 70593 Stuttgart, Germany.
VoL 6, No.2, 1996

village community with real authority to manage their forests; and (5) implementing successful afforestation, particularly in
taking care of the trees after planting.
The rapid decline of forest resources obviously occurs in many developing countries in Asia as well as in Northern Thailand. This leads to a decrease in wood energy supply of Northern Thailand. In other regions of Thailand, most rural Thai households obtained firewood and charcoal from their own lands but the northern region was an exception. Here the source of wood energy comes primarily from forests (Department of Energy Affairs, 1992). The declining of forests results in a reduction of wood energy supply while the consumption of wood energy still shows an increase in quantity although the rate of increase was not great. This has widened the gap between sustainable supply and actual consumption of wood energy in Northern Thailand. Consequently, economic losses and environmental impacts occur to family-household and the region. Since wood is and will remain an important energy source for rural Thai people in the future for at least the next two decades (Sabhasri and Wibulswas, 1992), strategies to make wood a viable and sustainable energy source for northern Thai people is vital (FAO, 1993). This will certainly contribute to improving the environment as well as the standard living of people. However, little has been done to examine the solutions to this problem from the perspective of family-household and from a holistic view.
This paper intended to identify the potential improvements of wood energy resources by considering the preference from a family-household viewpoint. Furthermore, future impact. of improved wood energy strategies will be examined on farm and family incomes as well as on social costs. Conceptual Framework
Farming and regional systems approach4 can be used as an appropriate approach to improve the present wood energy situation. Doppler (1994) defines this concept as:
Farming systems approach: at micro level, the decisions for all activities in the farm, family and household are considered and accomplished as a
4 Farming and regional systems approach is a holistic and behavioral approach which can be applied in farming related or rural development projects. It provides the philosophy, the concept and the strategy for developing and introducing solutions to problems at farm-family-househqld and regional levels (DOPPLER, 1991).
Journalfor Farming Systems Research-Extension

system. This includes the following ideas: (1) the complexity of the real world is increasingly considered, (2) dynamic and sustainable views are essential, (3) objectives and decisions of families are involved, and (4)
participation of target groups in problem solving are included.
" Village systems approach: at the village level, the important elements
are (1) the living standard of individual families, including non-farming families, and (2) common village resources including physical, socioeconomic and social structure regarding their change and investment
" Regional and rural systems approach: at the regional level, a complex
idea is developed by considering physical, economic, social, administrative and cultural points of view. This approach is comprised of (1) the decision-making of families, villages and regions, (2) all
relevant resources, and (3) a wide range of different disciplines.
The research followed the farming and regional systems approaches of Doppler (1994) by collecting the information at micro, village and regional levels (Figure. 1). Resources, innovations and solutions to problems are also included. Then two sources of data: secondary and primary data were collected. Data bank from questionnaires were designed. Finally, statistical analysis was done for example descriptive statistics, correlation, regression and linear programming.
In order to obtain the appropriate information on wood energy resources in Northern Thailand, Phayao province was selected as the study area (Figure 2) Since the availability and use of energy differed for each family depending on their location, their access to energy and the availability of energy, three main groups of families in the study area were classified:
1) The first class is entitled "rural forest families". They are the families located in or nearby the mountain forests. Along with some agricultural activity they are the main producers of wood energy and have easy access to the wood from the forest for their own energy needs.
2) The second class is called "rural agricultural families". They are the families located in the plains area. They are mainly agriculturists and therefore have many agricultural residues as an alternative energy to wood fuel.
VoL 6, No. 2, 1996

Figure 1: Procedure of applying micro, village and regional systems
approach (Source: Doppler, 1944)
collection of information
regional village micro level environment innovations
level level (farming and and resources and solutions
non-farming to problems
data bank
village farm- non-farm job
systems household enterprises employment
regional resource analysis of analysis of
resources village in systems non-farming
ana ysis frastructures analysis families
analysis of village livelihood systems
of failure
and success
simulation of future development and testing strategies innovations
changes in physical, economic, social, regional village farming non-farming organizational, adminimpact impact systems systems istrative and political
analysis analysis impact impact environment
analysis analysis
Journalfor Farming Systems Research-Extension

3) The third class is named "urban families". They are the families living in the municipal and sanitary areas in Muang and Dokkamtai districts and had ready access to modem energy supplies (petroleum products and electricity).
A survey of these three groups of families was done in 1993. Sixty families per class were interviewed, in total 180 samples. Figure 2: Study area, Phayao province, Northern Thailand
Thaland Lo
+ +,
Map of Thailand
Study area:
Muang and Dokkamtai districts SUrban families SRural agricultural families SRural forest families Phayao Lake 10 km
Vol 6, No.2, 1996

1. Socio-Economic Data and the Standard of Living of the Families
The standard of living of the families in the study area was generally satisfactory. The supply situation of food, health, water and housing was, in general, sufficient for rural forest, agricultural and urban families. However, there were common problems in all families in terms of family resources deterioration, for instances, degradation of soil fertility on farn land, poor management on and lack of irrigation water systems for agriculture and unemployment during slack periods of cultivation. In terms of economic success of the families, due to the rice blast disease in 1992/93, the rice yield was exceptionally low resulting in negative farm incomes. Consequently, offfarm income played an important role in this year. The average family income of all families reached a minimum income to satisfy' the basic requirements. This minimum income requirement of a family (the poverty line) is understood as the amount of income required for basic needs. However, shortages of cash occurred almost throughout whole year for rural forest families and some periods for rural agricultural families.
2. Household energy consumption and production
Firewood and charcoal, mainly from forests, were the main source of homeproduced energy for household energy requirement of rural forest and agricultural families for which they did not need cash to purchase. This implies that moving from home produced energy to the modem energy sources would require cash. This would be difficult where the shortages of cash still occur throughout the year. Modem energy (petroleum products and electricity) was the main source for urban families and wood played a minor role here. Based on an aggregated calculation between sustainable supply and actual consumption of wood energy in Phayao province, over-harvesting of wood energy from forests occurred resulting in a non-sustainable use of wood energy resource. This non-sustainable use (the gap between actual consumption and sustainable supply) equals to the amount of rice which could satisfy about half of the total population in the province a year. This loss could be saved by improving the wood energy situation.
3. Economic and environmental impacts of non-sustainable wood energy use
Economic and environmental impacts of non-sustainable use of wood energy resources are substantial. This has resulted in an increase of wood scarcity and prolonging the distance from house to the place of wood collection. Consequently, the labour requirement for wood gathering increased and less labour is available for farming. In addition, if wood supply becomes sufficiently scarce to the point where petroleum product substitutes become cheaper, cash of
Journaifor Farming Systems Research-Extension

