Farming systems research and extension at CATIE 1975-1985. Notes and observations

Material Information

Farming systems research and extension at CATIE 1975-1985. Notes and observations
Jones, James C.
Centro Agronomico Tropical de Investigacion y Ensenanza


Subjects / Keywords:
Caribbean ( LCSH )
Farming ( LCSH )
Agriculture ( LCSH )
Farm life ( LCSH )
Spatial Coverage:


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.

Full Text
A Report Prepared for Centro Agrono'mico Tropical
de Investigacidn y Ensefianza
By James C. Jones, Consultant July, 1985

Acknowledgements ...................................... ...... iv
Aims and Approach ............................................. vi
Small-Farm Production Systems Research at CATIE:
A Brief History ............................................ 1
Guatemala ................................................. ..... 5
History and Setting ........ ..... .. ..................5
CATIE Activity in the Field and its Impact ............. 6
Other CATIE Activity and its Impact ................... 7
The CATIE Methodology .............................. .. 9
Institutional Relations and Operational Problems......10
El Salvador .................................................. 12
History and Setting .............. ......................... 12
CATIE Activity in the Field and its Impact ............ 14
Other CATIE Activity and its Impact .................. 19
The CATIE Methodology ................................. 21
Institutional Relations and Operational Problems ...... 21
Honduras ....... ..... 0....................... ................. .24
History and Setting ........... ........................ .24
CATIE Activity in the Field and its Impact ............ 27
The CATIE Methodology ................................. 30
Institutional Relations and Operational Problems ...... 31
Panama ....................................................... 33
History and Setting ............. .................... .... 33
CATIE Activity in the Field and its Impact ............ 34
The CATIE Methodology ................................. 37
Institutional Relations and Operational Problems ...... 39
Costa Rica ................................................... 41
History and Setting ................................... 41
CATIE Activity in the Field and its Impact ............42
The CATIE Methodology.............. ... .............. 45
Institutional Relations and Operational Problems......46

Analysis and Conclusions ..................................... 48
CATIE, National Sovereignty, and the Demands
of Farming Systems Research and Extension .......... 49
The Dual Track: Basic Research Versus Pragmatism ...... 52
Research on Cropping Associations
as Systems Research ................................ 53
On Donor Pressures and Interventions .... 54
On Validation and Transfer ........................
On. the Nature of Technological Alternatives ........... 56
On Site Selection ...................................... 57
Acronyms Used ................................................. 58

The research for this report would not have been possible without the generous assistance of many people. My thanks to Rodrigo Tarte, Director General of CATIE, for hi recognition of the difficulties of the endeavor; to Romeo Martinez, who supervised the assignment as head of the Crop Production Department, for his friendship and general support throughout; and to Saritza Chavez, secretary in the Crop Production Department, for serving with efficiency, patience, and kindness as logistical coordinator for my country visits and other activities. I am also grateful to Carlos Burgos, Luis Navarro, Jose Arze, Franklin Rosales, and Donald Kass of the CATIE technical staff in Turrialba for responding to my ideas and for discussing CATIE-ROCAP project activities with me in a free and open fashion.
My assignment undoubtedly created true suffering for the CATIE country residents and technical personnel, for to those unfortunate souls fell the daunting task of playing cicerone to one more intrusive gringo. They arranged my local travel, scheduled private interviews for me with nationals, kindly honored my occasional requests to deviate from a prepared program, and responded to my probing queries with a grace unique to the Latin world--all as if they had done it many times before. My heartfelt thanks, therefore, to Bladimiro Villeda in Guatemala, Joaquin Larios in El Salvador, Roger Meneses in Honduras, and Washington Bejarano in Panama.
Numerous nationals who gave me time from often busy
schedules to discuss CATIE-ROCAP project activities in their countries deserve special recognition and thanks. Much of the content of this report is based on their observations and comments. Their names should remain anonymous, but they include administrators and technical people at national, regional, and local levels. And they also include a few farmers from El Salvador, Honduras, and Panama.
I am indebted to several other persons at CATIE who helped me. They include Dora Maria Flores for ferreting out materials; Amyel Locatelli, whose flawless English was brought to bear on the typing of early drafts; and Maricela Chavez for general secretarial support. Thanks also to Carlos Luis Araya and Emilia Solis, who assisted with the arrangenpents for my visit to field sites in-Costa Rica; and to Ing. Solis again for obtaining copies

of research and extension organization charts for the several Central American countries.
And my gratitude goes also to Rufo Bazan and Jorge Soria, formerly with CATIE and now with IICA in San Jose, for the time they freely gave to discuss with me the early history of multiple cropping research at CATIE. Without their assistance, and that of several other persons, I would not have been able to provide a historical perspective on the CATIE-ROCAP projects.
And last, I express my gratitude to ROCAP for approving. the consultancy, and to the University of Florida for releasing me from painful managerial duties in Gainesville to execute the assignment.
I alone, of course, am responsible for any errors of fact, judgement, or interpretation appearing in the report.

The research for this report was conducted under the
auspices of the Centro Agronomico Tropical de Investigacion y Ensefianza(CATIE), with funds(grant no. 596-0000-6-00-4091-00) from the Regional Office for Central America and Panama'(ROCAP), United States Agency for International Development(USAID). The research was commissioned to assess the cumulative impact of two CATIE-ROCAP farming systems research and development projects, the first from 1975 to 1979, the second from 1979 to June of 1985. The projects have been active in the countries of Guatemala, El Salvador, Honduras, Panama, Costa Rica, and Nicaragua. The research entailed visits to five of those countries; Nicaragua was excluded, although the projects have been quite active there and, according to CATIE staff, have made a substantial impact. This report is to be a basic reference document for a final evaluation of the second project. That evaluation is now scheduled to occur during late July and August and to be conducted by USAID's Farming Systems Support Project, managed by the University of Florida.
In more precise terms, the charge of the research here described was to assess the impact of the two projects through measuring the degree of adoption of the CATIE methodology by national research and extension institutions, and the degree to which improved technologies for small farmers developed by that methodology had been promoted by those same institutions.
But I soon realized that it would not be easy to assess project impact using those measures, at least in any systematic way in the limited time allotted to the assignment. At the same time, I was becoming aware of the important influence on the projects of the national institutional settings where they had been active, and of some of the operational problems that CATIE had had in those settings. I therefore enlarged the assigned scope of work on the march to include those influences and problems.
The report opens with a brief chapter on the history of production systems research at CATIE, followed by chapters on each of the five countries, which appear in the sequence that I visited them. Each country chapter includes separate sections on CATIE activities in the field and the CATIE methodology. And each includes an opening section on history and setting, a later

one on institutional relations and operational problems. A final chapter provides an analysis and conclusions. I have striven throughout the report to bring life to the projects, to place their activities in a human and institutional setting, and to capture the historical flow of their events.
I would hope that this report might be useful not only to the evaluation team, but also to USAID and CATIE management in their understanding of the demands of farming systems research, and especially of some of the problems that arise when an international research institute conducts this kind of research in national settings.
I gathered the information for this report alone,
conducting interviews in Spanish with a wide assortment of people in five countries. The report is based almost entirely on those interviews. I sought individuals who had been key actors in CATIE country activities, or at least were close enough to those activities to be reliable observers. I thus talked with directors of research and extension at the national level and with their counterparts at the regional level, where CATIE had worked. I also talked with CATIE administrative and technical staff in the countries and in Turrialba. For a historical perspective, when that was not to be had from individuals involved in CATIE activities at the time of my visit, I consulted with knowledgeable persons as I could find them.
I visited at least one CATIE field site in each of the
five countries, selected for reasons of convenience, or security, or because CATIE felt it had had the greatest impact in that region. The schedule for the country visits was hectic and demanding. I left Gainesville for Costa Rica on May 12 and remained at CATIE until May 15, reading materials and making the necessary logistical arrangements for the five-country tour. On May 15, I left San Jose for Guatemala, where I visited with officials at ICTA headquarters and spent a day in Chimaltenango, talking with officials there. Then to El Salvador on May 19, where I met with CENTA officials and flew north to San Miguel on May 20, thence to Jocoro by land. I spent a day interviewing CENTA officials--and one farmer--in the San Miguel-Jocoro area before returning to San Salvador by plane that evening.
I flew to Honduras on May 22, talked with SNRA officials in Tegucigalpa the same day, then journeyed by car the following day to Comayagua, where I met with regional officials and visited several CATIE research sites in the valley. On Saturday, May 25, I left Honduras for Panama"City, thence by air to David on Sunday. I visited CATIE research sites in El Progreso on Monday and talked with IDIAP officials both there and in David. Then by car to Guarumal, on the southern coast of Panama, on Tuesday, June 28. Again, meetings with IDIAP officials at the Sona station and visits to collective farms where CATIE had been

active. I continued by car in the afternoon to Santiago, location of IDIAP seat for the Central Region, and talked with officials there before driving to Panama City, arriving in the evening of the same day. I met with IDIAP officials at central headquarters most of the following morning, then caught a noon flight to San Jose, Costa Rica, for an interview at IICA. I returned to Turrialba in the evening and devoted the following two days--May 30 and 31--to visiting CATIE sites in Limon Province, eastern Costa Rica, and to talking with government officials in San Jose. I remained the following week in Turrialba interviewing CATIE staff and sorting through a sea of information from five countries. I managed to produce a rough draft of two country reports before leaving Costa Rica for Gainesville on Sunday, June 9.
In an ideal world, the procedure just described leaves much to be desired. It would have been desirable to begin the interviews, then to follow leads as they emerged. And it would have been desirable to visit more sites, perhaps to read more before collecting field information. But a lack of time and resources made all of this impossible.
Certain things should be said about the report. First, it is biased toward the second CATIE-ROCAP project, despite my charge to cover both. But this was inevitable: the first project had receded in time and was thus less accessible verbally. Also, many of the key actors had disappeared from the scene. Coverage of the period 1975 to 1979, therefore, is incomplete.- Second, project training activities are dealt with only summarily. I pursued the topic only in Guatemala, where I decided that the information I was getting did not justify the time and effort required to obtain it. Yet CATIE has devoted considerable effort to the training of nationals both in Turrialba and in countries of the region. And third, little on CATIE internal dynamics appears in the report. But these dynamics have influenced the impact of the CATIE-ROCAP projects in the field. The Animal Production Department, for example, has not always favored systems research and was reluctant to participate in the research on mixed systems.
James C. Jones
Gainesville, Florida
July 19, 1985

The Center Agrondmico Tropical de Investigacion y
Ensefianza(CATIE), a creation of the Instituto Interamericano de Ciencias-Agrfcolas(IICA; today the Instituto Americano de Cooperacion para la Agricultura) and the Government of Costa Rica, was founded in 1973 with seat in Turrialba, Costa Rica. CATIE was really the reorientation and restructuring of IICA's Training and Research Center at Turrialba, an agricultural research and training arm of the Organization of American States(OAS) to serve the Americas. Among other things, the reorientation involved a narrowing of geographic focus to the tropics of the Central American Isthmus and the Antilles.
Systems research at CATIE, first managed by the Tropical
Crops and Soils Department(today the Crop Production Department), began with an interest in multiple cropping just prior to the founding of CATIE. A small group of researchers were concerned that agricultural research in Central America and the Caribbean was not addressing the needs of small farmers, who produced most of the food for the region. But they also knew that those farmers widely practiced polyculture, or multiple cropping, yet agricultural research, which had worked entirely with monoculture, knew little about cultivating in associations.
To redress this deficiency, the group of researchers
initiated the famous Central Experiment in 1972. Conducted on the station in Turrialba, the experiment researched associations of regional crops--corn, beans, rice, cassava, and sweet potatoes. The crops were studied in monoculture as well as in numerous combinations and rotational patterns. Input levels were also varied in recognition that small farmers of the region worked with very limited resources. The experiment was concerned not only with yields, but also with creating stable cropping systems, for most small farmers had little land and had to work it continuously. The more promising associations and technologies from the Central Experiment were singled out for further research in "satellite experiments," also conducted in Turrialba. The Central Experiment, or at least parts of it, was continued until about 1978.
This early CATIE research effort, overwhelmingly
agronomic, has been both maligned and praised over the years. Detractors of the Central Experiment say that CATIE researchers, except for recognizing that regional farmers had limited land and cultivated in associations, did not understand small-farm cultivation practices and certainly did not use them as points of departure for developing improved technologies. Those critics

further cite the complex agronomic research designs of the experiment and the bewildering statistical rigor they required, and argue that the results did not justify the effort.
Defenders of the Central Experiment often wax eloquent in describing it. They take pride in the rigor of its design and argue that CATIE researchers of the time had no experience with cropping associations and needed first to understand their biophysical interactions. As one researcher involved in the effort put it, "The Central Experiment was a laboratory where we worked with biophysical components. It was never meant to be anything else."
The Central Experiment is important historically because it set the tone for subsequent CATIE endeavors. It focused research on a narrow system of biophysical interactions and sought to simulate, at least in a general way, farmer conditions on the Turrialba station. It can be seen, therefore, as a first step in the direction of on-farm systems research. There is no question that the experiment, in which biometricians played a key role, was scientifically rigorous. According to one researcher of the time, researchers often found themselves marooned in statistical analysis, for the design was too complex for the use of conventional agronomic methods of analysis. The Central Experiment, reminiscent of the extrapolation research that would b6 conducted nearly a decade later, marks the beginnning of a dual track that CATIE has followed over the years. There was an obvious concern for scientific rigor, a desire to do basic research and to contribute to the store of knowledge, yet the ultimate objective, to develop technologies suitable for small farms of the region, was profoundly pragmatic.
With the founding of CATIE, the constituent departments were required by the directorship to propose lines of research inquiry. The Department of Tropical Crops and Soils proposed multicropping on small farms as one line, new crops to diversify local economies as another. The crop diversification line was ultimately rejected by CATIE management in favor of research on multicropping to improve regional food production.
Meanwhile, the Latin American and Carribean Bureau of
the United States Agency for International Development(USAID) had learned of CATIE's interest in multicropping. AID was already following--and supporting--the Asian cropping systems work at the International Rice Research Institute(IRRI) in the Philippines, and the Bureau wanted to initiate a similar enterprise in the Americas. AID communicated this desire to the directorship of CATIE, thus further fueling enthusiasm there for the new line of research inquiry.
CATIE sponsored a regional conference in 1974 to discuss the potential of multicropping research in the Central American region. Supported by USAID and IICA, the conference was attended by dignitaries from AID/Washington, representatives of the

