FAR LTG a SYSTEMS RESEARCH:
THEORY AND PRACTICE IN MALAWI
by Dr. Art Hansen, farming systems analyst with the
University of Florida/United States Agency for Inter-
national Development (USAID) agricultural research
project. Paper was initially presented at Chitedze
Research Station 25 September 1981.
The purpose of this paper is to describe the farming systems approach
and how it is being developed within Malawi. The fanning systems research
program is based within the Department of Agricultural Research of the Ministry
of Agriculture as an important component of the total research effort, but the
program also links the Department with other departments within the Ministry,
particularly the Department of Agricultural Development and the Division of
Planning and Evaluation, and with the research faculty at Bunda College.
Professional and technical staff within the Department of Agricultural
Research need to know more about the farming systems research program because
it is part of the restructuring toward more adaptive research. The restruc-
turing means some significant changes in the allocation of Departmental resources
and in the work orientation of research staff. In addition to this introductory
paper, a series of in-service courses in farming systems research will be given
for Departmental staff and other interested Ministry and Bunda College staff.
During the courses the theory, methods and issues will be covered in more detail.
This paper is more of a brief comprehensive overview.
The farming systems program is part of the United States Agency for
International Development (USAID) funded project to strengthen the Department
of Agricultural Research. This project is supervised by the University of
Florida. The purpose of the five year project (1979-1984) is clearly stated
in the agreement signed by the two governments.
"The purpose of the Project is to strengthen the capability of the
Department of Agricultural Research (D.A.R.) to provide socially
acceptable and economically sound research recommendations to the
extension service for smallholder crop and livestock production...
Emphasis will be placed on assistance to improve and strengthen
the systems for research coordination in the selection, implementa-
tion, and management of research projects of optimum value to small-
holders. Special attention will be given to the needs for continuous
liaison between research and extension functions to achieve transfers
of research results to smallholders."
The main points of this project are, therefore, to help the Department
provide research recommendations that will help the extension service address
small farmer priorities. The recommendations must be relevant and appropriate
to small farmer conditions and must be acceptable to the small farmers. This
is the reason for the training and scholarships for Departmental research
staff, for the presence of University of Florida technical assistance staff,
and for the vehicles, housing, and laboratory equipment purchased by USAID.
Farming systems research is designed specifically to help the Depart-
ment realize these goals of smallholder recommendations. The farming systems
approach helps identify high priority problems of small farmers, understand
the critical constraints and opportunities in the existing farming patterns,
and develop farm-tested recommendations that are appropriate and acceptable.
Farming systems research is primarily carried out on smallholder farms.
The focus of research and the central concept of this approach is the "farming
system." This concept has been defined in different ways by various people,
but the definition I prefer is the following (from Hansen, et.al. 1981).
A farming system is not simply a collection of crops and/or animals
to which one can apply an input and expect immediate results. A farming
system is a complicated interwoven mesh of resources and factors agronomicc,
economic, social, cultural, physical, etc.) which are managed to a greater
or lesser extent by a farmer. Utilizing the technology known to the farmer,
this person or family unit attempts to increase or maximize the farmer's or
farm household's utility within a given context' of accepted preferences,
aspirations, and socioeconomic conditions. The farmer's unique understanding
and interpretation of the immediate environment, both natural and socioeconomic,
is instrumental in creating the farming system. The term "utility" in this
definition refers to a broad range of satisfactions. In the case of Malawi's
small farmers, utility definitely includes the provision of foodstuffs, both
for nsima (the staple dish) and ndiwo (the accompanying side dishes), as well
as the provision of some cash from the sale of crops and/or animals.
Each farmer interprets the opportunities and constraints of the
ecological (climate and soils, crops and animals, pests and plagues) and social
(prices and markets, political policies, cultural values, and uses for labor)
environment in which he or she lives, and each farmer then utilizes some of the
resources that are available to produce a certain mix of crops and/or animals.
The farming system that results is the interaction of environment and resource
allocation, integrated by the farmer's management decisions and work. Environ-
mental variables (rainfall, pests, input availability, etc.) may determine each
year whether the farming system successfully satisfies the farmer's desires,
but the form of the system itself is determined by the farmer's attempts to cope
with the anticipated environment. A drought one year will affect crop yield,
but where droughts are common farmers anticipate them by including sorghum,
millet and cassava into their fanning systems or by early or dry planting.
At one level each farm may be considered a unique farming system. At a
more general or abstract level, there are a number of similarities among indi-
vidual farms and farmers, and individual systems may be grouped into fairly
homogeneous categories. The farming systems research program in Malawi (and
other programs in other countries) works with categories rather than individual
farms because there are not enough research resources for individual work.
