SVolume Five, Number One
First Quarter, 1987
W Farming Systems
Support Project Newsletter
Farming Systems Research
197 A n c h i
plnnn an deivr of train
ing. prgrm ofte S
D. L. Gait and S. B. Mathema2
As the practice of farming systems research (FSR)
continues to evolve and mature, farmers' constraints,
problems, decision-making criteria, risk averseness,
and opinions are being taken into consideration to a
greater and greater degree. Part of the growing aware-
ness that farmers should be allied as partners with both
extensionists and researchers has been due to a natural
realization that technically-proposed solutions to
farmers' existing crop and livestock systems did not
improve the situation of the farm household-especial-
ly the women decision-makers, or were alternatives
which were totally unacceptable to farmers from the
beginning. In addition, other social scientists began to
complement the work begun by agricultural economists.
Especially active have been anthropologists and rural
sociologists in the analysis of household and village-
level economic, social, cultural and religious situations,
norms, taboos, practices and beliefs, and in the con-
sideration of nutritional constraints and traditions.
The addition of these groups of social scientists to
the traditional combination of biological scientists and
agricultural economists is leading to a much more
sophisticated approach to farmers-both male and
SExcerpt from a discussion paper prepared for the 17th meeting of the
Asian Farming Systems Working Group, Oct. 6-11, 1986, held at IRRI,
Philippines.The complete and fully referenced text of this paper, includ-
ing a case study on cropping systems and farming systems research in
Nepal, is available on request from the FSSP.
2The authors are, respectively: Socio-economist, Department of Agri-
culture (DOA)/Agricultural Research and Production Project (ARRP);
and Chief, Socio-Economic Research and Extension Division (SERED),
DOA, Ministry of Agriculture (MOA), Kathmandu, Nepal.
female decision-makers in agricultural systems-as
partners in the processes of agricultural research and
extension. Various approaches to include farmers as
partners in research and extension are evolving world-
wide. Some of them include (1) the "farmer-back-to-
farmer" approach, (2) the "farmer-first-and-last"
approach, and (3) the indigenous (i.e., grass roots)
agricultural revolution approach.
All these evolutionary approaches have in common
the ability to elevate the actual farmer crop and/or
livestock decision-makers to full partners in the Farm-
ing Systems Research and Extension (FSR&E) process.
Where women constitute the major decision-makers
in the component of the farm system in which inter-
ventions are proposed, their opinions and suggestions
are actively solicited. Where village-level decision-
making is the norm or cannot be avoided, village
leaders and other key informants must be brought into
the system intervention process.
Following is a series of incomplete, but frequently-
asked, questions concerning the trend of including
farmers as partners in the research-extension process.
While some attempt is made to answer these questions,
the answers are very incomplete and are meant to
stimulate much more thought and discussion. Given
the varied conditions throughout Asia and between
Asia and the rest of the world, consensus is neither
sought nor possible. However, what FSR approaches
need, in general, is shorter and more cost-effective
methods for including farmers as co-participants in
the research and extension processes.
QUESTIONS ABOUT GREATER FARMER
PARTICIPATION IN FSR:
The following are just a few of the questions that
researchers and extension workers ask about more
intense relationships with farmers. Many others could
and should be asked. Other answers than those included
here are correct, especially given the multitude of
differing agricultural practices, agro-climatic realities,
institutional arrangements, trained man-power avail-
ability for research and extension, and social, economic,
cultural and religious aspects of societies. Some of our
"correct" answers may be absolutely incorrect under
other conditions or settings.
Why the increased interest in involving
farmers more in research decision-making?
The green revolution considered the condi-
tions of many farmer sub-groups, and its
resultant improved technology impacted
positively a significant group of farmers. However, this
impact has been confined almost entirely to a "special"
group of farmers. This special group is characterized
to a very high degree as
a. Being dominated by male decision-makers;
b. Having timely and guaranteed access to those
agricultural inputs necessary to best utiilize the
package of improved technology;
c. Having access to relatively secure sources of
water: those living in large-scale flood irrigation
plains or in close proximity to tubewell projects;
d. Possessing the "better" quality soils;
e. Residing on lands with less, rather than more,
relative slope; and
f. Residing in or near villages or communities
relativley well supplied with different types of
infrastrucutre (i.e., roads, telecommunication).
While there are exceptions to each of these items which
characterize those most affected by the green revolu-
tion, there is no doubt that they do describe the
majority of green revolution beneficiaries.
When people ask why the green revolution has
paused after serving this group of farmers, most of
us reply that the rest of the farming population is
subject to different and adverse conditions. While
most of us agree with such an answer, few of us agree
as to its solution. Most researchers divide into two
groups, dominated by their conceptual frameworks.
The first group contends that research and exten-
sion must proceed in traditional ways, including a high
degree of dependence upon high levels of petrochemical
inputs to complement improved cultivars or livestock
breeds. Meanwhile, this group insists that the institu-
tional setting differences which face these farmers left
behind must be changed by policy-makers at national
levels before any of their research technologies will be
appropriate for these disenfranchised groups of farmers.
The second group believes that an intermediate
(stop-gap) approach is needed to address the immediate
concerns of these farmers. This group led in the
evolution of the FSR approach. It assumes that, in
many (some would say most) situations, macro-policy-
makers cannot (or do not have the interest to) change
those institutional conditions under which these groups
of farmers operate. This group assumes that macro-
political institutional conditions are exogenous in the
short run, so what can be done to improve farmers'
Furthermore, many international agricultural research
institutes (IARCs) and country programs have realized
that no matter how institutional barriers are manipu-
lated, the given elevation, average number of degree-
days, maximum and minimum temperatures, slope,
aspect, average and varied seasonal rainfall patterns,
soil texture and fertility, existing and natural vegeta-
tion, and degree of male-female participation in dif-
ferent crop and livestock components of any given
agricultural system are not about to change quickly.
Since these factors determine, to a very great degree,
the acceptability of any given agricultural technology,
this group feels that research must be conducted
under the same, or at least very similar, conditions.
The key to such research and eventual dissemination
of relevant technologies is the identification of relative-
ly large homogeneous groups of target farmers, where
"homogeneous" may be defined as agro-biotic, agro-
climatic, geo-political, or culturo-ethnic, among others.
2How can researchers, extension workers, and
farmers be equal partners?--farmers are
uneducated! While there are many illiterate
farmers in the world, there are very few
indeed who could be classified as uneducated.
If we narrowly define "education" as formal training,
the answer is, "yes, many farmers may be uneducated".
If, however, we define "education" as "the acquired
and experiential ability to understand a given agro-
climatic situation well enough to survive from year to
year under harsh environmental conditions and declin-
ing soil fertility;feeding a family near, at, or sometimes,
below, the subsistence level year after year; working
eight to fourteen hours a day; and, often, seeking
off-farm employment just to make ends meet", then
most farmers are quite "educated".
Which one of us would voluntarily switch places
with a farmer in the hills of Nepal, who feeds his or
her family of six persons on 0.6 ha. of mixed rainfed
and irrigated land? Or the Honduran farm family of
seven, renting 1.8 ha. of truely miserable soil, located
on a 60% slope with no terraces? Or the Botswanan
farmer, trying to raise a family of five on a rainfed,
single-cropped system of [maize + sorghum + melon],
when rainfall in a "very good" year consists of 400 mm
spread over seven or eight months? If the ability to
survive under conditions typified by these examples
counts for anything, farmers are definitely not unedu-
cated. Hence, researchers, extension workers and
farmers can always learn from one another by inter-
acting and coming to a point where they all agree on
making research worthwhile and relevant to these
types of prevailing conditions.
Do farmers really perform research? Defin-
itely. Before organized research or extension
existed, farmers were performing research
in their fields and in their herds. All early
breeding work, which took place before the
era of Gregor Mendel, was carried out by farmers.
Even by 1974, when Cornell University had been
breeding apple varieties for about 100 years, 94% of
the apples produced by commercial growers in the
state of New York were farmer-developed varieties.
Similarly, in the state of California, most of the com-
mercial clingstone peach varieties used for canning are
In eastern Nepal, one of the most successful im-
proved varieties of wheat, which has been renamed
and multiplied by researchers at the Pakhribas Agricul-
tural Centre (PAC), has its origins in a Nepali farmer's
field. The farmer, in turn, hand carried the variety
into Nepal several years ago from another farmer's
field in northern India, where he had been working,
and had been impressed with its growth and charac-
teristics. In the Tarai belt of Nepal, many farmers
obtain improved seeds from the Indian borders, then
experiment with them in their fields without knowing
anything about the characteristics of the crop. Nepali
agricultural farms and stations lag behind the farmers
in testing and recognizing the same varieties that
farmers are already cultivating.
Farmers do research. Organized research and exten-
sion exist to make these traditional farmer research
processes more efficient, not to usurp all farmer
initiative in these areas.
4What can researchers learn from farmers?
From some of the answers to the previous
question, one thing that researchers can
learn is to access and work with the investi-
gative ("progressive") farmers near their
experiment stations. Researchers can also learn the
following from farmers:
a. The prevailing conditions which exist in farmers'
fields and/or herds, so that they may be used
for screening technology more effectively;
b. Survival techniques adopted over the years by
farm families-techniques normally referred to as
"sub-optimal" by agronomists or animal scientists
-which allow at least continued subsistence by
spreading and reducing overall risk;
c. The predominant problems and constraints as
faced, and defined, by the farm family, with
respect to each crop or livestock component in
the farm household system, so that these prob-
lems can be attacked by researcher and farmer
designed trials in farmers' fields;
d. Farmer reaction to such trials, by finding out what
they think of the technological interventions, as
well as their reasons behind these reactions so
that trials can also be evaluated using farmer
e. The importance of women, the elderly, and chil-
dren, in both various aspects of the agricultural
production cycles and in household agricultural
decision-making processes; and, finally, but just
f. Farmer field innovations, which researchers and
extension workers may be able to multiply
throughout the area (the PAC example from
eastern Nepal), or transfer to another similar agro-
climatic area, with or without minor adaptations,
such as has been done by adding peanut after rice
in Northeast Thailand through Khon Kaen Uni-
versity's Farming Systems Research Project.
5 Can extension do its job efficiently if it takes
the time to learn from farmers? Yes. In fact,
extension should be able to perform more
efficiently if it takes full advantage of farm-
ers, especially groups of farmers, to assist
with the technology multiplication and dissemination
phases. Nothing is more successful than neighbor-to-
neighbor or relative-to-relative transfer of technology.
