- */62, 232-
THE ROLE OF ON-FARM RESEARCH IN TECHNOLOGY GENERATION
ROBERT K. WAUGH
INSTITUTE OF FOOD AND AGRICULTURAL SCIENCES
FARMING SYSTEMS RESEARCH AND EXTENSION (FSR/E) PROGRAM
UNIVERSITY OF FLORIDA
Gainesville, Florida 32611
."epu^ "I ^^Q^Zt ti^ ^' R Ky^ ^^-^ o"-- "i-,
14 /t a ,-CA- pv-
4 I4 -I ) _. '
^ i "A"RAA. C4 USZIA
THE ROLE OF ON-FARM RESEARCH IN TECHNOLOGY GENERATION
Robert K. Waugh
Technology for agriculture in Latin America, and in many other
areas of the world, can be considered as in its third phase, and without
much doubt is now entering a fourth.
The first phase was pre-World War II. It consisted principally of
the establishment of some experiment stations and educational and train-
ing programs. The introduction of technology for commercial export
crops by international corporations began in this first phase. I have
called this the introductory phase of technology.
Following World War II, a second phase was implemented based upon
the premise that the success of the United States in applying technology
for increasing food production could be repeated in developing countries
through extension systems which would inform the farmer. I have called
this the transfer phase.
When the transfer strategy did not give the desired (nor expected)
results (Rice, 1974), a third phase was initiated in which applied
research and institution building was emphasized. Applied and adaptive
research was given attention, especially in plant improvement, agronomic
practices and livestock management. Cadres of nationals were trained
and organized into national programs supported through foreign and
national agencies of government. The activities were not only national;
regional and international organizations were developed. This third
phase has had some successes. Total food production in the developing
countries has recently been increasing more rapidly than in the undev-
eloped (Barr, 1981; Waugh, 1980). But food production per capital has
not been rising as rapidly in the developing countries as in the devel-
oped; large segments of the world population are passed by and have not
benefitted very much from technology.
It is evident that research in the developing countries is entering
a fourth phase which I am calling client oriented, because it is char-
acterized by its focus upon a specific clientele, with emphasis upon the
small and limited resource farmer. Much of the research under the new
focus is conducted on-farm; cultural and economic aspects are given
consideration along with the biological in the orientation of research
and the transfer of results. The farmer becomes a participant client.
While it is recognized that agriculture cannot be a solution to all
of the rural problems, both production oriented researchers and develop-
ment agencies have become increasingly interested in the new approach;
they believe that research can make additional contributions to rural
This new focus is frequently referred to as Farming Systems Research
(FSR), especially in the United States. The approach is relatively new;
it is in an evolutionary stage of development; there is still consider-
able divergence of opinion about FSR. But among its proponents there is
consensus that on-farm research is an important part of it. There is
also consensus that transfer, in some manner, must be an integrated part
of the farm focused technological system. There is no doubt that but
the on-farm research has a strong transfer effect, but this effect of
the research has limited coverage if not linked to a mass transfer
system. ICTA, in Guatemala, is one of the examples where an entire
institute has focused upon the farm by direct linkage between commodity
and discipline research and the on-farm research,<, nICTA is now working
to develop linkages with the extension service(Waugh and Villeda, 1981),
largely through joint participation between ICTA and DIGESA in the
supervision of farmer managed tests in the evaluation phase of on-farm
research. Honduras has been working toward a similar relationship
between research and extension (Secretarla de Recursos Naturales,
Honduras, 1981: Waugh, 1981). The University of Florida has initiated
a Farming Systems Research and Extension project in North Florida.
ON-FARM RESEARCH: A CENTRAL POINT IN A TECHNOLOGICAL SYSTEM
Research and extension have both been criticized for having traits
that reduce their effectiveness in meeting development objectives. Most
of us accept that there is some basis of truth in the criticism, es-
pecially for certain conditions such as those of the small, limited
resource farmer. Also most of us will agree that on-farm research when
used 1) to integrate production components into farmers' production
systems, 2) for evaluation of production systems under farming condit-
ions, and 3) for its direct contributions to the generation of tech-
nology, can go a long way in correcting the traits that have been criti-
cized. Furthermore we agree that on-farm research, conducted in a
manner that not only studies the biological systems used in farming, but
also farming as an agro-socioeconomic system can further improve the
effectiveness of research.