international donor community, the heads of national research institutions in the region, representatives of Asian cropping systems research, and distinguished researchers working in Central America. To my knowledge, it was the first international conference on cropping systems in the Americas. CATIE used the occasion to show and explain its Central Experiment to conference participants.
Shortly after the conference, USAID announced to CATIE the availability of funds for multicropping research in the Central American region. This immediately led to the design of the Small Farmer Cropping Systems Project. After a brief turf squabble between AID/Washington and the AID Regional Office for Central America and Panamg(ROCAP), it was decided that the new project would be-managed from the AID side by ROCAP. Referred to in this report as the first CATIE-ROCAP project, the effort marks the beginning of a relationship between ROCAP and CATIE that would last for ten years, from 1975 until 1985.
The influence of the Asian cropping systems tradition in that relationship has been significant. It was USAID that first brought that influence to bear on CATIE and that has continued to promote it over the years. Several of the key advisors to CATIE procured by AID have come out of the Asian tradition. As regards the influence of research disciplines, soil scientists seem to have enjoyed a disproportionate share, especially during the decade of the 1970's.
Funds from the new project enabled CATIE to leave the Turrialba station to test its technologies and methods in countries of the region. But ROCAP first required CATIE to secure memoranda of understanding with those countries for research on the agricultural production systems of small farmers. Nicaragua, Costa Rica, and Honduras signed memoranda in 1975 and work began soon thereafter; Guatemala and El Salvador signed in 1976, but because of misunderstandings, work did not begin in El Salvador until 1977 and in Guatemala until 1978. CATIE did not effectively commence systems research in Panama until the beginning of the second CATIE-ROCAP project in 1979.
The primary objective of the first project, designed to work with annual crops, was to develop a regional approach to research that would improve small-farm cropping systems. In addition to this mandate for methodology, the project was to design ten technologies for areas of Central America. These came to be called technological packages, or "tech-packs," although CATIE preferred the term "alternatives."
Further information on this first CATIE-ROCAP project and its activities from 1975 until 1979 can be found in the project paper, in the final project report(1979), and in the final evaluation(1980). It bears emphasizing that a major product of the project was, and was intended to be, methodology--a methodology for doing cropping systems research that came to

include the stages of site selection, characterization, design, and on-farm testing. In accordance with the project design, technologies emerging from the testing stage were delivered to a national extension service for promotion. But evidence suggested that few farmers were adopting the technologies, a matter of concern to ROCAP as the project drew to a close in 1979. The final project evaluation cited the lack of an extension component as a weakness of the project.
The second CATIE-ROCAP project, the Small Farm Production Systems Project, began in 1979 and was a programmatic sequel to the first project. There was a continuing concern for methodology, again to be a major product of the effort, which was widened to include perennial crops and livestock. In contrast with the first project, technology transfer was a major preoccupation of the second. There were two parts to this preoccupation: first, CATIE was charged by ROCAP with developing a methodology for extending to farmers the technologies developed during the first project; and second, CATIE was to develop a methodology for "extrapolating" technologies from one agroecological zone to another. I deal with extrapolation at some length under El Salvador, to a less extent under Guatemala. As regards extension methodology, CATIE equivocated somewhat here, arguing with ROCAP that they were a research institution and had neither the expertise nor the resources to pursue the matter adequately. But ROCAP pressure mounted, and in 1982 a fifth stage--Validation and Transfer--was added to the CATIE systems research methodology, and CATIE hurriedly began to validate some of the technological alternatives developed earlier. The project was amended at the time to accommodate the new effort. The validation stage was really a compromise solution to the CATIE-ROCAP argument, which I deal with more fully in the last section of this report.
The project was extended in 1983 to June of 1985 in order to include research on mixed systems--crops and livestock.

History and Setting
The Instituto de Ciencia y Tecnologfa Agri colas(ICTA) was founded in 1973 as the national institution charged with agricultural research. ICTA was initially concerned with national food crops and did not begin livestock research until 1979. The Direccion General de Servicios Agrlcolas(DIGESA) is charged with crop extension in the country, and the Direccin General de Servicios Pecuarios(DIGESEPE) handles livestock extension. Just how extension relates(or should relate) to research is a controversial issue in Guatemala.
Like ICTA, CATIE was also founded in 1973 and in many ways the two institutions have parallel histories that have placed them on a collision course. These histories have conditioned relations between the institutions and must be understood in order to make sense of those relations today.
Both CATIE and ICTA early recognized the need to work with limited-resource farmers in order to increase national food production. From their founding until about 1978, both institutions were developing research methodologies that would help them achieve their goals, and both were meanwhile seeking identities as fledgling institutions. The two were part of the ferment of the times, of the 1970's, when researchers in Asia, Africa, and elsewhere in Latin America were also developing research methodologies for generating technologies for limited-resource farmers. ICTA moved toward a tradition rooted in Plan Puebla, while CATIE drifted in the direction of Asian cropping systems research as developed at the International Rice Research Institute(IRRI) in the Philippines.
Yet the methodologies of the two institutions have much in common: both have roughly the same stages--diagnosis, design, testing, and extension--and both do research on farmers' fields using a systems perspective. But there are also differences, not. as great today as they once were, and these have been sources of conflict between CATIE and ICTA over the years.
ICTA rejected early overtures by CATIE to work in
Guatemala. CATIE said that it had something to offer ICTA; ICTA said that it did not. ICTA had seen the Central Experiment in Turrialba and was not impressed: the experiment was too complex, there were too many variables under manipulation, and the results

seemed trivial relative to the effort. ICTA did not want CATIE testing hypotheses and developing a methodology in Guatemala. Tension between the two institutions mounted, as they did between ICTA and ROCAP, for ROCAP had made funds available to CATIE to work in the countries of the region, but first there had to be signed agreements between those countries and CATIE. Also, ROCAP felt that ICTA's research methods were not sufficiently rigorous. CATIE did not begin systems research in Guatemala until early 1978.
CATIE Activity in the Field and its Impact
CATIE has arguably had some impact with new technologies in two areas of Guatemala: in Chimaltenango Department on the lower reaches of the Central Highlands, about an hour from Guatemala City; and in Nuevra Concepcidn, Escuintla Department, in the south of the country. I visited only the area of Chimaltenango.
The Wide-Furrow Technology in Chimaltenango
ICTA began working with the traditional
milpa(mafz-frijol-haba) system of Chimaltenango in 1976. The objective was to increase the yields of corn and beans on highland farms with limited land. The double-furrow(surco doble) system was first developed. That system involved a change in the spacing of corn--to a system with two furrows on either side of an alley, which was to be used for intercropping potatoes or wheat.
A CATIE resident researcher entered the area in early 1978 to work with an ICTA technology testing team in Region V. He began with the ICTA idea, that of changing the corn spacing, but varied the theme to include the planting of crops such as carrots, potatoes, beans, and broccoli in the alley between the corn rows. He also further varied the corn spacing, using ohly one furrow of corn on either side of the alley instead of two. His fertilization and other agronomic practices for corn and vegetables varied little from those traditional in the area. A later comparison of the net income of the double-furrow and the wide-furrow(surco ancho) systems, with the same crops in the alleys, revealed the wide-furrow system to be superior.
The new alternative was just entering the validation stage when a violent civil disturbance forced a halt to all field operations. ROCAP authorized CATIE to withdraw its resident researcher, who left the country in late 1981 for Turrialba, with no chance to continue the work.
This technology has recently awakened some interest. It

is today described in a public flyer printed by ICTA, with instructions to the reader to contact DIGESA for further information. According to my sources, the technology may need some minor adjustments, but is otherwise ready for diffusion. Both DIGESA and a local cooperative are interested in it. If a large BID project for research, extension, and seed production currently pending before the Guatemalan Government is approved, the wide-furrow system will likely be promoted.
Dual-Purpose Cattle Technology in Nueva Concepcion
Although carried out under an ICTA banner, this
technology(Modulo de Produccion Bovina de Doble Propdsito) is widely recognized in Guatemala as a predominantly CATIE contribution. The technology is highly acclaimed as of this writing, and there is much agreement that it represents the major contribution of CATIE in the country to date.
ICTA began livestock research in the area of Nueva
Concepcion in 1979, and soon thereafter, realizing its weakness in this realm, entered into a cooperative agreement with CATIE. The two institutions worked together and developed a dual-purpose cattle technology, which was ready for validation in 1983. That validation is now being conducted by ICTA, DIGESEPE, and the Banco Nacional de Desarrollo Agrfcola(BANDESA). As with the wide-furrow technology, plans are to further extend the livestock alternative through the BID project now before the Guatemalan Government.
The success of the work in Nueva Concepcion has led ICTA
to plan the development of a similar dual-purpose alternative for the dry areas of the country. Slated to direct that effort is an ICTA researcher trained by CATIE. According to one well-informed source, the work in Nueva Concepcion has stimulated ICTA to prepare a proposal for the creation of a livestock research unit within the institute.
Following the completion of the development of the
dual-purpose alternative, the CATIE-ROCAP project extended the research in Nueva Concepcion to include crops. One alternative has been developed for a mixed system, but the results were not available when I was in Guatemala. This work, however, has had neither the endorsement nor the support of ICTA.
Other CATIE Activity and its Impact
The Farming Systems Research Network
CATIE initiated the formation of national farming systems

research networks in Guatemala and El Salvador in 1983, and planned to organize them in other countries of the region if funds could be obtained. These national networks were to form part of a regional network that would exchange information. Networking activity was part of the CATIE outreach program at the time.
The network in Guatemala has about 100 members, most of them ICTA researchers working in the area of Quetzaltenango, where the association is based. The group has a steering committee but no regularly scheduled meetings. Indeed, only three meetings have been held since the network was founded. None of the ICTA people I talked with mentioned the network. It was formed at the initiative of CATIE, so it seems, and then abandoned.
The impact of CATIE in this area is difficult to gauge. CATIE has provided degree training at the master's level to numerous Guatemalans, several of whom specialized in agricultural production systems. The current Technical Director of ICTA is a CATIE graduate as are several other ICTA employees. Six professors of agronomy on the faculty at the University oE San Carlos have master's degrees from CATIE; at least two of them specialized in production systems and are now teaching systems courses at the university. The CATIE graduates at San Carlos are seeking closer relations between their university and CATIE. An agreement between the two institutions was recently drafted but has not yet been signed.
The CATIE graduates with whom I talked in Guatemala were in the main satisfied with the training they had received; at least one distinguished ICTA researcher had nothing but praise for the CATIE training program. The only negative comment made to me, and not voiced with much feeling, was that CATIE instructors are sometimes remiss in their teaching and advisory duties owing to travel and other responsibilities.
A few comments compared CATIE training with that received in the United States. CATIE training, it was said, is more general than that offered by the American universities and thus enables researchers to be more versatil and to better function in the environment of Guatemala. Also, CATIE trains students in Spanish and in a tropical agricultural setting, where they will be working.
Guatemala was one of several countries in the region

selected for experiments to develop a methodology to extrapolate technologies from one ecological zone to another. The development of such a methodology is stipulated in the second CATIE-ROCAP project. The extrapolation research in Guatemala was conducted in the Departments of Jutiapa and Santa Rosa from April of 1982 until May of 1984. Experiments were to be initiated also in the southern Guatemala-Baja Verapaz region, but the political violence there foiled the plans. The objective of this multi-country research was to develop an extrapolation methodology for the semi-arid tropical areas of Central America. A Corn-sorghum association was used in the experiments. The effort really centered on Tejutla in El Salvador, where the experimental association was first generated. The idea was to test it in zones ecologically similar to Tejutla elsewhere in Central America and where the same association occured. By distributing the test sites over a given zone in such a way that the full range of variation in soil and climate was represented for that zone, the critical ecological variables determining the behavior of the associations could be isolated. Or so the theory went. Since a significant amount of energy and resources has been allocated to the work on extrapolation, I deal more fully with the topic, including the historical origins of the idea, under El Salvador.
A lengthy report was published in 1983 on the Guatemalan
effort entitled: Extrapolacion. Informe Anual 1983. A companion volume, Metodologfa del Andlisis Fisiogr~fico Aplicado a Extrapolacidh de Sistemas de Cultivos, was published in 1984. This work is a study in the application of aerial photography to physiographic analysis conducted to produce maps useful in extrapolation.
I was unable to find anyone in Guatemala who knew anything about this work, which is highly technical. I do not think it went beyond the production of the reports.
The CATIE Methodology
CATIE has not influenced Guatemala in the area of
methodology. There is simply no reason for ICTA to turn to CATIE here. Indeed, methodological differences, or sometimes the perception thereof, have been a source of tension between the two institutions over the years. And those differences have placed a strain on CATIE researchers working in Guatemala, for ICTA has not been tolerant of research methods other than their own. A brief foray into a few of the more salient differences may be instructive.
CATIE has often worked with crop complexes as such, those complexes approximating in varying degrees complexes already existing at a selected site. ICTA also starts with an actual complex and seeks-to improve it, but, in contrast with CATIE,

works with only a component or two, while maintaining a systems perspective. What I just described are tendencies, or ways the two institutions tend to approach their tasks. The differences are subtle, and researchers from both institutions may deviate from the tendencies in the field. Other differences are more obvious. CATIE, for example, expends more effort in the collection of data than does ICTA, certainly at the charactrization, or diagnostic stage. And CATIE places more emphasis on the biophysical aspects of farming systems research than does ICTA.
Institutional Relations and Operational Problems
Much friction between CATIE and ICTA over the years is
related to what I will call sovereignty. ICTA is the sovereign entity charged with agricultural research in Guatemala and has guarded that charge with a jealousy unusual among research institutions in the region, a jealousy undoubtedly reinforced by the international attention the institute has received since its founding. All agricultural research in Guatemala is to be coordinated by ICTA, and all research involving foreign entities is to be a collaborative effort with ICTA. Furthermore, that research is to be conducted under an ICTA banner, and the results are to be published with the ICTA seal.
With all of this CATIE has clashed over the years. When CATIE began work in the country in early 1978, its site had been selected for it--ICTA's Region V, the area of Chimaltenango. CATIE wanted to work in the lowlands rather than in the highlands, where it had no experience. It was difficult for ICTA to fit the new CATIE resident researcher into its activities during 1978, which were already programmed when he arrived. And there is evidence that he was not taken seriously during his first year. But his situation improved substantially in 1979.
Project "ownership" and the control of CATIE operating
funds were issues of contention between ICTA and CATIE during the early years. Both institutions sought to own and manage the projects on which they worked, and ICTA resented CATIE's control of its operating funds. The large budget that CATIE enjoyed did not go unnoticed either, for ICTA required them to provide substantial material resources to cooperative endeavors.
ICTA researchers thus felt that CATIE was too independent, that it made its decisions without regard for ICTA, and that its researchers took their orders from Turrialba rather than from Guatemala.
Additional tensions have arisen from the different positions occupied by the two institutions along a theory-practice dimension. ICTA researchers perceive themselves as pragmatic, CATIE researchers as theoretical. They say that

CATIE researchers want to publish, to do science for the sake of science. "CATIE researchers are thinkers," I was told, "and this is a luxury we can't afford. We must be doers." And ICTA researchers cannot understand the heavy emphasis placed by CATIE on developing methodology. "You can't eat methodology," said one of them.
The increasing political violence that finally drove CATIE out of the Chimaltenango area in late 1981 was serious and greatly disrupted research activities. The local peasantry was caught in a conflict between guerrilla forces and the Guatemalan Army. At least one farmer collaborator in San Martin Jilotepeque was killed in late 1980, and others abandoned their farms. ICTA agronomists grew reluctant to enter the area. In Alta Verapaz Department, where CATIE was beginning livestock research of a kind already underway in Nueva Concepcion, an ICTA agronomist paid by them was killed when his CATIE-marked vehicle was raked by machine-gun fire. These incidents struck fear into CATIE and ICTA, and research in Alta Verapaz and Chimaltenango was soon abandoned.