In fact, this is the key point of farmer-managed trials,
supervised by research and extension. When neighbors
and/or relatives of farmers with trials become con-
vinced that the circumstances facing them are the same
(that the technology has not simply been "given" to
someone; that their major agro-climatic and socio-
economic conditions are approximately the same), and
they see that the benefits to the host farm family
outweigh the costs, adoption of the innovation by
these other farm families is generally quick. One of
the major roles of organized extension is to facilitate
the multiplication of these types of verification,
demonstration, or production block trials in concert
with research and farmers.
6 s it enough to involve farmer participation
in design of trial interventions? No, but it is
certainly better than simply using a farmers'
fields as mini experiment stations and not
consulting them on trials put on their lands.
However, farmers should be involved as full co-partici-
pants in as many steps of the FSR process as possible.
Thus, farmers should be involved in
a. Recurrent diagnosis (or tailored, follbw-up sur-
b. Trial design, refinement and redesign; and
c. Innovative technology multiplication and dissem-
What is needed is
a. A change of attitude on the parts of all agricul-
tural scientists as to the potential value of farmer's
contributions to agricultural science; and
b. Enough time and manpower to carry out the
necessary farmer-researcher-extension contact.
The first can usually be accomplished by facilitating
visits to farms, and interactions with farmers, by
researchers and agricultural policy-makers in their own
country, region or even village. Furthermore, extension
workers can be taught to view farmers as partners in
the agricultural development process, not only as clients
to fill contact quotas, or recalcitrant students needing
to be lectured to or taught lessons.
The second need may be addressed in a variety of
ways. One of the most effective is by means of directly
including more agricultural economists, anthropologists
and rural sociologists in national agricultural research
and extension programs. While some national institu-
tions may have difficulties with this concept at first,
the dividends that usually arise from having social
scientists supplement, and work jointly with, biologi-
cal scientists in farmers' fields generally override
traditional institutional separation of the social from
the biological scientists.
7What level of farmer participation and
agreement is important? Must it be at the
individual, group, village, or higher political
sub-unit level? The answer to this question,
perhaps more than any other yet posed,
depends on the situation prevalent in a given country
and society. In countries such as the U.S.S.R., the state
does most of the agricultural planning and decision-
making. In others, such as the People's Republic of
China, agreement on agricultural production must be
reached at the block or cell level. At the other extreme,
in some countries, such as the U.S.A., it is rare for
farmers to reach joint agreement on anything. Most
nation's farmers find themselves between these polit-
ical extremes. There are many examples of countries
where group and/or village cooperation and joint
decisions are required.
In Nepal, Village Panchayats (governing bodies)
consist of nine wards, each of which elects a chair-
man and four ward representatives. The Panchayat
itself is lead by the Pradhan Panch, who is assisted
by a vice Pradhan Panch. Theoretically, six village-
level organizations, one of which is a farmers' or-
ganization, also exist. Thus, to implement effective
research in a given target area or Panchayat, these
village-level decision-makers must be notified and
included in all planning processes. These leaders
must also agree upon the objectives of any field
The important point being made here is that a
sufficient number of farmers, facing relatively homo-
geneous conditions, must be in agreement with the
objectives of any research/extension effort, or very
little progress can be made. The number of full
collaborative farm households should never begin
at less than ten per homogeneous target group, and
should grow rapidly to a much larger number as
8Do all levels of household decision-making
need to be monitored? How often should
such monitoring occur? The most detailed
farm-level data is required in cases where
researchers wish to be able to formulate
linear programming models from them. To perform
linear programming at this level, all farm household
transactions should be monitored for at least one year
on a daily basis (daily budgeting). However, for this to
occur successfully, the requirements of data quality,
analytical capacity, and manpower needs are presently
beyond the capabilities of most country's national
programs. While linear programming will be more
common in the future (i.e., IRRI's cropping system
sites in Central Luzon, Philippines), most national
research programs today must make do with less
complete monitoring techniques and less sophisticated
models deriving therefrom.
The objectives of the use of the data determine the
frequency and intensity of monitoring, and/or the
methods used to obtain the farm or village-level data.
Farm labor can be taken as an example. If the research
program wishes only to avoid labor peaks when intro-
ducing their new, innovative technology, a key infor-
mant survey in the target area is sufficient to determine
when these peak labor times occur. However, if the
research program requires more specific information
about labor use (such as the distribution of crop/
livestock labor activities among males, females, the
elderly and children within the household or village),
there is no substitute for either several representative
farmer group meetings or the maintenance of farm
records on a statistically representative sample of
farm households (10-40, depending on the relative
homogeneity of the agricultural system and the agro-
Not all household decisions must be monitored,
especially as an FSR effort first begins in a new target
area. Instead, FSR programs should go slowly and
add socio-economic monitoring at a more detailed
level as the need arises and only when no other alter-
native exists for obtaining the information more
rapidly or more efficiently. Both the collection and
analysis of farm-level socio-economic data are time-
and manpower-consuming activities. To date, several
national approaches to FSR have floundered around
on the over-collection of socio-economic data and
the under-implementation of systematic farm-level
trials. Socio-economic data collection must com-
plement, not lead, FSR&E.
Agricultural research programs must be more and
more innovative in the collection of socio-economic
data. National programs must be more aggressive in
encouraging selected graduate students from their
agricultural universities and colleges to pursue thesis
topics with farm household foci, with the support and
collaboration of the national research scientists.
National graduate students can be an excellent source
of manpower to obtain the detailed data required for
more and more sophisticated understanding of the
socio-culturo-economic conditions facing farm house-
holds. This potential resource currently is vastly
What is the refined role of research in the new
Researchers should be able to look forward
to a more efficient utilization of their new
technological innovations by farmers, if they
add to their techniques of technology evaluation those
important constraints facing major systems compon-
ents of important, homogeneous farmer target groups.
Take plant breeding as an example. Breeders should
encourage earlier farm-level testing of new cultivars in
their breeding pipelines before official varietal release.
To do this efficiently, breeders need practical metho-
dological suggestions to assist them to tailor their
breeding pipeline to the diverse needs of different
groups of farmer clients.
Researchers should also view both extension workers
and farmers as equal partners in the technology devel-
opment and dissemination process. Researchers must
consider how extension workers can multiply their
individual research efforts, making their efforts reach
many more farms on a trial basis. For this to occur
more efficiently, extension workers should be included
more frequently in training courses tailored to answer
such questions as, "why are trials replicated?" "what
are the objectives of these trials?" "why are we measur-
ing plant height and days to maturity?" etc. All too
often, extension workers are simply told to set out a
certain number of trials, to "guard them with their
lives", to take harvest data, and then to send results
back to a research program. What is in this process
for them? Does such an approach make them feel like
co-professionals; Hardly. Finally, researchers must
realize that farmers have been performing research
in their fields, herds and homes long before formal
research programs were funded at the national level.
O What is the refined role of extension in the
new farmer-researcher-extension relationship?
Extension in third-world countries is seldom,
if ever, professionally rewarding. While
researchers in national agricultural programs
rarely have adequate levels of financial resources or
manpower with which to carry out their research pro-
grams most efficiently, extension is usually one more
power of magnitude further removed from an ideal
financial and professional situation. Usually, in com-
parison with research, extension suffers from
a. Lower levels of education;
b. Lower salary levels;
c. Lower and less frequent per diem allowances;
d. Less (or, in some cases, no) mobility;
e. Fewer opportunities for either formal or informal
f. Less potential for professional or civil service
g. Greater levels of frustration; and
h. Lower levels of motivation
in any given national research and extension program.
Seldom are extension workers involved in farm-level
trial design, even if they will be the ones asked to carry
out trial implementation and analysis. More serious
and widespread is the fact that extension workers may
never have received from research any innovative
technology which offered the farmers in his or her
district a real improvement over local practices or
Under conditions such as these, it is largely up to
a. Admit to themselves and to extension that they
have been unable to provide the technological
breakthrough innovations originally envisioned;
b. Offer to include extension representatives in
planning and trial design, not just implementation,
in their extension districts; and
c. Offer to act as partners in the testing and verifi-
cation of new technology with farmers in these
same extension areas.
Everyone stands to gain from such a change in
approach. Researchers give in-field and informal class-
room training to extension workers, who in turn learn
something practical about trial design and analysis.
Extension workers, in turn, provide researchers with
an improved (better trained) manpower resource,
within which farm-level research efforts can be multi-
plied most efficiently. Furthermore, farmers benefit,
as extension workers begin to test (with researchers)
agricultural system innovations which are more rele-
vant to their needs and given situations. This, in turn,
maximizes the likelihood of innovative breakthrough
in any given agricultural system. Such a breakthrough,
in turn, should lead to an improvement in the condi-
tions of the affected rural farming population. This
improvement in conditions, in turn, should enable
the concerned nation to accomplish at least one of the
following: (1) save scarce foreign exchange because
less critical food crop or livestock imports are required;
(2) shift such critical imports to other areas where the
need is greater and which have not yet been affected
by a systematic approach to research and extension;
or (3) add more, or greater quantities of the same,
food commodities and/or livestock products to the
export list. Options (1) and (3) improve the national
balance-of-payments situation, while option (2) cannot
help but improve the internal political situation while
averting starvation. E
Farming Systems Research and Extension:
Achievements and Future'
H. G. Zandstra2
At this, the sixth farming systems research and extension sym-
posium, I am most grateful for the opportunity to return to
Kansas State University's lovely campus. I am especially grateful
for the opportunity to meet with so many colleagues active in
farming systems work.
This symposium and the pre-symposium workshops show the
extent to which FSR/E has captured the hearts and minds of
agricultural scientists. For, finally, here is a clear route to our
ultimate goal: increasing food production and bettering the liv-
ing conditions of low-resource farmers, particularly those in the
Third World. The key to that goal is the installation of research
on the farm, not dramatic demonstration of new technology in
the rarefied atmosphere of the research station.
Successful installation of research requires a different approach,
a community-based approach that is at once ecological-dealing
with problems within their agricultural, social and political
milieu-and also eclectic-drawing from a cross-section of dis-
ciplines. An examination of the origins and achievements of
more than 20 years of progressive research shows the path we
have taken to farming systems research and extension today.
Though it's difficult to trace the genesis of FSR/E with a great
deal of precision, everyone probably agrees that, by the late
1960s, the notion of inefficient resource use causing low prod-
ductivity was put to rest by such economists as Hopper (1957)
and Chennareddy (1967). Their work supported Shultz's con-
tention that poor farmers were actually quite efficient.
It was then, too, that we began to realize the limits of tech-
nology transfer from Western to Third-World systems. In fact,
the industrialized countries' research-extension system itself
produced few measurable results in developing countries (Rice,
But there was good news. Beginning with such pioneers as
Comilla (1959) in what is now Bangladesh, CADU in Ethiopia
(1967), Puebla in Mexico (1967), and Caqueza in Colombia
(1970), researchers sought ways of allowing rural communities
to capture the economic gains implicit in new technologies.