But on-farm research can do even more for the improvement of agri-
cultural research because it allows us to establish a system with a
continuum of sequential. activities from the generation of technology to
its use by farmers in their producing systems.
On-farm research can become a central point for developing a tech-
nological system to serve farmers which can open up new possibilities for
improving the effectiveness of research from several points of view:
1. On-farm research can serve as a focal activity for on-going
research and extension and the improvement of both.
Lack of coordination and collaboration between research and exten-
sion has long been pointed out as a weak link in connecting technology
generation with transfer and application. The report of Rice (1974)
strongly indicates that a transfer system independent from research may
have very little impact. Evenson (1978) points out that technology
search isolated from organized research is subject to exhaustion.
2. On-farm research makes reductive research more purposeful and
also serves as a basis for evaluating the output from discipline and
Research is criticized for being too reductive. I can see no good
substitute methodology for the reductive nature of some research because
the biological systems which operate in farming are too complex and the
conditions too variable to be understood using holistic methodology.
However on-farm research can function to integrate the results from
commodity and discipline research and evaluate them under a holistic
3. What is learned through on-farm research can serve as a basis
for the orientation of reductive commodity and discipline research and
the selection of priorities.
Research is often accused of being focused upon the lesser im-
portant aspects of farming while more important ones are neglected.
4. On-farm research can make research more plausbile to decision
makers of government.
Research almost never is well understood by decisions makers or the
political forces that occur within government.
5. On-farm research can furnish information, and can introduce
checks and balances (evaluations) that can improve management.
Management of agricultural research is often considered weak and
6. On-farm research can be a hands-on experience to improve the
effectiveness and the image of researchers and extensionists.
Researchers are viewed as inhabitants of ivory towers who do not
understand the reality of farming. Extensionists are frequently accused
of knowing very little about new technologies--and making recipe recom-
7. A new dimension can be added to biological research and make it
Traditional research conducts on-farm (perhaps off-station would be
a better term) research. However on-farm research under a systems
approach takes on a quality different from the on-farm (off-stat'on)
regional yield trials, for example, conducted traditionally. In the
latter trials the biological system under study, such as a new variety,
is exposed to a wider ecological range than is found on the experiment
station, but several of the non-experimental variables such as plant
population, soil nutrients, weeds etc. are usually controlled. These
kinds of trials are important for selecting among advanced breeding
lines and the farm oriented approach must continue to use this kind of
trial. But it goes further by also evaluating the biological system
when the non-experiemental variables, including management, are allowed
to fluctuate within a range that might be expected under the farmers'
conditions of production. The traditional system gives as estimate of
what would happen if farmers were to control variables as the researcher
does. It furnishes an estimate of potential when the experimental
conditions are met by the farmer. On-farm research can give an estimate
of results if farmers were to use the new variety. While both of these
estimates are important to farmers, planners and researchers, the second
is frequently (usually) missing.
These are examples of where on-farm research, conducted under the
systems mode, can contribute to many aspects of overall research activ-
What I conclude is that on-farm research needs to be brought into a
research and extension system--a system that is organized, coordinated
and directed--a system that permits identification of objectives and
goals--a system that identifies progressive steps toward the goals so
that progress can be evaluated--a system that can be managed with the
objective of improving the effectiveness of technology and the ef-
ficienty of its generation and transfer.
With the objective of bringing on-farm research into an organized
and institutionalized system it must be given more order: objectives,
methodologies and design. For commodity and discipline research this
has been accomplished over a period of time. We have a good start with
on-farm research but it still needs further definition.