the family will be increasingly required to pay for market fuel instead of home produced energy. As a result, the poor families will need to reduce cooking time or the number of cooked meals which would reduce their nutritional and health levels because of infrequent meals, undercooked food and changes in dietary pattern (Meier and Munasinghe, 1987). This has a direct impact on the standard of living and the labour force of the families. Since cash and labour are very important resources in farming, this would also affect the development of farming systems.
The over-harvesting of wood energy creates serious environmental impacts on the region. For instance, excessive pruning of the branches will reduce a tree's capacity for growth; removal of the more easily felled'younger trees will reduce the regenerative ability of the forest; excessive opening of the canopy through the removal of too many trees will make the forest susceptible to damage from wind and sun and affect wildlife; the removal of all residues, even to the point in some areas of sweeping up the leaves, will remove the nutrients that should return to the soil to maintain its fertility; removal of stumps, bushes and shrubs will destroy much of what remains of the soil's protective cover and binding structure; and eventually, the whole forest will be felled and disappear (El-Hinnawi and Biswa, 1981).
4. Potential improvements of wood energy resources
There are two main potential. improvements of the present non-sustainable wood energy resources in Northern Thailand.
1) reducing wood energy consumption through:
1.1 shifting to modern energy such as Liquified Petroleum Gas (LPG)
Shifting to LPG use was not preferred by most of the wood respondents, especially to rural forest families (Table 1). The main reason against shifting to LPG in the near future was that LPG is unaffordable as more cash is involved and must be continually available. A second reason was the fear of LPG explosions. The use of LPG requires some knowledge and management and the villages often heard about gas explosion. For them, wood energy is the safest fuel for cooking. This indicates a high preference for wood and an unnecessary policy to subsidise LPG. However, some respondents would like to try LPG if they earn higher income. This is due to the reasons of convenience, cleanliness and increase in a shortage of wood. Therefore, a strategy to increase family income would make LPG possible to substitute for wood.
1.2 improved cooking devices such as introducing more efficient woodstoves
More efficient woodstoves were rarely employed in the study area since normal woodstoves (Anglo) are cheap and easy to buy from the market. These
VoL 6, No.2, 1996

more efficient woodstoves, however, have many advantages over normal woodstoves such as saving wood quantity due to complete combustion (Voravech, 1992) and many benefits to individual such as saving cooking time and less smoke and to society such as reducing the wood energy shortage and environmental impact. If more efficient woodstoves are cheap and easy to buy from the market, the respondents are willing to use them. So, the introduction of more efficient wood stoves could help to reduce the demand for wood energy.
1.3 introducing alternative energies such as biogas
There was no respondent in the samples who utilised biogas. When asked about biogas, some respondents refused to apply biogas due to an unclear understanding of biogas and some did not know about biogas at all. Those with a better income status such as village heads, however, reported a desire to use biogas. Since biogas provides several advantages not only to the environment but also to household use in terms of cleanliness and faster cooking (Bumronggittikul et. al., 1993), information distribution of biogas technology on the village level is important. This must be improved to make biogas possible as an alternative energy to wood.
2) increasing wood energy supply through;
2.1 reforestation
Since the reforestation program has been implemented, the forest areas in Thailand are, in fact, still decreasing. The contribution of reforestation to wood energy is actually low. However, the development of reforestation is still required since most respondents felt that reforestation is the best solution to the wood energy problem. A successful reforestation effort is crucial. This could be achieved through both private and government agencies; involving villagers is particularly vital throughout the process (Flaherty and Filipchuk, 1993).
2.2 community forests
Community forest practices are based on the villagers' knowledge of local ecosystems. Unfortunately, this knowledge has never been accepted or valued by state organisations (Chantawong et. al., 1992, p. 156). For instance, Ban Hau Kaew Luang community forest in Phayao province faced the problem of the overuse of charcoal and firewood from the forest by other nearby villages. The local community had no real power or law to protect the forest and there was no real support from the government. Nonetheless, theoretically community forests have great potential to improve the present wood energy
Journalfor Farming Systems Research-Extension

Table 1: Opinions of wood energy respondents on a future change
from using wood energy to Liquified Petroleum Gas (LPG), the three family groups, Phayao province, Northern Thailand, 1992/93
Rural Rural Urban All
Items forest agric. fams. fams.
fams. families
n=45 n=38 n=6 n=89
% of families not wanting to change 70 55 50 62
to LPG in the future
Reasons ()
-LPG is dangerous and wood 22.6 4.8 0 14.5
energy is safer 58.1 81 66.7 67.3
-no money to change to LPG
-no problem with wood energy 3.2 9.5 0 5.5
-convenience and familiar with 9.7 4.8 33.3 9.1
wood energy use
-wood energy is important for 6.5 0 0 3.6
% of families wanting to introduce
LPG in the future 30 45 50 38
Reasons (%)
-convenient and clean 50 47.1 66.7 50
-shortage of wood energy 35.7 29.4 33.3 32.3
-lack of labour for wood collection 0 11.8 0 5.9
-time savings 14.3 11.8 0 11.8
situation if real authority is given to the village community to manage and take care the forest, as recommended by several studies (Leepreecha, Thongdeeloed and Benjavittayatham, 1991 and Chantawong, et. al., 1992). Extension and promotion of the concept of community forests is needed for both state officers and villagers.
VoL 6, No.2, 1996

2.3 Agroforestry, particularly an introduction of trees to farming systems
Agroforestry, in its many forms, has the potential of making vital contributions to solving wood energy problems. For instance, the introduction of multipurpose tree species to farming can be considered. McGranahan (1991, p. 1284), studied the case of Java and Indonesia, found that agroforestry functioned well to maintain sustainable wood energy use. Java is widely known for its productive household agroforestry, which provides rural cultivators with most of their fuelwood and other forest products. As a result, only a minority of forest dwellers were-affected by the decline in forest foraging. In addition, Clarke and Shrestha (1989b, p. 50) suggested that growing trees together with food crops on a farmland offers great potential for balancing the growing scarcity of woodfuels in rural areas. Furthermore, Soussan, O'Keefe and Mercer (1992) suggested that localised fuelwood problems and policy issues can be addressed through managing natural woodland areas and encouraging multi-purpose tree cultivation on farm-land, since the farmers' appreciation of tree species was strongly related to agricultural issues. Based on interview results, Tamarind and Longhand were ranked first by the farmers in the study area as good multipurpose trees for wood energy. In the light of this finding, the introduction of fruit trees such as Tamarind and Longhand are suggested to increase wood energy supply source for household energy requirement in the future
5. Future impacts of improved wood energy strategies
Doppler (1994) stated, "Analyses of the development of the systems, the development of solutions to problems, the modelling and measuring of future impacts of changes require the combination and integration of approaches and methods developed at various levels ... at farm/household, village and regional levels." Since there is often a difference between the family-household and regional needs, this may lead to conflicts related to resource use. Therefore, two models are developed: the family and regional models. The social costs are not considered at any level of wood energy use in the family model (with social costs). However, social costs arising from non-sustainable use of wood energy from forest resources are costs to the region. These costs are then included in the regional model.
The results of the future impact analysis of the selected strategies to improve wood energy at the family (without social costs) and regional levels (with social costs) are summarised as follows:
* There is a difference in family income between the family models
(without social costs) and the regional model (with social costs). The income per family of rural forest, agricultural and urban families is reduced when social costs are taken into consideration. This income
Journalfor Farming Systems Research-Extension