From this "biological soup" of integrated rural development
projects evolved many techniques for on-farm evaluations. For
instance, over a very short time the Puebla project evolved from
a commodity-based maize program to one that recognized the
validity of local farmers' multi-cropping techniques. It increas-
ingly focused on the whole farm family. The Caqueza project,
too, emphasized multi-cropping and included livestock in its
Over time, a greater integration of social and biological
sciences was achieved. A strong emphasis on diagnosis and
continuing evaluation of farmers' reactions to projects led to
such techniques as adoption studies, risk-sharing plans for
introducing high-input technology (Zandstra et al, 1975),
farm monitoring, and rapid appraisal, which recognized land
SKeynote address for the symposium "Farming Systems Research and
Extension: Food and Feed", October 5-8, 1986. Manhattan, Kansas.
2 Director of the Agriculture, Food and Nutrition Sciences Division,
International Development Research Centre, Ottawa, Canada. The
views expressed in this paper are those of the author and not neces-
sarily those of IDRC. The assistance of Chris Mercer of IDRC and
Louise Fresco of the University of Wageningen is greatly acknowledged.
types and production systems (Cobos and Gongora, 1977).
These developments and others spawned ICTA's (the Institute
of Agricultural Science an Technology in Guatemala) agricul-
tural research model. The ICTA model shows the importance
of farmer participation and the use of community organiza-
tions in research, and contributed greatly to the wider use of
rapid-appraisal techniques (Hildebrand, 1981).
Early FSR-like activities ranged from predominantly de-
scriptive to strictly experimental (Gilbert et al, 1980, and
Whyte, 1981). However, there was considerable sharing of
experiences. In the early years, information sharing and training
by the Puebla project's Antonio Turrent, Leonardo Jimenez
and Reggie Laird boosted Latin American on-farm research.
In 1975, the first Asian Cropping Systems Working Group
brought together influences from African farming systems
work at Amadu Bella University, Asian multi-cropping work
from the International Rice Research Institute (IRRI), and
the Latin American Puebla and Caqueza projects.
At the same time, researchers were contributing to classifi-
cation systems (Ruthenberg, 1980), and understanding the
ecological and socio-economic determinants within and be-
tween different farming activities (Norman, 1974). Bradfield
and Harwood's Asian work on multiple cropping and intensi-
fied cropping led to many of the agronomy and crop physiology
concepts now used and confirmed FSR's strong orientation
toward technological change.
Towards a Consensus on Farming Systems Approaches?
By 1980, FSR approaches began to converge. A major contri-
bution towards this was the "Stripe Review" commissioned
by the Technical Advisory Committee of the CGIAR (Con-
sultative Group for International Agricultural Research) (Dillon
et al, 1978). The USAID funded Farming Systems Support
project greatly increased communication among researchers
and encouraged a greater understanding of similarities and
differences. This may also have spawned recent papers on
FSR nomenclature: Fresco (1984), Simmonds (1985), Sands
(1985), and Stoop (1986). These were discussed in further
detail at the TAC sponsored Intercenter Workshop on Farm-
ing Systems Research (ICRISAT, February 1986) where a
somewhat delicate consensus was reached along the following
lines (Arnold, 1986):
FSR, as a generic term, would refer to research with a
"Farming Systems Perspective" or a "Farming Systems
Approach". The latter two terms were preferred. There
was a general, be it not uniform, desire to adopt the follow-
ing three terms for further differentiation of concepts and
work related to them:
"Farming Systems Analyses" (FSA) would be used to
describe the deep analysis of existing farming systems,
including all the socio-economic aspects. It would be
limited to on-farm studies and data analysis.
"Farming Systems Adaptive Research" (FSAR) would
include elements of FSA but would also involve on-farm
and on-station research. Feedback from on-farm research
would contribute to the design of on-station experiments,
thereby developing technology closely adapted to existing
"New Farming Systems Development" (NFSD) would
eventually encompass all aspects of both FSA and FSAR
but would be based initially on on-station experiments aimed
at devising novel production systems, including agroforestry."
I will resist the temptation to comment on the merits of this
nomeclature, except to say that it is the most recent (I think).
Concluding, it is clear that development of FSR approaches
has been gradual and in response to regional, institutional and
disciplinary influences. Perhaps I can help set the stage for this
week's discussions by reviewing some of the issues of incorpora-
ting the FS approach into national research structures, the
experience with livestock, the relation to commodity research,
and others which continue to affect FSR approaches and which
presage the future evolution of farming systems research.
FSAR practitioners have always stressed the importance of
farmers' participation. The introduction of research teams in
villages requires careful community briefing about the objectives
and approaches of FSAR. To do so, meetings are held in each
village and with farmers' groups involved in the diagnosis and
Increasingly, farmers contribute to the design phase, as is the
case in Latin American livestock systems research projects. In
Asia, farmers are confronted with research designs and hypoth-
eses, and are asked whether they would prefer to modify treat-
ment methods and evaluation criteria.
Farmer-managed trials have played an important role in FSAR.
The extent to which farmers apply treatments or participate in
research designs or in the choice of research objectives still varies
greatly. It often depends on the expertise available, the farmers'
level of education, and on the complexity of the procedures. In
some cases, the desire to increase farmer participation has led to
a diminished emphasis on experimental techniques that allow
confident comparisons between alternatives. This, of course,
leads to a compounding of uncertainties, certain to render the
research inconclusive. Careful experimental design and execu-
tion are no less a requirement with farmer participation than
The shift towards farmer participation should continue to
receive emphasis. The increased use of rapid-appraisal tech-
niques, which feed research team interpretations back to farmer
and key informant groups, is encouraging.
Gender and group interests should receive more attention in
the design and ex-ante performance and impact analyses of
alternative technology. This means that appraisal and monitor-
ing must provide gender and community group differentiated
information about resource use and benefits. These methodo-
logical changes are taking place, but there are still few cases
where research objectives and technological or policy choices
have been changed as a result.
As part FSAR's commitment to community-based research,
greater attempts should also be made to employ field assistants
from the community. The inclusion of community extension
workers on FSAR teams can help pave the way for a stronger
community base. The location of the research team within the
farm community is also an effective means to increase inter-
action and acceptance.
Though farmer participation is essential, the farming sys-
tems approach cannot depend solely on what farmers already
know. There must be a combination of the experiential know-
ledge of the farm community (in particular its technological
history and environmental insights) with the biological and
technical knowledge of the researchers. One has only to think
of the frequent success of lateral transfer of farming tech-
nologies-even across oceans-to recognize the importance
of new technological insights. It is often the knowledge of
what is possible and what alternatives are available that is
limited in the farm community.
Developing a Strong Research Capability
The FSAR perspective should be integrated with existing
national research systems, but this is often easier said than
done. Stoop (1986) treats the difficulties associated with the
introduction of a farming systems approach into national
research and extension systems. He concludes that a policy
gradually introducing on-farm research with a systems per-
spective is more likely to succeed than using large, externally
funded farming systems units that are not integrated into the
In this respect, the use of one to three small research teams
composed of BSc-level research and extension staff from the
research site has been effective. These teams could be coordin-
ated at a national level. At that level, contacts with commodity
research groups and support from social scientists and soil and
climatological research groups can be established. These can
be formalized as part-time participants in the national FSAR
team, or as a technical advisory unit to the FSAR project.
Experience in Asia and several African countries shows that
research teams should have limited responsibility in terms of
the target area or agrological production zone. As much as
possible, the FSAR teams must have a reasonable degree of
independence in their day-to-day activities. They should be
responsible for the design of their research (appraisal, monitor-
ing and trials) and for conducting analyses of research results.
The location of initial FSAR teams should allow access by
supporting scientific staff (for example, from a nearby research
While the FSAR approach should be implemented by line
agencies for agricultural research and extension, other organi-
zations can make important contributions. A wider coverage
and greater decentralization can be achieved when national
NGO's and universities participate. These linkages should
extend to contacts with credit and marketing institutions and
groups active in post-production research, so that the oppor-
tunities for removing institutional constraints can be more
A great limitation to widespread adoption of FSAR is the
lack of trained field staff. The demand for training at the
regional level (for coordinating staff and trainers) and at the
national level (for field staff) far exceeds the installed capacity.
International centers should consolidate and strengthen their
training efforts, and more in-country training of the sort offered
by the FSSP in West Africa and by the CIMMYT group in East
and southern Africa should be encouraged.
Where absolutely needed and requested, foreign specialists
can be added to the advisory unit. One or two should be suffi-
cient and they should report to national program leaders. Their
function should be limited to training and advice; institutional
and managerial decisions should be left with the national
Commodity Research and FSAR
The growing interdependence of FSAR and commodity pro-
grams testifies to the importance of maintaining good links
between the two research activities. FSAR depends greatly on
the availability of a wide range of genetic materials. The FSAR
perspective helps commodity researchers to understand the
demands that will be placed on their materials.
It is not coincidental that CIMMYT's maize program devel-
oped a strong FSAR approach. They needed more information
about site-specific and systems-conditioned demands on maize
varieties. The association of IRRI's systems programs with
rainfed lowland and upland rice-growing regions did not arise
because IRRI badly wanted to work with non-rice crops. It
was because these were the environments in which the perform-
ance of rice varieties (and thereby varietal requirements) were
strongly conditioned by land type, growing season, and other
crops in the rotation.
There is no doubt that national agricultural research systems
require an overall land/resource-based FSAR approach (as ex-
plained by Chigaru and Avila, 1986)-not a commodity or
input-constrained approach to making the most of resources.
This approach deals effectively with multi-cropped mixed
farms and is pretty well necessary if research is to consider
policy implications. Much of farming systems research takes
place on marginal land. Those studying new farming systems
for the humid tropics of highly eroded tropical highlands are
well aware of the importance of sustainability of production
systems. Research on ways to regenerate lost production po-
tentials of marginal land is better served by an overall resource-
based approach. It could even be argued that a strong commodity
orientation risks a neglect of environmental concerns.
To make the most of both commodity research and FSAR,
we need a well-developed feedback system. Communicating
FSAR insights from diagnostic work, on-farm testing, and
project monitoring helps commodity researchers to refine
their product specifications. Eventually, such specifications
as breeding objectives will reflect on-farm limitations and
potential more clearly.
In return, FSAR/commodity cooperation will help define
screening conditions (seasons, type of land preparation, input
levels, land type, and so forth) for component technology
A good illustration of the values of such cooperation is in
the workshop proceedings on "Crop improvement" in Eastern
and Southern Africa: Research Objectives and On-farm Testing".
Here, Kirkby (1984) analyzes the relationship between types
of research and deals with the views of commodity program
leaders on the subject.
It is indeed encouraging to note commodity program leaders'
increased interest in participating in the research design of FSAR
programs. Several national research systems now have FSAR
advisory or coordinating teams to which related commodity
program leaders belong.