History and Setting
El Salvador is a country at war with itself. The war has profoundly conditioned agricultural research and extension activities, and CATIE's part in them, since the beginning of the insurrectionary movement in the late 1970's. The conflict has not only diverted resources from agricultural research and extension, so obvious in the remoter northern zones of the country, but has also created a highly unstable political and institutional environment within the public sector.
The organizational scheme for agricultural research in El Salvador can accommodate systems research. Activities within the Ministerio de Agricultura y Ganaderia are organized through five centros, or centers, charged with natural resources, crop technology, livestock development, fisheries development, and agricultural training. Systems research in the country is conducted by the Centro de Tecnologia Agrfcola, or CENTA. It is mostly with CENTA that CATIE has interacted and collaborated since 1977.
Created in about 1970, CENTA is today organized into four divisions responsible for crop research, crop extension, seed technology, and seed certification. The crop research division(Divisidon de Investigacion Agrfcola) is further organized into several departments among which are the Department of Production Systems for Small Farmers(Departamento de Sistemas de Produccidn para Pequefios Agricultores), formed in 1984, and the Department of Economics and Statistics(Departamento de Economia y Estadfstica). This latter department manages a program called Technology Validation and Transfer, initiated in 1983 but which follows a precedent established in 1979. The crop extension division of CENTA is further organized at the regional level, with a director and assistant director based in each of the four regions, where extension agencies are scattered.
CENTA has been rocked by instability and financial
crises since the coup d'etat in 1979. It has had six directors since that time, and the position was vacant at the time of my visit. Virtually all of CENTA's personnel and material resources were diverted to implement the agrarian reform in 1980. Several

vehicles were either taken or burned by the guerrillas, and little operating money remained for research and extension. In 1982, the entire Ministry of Agriculture was restructured as part of a decentralization effort. Under the new structure, all activities were to be handled through three institutes, and the country was divided into four regions. CENTA lost its name and was absorbed by one of those institutes, the Salvadoran Institute for Agricultural and Fisheries Research(ISIAP). ISIAP was to be responsible for research over a wide spectrum--livestock, forests and fisheries, and all crops, including coffee--as well as for seed production and certification. The reorganization placed agricultural extension in a separate institute. This was the first time that crop extension had been divorced administratively from research since the two were united under CENTA in 1973.
Another reorganization of the Ministry a year later, in January of 1983, returned the previous structure, including the CENTA name, but with the difference that coffee, livestock, and fisheries research would not be conducted by CENTA. It is the 1983 structure that is first described above. I was told that there was on paper a plan to reorganize the Ministry once more, again to achieve decentralization.
Research and extension are highly politicized in El
Salvador, and there is much personnel instability, especially at the higher levels.
The history of farming systems research in the country
goes back to the early 1970's and. the CENTA economics department, created in 1972. A pioneer in Asian cropping systems research at IRRI passed through El Salvador on his way to CATIE in 1973 and gave a seminar on multiple cropping to the economics department. The seminar stimulated considerable enthusiasm within the department, which by 1984 was experimenting with associations of corn, beans, and radishes, with either tomatoes or cucumbers twining around the corn stalks. The department did its first diagnostic survey in 1976, in the arid northeastern region of the country. It was then that CENTA researchers became aware of thp extent of small-farm agriculture in the region, and of how little research had to offer them.
In 1979, a multidisciplinary team called
INVEXT(Investigacion y Extension) was created within CENTA to test in different parts of the country the most promising technologies developed by research. The team was composed of both researchers and extensionists, and was supported by a CENTA-BID project as well as by CIMMYT and CATIE. This was an important effort historically to coordinate research and extension for work in the field with small farmers. The INVEXT team set a precedent for the Technology Validation and Transfer program, initiated by the CENTA economics department in 1983 to address a perceived need for an intermediate stage between

research and extension. When CENTA began implementing the agrarian reform in 1980, all other activities, including those of the INVEXT team, declined sharply. CATIE, working alone, continued with some of the testing.
The Department of Production Systems for Small Farms was created within CENTA in 1984, as already noted.
CATIE Activity in the Field and its Impact
All of the farming systems work under both CATIE-ROCAP
projects has been conducted in a belt running across northern El Salvador. The northern zone is agroecologically harsh: its clay soils, the poorest in the country, contrast sharply with the alluvial soils along the southern coast, and the rich volcanic soils of the coffee zone. Corn and sorghum is a common association. The north also receives the brunt of the can'cula, a devastatingly dry period within the rainy season that affects about a third of El Salvador. The growing rains cease during June, when crops are in the ground, and again during late July and early August. The onset, duration, and intensity of this climatic phenomenon vary from year to year, and from one area to another in a given year. CATIE conducted and published an extensive study of this major challenge to agricultural technology in quantitative and probabilistic terms. It likewise conducted and published a detailed study of soils in the north.
The northern zone is also marginal politically and
socioeconomically. Heavily populated by small farmers and far from San Salvador, the north has been an area traditionally neglected by the government. The insurrectionary movement began in the north, where the force of it remains today. Realizing the extreme socioeconomic marginality of the zone, the government made it a focus for rural development from 1978 to 1980 and initiated a program(Bienestar para Todos--"Wellbeing for All") the main emphasis of which was agricultural technology development and transfer. Multidisciplinary farming systems research teams were formed to do agrosocioeconomic surveys in the region. CATIE collaborated closely with CENTA in this effort, beginning in 1978 under the first CATIE-ROCAP project and continuing in 1979 and thereafter under the second one. In 1980, CENTA personnel were diverted to assist with the agrarian reform, leaving CATIE alone in the northern zone with little CENTA support.
- La Trompina and Tejutla
The first CATIE resident in El Salvador arrived in

November of 1977. CENTA had asked that operations to increase agricultural production begin in an area known as La Trompina Alta, to the northeast of the town of Jocoro, in the Department of Morazan. After briefly assessing conditions around Santa Rosa de Lima, the CATIE team concluded that there was little they could do there and asked CENTA to let them work elsewhere. The request was both honored and denied: CATIE had to continue working in La Trompina, but could choose another site and begin operations there. That other site was Tejutla, in the Department of Chalatenango. For nearly two years, therefore, CATIE worked simultaneously in both La Trompina and Tejutla, but with greater emphasis on Tejutla. Work was also initiated in Jocoro in 1979. La Trompina was abandoned in 1980, and all CATIE efforts thereafter were focused on Tejutla and Jocoro, but mainly on Tejutla.
An area of sparse vegetation and highly eroded rocky
soils, the area of La Trompina was densely populated by small farmers and severely affected by the canicula. The residents were marginal people: land distribution was markedly unequal and limited resources forced entire families to migrate seasonally from the area to harvest coffee and cotton on the plantations. There were always stories of sickness and death from pesticide poisoning on the cotton plantations, where pesticides were in heavy use.
The CATIE work in La Trompina included a rapid survey and some experiments with associations such as corn-sorghum, corn-pigeon pea, corn-sisal, corn-bean, and corn-cowpea. But little was accomplished. Most of the local crop varieties proved superior to new ones, and social tensions were rife: there was fighting between two peasant groups, with both of whom CATIE worked, and the guerrillas were meanwhile organizing in the area. Amid numerous hardships and the mounting violence, CATIE withdrew during 1979 and 1980, after a long two years, and moved to Tejutla.
In the Department of Chalatenango, sixty-four kilometers north of San Salvador and well to the west-northwest of La Trompina, the municipio of Tejutla is controlled today by rebel armies. The area suffers many of the asperities of land and climate found in La Trompina. The terrain is sloping and the soils are shallow, rocky, highly eroded and of low fertility. Most of the farmers were poor; the average farm size was 1.4 hectares. The traditional farming system centered on an association of corn and sorghum.
CATIE-CENTA began operations there in 1978, selecting the area because of the presence of knowledgeable extension personnel, good access roads, and a high concentration of limited-resource farmers cultivating annual crops. CATIE expended considerable effort in developing methodology in

Tejutla, and in training Salvadorans. The extrapolation experiments conducted over much of Central America, discussed in the following section, center on work there.
After an initial and thorough characterization of the
area, CATIE in 1984 and which gives much emphasizes to agroecology, experimental work was begun around the traditional corn-sorghum system. New varieties of both crops were tried and fertilizer levels and timing of application were varied. The new corn variety did provide greater yields, but the new sorghum variety was not an improvement over its creole predecesor. Legumes such as beans, cowpeas, and pigeon peas were experimentally introduced to the corn-sorghum system to increase the intensity of land use, as were new crops such as forage sorghum, pineapples, plantains, avocados, leucaena, citrons, and cultivated pastures. The CATIE-CENTA team also tried to improve the rustic cultivation tools.
CATIE had just finished validating a corn-sorghum
association when guerrilla armies entered the area in 1982. The research team departed and there has been no further work there since. Tejutla is today in a rebel-occupied zone, where there are no CENTA offices. No extension work has been conducted since the alternative was validated, so I was unable to obtain reliable information on the impact of the CATIE work there.
Located in the south of Morazan Department, Jocoro is a municipio of San Francisco District. I visited there, going first by air from San Salvador to the town of San Miguel, seat of Agricultural Region lV, thence by road to Jocoro, a distance of some twenty-five kilometers. All airfields were under heavy guard. The area in and around San Miguel is at present controlled by the Salvadoran Army, although the town has twice fallen to the rebels for brief periods of time. Like San Salvador, it was brimming with refugees from the countryside. The highway between San Miguel and Jocoro was heavily guarded, especially the bridges, and soldiers stopped the few circulating vehicles at several points to check documents and conduct searches. An occasional helicopter lumbered overhead, patrolling the countryside. At one point on the road, a couple of soldiers relieved the tedium by firing automatic rifles at vultures in a nearby tree.
In Jocoro the canicula is severe and the acid soils.are rocky, of shallow depth, and with little capacity for water retention. Yet I was told that farmers there had more resources than those in either La Trompina or Tejutla. Many of them have cattle and consume as well as market dairy products, thereby

reducing their reliance on the traditional corn-sorghum association. But they are still marginal, like the land on which they live. Jocoro was selected for CENTA-CATIE work by the Ministry of Agriculture because it fulfilled the criteria of easy access, the presence of extension agents from both CENTA and the Center for Livestock Development, and homogeneity with respect to agroecology, the density of small farms, and socioeconomic characteristics.
The war lingered in the minds of those research and extension officials with whom I talked. It was variously referred to as "the crisis," "the political problem," or as merely "the problem." The region was often described as "a conflictive zone." Obviously disquieted and made anxious by the war, some of the officials seemed to feel that they and the region had been abandoned by the government. They were far from San Salvador, they said, and thus received little attention and fewer resources.
CATIE first entered Jocoro in 1979, under the second CATIE-ROCAP project. As in Tejutla, an extensive characterization of the area was conducted, but not published until 1984. In Jocoro CATIE helped CENTA with the on-farm testing of corn varieties that led to the development of the drought-tolerant CENTA M3-B, released in 1982. The new variety was later--in 1984--validated on thirty-two farms in the area. The selection phase that led to the development of CENTA M3-B was coordinated by CIMMYT., in close collaboration with ICTA in Guatemala. The new variety is a blend of the eight best families of the ICTA B-I variety.
CATIE experimented with cowpeas and sorghum(CENTA S-2, a grain and forrage sorghum) before beginning research on mixed systems in 1982, and also with pigeon peas as a relay crop with corn. CENTA has received support from CATIE in a miscellany of waYs. Only last year, for example, CATIE helped them validate an association of climbing beans and corn: CATIE provided technical assistance and the bean seeds. And with seeds supplied by CATIE, the home demonstration agent in San Miguel has promoted, apparently with some success, both cowpeas and pigeon peas among housewives in the area; twenty-two women were experimenting with cowpeas during my visit, mixing them with corn to make tortillas. A current limitation to promotion on a greater scale is the lack of seeds.
Under much pressure from ROCAP, CATIE began to work with
mixed systems in Jocoro in 1982, after a delay caused by friction in Turrialba between the Departments of Crop and Animal Production. The research on mixed systems involved reorienting the previous crop research to address the problems of a more comprehensive system that now included dairy cattle. The major limitation to dairy production was the canicula. Until recent

times, cattle during this dry season were fed cotton-seed flour imported from the cotton areas along the coast. When the rebels disrupted cotton production as part of their campaign to deprive the government of foreign exchange, the cost of the flour became prohibitive.
The key to the mixed system alternative developed in
Jocoro, and under validation during my visit, is silage for the dry season, made using a mixture of cowpeas and sorghum(CENTA S-2). In addition to sorghum, the full package includes corn(CENTA M3-B) planted in association with cowpeas. As a legume, cowpeas restore nitrogen to the soil and supply vegetable protein to the household diet. A small section of land is fenced to serve as a silo, where the cowpeas and sorghum are cut and piled, then compacted by human feet. The mixture is covered with a sheet of plastic and allowed to ferment. The design of the technology sought both simplicity and economic viability. Only cowpeas are new, and farmers will have to fertilize the sorghum--they already fertilize corn. Livestock recommendations in the package include little that was not already recommended for the area--vaccinations and vermifuges, for example.
There are two potential problems with this alternative: the availability of seeds and credit. Seeds for corn(CENTA M3-B), sorghum(CENTA S-2), and cowpeas have not been readily available to date(On the day that I left El Salvador, a local newspaper announced a large shipment of CENTA M3-B seed corn to the north). For the experiments and validation trials, CATIE supplied the seeds--and other required inputs as well. The seed problem may prove especially refractory. According to one well-placed source, the seed technology unit of CENTA will not multiply seed without some guarantee of reimbursement. They have asked the extension service on occasion for prior deposits.
Whether the alternative will require the availability of credit is moot. If credit is needed, however, ittmay not be available. The loan programs of local credit banks are reputed to be unsuited to the needs of limited-resource farmers.
Candelaria de la Frontera
CATIE began operations in Candelaria de la Frontera in 1982, after guerrilla forces entered Tejutla. Located in the west of the country near the Guatemalan border, Candelaria was selected in part because it offered the prospect of a peaceful and sustained work effort, a luxury that CATIE had never enjoyed in El Salvador.
Achievements were modest during the year and a half of work there. The climate, soils, traditional agricultural

practices and economy of the zone were characterized. Variety trials were conducted with yuca, but there was a major problem with thievery--because farmers in the area wanted to experiment with the new crop, I was told. Also, scavenging pigs damaged the trials. Some work was also done with corn-bean associations, trying new varieties of both. But the alternatives were not significantly better than technologies already in use by local farmers.
I was not able to obtain further information on activities in Candelaria, though it was rumored that CENTA was conducting validation trials in the area, probably with corn-bean associations.
Other CATIE Activity and its Impact
The development of a methodology to extrapolate
agricultural technologies from one site to another has been one of the more controversial endeavors of CATIE under the second project. And it may have been the most esoteric. Initiated and strongly supported by ROCAP, the activity has been an important part of the project. Reliable information on extrapolation research was hard to obtain, and often contradictory. The effort, which was led by biometricians, began in El Salvador and spanned three other countries--Honduras, Guatemala(see my comments under Guatemala), and Nicaragua.
Extrapolation has its origins in the Central American Soil Fertility Project. That project worked with soil analogues--the term "extrapolation" did not come into use until the second CATIE-ROCAP project--and sought to characterize Central American soils and climate as a way to facilitate technology transfer. A technology Viable in one soil-climatic zone, it was reasoned, could more easily be transferred to another similar zone, the "analogue" of the first. Only the variables of soil and climate were considered. As the Soil Fertility Project drew to a close, ROCAP folded it into the cropping systems work of the new CATIE-ROCAP project, thus sowing the seeds for extrapolation.
That technologies might be extrapolated from one zone to another was a seductive idea. It offered a way out of what some felt to be the site-specific limitation of farming systems research, no small attraction to resource-poor national research institutions and economy-minded donors. The idea also squared nicely with the notion of technology transfer, a major emphasis of the second CATIE-ROCAP project.