reduction per family equals the value of the amount of paddy rice needed
to supply a person for a year.
"The introduction of a more efficient woodstove to the family-household
will have impact on reducing wood consumption by 30% and slightly increasing the farm and family incomes of rural forest, rural agricultural and urban families. At the regional level, more than 50% of total social costs will be saved, and total regional incomes as well as average income per family will increase. Comparing the regional and family models, family income increased more in the regional than in the family model, even for urban families. This suggests that the introduction of more efficient woodstoves will provide benefits not only for the users but also
for everyone in the society.
" The introduction of fruit trees to farming will have an impact on
significantly decreasing wood energy collection from outside the farm and also greatly reducing total energy expenses. Accordingly, farm and family incomes will rise. On the regional level, wood gathering from forest resources will be reduced by more than 50%. Social costs to the region will be completely eliminated and total regional incomes and the average income per family will increase. The difference in family income between the regional (with social costs) and family models (without social costs) indicates that the introduction of fruit trees to
farmiing will be beneficial even for the non-farming sector.
" Decreasing LPG prices will not contribute to wood savings unless large
*reductions in LPG prices (more than 75%/) are considered. This
underlines the high level of stability and preference for firewood and
* Without any improvements, the social costs in the next decade will rise
and the amount of sustainable supply will be reduced. Any measures that increase the wood supply of I kg at a cost of less than 5.18 Baht in 1998
or less than 6.61 Baht in 2003 will be highly beneficial to the region.
In conclusion, wood will remain the major source of rural energy in Northern Thailand. Since wood comes primarily from forests, the rapidly declining forests in Northern Thailand have reduced the wood energy supply resulting in the negative environmental and economic impacts to the region. Recommendations to improve this situation are:
VoL 6, No. 2, 1996

1. For the family. In order to avoid wood energy shortages, the family should introduce
ea more efficient woodstove to reduce the amount of wood required;
multipurpose tree plantings, such as fruit trees, on farm lands to increase the supply of wood.
2. For the region. In order to avoid the over-harvesting of wood from forests and to make wood energy a viable and sustainable resource, the regional policy makers should focus activities on
esubsidising local stove industries in the initial phase to make more efficient woodstoves more available and cheaper for the family;
*enhancing rural job opportunities to increase the family income making the substitution of modern energy possible;
*encouraging multipurpose-tree plantations on farm lands and woodlot plantations for rural industries, such as tobacco curling industries, through seed subsidies;
*promoting community forests by providing the village community with real authority to manage and take care of their forests;
*implementing successful afforestation, particularly in taking care of the trees after planting.
CHANTAWONG, M. and et al. (1992). People and Forests of Thailand, In: LEUNGARAMSRI, P.
and RAJESH, N. (ed.), The Future of People and Forests in Thailand After the Logging Ban,
Project for Ecological Recovery, Thailand, pp. 151-196.
CLARKE, H.R. and R.M. SHRESTHA (1989b). Traditional Energy Programs with an open Access
Forest Resource: Policy Implications, Energy Journal, Vol. 10, Iss. 4, pp.45-57.
Department of Energy Affairs (1992). The study on rural household energy consumption in Thailand,
Ministry of Science, Technology and Environment, Thailand (in Thai).
DOPPLER, W. (1991). Landwirtschafiliche Betriebs systeme in den Tropen und Subtropen, Eugen
Ulmer GmbH & Co., Germany.
DOPPLER, W. (1994). The Role of Quantitative methods in Integrating fanning village and regional
systems approaches, a paper presented in International Symposium on Systems-Oriented Research
in Agriculture and Rural Development, on 21-25 November 1994, in Montpellier, France.
DUMRONGGITTIKUL, P. and et al. (1993). Environmental Aspects of Biogas Technology in
Thailand, a paper of Thai-German Biogas Programme, Northern Regional Agricultural Extension
Office, Chiang Mai, Thailand.
EL-HINNAWI, E. and A.K. BISWAS (1981). Renewable Sources of Energy and the Environment,
Natural Resources and Environment, Vol. 6, pp. 191-197.
Journalfor Farming Systems Research-Extension

FAO (1993). Wood energy development: Planning, Policies and Strategies, vol. I, Report on the
RWEDP Regional Meetings on Wood energy planning and policies, Field Document No. 37a.
FLAHERTY, M.S. and V.R. FILIPCHUK (1993). Forest Management in Northern Thailand: a Rural
Thai Perspective, Geoforum, Vol. 24, No. 3, pp. 263-275.
forest resource in the upper northern Thailand, in collaboration between the upper northern region
NGO and Chiang Mai University, Thailand.
McGRANAHAN, G. (1991). Fuelwood Subsistence Foraging and the Decline of Common Property,
World Development, Vol. 19, No. 10, pp. 1275-1287.
MEIER, P. and M. MUNASINGHE (1987). Implementing a practical Fuelwood Conservation Policy
the case ofSri Lanka, Energy Policy, April 1987, pp. 12-134, 1987.
Need of families in Thailand, unpublished document, Office of Community Development of
Phayao province, Phayao province, Thailand, (in Thai).
SABHASRI, S. and P. WIBULSAWAS (1992). Thai energy sources and related environmental issues,
Energy Policy, Vol. 20, No. 6, pp. 522-532.
SOUSSAN, J., P. O'KEEFE and D.E. MERCER (1992). Finding local answers to Fuelwood Problems,
Atypological approach, Natural Resources Forum, Vol. 16 (2), pp. 91-101.
VORAVECH, P. (1992). The study of the high efficient improved stoves at Tau Tae village, a paper
presented in seminar on 23-24 March 1992, at Center of Community Forest in Asia and Pacific
Regions, K.U., Bangkok, Thailand.
Vol 6, No.2, 1996