Crop improvement programs should not necessarily depend
on the FSAR activities of others to provide the feedback they
need. On-farm evaluation with participation of farmers or
farmers' groups, taking into consideration socio-economic
factors, should be part and parcel of commodity improvement
programs. Obviously, there is a lot to be gained if such research
can be conducted at FSAR research sites and coordination with
Livestock in FSAR
Our symposium theme, "Food and Feed" is a good illustra-
tion of the interdependence, not just of commodity research
and FSAR, but also of the different disciplines within FSAR.
Though many FSAR activities concentrate on either livestock or
crop components, some treat both. Examples of livestock-based
programs include the Centre for Research and Training in Trop-
ical Agriculture (CATIE,Turrialba, Costa Rica);the International
Livestock Centre for Africa (ILCA) and a wide range of projects
supported by WIN ROCK International.
Much experience has been accumulated, particularly in Latin
America, where a formal collaborative research network has
grown out of the CATIE and IDRC support to livestock produc-
tion systems research (Li Pun and Ruiz, 1984). The Asian Farm-
ing Systems Research Network has since 1983 included a small
number of crop-livestock research projects. Papers from these
groups will be presented at this symposium.
Because of the predominance of mixed farms in small holder
agriculture, it's tempting to assume that all FSAR projects
should intervene in both crop and livestock components of
mixed farms. It is true that the diagnostic stage and farming
systems analyses must lead to an understanding of the roles
and transfer relationships between the two. However, beyond
this, it is quite feasible to concentrate on the sector in which
your institution has expertise.
Because the major constraint to livestock productivity is the
lack of a reliable year-round supply of quality feed, the search
for better feedstuffs has led to improved use of by-products,
intercropping, hedgerow cropping of fodder crops and trees,
and increasing forage yields and/or quality from food-crop
production. It is therefore important that livestock researchers
exploit opportunities for intervention in the crop sub-system.
Most of these interventions can be dealt with using FSAR
techniques from cropping systems research, although the use
of tree species presents special difficulties (Nitis et al, 1985).
Where interventions involve substantial changes in herd
management or require evaluating several technologies, whole-
farm evaluation appears the only approach. In some projects,
this is achieved by using researcher- or farmer-executed model
farms. However, for interactive and realistic results, it's
necessary to introduce the changes gradually to a number of
farms. These can then be compared to carefully paired check
farms that have been monitored in the same way.
Most projects using whole-farm testing use interdisciplinary
teams in the diagnosis, design of alternatives (generally one or
two only) and ex-ante analysis. Several have used simulation
models for ex-ante analysis and design activities, and all seek
reactions from farmers and extension workers several times
during the design process. The increased emphasis on systems
simulation and ex-ante analysis to compare potential inter-
ventions is an understandable adjustment, given the costs and
complexities of evaluating whole-farm crop-livestock alterna-
tives. It places, however, greater demands on the research
system and can limit the ability of non-specialized staff to
participate in decisions about research directions.
How should livestock production research be included
in national research systems? This naturally varies greatly
from country to country. In general, different research organ-
izations are responsible for crop and livestock research. It may
then be best to encourage independent activities and to seek
collaboration on selected research sites. Where both compon-
ents reside in the same organization, there is merit in combin-
ing the livestock and crop capability at the national level. This
has been achieved effectively in Zimbabwe, where the farming
systems unit consists of two livestock scientists, two agronom-
ists, one agricultural economist, and support staff. The research
is conducted at two research sites, each managed by a small
field team (Chigaru and Avila, 1986).
Agricultural Policy Support
Canadian farmers know how important agricultural policy
support is to their livelihood. Instead of ensuring farmers a
fair return on their investment, agricultural policies have aban-
doned them to an international marketplace marked by low
prices and subsidized competition.
In the Third World, political support for new agricultural
technologies is equally important. However, our methods of
attracting and building this support need refinement. FSAR
usually promotes policy changes through highly visible demon-
strations of potential benefits using pilot production projects.
This then leads to special projects that provide the necessary
additional credit, input and marketing supports. Although many
of these institutional changes were confined to the project area,
and were sometimes of a temporary nature, the use of improved
technology as an instrument for policy change merits greater
Even when projects attempt to foster such support, it is not
always forthcoming. This has led to lost production and income
on many occasions (Zandstra et al, 1975). To get away from the
problem of transient support, more formal links between FSAR
coordinating groups and policy planners are required. This will
allow the presentation of carefully documented findings in
non-threatening ways. The possibility of policy makers and
planners participating in national-level committees or units
should be more aggressively pursued.
The close association with the rural community of those
conducting FSAR places them in an excellent position to
predict or document the impact of technological and policy
changes on different groups. To contribute effectively to the
policy-making process, FSAR teams need, however, to take
store of their usual methods of data collection and presenta-
tion. These should give greater recognition to the institutional
costs of recommended changes, and to area-based estimates
of input requirements, production changes, market interven-
tions, and employment and income effects. For most national
FSAR activities, this will require a considerable strengthening
of analytical and communication capabilities.
Agricultural research has never been more important. Not
only because food availability and incomes of smallholders
in the Third World need to be increased, but more and more
because the world must protect or even improve the future
production capacity of its agricultural land.
The farming systems approach continues to evolve and
become more clearly defined. Future directions that merit
-A greater emphasis on long-term sustainability and
thereby closer links between the development of new
farming systems and their early on-farm evaluation.
-Better participation by farming families and different
beneficiary groups (e.g. gender or tenancy specified) in
the design and execution of research plans.
-Increased consideration of the relationship between
agricultural policy and technological change.
-Broader participation of commodity and disciplinary
(soil, engineering, etc.) research groups in technology
design. This should allow a more complete scan of
possibilities, such as food and fodder, trees, non-food
crops, and land or water-based livestock.
The farming systems approach has become widely accepted
in the last ten years. There are some promising institutional
implementations of FSAR to attest to this. Despite this, the
main constraint to the adoption of improved strategies for sus-
tainable rural resource use continues to be the lack of national
research capacities to exploit the technical and policy oppor-
tunities in specific rural regions. There is therefore, first and
foremost, a need for continued support for training and insti-
tutional implementation of FSAR.
Implementation of FSAR should normally be gradual. The
use of, initially, a few multidisciplinary field teams is advised.
These should be composed of BSc level research and extension
workers from the region and trained field staff from the villages
in the work area. Strong national level support is required from
systems research specialists who can provide methodological
inputs. A national FSR coordinating group or committee has
proven a most useful means for ensuring close collaboration
with policy-makers and leaders of commodity programs. Com-
modity programs should increase on-farm research, where
possible in close collaboration with FSAR teams. After three
or four years' work on two to three locations, national or state
programs are in a better position to estimate the eventual level
of investment in FSAR teams and their coordination that is
required and can be sustained.
Finally, it cannot be overstated that the implementation of
farming systems research is a long-term process. The initiatives
by International Agricultural Research Centers and donor
agencies have been instrumental in bringing about its wider
implementation. The Farming Systems Support Program has
resulted in an important consolidation of research methods
and experiences and strong support for in-country training. I
sincerely hope that this determination to support FSAR will
continue, as it is sorely needed if we are to make available the
benefits of research to resource-poor farmers.
Anderson, J. R. and J. B. Hardaker (1986). Policy Issues of Farming Sys-
tems Research. Presented at: Workshop on Farming Systems Research,
ICRISAT Center, Hyderabad, India in February 1986.
Arnold, M. H. (1986). Preliminary Draft Proceedings of the Workshop on
Farming Systems Research, 17-21 February, 1986. ICRISAT Center,
Chenareddy, V. (1967). Production efficiency in South Indian Agricul-
ture, Journal of Farm Economics (Lexington, KY) 49j: 810-820.
Chigaru, P. R. N. and M. Avila (1986). Farming Systems Research Ex-
perience in Eastern and Southern Africa. Presented at Workshop on
Farming Systems Research, ICRISAT Centre, Hyderabad, India in
Cobos, B. C. and C. S. Gongora (1977). Una Metodologia para identifica-
cion y analysis de sistemas de production agropecuarios en arias de
pequenos productores. Boletin Tecnico No. 61, ICA, Bogota, Colombia.
Dillon, J. L., Plucknett, D. L. and G. J. Vallaeys (1978). Farming Sys-
tems Research at the International Agricultural Research Centers.
Technical Advisory Committee of the CGIAR. Rome, Italy.
Fresco, L. (1984). Comparing Anglophone an Francophone Approaches
to Farming Systems Research and Extension. FSSP Network Paper
No. 1. Gainesville, Florida, U.S.A.
Gilbert, E. H., Norman, D. W. and F. D. Winch (1980). Farming systems
research: a critical appraisal. Rural Development paper No. 6, Michigan
State University, Michigan, U.S.A.
Harwood, R. R. (1974). "Resource Utilization Approach to Cropping
Systems Improvement". International Workshop on Farming Systems
pp. 249-60. ICRISAT, Hyderabad, India.
Hildebrand, P. E. (1981). "Combining disciplines in rapid appraisal: the
Sondeo approach". Agricultural Administration, 8:425-432.
Hopper, W. D. (1957). The economic organization of a village in North
Central India. Unpublished Ph.D. Dissertation, Ithaca, N.Y., Cornell
Kirkby, R. A. (ed), (1984). Crop Improvement in Eastern and Southern
Africa: Research Objectives and On-Farm Testing. A regional workshop
held in Nairobi, Kenya. July 20-22, 1983. IDRC. Ottawa, Canada.
Li Pun, H. and M. Ruiz (1984). "La Red de Proyectos de Investigacion en
Sistemas de Produccion Animal". In: Informe de al V reunion de Tra-
bajo sobre sistemas de production animal. IDRC, Bogota, Colombia.
Nitis, I. M., Lana, K. and M. Suarna (1985). "Three Strata System for
Cattle Feeds and Feeding in Dryland Farming Areas in Bali". Proceed-
ings of the 2nd Monitoring tour of Crop-Livestock Systems Research,
Nepal and Indonesia, August 1985. pp 245-71. Ministry of Agricul-
tural, Government of Nepal, Ministry of Agricultural, Indonesia, and
Norman, D. W. (1974). Rationalizing mixed cropping under indigenous
conditions: the example of Northern Nigeria. Journal of Development
Studies 11: 3-21.
Norman, D. W. (1978). Farming Systems Research to improve the liveli-
hood of small farmers. American Journal of Agricultural Economics
Rice, E. B. (1971). Extension in the Andes: An Evaluation of Official
U.S. Assistance to Agricultural Extension Services in Central and
South America. U.S. Agency for International Development, Wash-
ington, D.C., U.S.A.