The work to develop an extrapolation methodology centered on Tejutla and corn-sorghum associations. The presence of this association, common over much of northern El Salvador as well as in other dry areas of Central America, was thought to be a reliable indicator of limited-resource farms and subsistence agriculture with high risk in dry areas.
The theory behind extrapolation held that if the
environmental factors(mainly soils and climate) most critical to the biological behavior of this association could be determined with precision in Tejutla, then an improved production technology developed for the association there could be extrapolated to other similar agroecological zones--in El Salvador and other countries of the region--where the same association was found.
Just how socioeconomic factors figure in this theory, or whether they do, remains unclear. My efforts to elicit clarifying information from those individuals most closely associated with the activity--the few who remain at CATIE--produced conflicting responses. On the one hand, I was told that such factors did not figure, nor did they have to. Farmers everywhere with the corn-sorghum association were ipso facto marginal in socioeconomic terms, so that a viable alternative developed in Tejutla would be viable elsewhere, or at least approximately so; and what it lacked in viability could be corrected through local adjustments in the technology. On the other hand, I was told that efforts to .incorporate socioeconomic variables to the mathematical extrapolation model--and the effort did involve some rigorous mathematical modeling--made for an unmanageable complexity, and on this complexity the model foundered and the work was abandoned. According to one source close to the extrapolation work, the effort got too technical for everybody--but, continued the source, CATIE researchers learned much from the activity.
The effort, to which the corn-sorghum association was
merely incidental, did not to my knowledge produce a methodology for the extrapolation of technologies. I am not able to speak to the benefits of this research, which may have been considerable, but if so, they seem to have been mainly internal to CATIE.
The FSR Network
As in Guatemala, CATIE initiated the creation of a farming systems research network in El Salvador two years ago. Twenty-seven researchers belonged to the group at one time, and there have been two plenary meetings since its founding. A newsletter was published for a brief time, edited by the CATIE resident. But for lack of external support, the network has not been active for a year.

I gathered little information while in El Salvador on CATIE's considerable efforts to train Salvadorans. CATIE has conducted numerous short courses in the country, and before the agrarian reform diverted CENTA personnel, CATIE provided about eighteen months of in-service training to five CENTA employees working in Tejutla, paying their salaries during that time.
The CATIE Methodology
It is difficult to assess the impact of the CATIE
methodology. CATIE entered the country in the 1970's, followed by CIMMYT; and only recently, the sondeo, or rapid survey, made famous by ICTA in Guatemala, arrived in El Salvador. It is probably fair to say that the CATIE methodology has had the most impact. Certainly CATIE has made heroic efforts to promote it.
It would seem more than coincidental that the Departamento de Investigacion en Sistemas de Produccion para Pequenos Agricultores, created last year, bears the exact name of the CATIE-ROCAP project; and that the program Validacion y Transferencia de Tecnologia, of the CENTA economics department, bears the exact name of one of the phases of the CATIE methodology. Both names suggest indirect CATIE influence.
In the minds of several of the Salvadorans that I met,
CATIE is closely associated with characterization and validation activities, perhaps no surprise given the strong emphasis that CATIE has given to both in the country. As one CENTA official said, "The CATIE methodology is not for generating technology, but for validating and transferring it."
It is probably true that the CATIE presence in El Salvador has done much to sensitize Salvadorans to the need for a systems perspective and to its potential use in developing and testing technologies in farmers' fields. If so, this is no small achievement.
Institutional Relations and Operational Problems
It is not easy to conduct farming systems research in the vortex of a guerrilla war. The approach is challenged at its operational essence--mobility. Rebel forces enter an area where research on farmers' fields is underway and declare by radio a paro, or ban on movement. While the ban is in effect, it is not advisable to travel in order to monitor or harvest on-farm

trials, or to otherwise communicate with farmers. So the vital continuity of a farming systems research effort is lost. Even when the rebels do not declare such a ban, there is always the chance of an encounter with a roving guerrilla band in the countryside. Although such encounters are rarely violent, they do make for certain anxieties. At least one CENTA vehicle was taken by guerrillas in the San Miguel area--and has not been returned.
Beyond this, there is the issue of farmer confidence,
since peasant farmers are often pawns in the conflict between the government and the rebels. It can be very difficult for researchers to gain and maintain that confidence in such a setting. And there is also the problem of depopulation of the countryside, so obvious in the crowded cities and towns of El Salvador. Collaborating farmers sometimes abandon their farms while research trials are underway. The depopulation has further created a severe labor shortage in some of the rural areas of the north.
CATIE has had to work, frequently alone, in the most
inhospitable areas of El Salvador. There has too often been no time to complete what it set out to do: rebel armies enter, the violence mounts, and CATIE moves on--from La Trompina to Tejutla to Candelaria. The government has not always been able to provide support: in 1980, virtually all CENTA personnel were detailed to implement the agrarian reform. I was impressed with the numerous, even petty ways that CATIE has supported government efforts in the San Miguel-Jocoro area. CATIE has provided technical assistance, seeds, agricultural inputs, vehicles and their maintenance, gasoline, seasonal labor, and per diems. CATIE has filled a vacuum and done what the government could not have done for a chronic want of resources, human and material.
Institutional instability and frequent government
personnel changes have placed an added burden on CATIE. The management of institutional relations has consumed an inordinate amount of time. A new regional research or extension director arrives, and the CATIE resident must establish relations anew, explaining past and present activities and plans for the future. In the mixed system research in Jocoro, CATIE has worked not only with CENTA, but also with the Direccion General de Ganaderia, which handles livestock research and extension. The offices of these two branches of the Ministry are forty kilometers apart. And now CATIE wants to address the question of agricultural credit for the Jocoro area. Such management requires much time and energy of the CATIE resident, who must also function as a technical expert. The management problem is compounded by the notorious lack of coordination of research, extension, and seed technology in El Salvador.
Aside from some concern by CENTA that CATIE had moved its

national office from CENTA headquarters in the San Andres Valley to San Salvador, I encountered only a single CENTA criticism of CATIE. It was said that CATIE had special projects and operated independently of CENTA. But this criticism, uttered with much vigor in other countries of the region, was made almost parenthetically in El Salvador, and seemed to refer mostly to CATIE's work in Tejutla. But the criticism was quickly followed by an acknowledgement of CATIE's contribution of much-needed. resources at the local level, and a recognition that CATIE could never operate under the constraints of the heavy Salvadoran bureaucracy. At the local level, as I observed in the San Miguel-Jocoro area, CATIE has been extremely well received, in part because of this same material contribution, but also, perhaps, because of a feeling by regional government personnel that they have been forgotten by San Salvador.

History and Setting
The Secretarla de Recursos Naturales(SRN) is responsible for agricultural research and extension in Honduras. Crops fall to the Direccion General de Agricultura, cattle to the Direccion General de Ganaderia. There is no small-animal unit, but neither is there an official interest in small animals. Each direccion is composed of two units, one for research and the other for extension. In each of the eleven regions, there is a coordinator charged with administering all SRN activities in that region. This national structure, which dates from 1983, is replicated at the regional level, where the SRN regional coordinator has under him two research offices, one for crops and the other for cattle, and two extension offices, one for crops and the other for cattle. Extension agencies are scattered about the regions and some of them have resident researchers and extensionists. Most of CATIE's work in Honduras under the second CATIE-ROCAP project has been conducted in the region of Comayagua(Region 11). And CATIE has interacted mainly with SRN offices.
In Honduras, agricultural research and extension are plagued by low salaries, high personnel turnover, and job insecurity. The national head of crop extension had been in office only twelve days when I arrived, and Comayagua has had four regional directors over the last three years. Many employees work under three-month contracts only, which must be renewed through a lengthy process at the end of the term. Meanwhile, salaries fall months in arrears, or sometimes are not paid at all.
An agreement was signed between CATIE and the Government of Honduras in December of 1975. The first CATIE resident arrived six months later and established a base in San Pedro Sula, in the north of the country. Financed by the first CATIE-ROCAP project, research was conducted on multiple cropping in Yojoa, Agua Sucia, and Cuyamal. Cropping associations in the area were studied and efforts were initiated to develop a methodology for on-farm research. Work in this northern region was discontinued when the CATIE resident was transferred to Comayagua in 1978 by agreement with the SRN. In Comayagua, he was to provide support to the newly-formed Unidad Central de Investigacion Agricola.

The Unidad Central, or Central Unit, was the product of a series of fundamental changes in the organization of agricultural research and extension that took place during the 1970's. The SRN was restructured in 1975 and the Direccion General de Operaciones Agr'colas(DGOA) was formed. The DGOA included the Programa Nacional de Investigacion Agropecuaria(PNIA) and its extension counterpart, the Programa Nacional de Extension Agropecuaria(PNEA). With the restructuring, the SRN was decentralized to the regions and efforts were begun to make research and extension more responsive to regional needs.
The period from mid-1977 until mid-1980 was a time of
ferment and flourescence in the history of Honduran agricultural research and extension. As in the other countries of Central America, research theretofore had been organized by commodity and discipline and was conducted entirely on experiment stations. Special attention was given to plant breeding, and well trained researchers were few.
Six PH.D.'s, four foreigners and two Hondurans, arrived in Honduras in 1977 at the invitation of the SRN Secretary to advise on agricultural research matters. An innovative and open-minded man, the Secretary gave the six the general charge of making agricultural research in Honduras more productive. And he gave them much latitude to achieve this. The two Hondurans led the effort.
The group of six were to advise the experiment stations and did not initially have a systems orientation. After pondering how they might make research more productive, they decided on a systems strategy. The CATIE resident in Honduras, now interacting with the group, was influential in that decision.
The group established their base in Comayagua, to the north of Tegucigalpa, and formed the Unidad Central. They planned to develop a methodology for working with small farmers and to train Honduran researchers and extensionists in it. And they would also advise the research stations. The Unidad received financial support from the first CATIE-ROCAP project and from IDRC of Canada. Salaries of the Comayagua six were paid by the SRN, in national currency. Throughout their work in Honduras, the group was faced with an entrenched opposition to systems thinking and the methods they promoted.
There was considerable interaction between the Unidad
Central and ICTA in Guatemala. Guatemalans went to Honduras and Hondurans went to Guatemala. The Rockefeller Foundation posted an advisor to PNIA from 1979 to 1981, a man who had served in a similar capacity with ICTA during the formative years of that institution.
The Unidad soon conducted surveys in the Departments of

Comayagua, municipios of Comayagua, El Rosario, and San Jeronimo; Intibuca, municipio of La Esperanza, in the Honduran high country; and La Paz, municipio of La Paz. It was agreed that CATIE would work in all of those municipios except San Jeronimo.
From 1978 until 1981, the Unidad trained thirty Hondurans in the methods of farming systems research. The students, carefully selected from the agricultural college, spent nine months taking classes in Comayagua and applying their knowledge on farms of the region. They did theses under the supervision of the six Ph.D.'s and theL CATIE resident. CATIE also played an important training role at the Unidad Central, sending lecturers from Turrialba and conducting local seminars and workshops for the students.
From the middle of 1977 until the middle of 1980, the Unidad Central, supported by CATIE, worked with brilliance to reorganize the way agricultural research and extension were conducted in Honduras. They sought to reorient Hondurans to systems thinking, giving much emphasis to the training of students as a way to achieve the goal. Relations between CATIE and the Unidad were excellent.
But the flourescence radiating from Comayagua began to
darken during the second half of 1980. The next year and a half saw the progressive undoing of much of the work of the previous three years. The innovative SRN Secretary was replaced by one unsympathetic to systems research, to the reorganization underway, and to the Unidad Central. And the traditional forces, which had continuously offered strong resistance to the Unidad, assumed the reins'of power. The group of six had by now grown tired of job insecurity and lethargic bureaucracy; they worked under contracts that were renewed yearly, with delays of four and five months without pay. By 1982, they had abandoned the Unidad Central and gone their separate ways.
There was some dissent also on the CATIE team regarding the planting of horticultural crops. A specialist arrived in 1980, hired by the CATIE-ROCAP project to work with horticultural crops in the Comayagua Valley and La Esperanza. But he wanted to plant them in sole stands rather than in associations with other crops. CATIE and ROCAP insisted on associations and the specialist's contract was not renewed. There was also instability in the local CATIE leadership during the first three years of the second CATIE-ROCAP project, for there were four CATIE residents in Honduras between 1978 and 1982.
By February of 1982, when the current CATIE resident arrived in Honduras, CATIE was alone in La Esperanza and the Comayagua Valley, abandoned to the whims of an unfriendly SRN regime. There was no resident in Honduras during the six months prior to February of 1982. Honduran technicians in La Esperanza

and El Rosario worked without direction on the experiments during this interval. In so doing they clashed with the new powers in SRN, for their allegiance was to CATIE. When the new resident arrived in February, he found chaos in the fields: the trials had not been cared for, fertilizer had not been applied, and weeds were everywhere.
CATIE Activity in the Field and its Impact
I visited the areas of La Paz and San Jeronimo, both in the Comayagua Valley, traveling by car from Tegucigalpa, about 100 km. to the south of the town of Comayagua, where I talked with SRN officials. Much of the tortuous mountain road between Tegucigalpa and Comayagua winds through extensive pine forests before descending into the wide Comayagua Valley, a major agricultural zone of Honduras. Acacia trees aflame with orange flowers were numerous along the valley highway. A military air base spreads across the valley floor near Comayagua. A large cargo plane, the kind that carries heavy hardware, rested on the tarmac. Today, American soldiers are billeted in a jumble of new wooden structures on the base, and American military police assist Hondurans in patrolling the night spots in Comayagua on weekends. As in El Salvador, all bridges between Tegucigalpa and Comayagua were under guard.
Site of the first government-built irrigation scheme in
Honduras, the valley regularly experiences severe and continuous draught from December until June. According to one estimate, about ten percent of the arable land is under irrigation. The valley also has one of the country's model cooperatives, Fruta del Sol, now receiving support from USAID. Local market crops include tomatoes, onions, sweet corn, watermelon, and bell peppers. Among the export crops are cucumbers and acorn squash. Many farmers in the valley also have dual-purpose cattle.
La Esperanza and El Rosario
The new CATIE resident arrived in Comayagua in 1982 amid much pressure from ROCAP to validate the alternatives that had been developed up to that time in El Rosario, La Esperanza, and La Paz. This pressure was consistent with the new ROCAP interest in technology validation and transfer for the project, which was to end in late 1983.
I.n La Esperanza the project had worked with a corn-bean association, where the experimental variables were planting distance and the application of fertilizers and insecitides. The alternative was validated on thirty farms in 1982 and 1983.