T. E. Kleynhans 2
Dissatisfaction with inappropriate solutions to complex problems lead to the development of systems thinking and systems applications in various fields, including agriculture.
An overview of the evolution of systems thinking and systems applications is given. A systems awareness in itself provides no magic solution to complex problems. Inappropriate conceptualization of a social system such as a farm system will prevent effective cooperation among members of a multidisciplinary team assisting farmers. Inappropriate understanding of a system will also cause scientists to forget that the system itself is part of~ a larger system, thus ignoring the significant effect of non-farm earnings and the multiple goals of the farm household. This can lead to ineffective extension. An inappropriate conceptualization of the farm system also does not provide for a more radical change in the goals and structure of the farm system. Systems thinking does however provide interactive management methodology
for more radical changes.
Farm systems research and extension (FSR-E) was introduced into research and extension services as a means to improve the technology generation and dissemination process for small holder farmers (Anandajayasekeram, 1995). The shortcomings of the conventional research and extension procedures can be ascribed to a monodisciplinary approach which tries to explain complex
1This paper was presented at the AFSRE Symposium in Sri Lanka in
*November of 1996.
2 Department of Agricultural Economics, University of Stellenbosch, 7600, South Africa.
Vol 6, No. 2, 1996

phenomena in terms of simplistic, single cause-effect relations. This modus operandi assumes that the farmer's goal is given and the manipulation of the single cause (eg. more efficient input use or credit availability) is necessary and sufficient to accomplish the given goal. Inappropriate technology not fitting the variety of requirements of a complete environment and therefore low adoption rates is symptomatic of this approach. Dissatisfaction with inappropriate solutions to complex problems was not unique to agriculture as the development of systems thinking and systems applications in other fields indicate. As a starting point, this paper provides an overview of the evolution of systems thinking and systems applications. This framework suggests that a systems awareness in itself provides no magic solution to complex problems. Inappropriate conceptualization of a social system such as a farm system will prevent effective cooperation among members of a multidisciplinary team assisting farmers.
A second effect of such an inappropriate understanding of a system is that scientists concentrate so much on the integrated nature of the components of a system that they tend to forget that the system itself is part of a larger system. Conceptualizing the farm system as if it exists in isolation and ignoring the significant effect of non-farm earnings and the multiple goals of the farm household can lead to ineffective extension.
A third effect of an inappropriate conceptualization of the farm system is that it does not provide for a more radical change in the goals and structure of the farm system. A common feature of systems is their rigidity and resistance to change due to the integratedness of the components. Realizing this rigidity leads to the warning that FSR-E is not a quick fix or panacea for agricultural development and that not more than tinkering to the farm system can be expected (Shumba, 1996). Systems thinking does however provide a methodology for more radical changes. Interactive management which developed out of social systems thinking provides guidelines for development of new goals and new structures. Interactive management provides opportunities for FSR-E to progress beyond revealing client's existing needs and desires which are systemically linked to existing fanning patterns.
Aristotle declared more than 2,300 years ago that the whole is more than the mere sum of its parts. Jan Smuts, former Prime inister of South Africa and philosopher who established the term holism, contributed to the formalization of systems thinking in the 1920's (Smuts, 1926). Various efforts the past 50 years made powerful contribution to the tradition of science.
Journalfor Farming Systems Research-Extension

Today the systems movement consists of contributions from various problem fields which all use the term "system". The biologist Bertalanffy promoted the use of the term "system" in the 1940's. Together with the economist Boulding (1960, 1969), the physiologist Gerard and the mathematician Rapoport (1954), General Systems Theory (GST) was developed. GST offered a meta-methodology of holism which aspired to embrace different sciences by discovering concepts, laws and models applicable to systems of all types. GST was an ambitious attempt to establish systems thinking as an independent discipline because it argued that systems were worthy of study in their own right, even though the nature of the elements making them up (mechanical, biological, human) differed. It is crucial to note, however, that GST's claimed "generality" of approach usually embodied biological analogies (Lane & Jackson, 1995). Refer to Waelchli (1992) for a discussion of GST theses.
The shortcoming of the earlier approaches in systems thinking was that the beneficial use of systems ideas in specific fields of application showed greater progress than the development of an overarching theory. The difference between systems thinking in general and applications in various problem fields is explained with a schematic outline of activities within the systems movement (refer to Figure 1).
The first distinction drawn in Figure 1 is between the development of systems thinking as such, for example cybernetics (2.1) and the application of systems thinking in existing disciplines like geography (2.2). Within systems thinking, a distinction is. made between the development of systems ideas and their mutual relationships, like general systems theory (3.1) and problem solving applications of system ideas (3.2). The latter can be subdivided between "hard" systems applications such as systems engineering (4.1), the use of systems ideas as decision making aids like operations research (4.2) and "social" or "soft" systems applications on weakly structured problems (4.3) (Checkland, 1981).
Lane and Jackson (1995) identified two more recent streams in systems thinking: firstly, emancipatory systems thinking which deals with ways in which systems approaches can be used in coercive situations to assist less powerflil groups; secondly, critical systems thinking which bases itself on critical reflection and social awareness on complementarism and on ethical commitment. FSR-E can best be placed under'"the application of systems thinking in existing disciplines like geography (2.2).
Vol 6, No.2, 1996

Natural science Social science
Biology Economics and Socilogy
Theoretical development of systems thinking
Hierarchy thoy Information Work in 'soft' systems
thoY 3.2 I-Soft systems methodology
2.1 Problem-solving
movemet 2.2 Astes development
Idy as such or systems thinking In
id hessuch pInreal-world problems
Work in 'hard' systems Aid to decislan-making
Systems anginaring mentooogy RAND systems analysis
Computer systems analysis Operations research
and system engineering Management science
The system movement
Applications of systems
thinking In other disciplines
How the diagram is built tp:
4.1 Work In 'hard' systems f 3.1 Theoretical development (engineering systems
1. The systems ideas / and problem solving)
systems 3.2 Pbm-solving 4.2 Work in 'soft'systems
movement 2.2 Application in development i (human systems)
other disciplines In real-world problems
4.3 Aid to decision-making
2.1 3.1
2.1 3.2
1 ) 1 1 12 4.11
FIGURE 1: THE SHAPE OF THE SYSTEMS MOVEMENT (Arrows indicate major influences)
Journal for Farming Systers Research-Extension