Ruthenburg, H. (1960). Farming Systems in the Tropics. Third edition.
Oxford University Press, Oxford, UK.
Sands, D. M. (1985). A Review of Farming Systems Research. Paper pre-
pared for Dr. A. Von Der Osten, Executive Secretary, Technical
Advisory Committee of the CGIAR, Rome, Italy.
Schultz, T. W. (1964). Transforming Traditional Agriculture. Yale Uni-
versity Press, New Haven, Connecticut, U.S.A.
Simmonds, N. W. (1965). Farming Systems Research: A Review. World
Bank Technical Paper No. 43. World Bank, Washington, D.C.
Stoop, W. A. (1986). "Organization and Managerial Implications of a
'Farming Systems Approach' for NARs." Presented at Workshop on
Farming Systems Research, ICRISAT Center, Hyderabad, India, in
Whyte, W. F. (1981). Participatory approaches to agricultural research
and development. A state of the art paper. Rural Development
Committee, Cornell University, Ithaca, N.Y.
Zandstra, H. G., Swanberg, K. and C. A. Zulberti (1976). Removing
Constraints to Small Farm Production: The Caqueza Project. IDRC-
058e. Ottawa, Canada. (modified version of a paper presented at the
annual meeting of the Canadian Agricultural Economics Society,
Brandon, Manitoba, 22-25 June 1975).
Defining Recommendation Domains
For Livestock Research -
An Example From Thailand
2Shiriluk, Namkang, Jiraporn, Sompong, Charlito, Banda, Marcelo and R. J. Petheram3
The aim in Farming Systems Re-
search (FSR) is to develop changes
to existing farming systems, which
will result in benefits to farmers.
The researcher attempts, first, to
define opportunities for, and con-
straints to, improvement of the
'Reprinted with permission from DAP Project
News, No. 1, January 1987. This particular
article was recommended for inclusion in the
FSSP Newsletter by R. J. Petheram, Coordin-
ator, Draught Animal Power Project, Graduate
School of Tropical Veterinary Science, James
Cook University, Townsville 4811, Queensland,
SStaff and MSc students in Agricultural Systems
course, Asian Institute of Technology, Bangkok.
SGraduate School of Tropical Veterinary Science,
James Cook University, Townsville.
system, and then to work with
farmers to achieve the desired
changes. Assuming that a repre-
sentative farming area (e.g. a village)
has been selected as a research site,
an important step in FSR, is the
definition of "recommendation do-
mains" or (target) group of farmers
for whom research and recommenda-
tions are to be designed (Tripp 1986).
To illustrate this point, an example
is taken from the work of a FSR
program in Thailand. Following
initial data collection, the Ban Na
Subdistrict was selected to represent
a particular rainfed farming system,
and a rapid survey conducted to pro-
vide information for the classification
of farmer domains. The broad classi-
fiction of farmers derived from the
survey data, assumes that the FSR
program has a 'livestock focus'. It
is based on the ownership and use of
buffalo, which are the predominant
livestock species reared.
In Figure 1, various groups of
farmers can be identified, with
obviously different farming enter-
prises-and for whom research pri-
orities are likely to be very different.
The first division is based on livestock
ownership (or non-ownership) in
1980, and the second on livestock
ownership in 1986. Amongst the
livestock owners (in 1986) a third
division is made on source of draught
power-animals or tractor. Based on
these three divisions alone, four main
Figure 1 Classification of Farmers based on Buffalo Ownership and Power Source. Ban Na Subdistrict of Nakong Nayok, Thailand.
DOMAIN 1 DOMAIN 2 DOMAIN 3 DOMAIN 4
farmer "domains" are defined in
Each of the four farmer domains
inTable 1 could be further subdivided,
e.g. on main crop produced, land
ownership or size of family. Appro-
priate division, however, will depend
upon the aims of the research pro-
Table 1 Four (Livestock) Recommenda-
tion Domains in Ban Na District
1. Rearers of
2. Rearers of
of buffalo in
-experienced in buffalo
rearing. Interested in
power as well as other
-experienced in buffalo
interested in improv-
ing reproduction, and
in practices to mini-
-experienced in buffalo
rearing. Most who sold
buffalo did so to pur-
chase tractor, but
would like to have
kept buffalo, if they
had more capital and
could reduce labor
(and security prob-
-little experience or
interest in rearing
buffalo. Rice and
gram. In thecase of the FSR Program
in Thailand, livestock scientists were
able to define two main recommenda-
tion domains on which to concentrate
their efforts, i.e., Domains 1 and 2 in
Table 1. A larger sample of farmers
could then be chosen to represent
each of these important domains,
and surveys and monitoring studies
conducted to collect the information
needed to describe their systems and
the opportunities for, and con-
straints to, improvement.
In many tropical areas, social and
management factors can lead to
much higher variation in livestock
subsystems than those depicted in
this example. Rearers are often most
usefully categorized according to
their livestock feeding system, e.g.
grazing or handfeeding. Different
types of ownership or sharing
arrangements for livestock com-
monly have to be considered. For
example, stock owners may be
more interested in increasing growth
rate than in reproductive rates, while
sharers (who obtain a percentage of
offspring as payment for rearing) are
more interested in increasing repro-
Careful definition and selection of
recommendation domains allows the
focussing of research on important
groups of farmers at a site, and
efficient progress of research to the
stage where ideas for improvement
can be developed and tested on farms.
If research is attempted amongst
too mixed a population of farmers,
it can be extremely difficult to de-
cide on research priorities, or to
produce research results that are
applicable to any particular group.
In a multidisciplinary FSR pro-
gram, the usual procedure would
be to select the numerically largest
farmer domains, to study their
farming systems in some detail, and
then to base further research prior-
ities on their needs. However, most
FSR programs have some 'prede-
termined focus' (e.g. livestock or
crops) and this influences the way
that farmers are classified and hence
the definition of domains. Some
programs have, e.g. equity objec-
tives, which may restrict their
activities to working with the poor-
est groups of farmers. As the poor-
est farmers are invariably in the
majority domains in tropical farm-
ing systems, equity objectives are
often consistent with the aim of max-
imizing agricultural improvment.M
Tripp, R. (1986). Some common sense
about recommendation domains. FSSP
of Intensive French
Intermediate and Advanced ses-
sions are scheduled for May 30-
June 27, 1987 and June 28-July 25,
1987. The Institute is sponsored by
International Programs, Institute of
Food and Agricultural Sciences,
3028 McCarty Hall, University of
Florida, Gainesville, Florida 32611,
Telephone: (904) 392-1965.
Primary emphasis will be on French
communication skills; listening and
speaking. There will also be training
in reading for comprehension, so that
participants can expect concrete
results to be reflected in FSI language
proficiency tests. Students will norm-
ally achieve one step in the FSI
scoring ladder for each intensive
month's study: from 0 to 0+; from
0+ to 1; from 1 to 1+; from 1+ to 2;
from 2 to 2+. For a 3 or 4, the pro-
fessional and near-native levels of
competence, students should expect
less dramatic progress, committing
from three to six months of intensive
work for each step of advancement.
Participants will be placed in the
Novice, Intermediate or Advanced
class according to their level of oral
In addition to6 hours formal study
daily, participants will have all meals
and recreation in French. Specialists
will present slide-lectures on African
Studies. Francophone African gradu-
ate students will participate in field
trips to research projects and to
tourist sites in the region. In this way
participants will learn to communi-
cate in French in a broad range of
situations, day and night, 7 days a
week. They will see, hear and speak
no English for 4 weeks.
Per person cost is $1,700 for each
four week session. This includes
$1,000 tuition for 120 hours instruc-
tion and 248 hours of supervised
conversation practice. Room and
board, including three meals per day
for four weeks, is $700.
For application forms and addi-
tional information write:
Dr. Esther Y. Smith, Language Coordinator
Institute of Food and Agricultural Sciences
3028 McCarty Hall
University of Florida
Gainesville, Florida 32611
Selecting Representative Farmers
and Sites for On-Farm Experiments
During the diagnostic survey, researchers define
recommendation domains and select representative
farmers for on-farm trials. A representative farmer is a
member of a group of farmers within a recommenda-
tion domain, having similar circumstances for which
one can make more or less the same recommendations.
A field or a site is representative if its soil, terrain,
and previous cropping pattern are similar to the farms
of the recommendation domain. The representative-
ness of the implements and husbandry practices the
farmer uses are also important (CIMMYT, 1979).
At the Bako Agricultural Research Centre, certain
channels have been developed for selecting representa-
tive farmers and sites. Diagram 1 indicates the alterna-
tive channels which researchers use to reach farmers.
At each step the researchers must explain their research
program and on-farm experiments in particular. With
Ministry of Agriculture (MOA) staff, discussions
SAssistant Research Officer and FSR Agronomist, Bako Agricultural
Research Centre, P.O. Box 3, Bako (Shoa), Ethiopia.
include a review of aims and objectives of research,
past research activities, differences between on-station
and on-farm research, the need for on-farm research
extension and farmer linkages, and the objectives and
types of on-farm trials that are to be conducted in
their district. The MOA staff provide comments which
are used for modifying the trials where necessary.
The alternatives shown in Diagram 1 are the results
of experience and have helped much in time economy.
In the Bako area, on-farm trials are conducted with
two target groups-individual farmers and producer
cooperatives (groups of farmers who farm collectively).
Different channels are followed depending on the
degree of contact between researchers, MOA staff, and
farmers. The longest path is followed when the degree
of contact is weak and when on-farm experiments are
conducted for the first time in a district. In places where
farmers are knowledgeable about research, mostly
through FSR surveys, on-farm experiments and exten-
sion, the shortest route is the most economical one.
Diagram 1: Channels in Selecting Representative Farmers
and Sites at Bako Agr. Research Centre.
(Ind./coop) Ind farmer
Ext. Extension agent
Ind Individual farmers
Coop Producers' Cooperatives (Groups of farmers
who farm collectively).
PA Peasant Association
MOA Ministry of Agriculture
Editor's Note: In submitting this article to the FSSP Newsletter, Gemechu
Gedeno wrote: "I am sending you an experience based article for publi-
cation, hoping that it would be useful for FSSP readers." His article is
based on experiences of a farming systems team in conducting on-farm
experiments at Bako Agricultural Research Centre, Wonega Region,
Ethiopia. Field practitioners are invited to compare Mr. Gedeno's ex-
perience with their own.