A CONCEPTUAL MODEL
Figure 1 is a conceptual model of the farming systems approach to
the generation of agricultural technology and its transfer to (adoption
by) producers. No one (single) model can fit all situations, and almost
certainly will not satisfy everyone, especially since this approach is
new, and in a dynamic stage of development. Also there is always the
problem of the application of the model; in the present occasion we will
not all visualize its function, in exactly the same way. The important
thing at this time is to come as close together as possible in our
thinking about how such a system would operate, in order to focus upon
the job at hand of outlining some applications of design and analysis to
on-farm experimentation that would be carried out with a system of farm
The system: Who
The system visualizes that the principal groups that would be
active in a farm focused research and extension system would be:
1. Commodity and discipline researchers
2. On-farm area research and extension teams
3. Extension agents
FARMING SYSTEMS APPROACH TO
RESEARCH AND EXTENSION
Farming area activities (farms and farmers)
Y AND I PESN
CH REGIONAL EXTENSION PERSONNEL
INTEGRATION OF COMPONENTS:
EVALUATION AND INITIAL TRANSFER
MASS TRANSFER AND PRODUCTION
I COMMODITY AND
AND ANALYSIS OF FARMING AREAS; AGRO-SOCIOECONOMIC INFORMATION GATHERING
COMPONENTS: RESEARCHER FARMER MANAGED
INTEGRATION OF MANAGED TESTS AND
-- COMPONENTS-; -- AGRONOMIC AND -RESEARCHER
ALTERNATIVES ; ECONOMIC EVALUATION OF
EXPLORATORY EVALUATION ACCEPTABILITY
REGIONAL ON-FARM RESEARCH/EXTENSION TEAMS
The system also visualizes some overlap in the activities of these
The system: What and How
The phases of work to be carried out within the system can be
conceptualized as follows (what);
1. Management and planning
2. Characterization and analysis of farming areas agro-
socioeconomic information gathering
3. Component and discipline research
4. Integration of components into farming systems and study of
5. Evaluation of technology and initial transfer
6. Mass transfer and production
These concepts can be built into a sequential, cyclical, iterative
and overlapping system of work (how).
Management and planning. Management and planning are usually
recognized as necessary for the operation of institutions but not con-
sidered directly as part of the technological system itself They
should be an integral part of farm focused research and extension and
not simply the means through which research and extension are institu-
The essential quality of management is responsibility,,, This re-
sponsibility can be considered under three areas: 1) policy, 2)
operations (how the organization functions) and 3) technical direction !("dc i o / "
Management is elusive; it is both the persons) doing the supervising
and the act of supervision. There is top management, but responsib-
ilities are delegated and these responsilibilities must be felt at all
levels because top management alone cannot implement a farm oriented
technological system. This must be done through heads of units or
groups because many of the attributes of the system can only be incor-
porated into functional programs and managed through individuals, each
of whom is responsible for a segment of the overall activities, and is
one of the major reasons why there must be an organized system.
Planning is closely related to technical guidance and direction but
is not limited to strictly technical matters. It also pertains to re-
sources and their distribution. Planning then, is projection of the
work plans and estimations of the resources necessary to carry out the
work plan. It must be done before the work is carried out. It is the
future tense of the who, what, how and where.
On-farm research is dynamic. It is sequential, cyclical, iterative
and overlapping. The work must be planned ahead of time, and the needs
anticipated, in order to bring together the group inputs and institu-
tional resources into coordinated action.
Characterization and analysis of farming areas; agro-socioeconomic
information gathering. One of the major needs of on-farm research, as
an integrated part of a farm focused system is the comprehension of the
biological, cultural and economic aspects of farms and farming within
the target area. The need is not limited to an overall picture of the
area at one instant of time to serve as an initial orientation for the
research, but also for continual updating so that comprehension will
continue to improve. This way the best possible information is used in
developing the successive steps of the research and extension process.
The characterization and analysis of farming areas and the gathering of
agro-socioeconomic information is a two phase process: 1) the initial
phase in which existing documented information, results from the rapid
reconnaissance (sondeo); survey and any previous research results are
used to plan the initial thrust, and 2) a continuing phase in which
information is updated and made more complete. This new information can
come from the experience of the on-farm team members in the area,
special studies and surveys, farm cropping records, and current re-
search. A frequent misconception of farming systems research is that it
consists of the initial characterization. What is important, of course,
is the subsequent generation of relevant technology. Characterization
of selected target areas as an action independent of the system will
contribute very little.
Component research. Component research contributes to the in-
dividual production factors such as variety, weed control, pest control,
plant nutrients etc. Commodity and discipline research are therefore
largely component research. Some component research is conducted on-
farm. In Figure 1 this is indicated by an extension of the component
research frame into the farming area.
Integration of components into farming systems and study of alter-
natives. Some integration of components into farming systems may be
done off-farm (on-station) but most will be conducted by the on-farm
research team. The integration of components may result in the
synthesis of new systems of production; some may be simple modification
of what is already commonly used. Thus alternatives may be presented to
farmers through the study of integration of components, but also alter-
natives initially may be introduced through "exploratory" trials. Most
of the experiments in this phase would be classified as "exploratory"
and "site specific agro-technical trials".