Another corn-bean alternative was developed in El Rosario, but the beans were planted in relay rather than in association. That alternative was also validated on thirty farms at about the same time, but later proved uneconomic. The part played by SRN personnel in La Esperanza and El Rosario from about 1980 through 1983 seems to have been minimal.
The CATIE crop work in La Esperanza, El Rosario, and La Paz from 1978 until the validation period in 1982 and 1983 is described in five publications, three of them detailing the technological alternatives developed for the three areas and two giving a characterization of the agroecology of those areas. The documents were published by CATIE in 1984.
La Paz
In this part of the Comayagua Valley, the CATIE-ROCAP
project worked with corn-sorghum associations. Sorghum(creole) is used mainly for livestock, but is consumed domestically when corn fails. It is not an ingredient of tortillas, as in parts of El Salvador. A corn-sorghum alternative was validated by the project in La Paz(and Palo Pintado) in 1982 and is described in one of the 1984 publications just mentioned. The Animal Production Department of CATIE has also done some work in La Paz with sugar cane and leucaena as cattle forage.
In accordance with the new ROCAP thrust, the project began to work in mixed systems in 1983. Again, there was pressure to develop an alternative for validation before the end of the mixed-system extension of the project in 1985. Work had already been conducted separately on cropping and livestock systems in La Paz, so it was now a question of *omehow combining the two efforts.
The unavailability of forage during the long summer dry
season is the major constraint to cattle production in the area. The mixed alternative, planned and developed with much help from Turrialba, first involved both cattle and pigs. The cattle were to be fed with a cut mixture of sugar cane and leucaena, the pigs with residues from the corn-sorghum association described above, supplemented by pigeon peas for protein. CATIE had already conducted experiments in Turrialba and Honduras on the use of a cut forage mixture of leucaena and sugar cane, and the corn-sorghum association had just been validated in La Paz. The elements could now be combined in a mixed system. First, though, there had to be some local varietal testing of pigeon peas.
Work on the mixed system was temporarily-halted in 1983, when the head of the CATIE Animal Production Department refused to continue work in mixed systems in spite of ROCAP pressure to

do so and ROCAP money to fund it. Also, according to one report, researchers realized about the same time that local pigs were free roaming and not penned, as the alternative mistakenly assumed. Once the CATIE internal problem was resolved and the pig part of the alternative discarded, worked resumed, and by 1984 the alternative was being tested on three farms. Those tests continue on the same three farms today. The technology involves the feeding of dual-purpose cattle on a cut mixture of sugar cane and leucaena in the dry season, after the animals finish grazing the corn stubble. The data from the trials were being analyzed in Turrialba during my visit.
Of the CATIE technologies I encountered, this one may well be having the greatest impact as measured by farmer interest and acceptance, and by promotion by the local extension service. According to extension officials, the demand by local farmers for leucaena exceeded its supply. Leucaena is a CATIE introduction to the valley.
I visited two of the three farms where the trials are
underway and talked at length with one of the farmers, working in a corral at the time. Wooden feeding troughs stood in the corral; in them, and strewn over the ground, were the remains of leucaena and sugar cane. A cutting machine, developed by another project in the area and adopted by CATIE for use in the mixed system, stood in one corner of the corral. The cattle milling about the enclosure appeared healthy and content. Just beyond the fence was a field of leucaena. The farmer was enthusiastic about the technology, but planned to alter it somewhat in the future, planting more leucaena and changing its planting time. I asked him what his neighbors were doing with their cattle at this time. "They avert their eyes from the suffering animals," he said. The local practice is to allow cattle to roam the area freely as the dry season progresses, scavenging for what they can find. I saw several of these errant beasts, lean and untended, quite different in appearance from those in the farmer's corral.
San Jeronimo
CATIE-ROCAP project activity in San Jeronimo has been one of support to SRN research and extension activities there. San Jeronimo was one of the communities selected by the Unidad Central in the late 1970's, but it was not assigned to CATIE, which did not enter the community until 1983, well into the second CATIE-ROCAP project. The support role of CATIE in San Jer6nimo represents a significant departure from CATIE's usual modus operandi in Honduras, whereby it has worked in isolation and managed its own research operations.
The community of San Jeronimo may be unique for Honduras.

It is one of the few places today, I was told, where research and extension operate in a concerted fashion. It is of more than passing interest that the extensionist has been in the community since 1978, the researcher, one of the thirty trained in systems work by the Unidad Central during the fluorescent period, since 1980.
The focus of CATIE support in San Jeronimo has been on the validation and transfer of locally developed technology for rice production.
The material elements of the package include fertilizers, especially urea, and herbicides and insecticides. The recommendation calls for a new fertilizer formula and the early application of herbicides, thus eleminating the need for a second application. The validation was conducted on twenty-two farms in 1983 and 1984. I was told that about half of those twenty-two farmers continue to use the alternative. Some of the poorer farmers in the area are unable to obtain the necessary credit to purchase the inputs since they do not have enough cattle, which the bank requires as collateral for a loan. Also, the new fertilizer formula is not always available.
CATIE has supported the validation effort by providing technical assistance in setting up the trialsand material assistance for implementing them--agrochemicals, vehicles and their maintenance, gasoline for the vehicles, and salaries for seasonal labor. CATIE has also provided support for field days.
The local researcher and extensionist would like to .continue with the validation trials this year, using other farmers and trying different rice varieties. But they are concerned about how they can do this because the CATIE-ROCAP project is scheduled to end in June. Without CATIE, they say they will have to conduct fewer trials and to reduce the area planted to them. And they do not know how they will address the need for mobility to establish and monitor the trials. Here, the material contribution of CATIE to the validation effort is revealed in stark terms.
The CATIE Methodology
As regards influence on systems methodology employed in
Honduras, my talks with Hondurans suggested an ICTA influence, at least in terminology, for the terms they most often used for the sequence of activities in the farming systems research process were those of ICTA in Guatemala. Given the frequent interaction between PNIA and ICTA during the flourescent period, this is not surprising.- The real question is whether any systems methodology has been much practiced by Hondurans in recent years.

The flourescent period was the result of a rare confluence of human enthusiasm and brilliance, international monies, and the right political moment. Such a confluence has not been seen since late 1981, when agricultural research and extension began a rapid descent into a dark age. Some of those individuals trained by the Unidad Central are scattered about the country, like the researcher in San Jeronimo, and a few work in the bureaucracy in Tegucigalpa. But they can do little in the current political and economic environment, and remain instead flickering reminders of a brighter time. CATIE could nQt have had much influence on systems methodology since 1982, for most of its work has been independent of the SRN.
Should more good yet come of the flourescent period,
through a peculiar quirk of history, CATIE will have contributed to it. CATIE provided much support to the Unidad Central, training and otherwise, and the first CATIE resident in Honduras greatly influenced the systems thinking of the young leaders of the Unidad in Comayagua.
Institutional Relations and Operational Problems
As in other countries of the region, the sentiment was
voiced that CATIE should support national projects and programs rather than have its own, where there is little interaction with national institutions. Further, that support should respond to the needs of the country rather than those of the funding agency. The suggested solution to this problem was to admit Hondurans to the planning of projects. CATIE was further criticized for being inflexible in its planning, for not being able to alter plans once made, and for placing activities in a rigid time frame.
The criticism that CATIE operates independently of
national programs, perhaps the most serious one, was forceful and graphic. Indeed, two circles were drawn for me on a piece of paper, about two inches apart, one denoting CATIE's work site, the other an SRN work site. I conclude that there has been considerable tension between CATIE and SRN on this point, but did not have time to pursue the matter adequately. The criticism probably refers mainly to the period following the demise of the Unidad Central and the departure from the scene of the protective SRN Secretary and the dynamic young leaders. CATIE was left alone when the climate changed and power was assumed by forces inimical to the systems approach and to CATIE. The new regime, for example, pressured for CATIE to abandon its work in La Esperanza and El Rosario and move to Comayagua, but CATIE insisted on remaining in those communities until its validation work could be completed. The recent change in the CATIE modus operandi to a support role in San Jeronimo may be an attempt to improve relations.

Of the research and extension institutions in the five countries I visited, those of Honduras were probably the most impecunious. It may be no coincidence that I was twice reminded that the government would like to have CATIE's equipment--including a vehicle and some office furniture--at the end of the project. Perhaps the case of San Jeronimo is instructive: even there, where research and extension cooperate effectively, little could have been done without CATIE's material resources.
The poverty of national institutions, lethargic
bureaucracy--one administrator professed admiration for CATIE's fiscal flexibility in the field--and SRI personnel instability all tend to inhibit cooperation and favor CATIE's independent stance. The demands of farming systems research cannot easily accommodate those kinds of constraints.

History and Setting
The Instituto de Investigacion Agropecuaria de
Panama"(IDIAP) was founded in 1979 as a semi-autonomous institute within the Ministerio de Desarrollo Agrfcola(MIDA) and charged with agricultural research in Panama. Prior to the creation of IDIAP, agricultural research was organized along the traditional lines of discipline and commodity.
Within IDIAP, there is a unit for crop research and one for livestock research. The country is divided into three regions--Eastern, Central, and Western--each with an IDIAP regional director and under him coordinators for crop and livestock research. About three years ago, IDIAP began to decentralize decision-making to those regions.
There has been no extension service as such in Panama
since about 1968. Neither has there been an effective extension function since that time, say many Panamanians,.,who feel that this has been a major problem. The Production Department of MIDA has handled what extension that has been done, but they have worked almost entirely with collective farms(asentamientos campesinos) similar to the Mexican ejidos and created by the agrarian reform initiated by General Torrijos in 1964.
About six months ago, a restructuring of MIDA brought into existence the Servicio Nacional de Extensi6n Agropecuaria(SENAGRO) to take charge of agricultural extension in Panama. An agreement between SENAGRO and IDIAP is now being negotiated to facilitate cooperation between research and extension. There is local speculation that the creation of SENAGRO is the result of a seven-million-dollar USAID credit for technology, transfer in Chiriqui Province, where most of the food for Panama is produced.
CATIE did not commence systems research in Panami in
earnest until 1979, with the beginning of the second CATIE-ROCAP project and the creation of IDIAP. All project institutional interactions since that time have been with IDIAP. CATIE played an early advisory role to the new institute, helping it to define a clientele, to decide where to work and what to do. CATIE exercised a considerable influence on IDIAP at the time and exposed it to a methodology for generating technology for small farmers.

CATIE Activity in the Field and its Impact
CATIE has done research on both cropping and livestock
systems in Panama since 1979. From its base of operations in the city of David, capital of Chiriqul Province in the Western Region of the country, CATIE has worked in both Chiriqui and Veraguas Provinces. The work in Chiriqui has been with crops in the corregimiento of Progreso, near the Costa Rican border, and with cattle in the nearby corregimiento of Bugaba, both in Baru District. In Veraguas, Central Region, CATIE has worked in four corregimientos: Guarumal, Rio Grande, Quebrada Grande, and La Piedad, all in Sona' District. Both crop and livestock research have been conducted in this area, though mainly crop research.
I flew from PanamS to David, where I talked with IDIAP regional officials and visited nearby research sites(farms) in both Progreso and Bugaba; then went by car to the Guarumal area, on the Pacific Coast, to visit collective farms and talk with IDIAP officials at the Sona research station. From there I went by car to Santiago, capital of Veraguas Province and IDIAP seat for the Central Region, to talk again with officials. I returned to Panama by car the same day and met with IDIAP officials for the last time the following morning at central headquarters before leaving the country at noon for Costa Rica.
IDIAP had just received a large USAID grant in 1979, when CATIE entered the country under the second CATIE-ROCAP project, and had to select eight priority areas for agricultural research. CATIE and IDIAP together selected the sites, two of which, Chiriqui and Veraguas, were chosen for cooperative research between the institutions. The decision of whether to do crop or livestock research was based on which system, crop or livestock, was already predominant in the areas. Site selection ended in early 1980 and the characterization phase began. At IDIAP's request, CATIE did survey work in all eight areas before finishing the characterization of the Chiriqui and Veraguas sites in July of 1980. Research trials were in progress at those sites in 1981.
Chiriguf Province
It is estimated that sixty percent of the food for Panama' comes from Chiriqui Province--milk, basic grains, vegetables, coffee, beef, plantains, bananas, and sugar. About half of all IDIAP employees are based there, and more than half of IDIAP research is conducted there.
The CATIE-IDIAP team validated a cropping alternative,
rice followed by sorghum, in Progreso in 1983. Rice and sorghum is the traditional system there, and rice is an important cash crop. The major components of the alternative include chemical