PROBLEMS, THEIR CAUSES, AND LIKELY
SOLUTIONS IN SELECTING REPRESENTATIVE
FARMERS AND SITES
The problems faced by FSR programs in selecting
representative farmers and sites are variable. Some of
these problems are target group specific, that is, the
type of problem faced depends on the farming system
in which farmers operate. The following discussion
considers problems, their causes, and likely solutions
in selecting representative farmers and sites. The prob-
lems are grouped into three types: those arising from
a) field characteristics, b) from hazards, and c) from
A. FIELD CHARACTERISTICS
1. Field hetrogeneity: This problem is common to FSR
teams in Ethiopia, although varying in intensity. The
main causes around Bako include crop residue
accumulation at points in the field, termite mounds,
stumps, and uneven manure application. Crop residue
points are caused by piles of maize stover, teff
(Eragrostis teff) straw and noog (Guzotia absynica)
stalks, which are grazed by cattle and unevenly
burnt. The practice of destroying termite mounds
around Bako is uncommon. These and similar prob-
lems can be avoided by thorough field examination
and discussion of the problems with the farmers.
2. Shade trees: This problem is more serious around
Bako than at other sites where FSR teams are
operating in Ethiopia. It has been mentioned at
many forums that this problem is exacerbated by
large plots. Two years back it was impossible to
avoid the problem of shade trees, since plot sizes of
1000 and 2000m2 were used. We now use plots of
about 20 to 200m2 which are more appropriate for
on-farm experiments. However, we still experience
shade tree problems. We minimize the problems
through compromises such as debranching, cutting
down of small trees, and using fewer treatments
and less number of replications per site.
1. Wildlife attack: This problem is serious for both
farmers and on-station researchers at Bako. This has
been recognized by researchers since the establish-
ment of the station, but no solution other than
intensive guarding, has been offered. Although
farmers guard their crops, intensive guarding is very
difficult, given the labor shortage.
The following are found to be helpful in reducing
-Placing the trial in the middle of the field and
away from forest areas.
-Planting at the same time as the farmers.
-Using varieties with similar maturity periods as
those of the farmers.
-Using farmers' planting depth, which is deeper
than the recommended depth.
Unfortunately, placing the trial in the middle of
the field has inconveniences such as walking to the
trials during the wet season. In order to plant at the
same date as the farmer (in order to distribute the
attack) frequent visits to the farmers are required
and the overlapping of appointments is an additional
drawback since most farmers in an area plant at the
Deeper planting also has disadvantages, particularly
delayed emergence and yield reduction. However,
these compromises and inconveniences are generally
acceptable to farmers.
2. Livestock attack: This is experienced at most sites
where the trial is located next to pasture land, roads
or paths. Livestock also tend to attack trials which
are planted earlier than the nearby fields. For ex-
ample, during the 1986 season, goats caused great
damage at one site near Bako which was about one
hundred meters from pasture land and which had
been planted earlier than the field around the trial.
Thus it is important to select sites away from pas-
ture land and to plant at the same time as surround-
3. Accessibility: This is one of the most important
criteria for selecting representative sites and causes
important problems for the Bako FSR team. The
problem is especially serious with regard to one of
the target groups, the producer cooperatives. In
some cases the cooperatives are accessible while
their fields are not. Such cases are taken as acceptable
when the site is in walking distance of 15 minutes
from the main road. The situation is more difficult
when the roads to producers cooperatives are not
In these cases the trials are visited when the road
is not muddy or slippery. In other cases, some pro-
ducer cooperatives cannot be selected for on-farm
trials because they are inaccessible.
C. FARMER MANAGEMENT
1. Fitting the trials into the farmers' panel: This prob-
lem is specific to individual farmers, since their fields
are smaller than those of producer cooperatives.
The fields of individual farmers around Bako are
long narrow strips running down the slope and they
are usually only 60 meters wide. Thus for trials with
blocks longer than 60 meters it is difficult to find
sites that fit the trial. When such a problem is faced,
helpful measures include:
-Reducing the number of treatments.
-Selecting the field before selecting the farmer,
-Reducing the plot size (which should still be
in range of the recommended size).
Selecting the site before selecting the farmer is
sometimes a waste of time, particular when the
farmer is not cooperative, not representative, or
when the field is planned for a crop other than the
trial crop. However, selecting the field before
selecting the farmer has been found beneficial.
2. Farmers plant the selected site for their own pro-
duction before researchers come to plant the trial.
This is especially a problem on producer coopera-
tives where there are many decision makers. Al-
though the chairperson is nominally in charge,
there is sometimes a problem of communications
and reluctancy among the members who actually
implement the field operations. In other countries,
similar problems are observed in families with more
than one decision maker. A related problem is that
sometimes farmers change the selected site for
another site without consulting the researchers.
In the case of producer cooperatives, selecting
the sites in the presence of cooperative officers and
group leaders minimizes the risk that the trial will
be planted before the appropriate time. Identifying
the group leader (the person in charge of field opera-
tions for the field in which trials will be planted) is
helpful in avoiding risks such as the above.
3. Paths through trial fields. This problem is often
unforeseen during site selection since farmers may
make paths through the trial after the trial has
already been planted. This is due to searching for
short cuts to water points or nearby fields.
In case of individual farmers this can be solved
by discussing the problem with the farmer before
the trial is planted. In producer cooperatives it is
a more difficult problem, especially since the paths
may be made by nearby, non-member farmers.
Nonetheless, mentioning the problem when select-
ing the site will help minimize the problem.
4. Unwillingness to participate in on-farm experiments
The problem is common to many farming systems
programs. It is specially a problem when farmers are
not knowledgeable about research, as in the case of
many female family heads in Bako area. In other
cases, farmers claim that experiments require too
much of their time, citing their experiences planting
demonstration plots. They also fear that they will
be punished if any damage occurs to the trials. Some
farmers are unwilling to host on-farm trials if they
had no intention to plant the crop which researchers
want to test. This is particularly a problem if the
crop is a new crop or a minor one in the farming
system. A comprehensive orientation about on-farm
research, what is expected from farmers and the
objectives of the trials, is often sufficient for con-
vincing farmers to participate.
Another group of unwilling farmers are those
whose fields have been used for non-FSR experi-
ments where all the labor inputs were provided
from researchers. These farmers expect the same
from FSR researchers and are not willing when
they are requested to provide labor for maintaining
These non-FSR trials on farmers fields may spoil
not only the relationship with the individual farmer
but with all farmers in the area. Thus there has to be
coordination between commodity researchers and
FSR researchers before going out to conduct on-farm
5. Extension and local leaders' biases in selecting
farmers. Extension agents and local leaders often
select unrepresentative farmers such as their close
friends or the best farmer to host on-farm trials.
This is not a prominent problem in Bako because it
is tackled during discussions and orientation of
extension staff and local leaders. It is important to
make clear what the term "representative farmer"
means and the criteria for selecting them.
Farming systems research is a dynamic learning
process. Its flexibility in methodology allows one
to cope with problems that might not have been
foreseen. A variety of problems might be faced re-
quiring immediate and appropriate solution. If care
is not taken, the above mentioned difficulties result
in experimental biases, which may spoil the experi-
ment. Documentation of the problems faced and
the solutions attempted at each point contributes
to improving FSR.
CIMMYT, 1979: An Occasional Series of Papers and Notes
on Methodologies and Procedures useful in FSR and in
the Economic Interpretation of Agricultural Experiments.
Some Issues in off-station Experimentation. No. 4, Nov.
1979, Nairobi, Kenya.
Tanner, D. G. 1968: Notes on Experimental Design (unpub-
The FSSP newsletter is published quarterly by the Farming Systems Support Project (FSSP), which is funded by AID Contract No. DAN-4099-A-00-
2083-00 and administered by the Institute of Food and Agricultural Sciences (IFAS), University of Florida, Gainesville, Fla. 32611. IFAS Is an Equal
Employment Opportunity Affirmative Action Employer. The FSSP newsletter encourages the contribution of stories, pictures and ideas, which should be
sent to FSSP Editor, 3028 McCarty Hall, University of Florida, Gainesville, FL 32611.
Sd mU EnarvudtI@nl
An Anthroplogists Role in the Tropsoils Project
Carol J. Pierce Colfer2
Six months have elapsed since I left Sitiung, West
Sumatra, a village, where I had lived and worked for
three years, most recently as Team Leader, and pre-
viously as Farming Systems Researcher on the Trop-
soils Project. I am prompted to write about my role
in the project for two reasons: 1) the sense that we
we were able to work unusually effectively as a team,
and 2) the recognition that "farming systems research"
is implemented in very different ways on different
I was originally hired to try to ensure that we made
the match between soil management technology and
people in a generally beneficial manner. Looking back
at my role over the years, I believe I have made five
basic kinds of contributions.3 These are:
I. Responsibility for Making the "Soil-People Match"'
Although there may be general agreement that hu-
man factors are important, when a team is composed
purely of agricultural scientists, there is a tendency
'Tropsoils is another name for the USAID, centrally funded Soil Manage-
ment CRSP(Collaborative Research Support Program), operating in
several countries. The following discussion applies only to the project in
Indonesia, which involves the University of Hawaii, the Centre for Soils
Research (Bogor), and North Carolina State University in Raleigh. Pro-
ject papers mentioned in this paper are available from Tropsoils Project,
Sitiung I, Blok B, PO Box 2, West Sumatra, Indonesia.
2 Dr. Colfer was formerly an Anthropologist on the Tropsoils Project. Her
addresses are: Joint Commission, MUSCAT ID, Department of State,
to neglect this aspect. Scientists, almost by definition,
have their particular bailiwicks about which they are
interested and in which they are trained. They have
plenty of their own work to do: and paying attention
to human factors, though desirable, is time consuming.
They consider such work outside their realm of exper-
tise, and therefore-frequently-just don't do it. The
simple fact that I was part of the team, with the specific
responsibility to attend to such matters, contributed
to various research decisions which were responsive
to human needs.
For instance, our work on home gardens (Agus et al.
1986; Colfer et al. 1985; Dudley and Hidayat 1986) is
unlikely to have been undertaken in my absence.
Despite the different opportunities and constraints
affecting home gardens vis-a-vis upland fields, there is
often a perception within agriculture that the former
are of marginal importance. The economic and nutri-
tional significance of home gardens, as well as the
opportunity to work more effectively there with
women farmers homee we have found to be very
Washington, D.C. 20520 or P.O. Box 467, MUSCAT, Oman.
'The following account is not intended to imply that I was soley respon-
sible for the experiments done on these various topics. I had an influence
in the direction of the research; but the enthusiasm, receptivity and
knowledge of other team members were equally essential. Mike Wade,
Fahmuddin Agus, Carl and Stacy Evensen, Dan Gill, Stephenie Kan,
Djoko Santoso were all extremely active participants; and many others
contributed meaningfully as well.
involved in agriculture), came to our attention because
of our surveys and work with farmers.