Evaluation of the technology and initial transfer. These trials
are of two principal kinds: (a) researcher or extensionist managed and
(b) farmer managed with researcher or extensionist supervision. In the
first case the trials would be "regional agro-technical" or "regional
In the components integration phase, trials are designed mainly for
evaluation of biological materials or biological systems. Several non-
experimental variables are controlled. In the researcher evaluation of
components within the farmer's production system non-experimental
variables are allowed to fluctuate within a range that might be normally
expected when used by farmers.
In "farmer managed tests" new technology is subjected to more
variation including the variations from the farmers' management.
Both of these types of trials have an initial transfer effect.
Farmers collabroate in conducting researcher managed trials and are
introduced to the technology. The farmer managed trial is a self demon-
stration and is further exposure of farmers to the technology. It is
used also as a basis of evaluation by the farmer.
Mass transfer and production. The previous steps are not designed
for effecting mass transfer of technology; there are indications that
the evaluation phase is highly effective as a transfer mechanism but
only reaches a relatively small number of farmers. However, since the
farmer's test does show promise for transfer as well as for evaluation,
there is some basis to believe that the use of the self-demonstration
might be extended through groups and other mass transfer methodologies.
The System: Where. This system is designed to function on both
the experiment station and on private farms but the on-farm part is
given considerable emphasis and becomes a key part of the system.
The on-farm part is designed to operate in specific and defined
regions. Thus it is assumed that there would need be only one each of
the commodity and discipline programs for a large area such as for the
entire country, and these programs would work with several defined areas
Agricultural technology in the developing countries is probably now
entering a new, client oriented phase. Some experience has been gained
with this new focus which promises to be more effective than the trad-
itional approach, especially in support of small and limited resource
farmers. There is considerable interest on the part of production
oriented researchers and development agencies in this new approach.
On-farm research is now recognized as an important part of this new
focus. Here it is suggested that on-farm research can be effective not
only in integrating results from commodity and discipline research in
generating new technologies, and in the evaluation of new technologies
under realistic farming conditions, but can also make contributions to
the improvement of research as a figurative bridge between traditional
research and extension through the development of a more effective
technological system for the support of rural development.
In order that on-farm research, in this new context, can further
improve the technological system of research and extension it needs to
be understood, and methodologies developed. A conceptualized system has
been presented for the purpose of developing further insight into the
role that it might play. It is understood that no one single model will
serve under all situations, but it hoped that the one presented will
serve to help us advance our understanding of the methodologies and the
experimental designs that might be employed.
1. Rice, E. B., 1974. Extension in the Andes. Cambridge: MIT Press.
2. Barr, T. N., 1981. The World Food Situation and Global Grain Pro-
spects. Science, Vol. 214, 1087-1095.
3. Waugh, R. K., 1980. La Investigacion Agricola en elArea de In-
fluencia del PCCMA y Su Proyeccion Hacia el Futuro. Invitational
paper, Reunion Annual del PCCMCA, Guatemala.
r y -A- 1 t1- 'y 4
4: Waugh, R. K., and B. Vflleda, 1981. Ocho Anos del ICTA. Man-
uscript in Spanish, about the Instituto de Ciencia y Tecnologia
5. Secretaria de Recursos Naturales. 1981. Fucionamiento del Pro-
grama Nacional de Investigacion Agropecuaria y Su Integracion en un
Sistema Tecnologico. Program Nacional de Investigacion Agro-
pecuaria, Secretaria de Recursos Naturales, Tegucigalpa, Honduras.
"6. Waugh, R. K. 1981. Structuring a Technological Linkage Between
Agricultural Research and Extension. Manuscript.
7. Memorandum K. R Tefertiller to C 0 Andrew, J. T. Woeste and F.
A. Wood, December 30, 1980. University of Florida, Gainesville.
Z8. Evenson, R. E 1978. The Organization of Research to Improve Crops
and Animals in Low-Income Countries. In Distortions of Agri-
cultural Incentives. Theodore M Schultz editor. Indiana Univ-
ersity Press, 223-245.
n I -., "
""i \- t "-
f ( S-
;-" -- .-3 .. / y .
-*rM^^ 1 ^ C
/ I [7