The evolution of systems thinking shows the development of more appropriate and powerful models as intellectual tools to understand complex social phenomena. Traditionally, two types of models have been used to understand these phenomena, namely mechanistic and organismic models. But in a world of accelerating changes, increasing uncertainty and growing complexity, these models are inadequate as guides to planning and decision making. Commonly prescribed remedies, inter alia a technological technofix, are increasingly ineffective and often make things worse, indicating that something is fundamentally wrong with the way we think about social systems. Holism or a systems awareness which emphasizes the presence of a variety of components and their interratedness is in itself not sufficient to understand complex systems.
Soft systems methodology does not assume that the problem is one of rational choice. The conventional management science models problem solving as a goal-orientated decision process with the crux of the problem being to identify and evaluate ways of achieving fixed and desired goals within the
*particular problem context. Soft systems methodology assumes that the designation of the problem situation, the definition of the problem, and the possible goals to be achieved are all problematic elements in themselves (Ledington, 1992).
Social systems thinking provides a typology of systems which enables one to
*judge the appropriateness of a system model to understand a specific kind of phenomenon. Conceptualizations of mechanistic, organismic and social systems are compared below.
The mechanistic model
Mechanistic models of the world conceptualize it as a machine that works with a regularity dictated by its internal structure and the causal laws of nature. The world, like a hermetically sealed clock, is taken to be made up of purposeless and passive parts that operate predictably. Any deviation from regularity is reacted to with changes that restore it; the system is believed to tend in the long run toward a static equilibrium.
This type of model is based on two assumptions: the world can be completely understood and such understanding can be obtained by analysis. Analysis is a three-step thought process. First, it takes apart that which it seeks to understand. Then it attempts to explain the behavior of the parts taken separately. Finally, it tries to aggregate understanding of the parts into an explanation of the whole, according to the doctrine of reductionism.
Vol 6, No. 2, 1996

An understanding of the whole requires establishing a relationship between the parts. The relationship is cause-effect assumed to be sufficient to explain all actions and interactions of the parts. One thing is taken to be the cause of another if it is both necessary and sufficient for the other. As identification of causes provides complete explanations of their effects, the environment is not required to explain anything. This environment-free concept of explanation assumes the environment to be static and that the system can operate-as a closed system.
Organizations are taken to be instruments of their owners with members (labourers and managers) with no purpose of their own. Mechanistically modeled organizations are structured hierarchically and are centrally controlled by a complete autonomous authority.
The operations of an ideal machine do not vary. Therefore, as long as its input does not vary, its output won't. The result is that controllers (and supporters) of a mechanistically modeled social system focus on inputs rather than outputs.
The implications for researchers and extension officers conceiving social systems mechanistically is that they diagnose problems in a simplistic way and regard some or other input e.g. a technological device as necessary and sufficient for the efficient operation of the farm. The fallacy of this inappropriate model is that researchers and extension officers easily accept the goal of the farm system as given and project their supply driven goal onto the farm situation. The reason for low adoption rates of technology is seen as irrationality on behalf of the farmer.
The Organismic model
A social system conceptualized as an organism has a purpose of its own: survival for which growth is taken to be essential. Such a system is assumed to be dependent on its environment for essential inputs (resources). In a changing environment the system must be capable of learning and adaptating in order to survive. In contrast with a mechanistic model, the organismic model seeks a dynamic equilibrium. In order to adapt, planning becomes prediction of environmental changes and preparation for them.
Research and extension based on an organismic conception of a farm accepts growth and optimal production as given goals and apply all available resources to accomplish this goal, without taking account the wider socioeconomic and ecological consequences. The well known farm problem of declining farm incomes due to technological advance and resultant production exceeding consumption, exacerbated by low price elasticities of demand for agricultural products can be ascribed to a production-and-growth-at-all-cost approach to research and development. Environmental degradation due to this approach also indicates the need for a framework of understanding the world
Journalfor Farming Systems Research-Extension

which allows for adaptation of goals and not only changing of means to accomplish fixed goals.
The Social System Model
A social system is a purposeful system and its parts are recognized as purposeful systems and they are part of larger social, purposeful systems. To understand a social system, one must know the ends or goals of the components, the system, and containing systems, and how these affect their interactions. A purposeful system can produce the same outcome in different ways in the same environment and can produce different outcomes in the same and different environments. It can change its ends under constant conditions. This ability to change ends under constant conditions is what exemplifies free will. Such systems can learn, adapt and create. Human beings are examples of such systems.
Purposeful systems have all the capabilities of goal-seeking and statemaintaining systems. Goal-seeking systems such as the farm system have the capabilities of state-maintaining systems. The converse is not true, therefore they form a hierarchy.
Another hierarchy relevant to this systems hierarchy is the hierarchy formed by information, knowledge and understanding. Information is descriptive; it contains answers to questions that begin with such words as 1 hat, which, who, how many, when and where. Knowledge is instructive; it is conveyed by answers to how-to questions. Understanding is explanatory; it is transmitted by answers to why questions. To understand a system is to be able to explain its properties and behaviour and to reveal why it is what it is and why it behaves the way it does. Information presupposes neither knowledge or understanding. Knowledge presupposes information and understanding presupposes both. One can survive without understanding, but not thrive. Without understanding one cannot control causes; only treat effects and suppress symptoms. With understanding one can design and create the future.
The cause-effect relationship is inadequate to understand the functioning of a social system. The producer-product relationship is required. A producer is only necessary, not sufficient for the product. Therefore, explanation become environment-full, recognizing multi-causality which is typical of complex phenomena.
The appropriate purpose of a system conceptualized as a social system is development. Development is defined as the process in which individuals increase their abilities and desires to satisfy' their own needs and legitimate desires and those of others. Because development involves an increase of ability through learning and one person cannot learn for another, one person or organization cannot develop another. One can only encourage and facilitate that development. There is only one type of development: self-development.
Vol 6, No. 2, 199 6

Effective development planning requires a major reorientation of organizations or governments that assume that they can develop their members. Development is not a matter of what an organization or government does, but of what it encourages and enables its members to do.
A thorough understanding of development as learning is necessary before an extension officer can really be an effective facilitator to farmers. With inappropriate conceptualization of social systems, facilitation becomes a mere buzz word with the result that extension officers still act according to a preconceived goal and concentrate in effect only on means planning. This will happen despite following the FSR-E principle to reveal the variety of goals and considerations that a fanner may have.
The quality of life that a person can realize is the joint product of his/her development and the resources available to him/her. Although this implies that limited resources may limit improvement of quality of life, it does not imply that they limit development. Desires and abilities can increase without an increase of resources. Development is potentiality for satisfaction of needs and desires, not the satisfaction (quality of life or standard of living) actually obtained.
The effect of limits or obstacles on purposeful individuals and systems can be evaded either by changing intent or by using better technology. A limited resource ceases to be limiting if one's need for it decreases due to development of desires or if one learn how to use it more effectively through the development of abilities. Goal or ends development, over and above means development is an important characteristic of social systems.
To understand the farm as a complex social system (which includes physical and biological dimensions), a whole new approach is needed which is not only analytical. Contrary to a widely held belief, a multidisciplinary approach does not fill this need if it boils down to only a collection of analytically orientated approaches. To understand a system its structure, processes and functions have to be examined. A system's structure is the way its work is divided among its parts and their efforts co-ordinated, that is, the relationships between its parts. Structure can only be understood if observed in the functioning of a system. Therefore, analysis, which reveals only the structure of a system, not its functioning, cannot provide understanding, only knowledge. Synthetic thinking is therefore required to complement analysis.
The danger of inappropriate ways of conceptualizing social systems lies in the underlying assumptions about the nature of a social system of which the researcher or extension officer is not aware. A person does not deliberately perceive a farm system as a mechanism or organism, but in effect understands it inappropriately and plans accordingly. Exposure to the typology of systems provided by social systems thinking may help scientists to identify and clarify their own assumptions and to accept a more effective and powerful frame of
Journatfor Farming Systems Research-Fxtension