insect and weed control, chemical fertilizers, and the use of improved varieties of both rice and sorghum. The team also conducted experiments in the area on corn, with planting time and spacing as the research variables.
I visited one farm in Progreso where validation trials
had been conducted and the technology package accepted, but there was nothing in the field at the time because the farmer was having trouble getting the necessary credit. This may have been due to the recent government policy restricting credit for rice as a way of reducing national production. Aside from the implications of the new credit policy, it is hard to assess the impact the technology has had on local farmers. There is no active extension service in the area, no one to promote. technology and monitor its acceptance by farmers. However, a study was conducted in Progreso by an IDIAP researcher(Ing. Franklin Becerra) on the adoption of rice technology and the economic benefits of rice research.
I was told that the farmer I met in Progreso was typical for the area and I had no reason to doubt that. Farmers there interact closely with the market for inputs and the sale of produce, and they often rely on local merchants for technical advice--not always reliable, so IDIAP researchers reminded me. The farmer I met seemed very aware of the institutional and political forces shaping his life, and receptive to the world beyond his household and community. The same could be said for the farmers I met in Bugaba, where livestock research was in progress. There are many farmers like this in Panama, I was told, more than in other countries of the region.
In Bugaba, site of the livestock research, a
diagnostic(diagn6stico estAtico, in CATIE terms) was conducted from January through March of 1980. The diagnostic involved a large number of survey variables and much quantification, with all the data entering a computer file. According to one source, much of the information has never been used. Forty farmers were sampled in the survey, some of whom were selected for on-farm trials.
The research was soon interrupted for a year(Sept.,
1981-Oct., 1982), when there was no resident CATIE livestock researcher. Work was resumed in 1982 and continued until an alternative with four components was developed. The components correspond to the way local farmers conceptually classify their animals: there is a calf component, a heifer component, a milk-cow component, and a dry-cow component. The components involve pasture management, or the use of improved pastures to improve animal nutrition and increase productivity. The project has worked with several grasses, including guinea grass(Panicum maximum), cudzu(Pueraria phaseoloides), and Digitaria swazilandensis, a close relative of pangola grass, common in Central America. The alternative also includes a commercially available mineral supplement and improved health care such as

parasite control, vaccinations, and brucellosis detection and elimination. Reproduction management is in the package as well--having cows and bulls together at the right time. The alternative was tested on five farms, with validation trials now underway on those same farms. But in contrast with the usual CATIE validation practice, where all inputs are given to the farmer, who provides only the labor, validation here requires the farmer to supply both inputs and labor.
On one of the farms I visited, a team led by an IDIAP researcher was busy tagging the ears of calves and recording data. There seemed to be good cooperation between CATIE and IDIAP in this livestock research endeavor.
Veraguas Province
The greatest concentration of collective farms(about twenty-two) in Panama is found in Veraguas Province. It is estimated that ninety percent of the agricultural production there comes from either those farms or one or two very large private farms. CATIE and IDIAP have worked on the collective farms.
The alternative validated by research in the Guarumal area in 1983 involves the planting of rice, followed by corn after a brief resting of the land. The package has the same components as the rice alternative for Progreso--chemical weed and insect control, chemical fertilizers, and an improved rice variety. According to local IDIAP researchers, farmers have taken readily to the fertilizers and herbicides.
I met an IDIAP researcher in Guarumal who bore the title "extension coordinator," a local position created by IDIAP in 1982 to link with MIDA extension, such as it is. I was reminded that it is the job of IDIAP to generate technology, not to extend it. But it may not be possible to extend the rice alternative now, given the recent government effort to reduce rice production through a tight credit policy. IDIAP researchers at the Sona station are looking for research alternatives for next year to fill the vacuum that will be caused by the drop in rice production.
In Guarumal, cattle are just as important, perhaps more so, than crops. CATIE-IDIAP began livestock research there in 1980, but ceased operations a year and a half later. The area was surveyed and research begun on dual-purpose cattle, traditional in Guarumal. The work was abandoned, so it seems, because farmers had little interest in commercial milk production. There was no tradition for it, despite the good roads. Now, IDIAP talks of continuing this research; the highway is on a milk-collection route and local farmers are reported to be interested in commercial production.

CATIE-IDIAP has also done experimental work in
Guarumal(and in Progreso) with corn--planting time and plant spacing--and with sorghum--weed and insect control, fertilization, planting time, and plant spacing. Some work was begun in Guarumal with forage legumes in anticipation of a ROCAP request(which was never made) to work with mixed systems.
The CATIE Methodology
In the Guarumal area, IDIAP researchers said that the
CATIE methodology provided them with a way to approach a problem and solve it. They plan to continue the scheme for validation and transfer of technology in Sona District once funds from MIDA's technology transfer project are available.
IDIAP Central Region wants to begin work soon with some of the marginal Amerindian groups in the area, where agricultural technology is inadequate and environmental degradation severe and refractory. IDIAP does not have the resources for this work and plans to seek CATIE's help as well as to follow the CATIE systems approach in confronting the problem.
A debate over farming systems research methodology has
engaged the passions of Panamanians in recent years. The two key actors on the methodology scene in Panama have been CATIE and CIMMYT, both playing to an animated IDIAP audience. The work of each is judged by Panamanians in light of work by the other. The debate focuses acutely on work by CIMMYT in Caizan Province over about four years, from 1979 until 1983. CATIE had begun work there in late 1977, before, and perhaps in anticipation of, the second CATIE-ROCAP project, which did not enter Panama until 1979. No reference is made in CIMMYT publications to this earlier CATIE work.
The CIMMYT work in Caizan improved the traditional corn system, whereby corn was sown broadcast with no weed control or use of chemical fertilizers. CIMMYT introduced a new corn variety, the use of agrochemicals, and the planting of corn in rows. The production increase was dramatic and the adoption rate by farmers approached 100 percent. This achievement drew much favorable attention to CIMMYT in Panama.
The CATIE response was swift. As their argument goes, it took neither ingenuity nor effort to achieve results given the traditional system, especially when there was a ready market for corn. Further, the area of influence of the new technology embraced only fifty or so farms concentrated on 600 hectares in a small valley. The results do not, therefore, point to the superior merits of any research methodology, and the achievement has been vastly inflated.
Although the debate centers on Caizan, there seem to be

complicating factors, which I did not have the time to sort through. Reports suggest, for example, that the early CATIE work in Panama'(in Caizal1?) sought with some persistence to promote cropping associations in areas where there was no tradition of them. It is true that much Panamanian agriculture follows a monocropping pattern, and is in this regard different from that of most of the limited-resource farms served by CATIE elsewhere in Central America. This aberrant monocropping tradition is itself an interesting phenomenon. But the important point is that CATIE is seen by many Panamanians as inseparably wedded to working with cropping associations, the contrary evidence of the alternatives validated in Progreso and Barumal notwithstanding.
An event related to the CATIE work in those two areas has some bearing here. At a time when work on the alternatives there was in full flower, an advisor to the CATIE Crop Production Department secured by ROCAP visited the CATIE work in Panama and filed an extremely negative report. The report charged that CATIE was not doing systems work, as stipulated by the CATIE-ROCAP project mandate, since the alternatives did not involve crop associations. The report must have generated a small furor, for a ROCAP official journeyed to Panama in the company of a distinguished IRRI cropping systems specialist to investigate the charge. The CATIE work was soon pronounced sound given the Panamanian setting and the matter was dismissed. This event not only exemplifies the kind of pressures that CATIE has had to contend with, but also the profound confusion that so often attends farming systems research, and the peculiar battering to which its practitioners are subjected in our time.
The CIMMYT work in Caizan and subsequent events have led several researchers in Panama to view CATIE as too theoretical, too occupied with the time-consuming and costly task of gathering a miscellany of data that have no apparent purpose. As one IDIAP researcher put it, "The CATIE methodology is lacking in dynamism and agility. CATIE's overwhelming concern to develop a methodology for the region blinds it to the needs and realities of Panama."
Today, both CIMMYT and CIAT are engaged in training IDIAP personnel in systems research. CIMMYT only recently began its call system in Panama. CATIE's training endeavors, vigorous and influential in the early years, have waned over time.
To conclude, CATIE's influence on farming systems research methodology in Panama was considerable at the turn of the decade, as IDIAP was casting about in search of a way to do agricultural research there. But that influence declined as competing alternatives entered the country and forced Panamanian researchers to consider seriously--in a way that I encountered in none of the other countries I visited--their methodological needs as they confront Panamanian realities.

Institutional Relations and Operational Problems
The debate over methodology has not been the only source of tension between IDIAP and CATIE in Panama. The management of CATIE funds for project operations has also figured. These funds are for operating and maintaining vehicles, for buying agricultural inputs, and for procuring seasonal labor. CATIE has reserved control over the disbursement of the funds for itself. But IDIAP officials maintain that they should control that disbursement. They say that three years ago, before IDIAP decentralized its decision-making to the regions, the CATIE policy made sense. Today it does not.
They say further that control of the funds gives CATIE too much independence, thus isolating them in a way that IDIAP does not always know what they are doing. This isolation also foments an unhealthy dependence on CATIE of IDIAP personnel assigned to work with them. The IDIAP researchers communicate more closely with CATIE than with their colleagues in IDIAP. Even their reports go first to CATIE. And CATIE cannot train the wider body of IDIAP researchers if it works in isolation. So the IDIAP argument runs.
The problem has a yet greater dimension, one related to the limited resources available to IDIAP, which seems to have fallen recently on hard times. But it may be related as well to the way those resources are managed. It is widely rumored that administrative support for research within IDIAP is weak. I was told that the budget for this year, for example, had not yet been released. At the regional operations level, there are often insufficient funds to operate and maintain vehicles. International donors are little help here, for they only provide costly equipment items such as vehicles; operational costs for gasoline, vehicle maintenance, and researcher per diems must be borne by national programs.
There is also an indication that the personnel support provided by IDIAP to collaborative efforts with CATIE has not always been either stable or in sufficient quantity. Three researchers in three years were assigned to one of the crop-research sites, each remaining only a year. Each had to be trained in turn. In the livestock research area of Bugaba, work began in 1982 with five IDIAP researchers, two of whom left within a year and a half and have not yet been replaced.
To operate in this setting, CATIE has often had to do more than its share. Gasoline and other resources have been provided to IDIAP researchers in order to keep operations going. In Panama' as in other countries, the requirements of farming systems research are often not understood by either national administrators or international donors. In the farming systems approach, much of the research is conducted on farmers' fields, which means that researchers must have the capacity for mobility.

Trials have to be established and frequently monitored. And once the process begins and farmers are involved, it must be continued; a progression of sequenced activities must occur in a timely fashion. If a given activity in that progression does not so occur, activities previous to it cease to have value, the farmers' confidence is lost, and the whole research endeavor fails. At the field operations level, therefore, research management must have great flexibility in the use of resources. This may be one reason why CATIE has been reluctant to relinquish control over the disbursement of operational funds.
The independence of CATIE that is resented by IDIAP
officials turns on the control of funds--on that and nothing more. This control undeniably gives CATIE a potential autonomy vis-a-vis IDIAP. But that autonomy may in actuality have been less than some of the complaints would suggest. In contrast with criticism leveled at CATIE by research administrators in some of the other countries I visited, at no time did anyone in Panama accuse CATIE of operating in a vacuum, indifferent to the interests of IDIAP and without the involvement of that institution in the planning and execution of projects. And my forays into the history of CATIE-IDIAP relations do indeed suggest a high degree of involvement of IDIAP in project undertakings. IDIAP was very involved in the livestock research underway in Bugaba, for example, where relations between the two institutions would appear to be excellent.
A lesser complaint, and one I encountered in other
countries, was that CATIE takes research data to Turrialba for analysis and storage. Nationals often do not have access to it, and any training value for them that the analysis might hold is lost. It is my understanding that the Bugaba livestock project recently made arrangements to do its data analysis in Panama, thus solving that problem.

History and Setting
Agricultural research and extension are organized along very traditional lines in Costa Rica. Each has its own administrative unit within the Ministerio de Agricultura y Ganader'a(MAG). The Direccin General de Investigaciones Agrfcolas(DGIA), which is charged with crop research, is organized into several departments by discipline and commodity--sugar cane, coffee, entomology, and soils, to name a few. Research is conducted on five experiment stations. There is also a Direccin General de Salud Animal y Producci6n Pecuaria, which seems to be mainly concerned with animal health. In contrast with research, which has no regional manifestation, extension is further organized into regional centers, each with agencies scattered about the region. Those centers project the interests of the Ministry in the countryside.
As the organizational scheme would suggest, there has been little effort to make research respond to the needs and opportunities of the different regions of the country. As one Costa Rican researcher put it, "If the minister in office has an interest in dairy production, then dairy production is promoted; if his interest is citrus, then that gets promoted." According to a politician in the know, the Consejo Nacional de Produccidn, working together with MAG, formulates government agricultural policy. The Consejo decides that there should be a production increase of a commodity in a given region, then authorizes the national bank to issue the necessary credit. The source goes on to observe that a certain group of producers--"friends of the government"--are always favored by policy and the work of MAG. And these "friends" are all larger farmers.
MAG has shown little interst in systems research, the existence of a cropping systems department within DGIA notwithstanding. And there are no systems courses offered at the University of Costa Rica. The few Costa Rican students who have majored in farming systems research at CATIE have been unable to exercise their interests within MAG.
This lack of interest in systems research may soon change, however, in consequence of the reorganization of research and extension now underway as the condition for receipt of a large