Similarly the important differences in indigenous
farming systems vis-a-vis transmigrant4 farming sys-
tems would likely have escaped notice (as they have
in every other agricultural project with which I am
familiar in Indonesia). There is a strong government
interest in transmigrants and considerable pressure to
attend to their agricultural needs. However, from an
equity point of view, as well as a potential source of
knowledge and experience about the local environment,
research on matters of concern to the indigenous
population is warranted.
The observation that different management practices
characterized the different ethnic groups contributed
to our interest in different tillage practices (Field
Research Briefs No. 3, 4, 7, 14, 18, and 24; Wade et al.
1985; Colfer et al. 1984a), and a strong recommenda-
tion for the initiation of research on tree crops (espec-
ially with a marketable product) in this environment
(Naim and Agus 1985; Colfer and Gill 1986).
Other team members, following up on our observa-
tion of farmer practices, noted differing patterns of
residue management among different farmers, and
designed experiments to look at this (e.g., Gill et al.
1986). Similarly initial researcher interest in green
manures was reinforced by observation of farmer
practices and interest (Field Research Briefs No. 2,
8, 13, 16, 19, 20, and 29). This gave added impetus
to the work on pasture grasses and leguminous trees
as sources of organic matter for use as a soil amend-
ment (Field Research Briefs No.19, 20).
II. Experience Dealing with Communities
The reticence of agricultural scientists to work
with people is only partly due to overwork. Another
important factor is their lack of experience and train-
ing in how to do it (reasonably enough!). One of my
contributions has been to provide guidance in ways to
deal with farmers in a constructive manner.
One of our most valuable activities, for making the
"match" between technology and people, has been our
collaborative work with farmers. In this work, we
developed tentative research designs, submitted them
to the farmers, and revised them so they were consis-
tent with the goals of both farmers and researchers. We
then monitored farmers implementation of our joint
plans (Colfer et al 1984a; Wade et al 1985; Agus et al.
Agricultural scientists are frequently trained to look
at their activities as quite uni-directional. That is, they
4"Transmigration" is a long standing Indonesian program to resettle
people from densely populated Java and Bali to the sparsely populated
"Outer Islands" (mainly Kalimantan, Sumatra, Sulawesi, and Irian Jaya).
The 100,000 ha area known as Sitiung, includes over 10,000 transmigrant
families as well as a comparable number of indigenous people (Minangka-
bau), living in scattered villages of 100-300 families.
are familiar with the usual extension systems where the
researcher does science, and then passes the product on
to the extension agent, who in turn passes it on to the
We were trying to operate under a different model:
one where both the farmers and the researchers brought
knowledge and experience of value to the joint en-
deavor. We were trying to collaborate, joining their
experience with our science. However, understanding
unfamiliar systems is not always easy. Farmers may
fear researchers; or they may be unduly respectufl of
educated people, and thus unwilling to share their views;
or they may simply delight in misleading outsiders.
Anthropologists are trained to understand alien sys-
tems. We know simple techniques for learning about
people's views and behavior. The most important
techniques are extremely simple, but they do have to
be known. For example, I briefed my coworkers, on
our first sortee into the communities of Sitiung:
-to listen and look, to record their observations,
-to adopt a nonjudgmental attitude, to refrain from
correcting or arguing with misinformation; rather
to note it down for later reference.
-to treat farmers as equal, recognizing and respect-
ing their different kinds of knowledge/experience.
-to notice as much as possible about their way of life
(e.g., division of labor, decision-making, health
status, food consumption patterns, family compos-
ition, anything that might bear on agricultural
And as we worked with the farmers, I sometimes
gave guidance, as needed or requested. For instance,
in meetings with farmers of three ethnic groups, I
noted that one Indonesian coworker tended to address
his remarks only to farmers of his own ethnic group;
whereas we actually wanted input from all. The "slight"
was unintentional, and he was happy to make eye con-
tact all around. I sometimes reminded my coworkers
of the necessity to check farmers' statements, by ask-
ing others or by observing what they actually did.
Despite women's observable involvement in agriculture,
coworkers sometimes forgot-prompting a comment
My ability to make this kind of contribution effec-
tively was greatly enhanced by the kind of working
relationship my coworkers and I developed; they
appeared genuinely to want such advice. I also think
the willingness of farmers to express their points of
view (which we very much needed to know) was
greatly enhanced by the sympathetic atmosphere we
were able to create by use of such simple techniques
as listening attentively, behaving respectfully, and
trying to comply with local custom to some degree.
II1. Taking a Holistic View.
Although agriculture is important to farmers,
they (like us) are enmeshed in a cultural system which
includes such diverse components as kinship, religion,
politics, education, health, and so on. Even within
agriculture, there are a variety of components of
Agricultural scientists are trained in experimental
research designs. They think in terms of plots, fertili-
zer rates, ECEC, and ANOVAS. That is their work. By
looking at the Sitiung context in a holistic manner, I
was able to balance this (necessary) preoccupation with
the specifics of soil science, with a a grounding of sorts
in the complexities of the real world.
Unlike some farming systems projects, our goal was
not to study or work with the whole farming system.
Rather it was to keep in mind the whole farming sys-
tem, while determining our soil management priorities.
We used soil management practices as the focus of
Tropsoils activity, and investigated other human spheres
insofar as they appeared to have a link to, or effect on,
soil management5 (cf. Vayda, Colfer and Brotokusumo
1980, for a parallel approach).
Examples of the ways in which I made this kind of
-The team is trying to decide whether it makes
sense to do an experiment on pasture grasses and
legumes. I point out the number of cattle and
goats that are stall fed in Sitiung I, and the amount
of time people spend finding fodder for their ani-
mals, in support of such an experiment.
-During a team discussion of fertilizer rates on a
new experiment, someone suggests including a
treatment with 6 tons of line/ha. I remind them
that farmers cannot afford that (though we may
still decide it's worthwhile for other reasons).
-Due to team labor shortage, we are considering
abandoning our "tree systems" trial. I remind them
of the reduced labor requirement for farmers of
tree crops, as well as the higher cash incomes,
erosion control, source of organic matter, and
lower risk, as compared to food crops.
-A team member suggests a new trial using cowpeas
as a crop because they are so aluminum tolerant.
I remind them that people don't really eat many
cowpeas, and can't get a very good price for them.
-One of our surveys has shown that nutritional
status in the area is marginal. Stacy Evensen (our
nutritionist) and I suggest we initiate some experi-
ments using vegetable crops, to enhance nutrition
while doing soil science. The suggestion is taken,
and a farmer-managed experiment comparing
barnyard manure, composting, and inorganic
fertilizers is initiated (Fahmuddin et al. 1986).
5 Building on this holistic approach to data collection (systemic, yet tied
to soil management I have done some work on integrating our informa-
tion on farmers into an "expert system". My interest in this possibility
derives from the apparent ability of expert systems to utilize very com-
plex kinds of information to provide non-"experts" with help in prob-
lem-solving. It may be a helpful tool for policymakers and scientists,
facilitating more informed decisionmaking.
IV. Provision of Specific Information.
Anthropologists are also trained in various research
methods for getting at specific kinds of information
about people. A number of "special studies" were
designed on site in response to information needs
perceived by team members.
-The agricultural scientists jumped right into experi-
ments on fertilizer and lime use, but they soon
realized they wanted to know how much cash
farmers had available for the purchase of such
inputs. We were also concerned that our project
not have deleterious effects on community nutri-
tion. But to determine that, we needed a measure
of nutritional status in Sitiung. We interviewed 80
families for two days and recorded their food
consumption during that time as well as income
information (Chapman 1984; Colfer et al. 1985).
-We began to suspect that the local farming system
(Minang) might have some important pointers for
us if only we understood it better. We arranged for
a four month study by an Andalas University stu-
dent (Herman Agus) in a nearby Minang village,
focusing on its tree farming practices (Naim and
Agus 1985). These results were of such interest
that I began working in Pulai, another Minang
village, doing participant observation, and a variety
of other studies to expand our understanding of
this very different farming system.
-Team members began to wonder what farmers saw
as the primary constraints to their production. So
again we planned and implemented a small, indepth
survey to find out (Evensen et al. 1985).
V. Research of Anthropological Interest.
Although I have seen my role in Tropsoils as sup-
portive, by and large, I also felt some commitment to
do some research of a more scientific nature-research
on global concerns in social science. One of the first
studies I initiated on coming to Sitiung, besides the
collaborative research with farmers, was a time alloca-
tion study, I knew we'd want to know what people
were doing, who was doing what, and how much time
was spent on which activities. The time allocation
study mentioned above has done double duty in this
regard. Since I have used the same method in four
communities and among four ethnic groups, useful
comparisons and generalizations are possible (Colfer
1983; Colfer et al. 1984b).
A Galileo study, which provided use with "cognitive
maps" of people's perceptions of and values relating to
soil, was not dictated by specific team needs so much
as by the scarcity of information on people's views of
soil. Such a study seemed a legitimate contribution to
soil management in and of itself, since people's behavior
toward the soil is influenced by their perceptions of it.
It also has significant possibilities for use in extension
(Colfer et al. 1986).
The related "Indigenous Knowledge Study" aims to
compare the "soil science" of the Minang with that of
the soil scientists (Colfer and Gill 1986). Do the Minang,
thoroughly familiar with this environment, recognize
important differentiations that have escaped our
notice? Do their methods of choosing good land mesh
with our own? Do they alter their soil management
practices by land type, and if so, how?
Let me close with a brief comment on the nature
of intra-team collaboration on the Tropsoils project,
prompted by the recent article by Rhoades, Horton
and Booth (1986) on the same topic. After indepen-
dently written sections on their respective roles in
interdisciplinary research at CIP, they emphasize the
constructive role of conflict in interdisciplinary re-
search, seeing the alternatives as "drifting into dis-
I have worked on teams which functioned within
the "constructive (and sometimes not so constructive)
conflict" mode, but I far prefer working in the more
cooperative kind of atmosphere we were fortunate
enough to develop in Sitiung. Although, like Rhoades
et al., we were not completely satisfied with the
intensity of our collaboration; we did a comparatively
good job, overall, in that sphere (going regularly to
farmers' fields and research plots, planning, solving
crises, and writing papers together).
Several factors may have contributed to the partic-
ular operating style we developed. First, we had a
"team building" period in Honolulu before we went
to the field which emphasized the existence of differ-
ences in approach between disciplines and the impor-
tance of our coming to understand each others' research
styles. The notion that each team member brings
strengths and weaknesses to the team effort was
emphasized (we all took the Meyers-Briggs, which
prompted discussion of our respective roles). I think
this experience was extremely helpful in creating a
receptive atmosphere for anthropological input. I
only had to defend my presence to outsiders. My
substantive input seemed to be reviewed by colleagues
with the same combination of interest and critique as
they reviewed each other's suggestions.
Second, the norms of social interaction in Indonesia,
unlike South America, do not permit open conflict.