understanding farm systems. (For a more complete discussion of the systems typology, refer to Gharajedaghi, 1985).
Cultural research has shown that individuals can be encouraged to acquire new technical competencies (artifacts) without doing violence to long-held beliefs and assumptions that emerge out of lengthy socialization processes. However, inviting individuals to embrace new theoretically and philosophical positions is significantly more problematic, (Brochlesby & Cummings, 1994). For a discussion of the meta-crisis of perception of professions that shape the human-made environment and the need to shift from the "object" level to the meta-level of facilitation of community aspirations, refer to Motloch & Woodfin, (1993).
A mechanistic or organismic conception of the farm system sees the farm as environment free and tries to optimize the use of resources available within the farm border. Experimental design and farm plans are based on the assumption that the farmer tries to maximise production, influenced to a greater or lesser extent by socio-cultural considerations. However, deviations in farming practices of small farmers in parts of South Africa, resulting in idle land and yield increasing technology (such as fertilizer) used in a labour saving way indicates other considerations. Due to the availability of non-farm income earnings opportunities and a higher return on labour inputs for some of the household members, farm households tend to maximize their total income from the farm and non-farm sources (Low, 1986). An inappropriate
conceptualization of the farm system puts the extension officer in a dilemma with regard to his/her ability to understand the broader context and considerations of farm households and to adapt agricultural production advice accordingly.
Low (1986) argues that although FSR-E recognizes that farmers have multiple objectives, these objectives are generally looked at in terms of the farming system alone with two undesirable results: "First it may encourage researchers to think of those who firm as primarily or solely farmers, and thereby underestimate the role of non-agricultural activities in the larger household economy. Secondly, an exclusive concentration on farming may ill equip FSR to address one of the major issues in agricultural development in Africa: the withdrawal of labour from agriculture due to rural-urban migration". He advises the adoption of a household economic perspective and attempt to see how diverse production activities are combined to maximize household utility.
Vol 6, No.2, 1996

Realizing the broader environment and wider variety of objectives of the farm household provides the challenge for extension officers to be able to facilitate the farmer to choose effectively his/her farming and non-farming operations. Although the agricultural extension officer does not have to be an expert on for instance processing, marketing or transport activities, he/she should be able to recognize worthwhile opportunities and realize the implications for farming practices in order to find an overall better mix of economic activities. Such challenges should be realized when designing curricula for training of extension officers.
A product, service, organization and therefore also a farming operation is designed effectively if it provides consumers or users with what they want, rather than merely remove what they do not want. But determining what they need or will want is an effort that does not often meet success. The reason for this is that consumers or users' needs and desires are elusive because consumers or users themselves generally have not consciously formulated what they are or how to fulfill them.
FSR-E aims to identify farmer's objectives to be taken into account in the design and transfer of technology. One does get the impression however that the information obtained is more often used to increase the efficiency of the existing production system which boils down to tinkering with the farm system. Shumba, 1996 stated that more than tinkering with the farm system cannot be expected from FSR-E to cool down expectations of FSR-E as a panacea for agricultural development problems. Mere tinkering with existing farm systems is probably the most realistic expectation of FSR-E if the typical rigidities of any system to change are taken into account. This is implied in the term systemm" and is due to the integratedness of the components of a system (Kauffman, 1980). However, this does not mean that systems based thinking can only offer marginal improvements in the form of tinkering with a system.
Social systems thinking provides a methodology called "Idealized design" of a system or "Idealized redesign" of an existing system to identify new or concealed needs and desires and to design new structures and processes to satisfy these needs and desires.
Idealized design provides an opportunity to a person (eg. a farmer) or a group of clients to start planning from a zero base situation and reinterpret his/her, whole situation (farming and possibly non-farm activities). The goal of "Idealized design" is to escape from mental obstacles of being used to conventional farming practices and lifestyles in order to develop new goals to become more effective in a changing environment.
Idealized design gives participants the opportunity to create an unconstrained design of their ideal fanning operation (or rather broader income
Joui'nalfor Farming Systems Research-Extension

generating operations). Participants are told not to be concerned with the feasibility of the designs they create, only with their desirability. This is based on the belief that the principal obstruction to creativity is a preoccupation with feasibility, a condition that is usually associated with self-imposed (rather than actual) constraints. The only two constraints that have to be taken into account are: the end result cannot involve any technology that does not currently exist and it must conform to the law.
Idealized design assumes that, given proper tools and facilitation, average participants are often best equipped to design, from a functional standpoint, those products or production systems that are required for situations with which they have become familiar. It is this input from participants early on in the development process that differentiates idealized design sessions from traditional focus groups and surveys.
The facilitator (who can be the extension officer) is supposed to guide but not to provide content to the session. He/she asks the participants to imagine that an existing product/farming operation with which they are familiar was destroyed overnight and that they have the opportunity to create something totally new in its place. They then engage in a brainstorming session to prepare a list of specifications (ideal characteristics) for the ideal product/farmdig operation to be designed. Specifications can include any feature desired by the participants, no matter how outrageous. The entire group debates the merits of each point raised and finally arrives at some decisions regarding the ideal. Participants then break into smaller groups to plan designs that will incorporate as many of the specifications as possible. The advantage and disadvantages of each design are then discussed in the larger group, after which the smaller groups reconvene to refine and change their designs. Tids process is repeated as many times as possible with the goal being to arrive at one design that incorporates all of the participants' idealizations. (For elaboration on the methodology of idealized design, refer to Ciccantelli & Magidson, 1993 and Ackoff, 1981).
A systems awareness per se is not new; every agricultural scientist has some or other concept of a systems on which his/her research and/or extension is based. To use the term "systems approach" as a buzz word so that people think they understand it a similar way is aqtually detrimental to effective cooperation in research and extension. Agricultural researchers and extension officers need clarity on their underlying assumptions to find out if they correspond with the essential characteristics of the farm system. The systems
Vol1 6, No. 2, 1996