BID loan. The reorganization, sanctioned by law in March of this year, combines research and extension into a single Direccion de Investigaci6n y Extensidon Agricola at the national level, with separate subdirecciones for research and extension. This structure descends to the regional level, creating regional centers with separate heads for research and extension. The new scheme not only integrates research and extension administratively, but decentralizes the research function to the regions. This is novel for Costa Rica.
Central to the reorganization of MAG is the Programa de Incremento de la Productividad Agricola, or PIPA. Financed by BID(70%) and the Government of Costa Rica(30%), the PIPA seeks to increase agricultural productivity through the work of its subprograms in research, technology transfer, seed production and distribution, and the supply of agricultural inputs. The PIPA will not implement programs in the countryside, but rather will fortify and amplify the agencies and programs of MAG. It will capitalize institutions that do not belong to MAG, such as the Department of Agronomy of the University of Costa Rica, and enter into cooperative agreements with agricultural credit and other agencies that bear on agricultural productivity. Perhaps most important, the PIPA is to guide and coordinate research and extension activities in the country.
Operating from a systems perspective, the PIPA is to make research and extension responsive to the needs of farmers by following a process that begins with understanding farm conditions and farmer problems(la problemitica del agricultor), moves to technology design and testing, much of which is to be on farmers' fields, and finally enters an extension phase.
According to my information, livestock research does not figure in the reorganization of MAG or in the plans of the PIPA.
CATIE Activity in the Field and its Impact
The CATIE-ROCAP project began operating in 1979 in Limon Province, eastern Costa Rica. Work centered on the canton of Pococi, districts of Guapiles, Jimenez, Cariari, Rita and Roxana; and on the canton of Guacimo, districts of Guacimo and Rio Jimenez. The cantons lie in the humid tropical lowlands and are well described in one of CATIE's characterization reports, published in 1984. I visited both cantons.
Although most of CATIE's farming systems work under the second CATIE-ROCAP project has been in the Pococi-Guacimo area, there was also activity in the canton of Puriscal, San Jose Province of Alajuela Department, from 1979 until 1982, when operations there were halted(CATIE remained in the area, but with

another project) because of excessive operating costs. This decision by CATIE may have been related to a marked reduction in the number of MAG personnel assigned to the collaborative effort in 1982. The project had entered Puriscal rather than continue work begun under the first CATIE-ROCAP project(l975-1979) in San Isidro, Puntarenas Department, in the Pacific region of Costa Rica to the south. As happened later in Puriscal, work in San Isidro became too costly, thus forcing CATIE and MAG to move. I was not able to visit the Puriscal area, considered to be one of the poorest in the country, but I did talk with persons involved in the work there.
Crop Research in San Isidro and Puriscal
The research in San Isidro, under the earlier CATIE-ROCAP project, was with corn and with the associations corn-beans and corn-cowpeas. According to one report, the work was not successful because the systems developed were too complex for farmers to manage. It was a time when CATIE was still learning to work with farmers and their systems.
The MAG-CATIE collaborative effort in the Puriscal area
from 1979 to 1982 was the nucleus of agricultural research there, where MAG was implementing the World Bank's Training and Visit System for extension. CATIE worked in Puriscal with associations, which had little tradition in the area. The associations were to help control erosion, a problem on the cultivated slopes. In spite of severe bird damage and the loss of some of the on-farm experimental repetitions, the project developed a corn alternative in 1981, with variety, planting distance, chemical weed control, and fertilization as components. The alternative was never validated. Research was also conducted on beans in 1981, with variety, seed treatment, and planting density as experimental variables. The experiments could not be completed, however, because participating farmers harvested the beans without control. Work declined sharply in 1982 owing to. a dearth of MAG personnel assigned to the effort. In that year, the project conducted corn variety trials on a single farm as well as a few experiments with bean varieties.
The Cropping Alternatives in Pococi-Guacimo
The earlier CATIE-ROCAP project had also worked in Pococi, beginning in 1976. Experiments were conducted on the system corn followed by corn, and on two associations, corn-yuca and corn-yuca followed by beans that climbed the yuca stalks.
Both corn and yuca are staples in the area; yuca is

especially important on the farms with extremely limited resources. Under the second CATIE-ROCAP project, the corn-corn system was validated on about forty-five farms during 1982 and 1983. The corn is a local variety, and the alternative calls for a change in plant spacing and for the application of fertilizers, herbicides, and insecticides. There is evidence that this alternative, or at least parts of it, has enjoyed some success with area farmers. The credit bank of the Instituto de Desarrollo Agrario(IDA) is using the alternative as the basis for credit to 125 farmers, most of them in Cariari, where the alterntive was validated. One reliable source estimated that seventy percent of the farmers in the Pococi-Guacimo area had adopted the alternative, while another gave an estimate of ninety-five percent, but qualified it by noting that some farmers are varying the recommendations, or adopting only parts of the package, mainly fertilizers and the new spacing, because of the high cost of the agrochemicals. There was also a complaint that banks in the area give only about half of the costs required to produce and market a hectare of corn. I was told that the MAG extension service is not promoting the CATIE alternative but rather a different one.
A corn-yuca alternative was also validated for the area in 1982 and 1983. The alternative involved a new variety of y uca, a change in plant spacing, and chemical weed control. The key component was to be the new variety, a rapidly maturing one that would substantially increase production. But there were two problems.
First, the new variety was widely rejected by farmers for domestic use(I had no time to pursue this). Pressure to validate the alternative(the project was to end in 1983), which was done using only cuttings from the trial plants--the only available--may have led researchers to overlook early farmer reactions to the variety. And second, a sudden downward swing in the price of yuca in 1984 made commercial production uneconomic. Just after the alternative was validated, local banks extended credit for yuca production. The larger farms in the area(I understood there to be a land tenure problem) planted large areas to yuca, thus flooding the market and driving the price down. The banks were indemnified by the government. Banks in the area today will not give loans for crops grown in association; whether this was true when the alternative was first validated, I do not know.
The Mixed System in Pococi-Guacimo
The work on mixed systems in the Pococi-Guacimo area began in 1982. The idea was to work with pigs, feeding them with produce from the farm. Trials were conducted with legumes,

including soybeans, to provide protein. There was no work with improved breeds. The alternative developed has been tested on three farms in Pococi and three in Guacimo. In addition, several farms have been monitored to serve as controls. The alternative has also been established on the nearby Los Diamantes experiment station. The results of the research were being analyzed during my visit.
The components of the alternative were described to me as follows: a swine(creole) herd of five females and one male; soybean meal purchased on the market; one hectare of sugar cane for feed when other sources such as corn, yuca, and bananas are unavailable; the residues from five hectares of corn grown for sale and home consumption; and the residues from two hectares of yuca grown for sale and home consumption.
The system was initially to include dairy cattle as well, with whey for the pigs. The improvement of milking corrals, the partitioning and improvement of pastures, and selective breeding practices were to be components of the alternative. But little work was done on the dairy module: there was no time for it and no local dairy tradition. Toward the end of 1984, a dairy installed a collecting vat in the area and some milk is now being produced. And there is today some interest in what would have been the dairy module of the mixed system.
It was the opinion of one area researcher that the pig
module would be more successful if an improved breed of pig were used. The creole pig has a high fat content and there is little market demand for it. Nonetheless, there is good evidence of an interest in the pig module by local farmers.
The CATIE Methodology
Those few MAG researchers and extensionists who have
worked closely with CATIE over the years have probably developed some appreciation of the CATIE methodology and its ability to make research and extension more responsive to the needs of farmers. But otherwise, the methodology has had little impact to date on the way research and extension are conducted in Costa Rica. This seems due in the main to a profound lack of receptivity to the systems approach on the part of the government. But as I said in the initial section of this country report, that may soon change.
The PIPA seeks to achieve its objectives by following a
systems approach--one very like that advocated by CATIE. I saw a hand sketch of this approach in the PIPA office of MAG. It includes diagnosis, the on-farm testing of technologies, and a stage called validation and transfer. The scheme, even the terms

for describing it, are strongly suggestive of the CATIE methodology.
Two challenges await the PIPA. First, it must solve the political problem of directing research and extension efforts at small farmers. And second, it faces the daunting task of reorienting traditional researchers and extensionists to systems thinking and to on-farm research.
In a MAG brochure describing the PIPA, CATIE figures prominently as a participant in research, in the training of extension-agents, and in the production and distribution of seeds. If the government can successfully meet the challenges above, and if CATIE and MAG can cooperate in a viable fashion, CATIE may yet have a considerable impact on the way agricultural research and extension are conducted in Costa Rica.
Institutional Relations and Operational Problems
Cooperation between MAG and CATIE over the years has been minimal; CATIE has operated with much independence, often hiring its own technical assistants. This has been true for the Guacimo-Pococi area, where relations appear to have been mostly cordial and a very limited cooperation has relied on personal relationships. MAG officials there expressed the need for a mechanism whereby they could know what CATIE is doing and the results of its research. They hoped that CATIE would make its research data available to them when it left the area.
The lack of human and material resources within MAG for systems research is noteworthy and has conditioned its cooperation with CATIE. MAG personnel posted to the collaborative effort in Puriscal included an agronomist at forty percent time and an extensionist at five percent. Another agronomist was promised full time, but was never sent. As a result, the project was understaffed and eventually had to leave the area.
MAG researchers complain today about the unavailability of vehicles and about the small per-diem allotment of 1,200 colones per month, barely enough for a week's lunches. The presence of CATIE in an area brings vehicles, gasoline, and agrochemicals--all for doing on-farm research. One Costa Rican said that if MAG extension were conducting validation trials, the farmers would have to purchase the inputs, for extension has no money for that. "When the CATIE-ROCAP project ends," he continued, "MAG will abandon systems work. CATIE works with small farmers, but MAG has no money for that."
The lack of resources is a problem in other countries of

the region too, but Costa Rica differs from them in that it has made no commitment to small farmers or to systems research. And this lack of commitment may explain in part the lack of available resources: the government's priorities lie elsewhere.
I also encountered the charge that CATIE's work had
benefitted CATIE more than it had Costa Rica. There may be some truth to this, in Costa Rica and elsewhere, if it refers to the first CATIE-ROCAP project, when CATIE was learning about small farmers in the region and was developing a methodology for working with them. But in light of the government's priorities in recent years, the charge rings hollow. One Costa Rican researcher summed it up quite well: "The Government of Costa Rica," he said, "simply does not know how to use CATIE."
CATIE has also had some problems on the strictly
operations side. There is evidence of the forcing of data collection, especially on the mixed-system farms, where farmers were visited daily, often at an inappropriate time, and became annoyed with the procedure. This may have been caused by pressure on CATIE researchers to meet deadlines. Those same pressures, coupled with a shortage of cuttings and a consequent reduction in the area planted to validation trials, may have led researchers to overlook farmers' reactions to the new yuca variety developed in Guacimo-Pococi.

In accordance with my assignment as described in an agreement between CATIE and ROCAP, I set out to look at the impact of the CATIE methodology(as measured by adoption by national research and extension institutions) as well as of technologies that had been developed by it and promoted by national institutions. I soon realized that that impact would be extremely difficult to assess, and consequently altered and enlarged the scope of my work.
The countries of Central America have been exposed to several methodological influences, as I have tried to show in this report. The result is sometimes a methodological hybrid, or even different methodologies holding sway in different regions of a country. Or maybe CATIE has sensitized researchers to the need to look at problems in a systems perspective. That too might qualify as an impact of the CATIE methodology. This is all to say that it can be exceedingly difficult to sort out methodological influences for a national setting. Also, CATIE operatives have sometimes been more flexible in practicing and promoting the so-called CATIE methodology than some of the sterile, formulaic descriptions of it would suggest. The national settings have forced this flexibility. Such is the case with the early CATIE work in the Chimaltenango area of Gdatemala, where the CATIE resident had to trim his sails to the ICTA methodological winds.
The promotion of technologies developed by the CATIE
methodology is no less problematic, and a caveat must be entered here. Whether a given technology has or has not been promoted does not necessarily reflect on CATIE. Panama, for example, has effectively had no extension service for fifteen years, and El Salvador has not been able to promote anything in the rebel-occupied zones, where CATIE has worked.
For information on the impact of CATIE methodology and technology, the reader is referred to my reports on the individual countries, each with its peculiar institutional cast and political problems. There, the two topics are dealt with under separate section headings.

CATIE, National Sovereignty, and the Demands
of Farming Systems Research and Extension
CATIE has experienced a series of problems in countries of the region that derive from conflicts between national sovereignty and the demands of the farming systems approach to agricultural research and extension. These problems must be faced by any international center that intervenes in national settings to implement this approach.
International research centers differ fundamentally from national research institutions, for they have no countries in which to gain experience and ply the systems craft. And to state the obvious, there are no international farmers.
In all of the countries that I visited, there was
criticism of CATIE at high administrative levels for its alleged autonomous behavior, either fiscal or programmatic. Expressed differently in each of the countries, the criticism was strong in Guatemala, Honduras, and Panama; less so in El Salvador and Costa Rica--CATIE does not coordinate its activities with those of national projects, CATIE works alone, CATIE has its own projects, CATIE manages its own operational funds. Thus ran the litany.
Much of this criticism should be seen in light of the
operational requirements of farming systems research: field teams must have the capacity for mobility--vehicles, their maintenance, spare parts, gasoline, and per diems must all be readily available. Farming systems research involves a timed sequence of activities that must not be interrupted. An interruption at some point in the sequence nullifies all previous steps, farmer confidence wanes, and the research effort succumbs. Once the farmer is engaged, it is imperative that the process continue.
National-researchers and extensionists most often do not enjoy this capacity for mobility. CATIE, therefore, has frequently provided the resources to make farming systems research and extension possible. And to be able to do this, CATIE has insisted on some autonomy--in the control of operational funds, for example. Insistence on this control has in turn provoked tensions between it and national administrators. National researchers and extensionists at the local level, where CATIE has operated, are poorly paid and lack resources to do any kind of research and extension, let alone systems research and extension. And these same personnel often feel neglected by superiors in their own organizations. Since CATIE has *resources, national personnel at this level become dependent on CATIE. CATIE becomes their patron, and to CATIE they render first allegiance. And this further alienates them from their own organizations, thus deepening tensions between their administrators and CATIE. The vital importance of those resources to the farming systems research effort, and national

institutions' chronic lack of them, may explain why CATIE is so often seen by nationals as a provider of material goods--not that nationals think CATIE unable to contribute in other ways too. It might be conjectured that if CATIE were not doing farming systems research, then there would not be this focus on its material resources. And there may also be an element of envy behind the administrators' criticism of CATIE. The whole problem is confounded by a lack of understanding on the part of national research and extension administrators--and donors--of what farming systems research and extension require with regard to operations and resources to finance them.
A couple of minor operational foibles have also annoyed nationals. First, CATIE customarily sends its research data to Turrialba for analysis, thus depriving nationals of any training benefits that might otherwise be theirs through participation in that analysis. CATIE was criticised in a couple of countries for this procedure. Second, I was told that CATIE does not pay sufficient attention to its field operatives, or to those persons who interact frequently with local farmers. The operatives feel that their advice is often ignored by--as one of them said--"office researchers." If this is true, it may be related to the haste with which CATIE sometimes works in order to keep to donor-imposed schedules. This topic is further discussed below.
Although CATIE has theoretically collaborated closely with national projects and programs, providing support to them in a subservient role, the reality has often been otherwise. CATIE has directed and implemented entire projects in areas either assigned to it by a host government, or selected jointly. National research and extension institutions have then-been placed in the position of supporting CATIE, which, as this report shows, they have not always done. There have been several instances where CATIE has been abandoned by national institutions and has had to work alone. In El Salvador, when implementation of the 1980 agrarian reform demanded the exclusive attention of all agricultural personnel, CATIE was left alone in the north. CATIE has often filled a vacuum in El Salvador, doing what national research and extension institutions there have not had the resources to do. CATIE was also left alone in Honduras when the flourescent period ended and political forces inimical to CATIE--and to farming systems research--assumed power. The case of Costa Rica is somewhat different, for there the government has been indifferent to systems research and to small-farm agriculture. But the result has been the same: CATIE has worked alone.
This awkward position of CATIE alone, or nearly so, deeply immersed in sovereign institutional and policy settings, bears on another requirement of farming systems research and extension: institutional and policy coordination. First, there must be coordination between research and extension, no small matter in some of the countries of the region. There must also be coordination among institutions that supply inputs such as seeds,

fertilizers, and other agrochemicals; and institutions that provide credit and market agricultural products. And there must be a favorable policy environment for small farmers, one that is committed to addressing their needs and that guarantees reasonable prices for both inputs and products.
The political economies of the countries in which CATIE has operated are profoundly unstable, as are the research and extension institutions, where personnel changes are frequent and unpredictable. Domestic political forces in those countries cannot achieve the requisite institutional and policy coordination for effective agricultural research and extension. Is it realistic to think that CATIE can achieve this?
The depth of CATIE's immersion in the institutional and policy environments of countries of the region warrants serious consideration. Any evaluation of CATIE's farming systems activities and achievements must take account of that depth, for it links CATIE's fortunes, which have not always been good, to forces that are entirely beyond CATIE's control. El Salvador again provides the example. The divisions of research, extension, and seed technology have not often operated in concerted fashion in the areas where CATIE has worked. In Jocoro, four separate offices have had to be coordinated: livestock research, crop research, livestock extension, and crop extension. CATIE has expended much effort in this coordination, and now wants to address the problem of inadequate credit programs for small farmers. Can CATIE manage all of this?
Once immersed in a national setting to do farming systems research, it seems that two courses of action are open to CATIE. First, there is the course taken in El Salvador, at least in Jocoro, where CATIE has sought to coordinate institutions and policies--to do, that is, what the government could not do. This seems a hopelessly impossible task. And there is a second course, one taken in Costa Rica and Panama, whereby CATIE does not seek the required institutional coordination but elects instead to operate on a narrow research front, developing a technology and delivering it to some government agency for disposition. Whether there is an effective extension function, or whether the agrochemicals and credit required by the technology are available, are not CATIE's problems. This, too, would seem an unsatisfactory way to proceed.
Under either course of action, if CATIE really operates
alone, with minimal to no participation by national institutions, there can be no on-the-job training benefits for nationals. And if national research and extension institutions do not have the resources(or do not allocate them), or the commitment, or the institutional coordinating capacity to engage in farming systems research and extension, what have CATIE's achievements proved? And what have the countries gained from it all?