Any attempt to engage in the kind of "constructive
conflict" so common in the eastern US would have
precluded meaningful Indonesian input.
Lastly (and this is the interpretation of Mike Wade,
my most consistent collaborator), we could have just
had a lucky mix of personalities.E
Agus, Fahmuddin, M. K. Wade, and J. Prawirasumantri. "Effects
of post clearing Methods on Soil Properties and Crop Produc-
tion. Field Research Brief No. 19, Tropsoils, Sitiung, West
Chapman, B. Memo No. 1, No. 2, and No. 3 (on nutrition and
income surveys), reports submitted to Tropsoils, June and
Colfer, C. J. P., "Women, Men and Time in the Forests of Kali-
mantan. Borneo Research Bulletin 13(2):75-85, 1981 (re-
printed by East-West Environment and Policy Institute, No.
Colfer, C. J. P., C. Evensen, S. Evensen, Fahmuddin Agus, D.
Gill, and A. Wade. "Transmigrants Gardens: A Neglected
Research Opportunity". In Proceedings Centre for Soils Re-
search Annual Technical Meetings, Bogor, February 1985.
Colfer, C. J. P., and D. Gill. "Indigenous Agricultural Models:
A Source of Scientific Insight". In draft form, 1986.
Colfer, C. J. P., B. Newton, and Herman Agus. "On People's
Perceptions of Soil, Sitiung, West Sumatra". In Proceedings
Centre for Soil Research Annual Technical Meetings, Bogor,
Colfer, C. J. P., M. K. Wade, Atin Kurdiana and Suwandi.
"Incorporating Farmer Conditions and Preferences into
Farming Systems Research". Paper presented at The World
Bank, Washington, D.C., October 1984a.
Colfer, C. J. P., R. Yost, Veronica Kasmini, and Atin Kurdiana.
"Time Allocation Studies: a Methodology for Food Produc-
tion Systems". Keynote Address at International Food Day
Workshop on Food Production Systems, Cornell University,
Dudley, Richard G. and Endang Hidayat. "A Survey of Small
Fishponds in Sitiung, West Sumatara, Indonesia". Tropsoils
Report, Sitiung, West Sumatra, April 1986.
Evensen, C., C. J. P. Colfer, M. K. Wade, and Herman Agus.
"Constraints to Crop Production in Sitiung, West Sumatra: A
Comparison of Researcher and Farmer Perceptions". Paper
presented at the International Farming Systems Workshop,
Sukarami, West Sumatra (Balitan), December 1985.
Fahmuddin Agus, S. Evensen, C. J. P. Colfer and Jamalinal.
"Farmers' and Crops' Responses to Different Sources of
Fertilizers: A Farmer-Managed Study on Home Gardens".
In draft form, 1986.
Gill, D. W., A. Kasno, and J. S. Sri Adininsgsih. "Response of
Rice and Soybeans to Potassium Fertilization, Residue
Management and Green Manuring in Sitiung". In Proceed-
ings Centre for Soil Research Annual Technical Meetings,
Bogor, March 1986.
Naim, Mochtar and Herman Agus. "Tree Crop Farming Sys-
tem Pattern in Koto Padang, Sawah Lunto/Sijunjung Dis-
trict, West Sumatra". Report submitted to Tropsoils, Jan-
Rhoades, Robert E., Douglas E. Horton, and Robert H. Booth.
"Anthropologist, Biological Scientist and Economist: The
Three Musketeers or Three Stooges of Farming Systems
Research?" In Social Sciences and Farming Systems Re-
search: Methodological Perspectives on Agricultural Devel-
opment (ed., Jones and Wallace), Boulder, Colorado, West-
view Press, 1986.
Smith, Kirk, and C. J. P. Colfer. "Cooks on the World Stage:
The Forgotton Actors". East-West Resources Systems
Institute Working Paper WF-83-5, 1983.
Vayda, A. P., C. J. P. Colfer and M. Brotokusumo. "Interac-
tions Between People and Forests in East Kalimantan". In
Impact of Science and Society, Vol. 20, No. 3:179-190, 1980.
Wade, M. K., Fahmuddin Agus, and C. J. P. Colfer. "The Con-
tribution of Farmer-managed Research in Technology
Development". Paper presented at the International Farm-
ing Systems Workshop, Sukarami, December 1985.
The Journal of Production Agri-
culture is published quarterly by the
American Society of Agronomy,
Crop Science Society of America,
and Soil Science Society of Ameri-
ca. The American Agricultural Eco-
nomics Association, American Forage
and Grassland Council, American
Society of Animal Science, and the
Weed Science Society of America
are cooperators in the publication.
Contributions to this journal trans-
fer production-oriented information
to a wide range of professional
agriculturists. Emphasis is placed on
research related to agronomy in
concert with other disciplines such
as animal science, weed science,
agricultural economics, entomology,
plant pathology, horticulture, and
Articles submitted by members
and nonmembers of ASA-CSSA-
SSSA will be considered for publi-
cation by the editorial board. Man-
uscripts may be volunteered or
invited; all manuscripts will be
subject to review. Page charges
will be assessed for all volunteered
Authors should submit four legible
copies of their manuscripts, typed
double spaced on line-numbereed
paper, to the editor: R. G. Hoeft,
Dept. of Agronomy, Univ. of Illinois,
1102 S. Goodwin Ave., Urbana, IL
The usual format of an article is
title, author's name and affiliation,
abstract, additional index words,
introduction, body, and references.
Articles must contain an abstract
and should contain an interpretive
Include the Latin binomial or tri-
nomial and authority for all plants,
insects, and pathogens at first men-
tion in the abstract and manuscript.
Use the English (inch-pound) system
of measurement unless a unit does
not exist for the item measured. SI
units may be included parenthetical-
ly. The ASA Publications Handbook
and Style Manual (1984) is the offi-
cial guide for preparing manuscripts.
Consult the CBE Style Manual (5th
ed., 1983) for style questions not
covered in the ASA manual.
Two main sections in the journal
are Research Articles and Review
Articles. Other sections are Notes,
Letters to the Editor, Book Reviews,
Examples of articles include:
* New crops for profitable agricul-
* Series of papers on soil tests and
* Pesticide application techniques
* The farm problem, government
farm programs and commercial
Subscription information will be
available at a later date. Look for the
premier issue of the Journal of Pro-
duction Agriculture the second half
The Project on African Agriculture:
Crisis and Transformation, sponsored
by the Joint Committee on African
Studies of the American Council of
Learned Societies and the Social
Science Research Council, aims to
establish a framework for interdis-
ciplinary analysis-particularly involv-
ing natural and social scientists-of
the agricultural crisis in Africa.
Through a program of fellowship
activities organized around strategi-
cally selected substantive issues, the
project seeks to develop new con-
cepts and methods and the skills and
resources needed to apply them to
analysis of the biological, socio-
economic and historical processes
which affect agricultural perform-
ance in Africa. The project has
selected access, control and use of
resources as its focal theme for
activities during the fellowship pro-
gram's first year-1987-88.
Financial Support and Activities
The typical fellowship award will
provide up to $15,000 for support
of activities during a period of 3 to
12 months. Fellowship support may
be given for:
* activities which will assist fellows
in developing research project
and proposals, including travel to
libraries, literature searches, collec-
tion of bibliographic materials,
visits to potential field sites, pilot
studies, contacts with other re-
searchers, and participation in
* activities which will assist fellows
in analyzing and writing up re-
search results, including consulta-
tion with other researchers, parti-
cipation in workshops to discuss
preliminary findings, support for
short periods of write-up, and
participation in conferences to
present and assess research results
* activities designed to provide spe-
cialized training needed to under-
take particular kinds of research.
Individual African researchers and
research teams comprised of African
and non-African researchers in the
following categories are eligible to
apply: (1) mid-career scholars based
in universities or research institutes,
(2) individuals who have recently
received graduate degrees (M.A.,
Ph.D., 3rd Cycle Doctorate or equi-
valent degrees); (3) professionals in
governmental posts (e.g., ministries
of agriculture) or other applied re-
search settings. The project partic-
ularly encourages proposals for col-
laborative activities between natural
and social scientists.
Two cohorts of fellows will be
selected during 1987 for the 1987-
88 program. The first cohort of
fellows will be selected in May; the
second cohort will be selected in
For additional information about
the fellowship program and applica-
tion procedures, as well as details on
other areas of support provided by
the project, write to: Fellowship
Program, Project on African Agricul-
ture, Social Science Research Council,
605 Third Avenue, New York, N.Y.
The University of Missouri-Columbia
Libraries has completed the first
phase of an extensive database on
International Agricultural Systems.
From 1979 through January, 1985,
the University of Missouri-Columbia
(UMC) Small Farms Study Library
was located in, and administered by,
the Department of Agricultural Eco-
nomics. In January, 1985, it was
moved to the Ellis Library under the
administrative guidance of the Social
Science Librarian. In October, 1985,
a computerized database was begun
and the collection was named the
International Agricultural Systems
The goal of the collection is to
store and disseminate information
on international agriculture in gen-
eral, and specifically on small farm-
ing systems in developing countries.
Two important sub-objectives of this
project are: a) to develop computer-
generated indices of the bibliographic
holdings of the Library; and, b) to
identify "fugitive" materials in other
locations which are of interest to
farming systems researchers both on
the UMC campus and on-site in
foreign countries. Resources are also
collected from fields relevant to
agriculture such as agronomy, animal
husbandry, communications, eco-
nomics, engineering, health and
nutrition, rural sociology, soil science,
and technology. There is also a strong
emphasis on materials pertinent to
women in development. A printout
file is maintained for current and
active search topics.
The current database was created
using the REVELATION database
applications software on an IBM-AT.
There are over 3900 records in the
system as of December 11, 1986,
containing over 27,000 keywords.
The indexing (hashing algorithm)
capabilities of the system permit one
to search for any character string or
combination of character strings
from any field or from any combin-
ation of fields. The structure of the
applications system permits all data
entry and searching tasks to be
completely menu-driven and sup-
ports case-sensitive help screens.
Each record is restricted in length
to 64k bytes; the number of records
in the database is limited only by the
disk space available. The applications
package can import or export data
in dBASE II, ASCII, Lotus 1-2-3, or
UMC is willing to lend or photo-
copy items from the collection
through normal Library interlibrary
channels. We encourage the use of
the materials either on campus,
through lending, or through requests,
For further information about the
collection, please contact:
June L. DeWeese
Social Science Librarian
2032 Ellis Library
University of Missouri-Columbia
Columbia, MO 65201
FARMING SYSTEMS SUPPORT PROJECT
INSTITUTE OF FOOD AND AGRICULTURAL SCIENCES
UNIVERSITY OF FLORIDA
3028 McCarty Hall
Gainesville, Florida 32611