80 T. E. KLEY xiTANS
typology provided by social systems thinking can be of value in this clarification process.
The socio-economic environment of the farm system affects the farm household's decision-making in an important way. An inappropriate
conceptualization of the farm system will cause the extension officer and researcher to ignore these influences and design technology and make recommendations as if the farm exists in isolation. The effectiveness of
research and extension should however be measured in terms of how it enhances a higher quality of life and standard of living in the broader environment. Technology design and recommendations will have to take cognisance of the urban-rural interaction and its effects on the availability of labour on the farm.
The changing environment demands adaptation not only of farming procedures through means planning aimed at increased efficiency, but also creating new goals for greater effectiveness. Given that conventional farming practices become obsolete, the question arises whether FSR-E has more to offer the farmer than just the support of means planning. Social systems thinking claims that in a rapidly changing environment tinkering with a farm system is not sufficient and that goal development can be effectively supported by applying idealized (re)design as an interactive planning method. This powerful planning tool can be adjusted to various circumstances by following a few simple guidelines.
Ackoff, RL. 1981. Creating the corporate future: Plan or be planned for. John Wiley &
Sons, New York.
Anandajayasekeram, P. 1995. Fanning systems research: Concepts, procedures and
challenges. Keynote address presented at the Fourth Regional Conference of the Southern African
Association for Farming Systems Research-Extension, Harare, Zimbabwe, 2-4 October 1995. Boulding, KE. 1966. Economic analysis. Harper & Row Publishers, New York. Boulding, KE. 1969. Economics as a moral science. American Economic Review, Vol. 59. Brochlesby, J. & Cummings, S. 1994. Combining hard, soft and critical methodologies in
systems research: The cultural constraints. Systems Research, 12 (3).
Checkland, P. 1981. Systems thinking, systems practice. John Wiley & Sons, New York. Ciecantelli, S. & Magidson, J. 1993. From experience: Consumer idealized design:
Involving consumers in the product development process. Journal of Production Innovation
Management Vol. 10.
Gharajedaghi, J. 1985. Toward a systems theory of organization. Intersystems
Publications, Seaside, California.
Kauffman, DL. 1980. Systems One: An introduction to systems thinking. The innovative
learning series. Future Systems, Minneapolis.
Journalfor Farming Systems Research-Extension

Lane, DC. & Jackson, MC. 1995. Only connect: An annotated bibliography reflecting the
breadth and diversity of systems thinking. Systems Research, 12 (3).
Ledingtin, P. 1992. Relevance, formality and process: Toward a theory of soft systems
practice. Systems Research, 9 (4).
Low, A. 1986. Agricultural development in Southern Africa: Farm household-economics and
the food crisis. David Philip, Cape Town.
Motloch, JL. & Woodfin, T. 1993. General Systems Theory, cultural change and a human
science foundation to planning and design. Systems Research, 10 (2). Rapoport, A. 1954. Operational philosophy. Harper & Brothers, New York. Shumba, E. 1996. Importance of farmer circumstances in decision making. Paper delivered
at a national workshop on FSR strategy development, Silverton, 1996. Smuts, JC. 1926. Holism and evolution. Macmillan, London. Waelchli, F. 1992. Eleven theses of General Systems Theory (GST). Systems Research, 9 (4).
Vol 6, No.2, 1996

Chambers, Robert. Whose Reality Counts? Putting the first last. London: Intermediate Technology Publications, 1997. 297 pp.
Robert Chambers' book Whose Reality Counts? is a gem. The subtitle of this book Putting the First Last is a logical sequence to one of his earlier books Rural Development: Putting the Last First (Longman, 1983). Chambers' analysis of events and attitudes is a sobering challenge to all professionals in development work, especially for those concerned with fanning systems research-extension.
Chambers points out that at the start of the 21st century "we, humankind, have more power and more control over things, and are more closely and instantly connected with each other, than ever before. At the same time, more people than ever before are wealthy beyond any reasonable need for a good life, and more are poor and vulnerable below any conceivable definition of decency. New power, knowledge and social and economic polarization coexist on an unprecedented and scandalous scale ....Hundreds of millions of people are worse off now than twenty years ago. That some nations should be rich and others poor can even seem inevitable as we watch, year by year, the indicators of well-being improve in some and decline in others, with lower incomes, fewer children in school, deteriorating services in health, mounting civil disorder, lower expectations of life, and greater vulnerability." (p. 1)
This is not a challenge Chambers takes lightly. He has led workshops on participatory rural appraisal (PRA) in many countries in Asia, Africa and Latin America in an attempt to help professionals in development work develop their own skills in putting the last first. Those who do that, therefore, put themselves last. Chambers has written many articles found in IDS Bulletins, World Development and other publications. Much from these earlier publications is combined in this book to address the over-riding issue of how those of us who work in the development field can put the reality of the rural men and women at the forefront.
The 'discovery' of what rural people know, what they need and want and how they put new technical knowledge into the mix of factors comprising their reality is the heart of Chambers' concern. This is no doubt increasingly
VoL 6, No2, 1996

important as fewer development professionals come from rural areas where they grew up intimately involved in livelihood activities siilar to the ones being done today by the world's most vulnerable women and men.
Chambers points out that the development assumptions of some professionals in the '50s and '60s have been exposed as misconceived, and "'with the easy wisdom of hindsight, naive." All of us have to appreciate his willingness to admit his own failings in these categories -- and to appreciate even more the benefits of learning from this experience. His chapters on "All Power Deceives" (Chapter 4) and "Learning to Learn" (Chapter 5) are built on his earlier writing. He points out in Chapter 4 that "Being clever, centrally placed, served by instant communications of unprecedented quality, and powerful does not, then, ensure being right or doing good" (p.99). Chapter 5 is somewhat a 'recipe' for PRA, parts of which will be familiar to readers from earlier Chambers' articles and lectures.
The focus of Chapter 8 "Poor People's Realities: Local, Complex, Diverse, Dynamic and Unpredictable" (LCDDU) is the heart of this book. These characteristics of the reality have always been there, but it took participatory approaches and practices (listening, giving the speaker the "stick", etc.) to bring it to the conscious level of development professionals. PRA has made a significant contribution. While I find the acronym 'LCDDU' annoying, it clearly establishes a base for understanding that there really can be no recipe which will address poor people's realities. (Although perhaps PRA is seen as the beginning of a recipe). In this chapter, the description of each of these characteristics and their multiple influences on poor people's realities is very useful.
In Chapter 10, "Putting the First Last" Chambers suggests strategies and tactics for institutional change and elaborates on these 6 steps: 1. commit with continuity 2. network with allies 3. start small and slow 4. fund flexibly 5. train, encourage and support grass-roots staff 6. Build out and up for grass roots success. (pp226 and 227).
He closes this chapter and the book with specific challenges to the various groups of people working as professionals in the development field: "The issue is whether we, as development professionals have the vision, guts and will to change our behaviour, to embrace and act out reversals, and:
* as economists and bureaucrats to decentralize, destandardize and support
local diversity;
* as staff in NGOs to continue to evolve, apply, share and spread
participatory approaches and methods;
" as teachers in Universities, training institutes and colleges, to go with our
students to local people to learn, to revise our curricula, to rewrite our
textbooks, to teach and lecture less, and more to help others learn;
* as staff in government organizations, not to talk down but to listen, learn
and facilitate, and to provide choices and responsive services;
Journalfor farming System Research-Extension