The Dual Track: Basic Research Versus Pragmatism
Over the years, CATIE has moved on a dual track with its farming systems program. One track is basic research, the other is pragmatic, applied research. The basic research track is best exemplified by the strong thrust to develop a methodology for farming systems research. Ardently supported by ROCAP, this thrust has been a major part of both CATIE-ROCAP projects. Developing methodology certainly has about it the aura-of science and is thus in the best tradition of an international research center. The work on extrapolation, and what some would argue to be the excessive collection of data for farming systems research, also fall within the tradition.
The other track, the applied track, is what farming systems research is all about. Its concern is eminently pragmatic, for it seeks to develop appropriate technology for limited-resource farmers, and to do so in the most expeditious, economic way possible. The-farming systems approach to research has about it a sense of urgency--indeed, it developed to satisfy an urgent need, one not being effectively addressed by the more traditional approach, which confined itself almost exclusively to station-based research conducted by biological scientists.
Associated with each of these tracks is a set of attitudes and self-perceptions. Those who practice science, it is often held, are following a higher calling in life. The science-minded often value methodological rigor for its own sake and pursue it without regard to any tangible returns. They have been trained in these values and their colleagues have provided much group reinforcement. Applied researchers, on the other hand, are more pragmatic: they do not value methodological rigor for its own sake, but rather require of all endeavor that it generate some reasonably predictable, practical result in the near term. Applied research is commonly seen as less scientific and, therefore, less demanding intellectually. The job of the applied researcher is not to create knowledge but to implement received knowledge. Accordingly, applied researchers often occupy a lower rung on the professional ladder.
It is my thesis that these two tracks, or traditions, have not been well articulated in the CATIE-ROCAP farming systems projects. And this lack of articulation has led some national research and extension workers to take a dim view of CATIE. The problem may well begin with the mandate of the CATIE-ROCAP projects--to develop a farming systems methodology for the region--, for CATIE, having no country, can only do this development work in the surrounding countries. Two national administrators described their countries as laboratories for CATIE, another described his as a platform that supported CATIE. In all cases, the basic question was: What is in it.for us?
The effort to develop methodology, to pursue science, has

often clashed with what national researchers are about: increasing national food production at the lowest cost, by whatever means. National researchers sometimes see CATIE as pursuing its own esoteric interests, which they do not share and which strike them as having little relevance to their needs. As I heard in Guatemala, "CATIE researchers are thinkers, not doers." Unlike national researchers, they are not forced to act. CATIE's superior economic position, and its well-paid personnel with secure jobs, may also influence the attitude of nationals toward it.
Research on Cropping Associations as Systems Research
As the first CATIE-ROCAP project began in 1975, the
agricultural research world was discovering the rationality--both economic and ecological--of cropping associations. Such associations were common enough on the small farms of Central America, and much of the early CATIE systems work sought to understand their biological interactions. That was a major objective of the famous Central Experiment in Turrialba.
There is the perception in both Guatemala and Panama that CATIE works only with associations. One Guatemalan researcher, trying somewhat clumsily to make this point, said that ICTA did not work with systems but rather with commodities. What he meant was that ICTA did not work with associations as such, but with components as parts of associations, or broader systems. Panamanians told me that CATIE tried to force associations there, in areas where there was no tradition for them. And a dispute in Honduras centered on whether vegetables were to be planted alone or in associations.
Although my evidence is largely indirect, I feel strongly that CATIE has, at certain times and places, promoted the intrinsic value of associations. Also, I think that CATIE has elected to operate on associations as such in its research interventions. It has sought to develop, that is, improved alternative associations. The association, in other words, becomes the research object. If an association is composed of components w,x,y,z, for example, CATIE's research object would be w,x,y,z. Much farming systems research, however, would work only with, say, x, but would do so in a way that the relations between x and w,y,z--and between x and any other important component--are considered.
If I am right about this commitment to associations, then CATIE has operated with a fairly distinctive notion of systems research for the Americas. Once at CATIE, I soon learned something of the way researchers there conceptualize their domain of intervention. According to a few key researchers, the inclusion of livestock(or forestry) distinguishes farming systems research from cropping systems research, where CATIE began in the

early 1970's. As I was told, only in recent times did CATIE acquire the expertise to work with whole farms.
At the risk of over simplification I would summarize the above as follows. An intervention that has as its research object only a single crop would not qualify as systems research. If the research object includes more than one crop, the effort qualifies as cropping systems research. And if livestock(or forestry components) is further added to that object, the intervention qualifies as farming systems research. I would not be so bold as to assert that all CATIE researchers share this view today, or that all of them ever did. But I would argue that the view was widespread during CATIE's formative period in the 1970's, and that it lingers today as a "deep structure" in the minds of several key researchers.
On Donor Pressures and Interventions
Both CATIE-ROCAP projects, under which nearly all of
CATIE's farming systems research has been conducted, have been USAID projects implemented by CATIE. Accordingly, ROCAP has exercised much influence on the projects, intervening programatically and pressuring CATIE periodically to achieve specified results by given dates.
Although the basic methodology for farming systems
research was developed during the first project, the second mandated a continuing concern with methodology, adding to it the new thrust of technology transfer. Extrapolation research and technology validation and transfer efforts have both been pushed heavily by ROCAP, as has also the work in mixed systems--crops and livestock--from 1983 until June of 1985.
ROCAP has monitored the project closely over the years and has intervened, not always in a positive way. I related the incident in Panama, where a ROCAP consultant challenged CATIE for not working with associations. I also briefly cited an incident from Honduras, where ROCXP pushed for vegetables to be grown in associations.
There is also evidence that CATIE's rush to validate
alternatives, or to achieve some other stipulated product by a given date, has sometimes resulted in careless workmanship, or has created tensions between CATIE field workers and local farmers. There may be something to learn here from CATIE's work with yuca in Costa Rica, or its work with mixed systems there.
Effective farming systems research requires great
flexibility on the part of research: technologies must be field tested and modified, sometimes drastically, as information from those tests cycles back to research. This process does not always square with the rigid AID project design, implementation,

and evaluation scheme. Indeed, farming systems research does not fit the five-year AID project term. CATIE was criticized in some of the countries for its inflexibility and the rigid time frames of its work. ROCAP pressure for results may account for much of this criticism.
This is all to say that any evaluation of CATIE in farming systems research is also an evaluation of-ROCAP, for the two have been inextricably intertwined in the endeavor.
On Validation and Transfer
The implementation of the validation and transfer stage of the CATIE methodology has been a major effort of the second CATIE-ROCAP project since 1982, when the activity entered the staged progression. It was conceived as transitional between the experimentation-cum-field-testing stage, and the final stage of dissemination to the masses of farmers. I argue here that this stage is problematic and may need further attention.
The stage involves the validation of a promising
technology--one that has passed the testing stage--on at least thirty farms(livestock validations seem to use fewer). Validation plots(for cropping technologies) are larger than test plots and the technology is managed entirely by farmers, with all material inputs furnished by CATIE. Harvests go to the collaborating farmers. During validation, the technology is carefully monitored for technical and economic feasibility under farmer management. If the results are favorable over two cropping cycles, the technology is delivered to the extension service for promotion. If they are not, it is returned to research for alterations.
Ideally, research and extension cooperate closely during the validation stage. The farmers' management of the technology is observed, especially his ability to follow the instructions. This observation aids extension in predicting adoption rates as well as in knowing how best to promote the technology.
The historical emergence of the validation stage is worth noting. An evaluation of the first CATIE-ROCAP project remarked the lack of adoption of technologies that had been developed, and concluded that more attention to technology transfer was called for. In this way, technology transfer became an important part of the second project.
My inquiries encountered little evidence of work by CATIE on extension during the early years of the second project. There was a plan to do research on farmer "communication channels" in Tejutla, El Salvador, but little seems to have come of it. By 1982, ROCAP pressure was strong for CATIE to develop an extension methodology. But CATIE resisted, arguing that it had neither the

money nor the expertise to do so(the extrapolation work excepted). After a period of deliberation between ROCAP and CATIE, the validation and transfer stage emerged as a compromise solution. The project agreement was then amended to accommodate it.
There seem to be several problems with the stage. First, it did not emerge as a freely reasoned step in a research and extension process, but rather as a compromise to satisfy two conflicting positions. CATIE is a research institute and wanted research, ROCAP had broader rural development interests and wanted extension.
Second, what is being validated if CATIE supplies the inputs? The issue here turns on economics. I did not have sufficient time to pursue this matter adequately, but it is my understanding that the economic circumstances of the collaborating households are known to researchers at the time of the validation trials. Further economic data--agrochemical and labor costs, product prices--are collected during the trials, so that calculations can ultimately be made regarding the economic feasibility of the technology. But any statements regarding this feasibility must be couched in terms of what the farmers could do--not because they actually did it, but because it was done with CATIE's help. Do successful validation trials mean that collaborating farmers will adopt the technology? Is the demonstration value of these trials such that farmers will undertake the risks of investing in the necessary agrochemicals? I doubt it in most cases, but the question ultimately remains an empirical one. I have my doubts in part because of the instability of product and input prices and the frequent unreliability of credit institutions. A technology valid one year may not be valid the next. This topic is taken up in the following section.
There may be yet another problem with the validation and transfer stage. The information that I gathered suggests that national research and extension institutions do not have the material and human resources to execute this stage. Nationals often commented on the material inputs supplied by CATIE for these validation trials and further remarked that local farmers would have to furnish them should national research and extension institutions conduct the trials. The validation stage may, therefore, lie beyond the economic reach of most research and extension institutions in the region.
On the Nature of Technological Alternatives
I encountered at least one improved technological
alternative in four of the five countries visited that was problematic because of either uneconomic ratios of input costs to product price, or unfavorable credit terms. And some of those

technologies had been validated. I was struck generally by the extensive use of agrochemicals in the technologies.
The problem here is the vulnerability of technologies that require capital--and credit institutions to provide it. Price swings of both inputs and products are common in the unstable national economies of the region. Credit terms also fluctuate, or may not be favorable to small-farm agriculture at all. And the required agrochemicals are not always available at any price.
There may be a psychology lesson here too. If research and extension are integrated and conducted under a single national leadership, then the research function necessarily shares in the fortunes of technologies--in their extension and their adoption by farmers--that it develops. There would be a strong incentive to provide a viable technology. As a foreign research institution only marginally involved(through the validation and transfer stage, say) in extension, CATIE likely has no-such incentive. Once an alternative has been validated, CATIE's job is done and the alternative is delivered to a national extension service. This matter of incentive may be just one more problem with an international agricultural research center's doing farming systems research in a national setting.
On Site Selection
Any judgement of the results of CATIE's farming systems endeavors must consider the sites where CATIE has worked. They have often been highly marginal agroecologically and socioeconomically. In El Salvador, CATIE has moved across the marginal northern zone, only a step ahead of advancing rebel armies and the mounting violence. In Panama, IDIAP will soon invite CATIE to work with a marginal Amerindian group living in an ecologically impoverished zone. CATIE faces a real challenge in such areas, where progress comes slowly if at all, and where there is always a high risk of failure.
I had little time to pursue the matter of site selection, but it is clear that decisions of where to work have involved host governments, CATIE, and ROCAP in varying degrees. In some cases CATIE was consulted in the selection process, in others it was not. It seems that CATIE has had little control over the selection of its work sites in El Salvador, for example.
There may have been areas where CATIE should not have worked, simply because the problems there were of a kind or magnitude that farming systems research held little prospect for solving. I know that the CATIE staff are concerned that the serious problems they have faced in some of the more difficult areas might not be appreciated. That concern is entirely reasonable.

BID--Banco Interamericano de Desarrollo CATIE--Centro Agron"Mico Tropical de Investigacicrn y Enseffanza CIAT--Centro Internacional de Agricultura Tropical CIMMYT--Centro Internacional de Mejoramiento de Ma'z y Trigo IICA--Instituto Interamericano de Cooperacio'n para la
Agricultura(of the OAS)
IIRI--International Rice Research Institute OAS--Organization of American States ROCAP--Regional Office(of USAID) for Central America and Panama USAID(or AID)--United States Agency for International Development
BANDESA--Banco Nacional de Desarrollo Agricola DIGESA--Direcci(5n General de Servicios Agrfcolas DIGESEPE--Direccion General de Servicios Pecuarios ICTA--Instituto de Ciencia y Tecnolog'a Agr'colas
El Salvador
CENTA--Centro de Tecnologia Agricola INVEXT--Investigaci6n y Extension ISIAP--Instituto Salvadorefio de Investigacion Agri"cola y Pesquera
DGOA--Direccio'n General de Operaciones Agri'colas
PNEA--Programa Nacional de Extension Agropecuaria PNIA--Programa Nacional de Investigacidn Agropecuaria SRN--Secretar'a deRecursos Naturales

IDIAP--Instituto de Investigacio'n Agropecuaria de Panamz MIDA--Ministerio de Desarrollo Agricola SENAGRO--Servicio Nacional de Extension Agropecuaria Costa Rica
DGIA--Direccion General de Investigaciones Agrfcolas MAG--Ministerio de Agricultura y Ganaderia PIPA--Programa de Incremento de la Productividad Agrico a