* as political leaders to promote and sustain decentalization, democratic
values, tolerance, peace and the equitable rule of law;
* as people, to be self-critically aware, to respect others, and to value truth,
trust and diversity;
* as uppers, to disempower ourselves, controlling only the minimum,
handing over the stick, devolving discretion, encouraging and rewarding
lowers' initiatives, and finding fulfillment and fun in enabling others to
express, analyse and act on their diverse realities". (pp. 236 and 237).
As an occasional working colleague of Robert Chambers, the most negative reaction I have is to his use of "uppers" and "lowers". While it may accurately define each group's view of themselves, it also very condescending, even if Chambers didn't intend it to be --and that may be the heart of the problem: we are trapped (by years of refinement and use) in condescending language, whatever our attitudes really are.
Chambers' ability, and willingness, to take blame or responsibility for early mistakes in development activities stands as an example for all of us. And his skill at playful rhymes adds spark to the whole text. This little excerpt from one is an example:
"Then the poor might reply:
Donor, we reject your song Top-down targetry is wrong
Floods of funds as in your verse
Corrupt and spoil and make things worse.
Keep your money. We will show
True development's from below." (p.225).
His stature in the development community which encourages various Foundation contributions and ITD to publish "Whose Reality Counts?" in paper and at a subsidized price will be greatly appreciated by all development practitioners at the field level. Robert Chambers is a thoughtful, humble man not afraid to keep learning and sharing what he has learned a great example to all of us.
N. A.
VoL 6, No2, 1996

Hedrick, Ulysses P., The Land of the Crooked Tree, New York: Oxford University Press, 1948, 350 pp. (republished and available in paper back by Great Lakes Books, Wayne State University Press, 1996)
The author of this book never used words like farming systems, or even sustainable development, or international development. He was the son of a farmer, who lived his first sixteen years working on the family farm, with his brother and father, as well as a younger sister and his mother. And since he died long before AFSRE was organized, he could not have known of the Journal for Farming Systems Research-Extension, or submitted this book for review.
But although Ulysses Prentiss Hedrick later went on to a college education, became a professional agriculturalist, and an internationally known horticulturalist, in this book he describes life on a small, mixed farm family ecosystem, seen through the eyes of the child and growing boy that he had been. Having read it by chance this past month, while assembling materials for this issue of JFSRE, I was struck with the quality of a description and analysis of a farming system which goes into more detail and greater depth than current literature usually permits.
As writers for this Journal know, our "criteria" statement for regular articles specifies that "Articles will address farms as whole systems, including their production and consumption, and involving the plants, livestock, and humans on the farming system, or the institutional support services for farming systems, including the policy structure." In this book, Dr. Hedrick does everything we request in that sentence.
By 1948, when this book was published, Hedrick was a senior scholar/practitioner, and Director Emeritus of the New York State Agricultural Experiment Station. He must have reflected back more than sixty years, (perhaps with notes?) to write the manuscript. In his preface, he explains that "The Land of the Crooked Tree, found on old maps as 'LArbre Croche,' is on the northern tip of the lower peninsula of Michigan and was one of the last forest regions in eastern America to be settled .... It is with the activities of the early settlers there, in the 1870's and 80's, after it was opened to homesteaders, that this book is concerned."
Individual chapters in the book include detailed and intimate descriptions of such aspects of farm life as Our Farm, Father's Garden, Nature's Gardens, Mother's Poultry, Our Livestock, Farm Crops, Farm Chores, When the Railroad Camne, Parker's Store, The Blacksmith Shop, Forest Products, Preachers and Churches, and much, much more. There are thirty-one chapters. If you are really interested in identifying and describing all of the components of a farm family ecosystem, as well as the linkages among them and linkages
Journalfor farming Systems Research-Extension

BooKRF-vmws 87
with relevant outside systems (for supply of inputs or marketing of outputs, for example), you will find them all in this book. Here are some excerpts:
From his first look at the farm, in 1874, at age four: "The farm before our eyes was a forest black as night. In it no ax had ever been lifted. F cept for the rough trail over which we had come, there was no sign that man had ever set foot on our land. The appalling stillness seemed neverto have been broken by human voices. (p.61) ... Although we began in a small way to clear land the first summer and kept doggedly at it month in and month out until the sons grew up and the father grew old, it was three years before we built a house on our land. Mother would not live in an unbroken forest where there were no roads, no water, no near neighbors, and no school for her children." (P. 63).
Regarding a time a few years later, he wrote: "Father's garden was wholly utilitarian. In it he grew plants to furnish food for the year around and to add to our small supply of cash. It was planted with all the vegetables that would grow in our climate. In his garden, Father was tireless. He worked from daylight to dark, doing as much as two men, and in a spurt multiplied himself by three." (p. 106). "The first task in our garden was seed-sowing; the second was weeding. It would be difficult to say which was the most wearisome. There was this difference. Father was always with me when there were seeds to be sown, to see that all was done well; I worked alone at the weeding." (p. 107)
"It would be hard to say which was Father's favorite vegetable. Onions came earliest, and onions, green or dry, we had almost the year round; ten bushels of dry onions were put in the cellar every winter. For early onions, we planted small sets, but our main crop came from seed." (p. 111) "When tourists began to come in large numbers, the demand for green peas could hardly be satisfied, and peas became the best money-making crop we grew." (p. 112)
"String beans, which Father and Mother called by their Virginia name ,snaps," grew well with us. Our season was too short for the lima beans of that period and we grew them but sparingly. The markets called for yellow wax beans, but for home use we greatly preferred green pods, growing, of course, only stringless kinds. (p. 113) "The poles were set in the ground two feet, and three strong cords, at heights of three feet, looped them together. Around each pole he -planted six beans. When planted in late May the beans quickly climbed to the top of the poles, and clambered along cords until the row of hills made a solid wall of foliage, flowers, and beans. In picking, one could grab a handful of tender succulent 'snaps' for autumn eating. We saved pole-bean seeds year after year." (p. 114)
That chapter goes on with even more detailed descriptions of such other vegetables as sweet corn, cabbage, tomatoes, cucumbers, strawberries, and watermelon. In each case, Hedrick describes the plants themselves; the techniques of planting, cultivating, and harvesting; which family members did which tasks, and in which seasons; as well as how the crop was used in the
VoL 6, N62,1996