The program must first identify the important categories in various areas and
learn how farms in each category operate (learn the resources, constraints,
goals, and relationships) and then devise and test alternative technological
possibilities that will permit farmers to improve their productivity and utility.
The central concept in this work, the farming system, has a complex
definition that includes many variables. This is because the management deci-
sions that Malawi's smallholders are actually making are complex decisions, and
the resulting systems are complex. Almost every decision the farmer makes
involves satisfying some goals at the expense of others. Almost every action
involves costs and benefits since the resources used could be used in several
ways, and applying resources today to one activity means they cannot be applied
to another. Another way to understand this is to see most farming decisions as
compromises. The farmer continues to balance everything he or she wishes to
accomplish against available resources and available time. Generally the farmer
must cut back on the optimum production practices for each specific crop or
animal enterprise (in terms of optimizing yield) in order to keep a number of
enterprises going to satisfy a number of goals.
The complexity and inclusiveness of the farming systems concept force
research and extension staff to consider the same complexity of interdependent
costs and benefits that confronts the fanner. This helps the staff understand
why farmers make certain decisions, so that the staff may devise recommendations
that the farmer can accept within the existing system. Much of the discrepancy
between the technologies practiced by Malawi's smallholders and the technologies
advocated by research and extension staff may be explained by the differences
between the "real world" environment of the smallholder with its complex inter-
dependencies and the controlled research plot environment where many variables
are held constant and many others assumed to be irrelevant. Research recom-
mendations are made with the assumptions that farmers want to maximize yield
per unit of land and that farmers will devote as many resources as are needed
for that specific enterprise. Sometimes those assumptions are true in fact,
and the recommendations permit the farmer to obtain very high yields. Often,
however, smallholders are attempting to maximize several goals, such as a
secure food supply with several ingredients (nsima and ndiwo crops, plus
animals or animal products for ndiwo) plus a cash income, and must compromise
on their allocation of resources to each specific enterprise. Because they
grow many crops which demand attention during the single growing season, the
smallholders must either hire additional labor or compromise on performing
some cultivation tasks which demand simultaneous applications of labor.
The farming systems approach considers a wider range of factors and
relationships than more traditional research and extension approaches which
commonly focus on single crops, animals, or other elements (pests, soils,
machinery, etc.). This does not mean that a farming systems program replaces
the single-factor programs. Both approaches are needed to complement each
other. The more narrowly focused programs pursue in depth specific technical
relationships, while the broader, more comprehensive farming systems program
examines the extent to which existing recommendations are appropriate for
smallholders, identifies high priority research projects for other research
programs, and helps establish procedures to test the adaptability of recommen-
dations to smallholder conditions and goals. The farming systems research
program is, therefore, a component of the total research effort that specifi-
cally addresses smallholder needs and constraints.
The relationship between the different types of research programs is
clearly illustrated in the diagram at the top of the next page. This diagram,
adapted by van Blokland from the Tropical Agricultural Center for Research and
Teaching (CATIE) in Costa Rica (see Hansen, et.al. 1981), shows how the farming
systems approach (the broad arrow cutting horizontally across the page) and the
traditional single-commodity or single-discipline approaches (running vertically)
Another way to conceptualize the difference between the two approaches
is to see the commodity-specific or discipline-specific approaches as being
more idealistic. They construct ideal or optimum biological solutions to the
problem of increasing one characteristic, yield. The farming systems approach
TRADITIONAL AND FARMING SYSTEMS APPROACHES
is more practical. It emphasizes adaptive research to discover which alter-
native solutions are more suitable to actual small farmer conditions, resources,
and constraints. This is adaptive research in a complex sense because it
includes adaptation to natural and socioeconomic circumstances as well as to
the interplay of enterprises on a farm.
There are four general steps in any farming systems research program:
diagnosis and description, design of alternative technologies, testing of the
alternatives, and extension (Gilbert, et.al. 1980). After describing this
general method, I will detail how the research is being developed in Malawi.
The first step is descriptive and diagnostic. There are several goals:
1. Identify and understand the existing local farming systems.
2. Identify recommendation domains, i.e., categories of farms and farmers
that are homogeneous enough so that one set of recommendations will fit.
3. Identify relationships within the systems where resources are not used
as efficiently as possible. These would include compromises on techni-
cally optimum production technologies.
This first step involves reviewing secondary data (information collected
by someone else) as well as conducting on-farm interviews and observations.
Secondary information gives the research team background information about soil
and rainfall patterns, population distribution, economic flows (crop and live-
stock sales, purchases of inputs, location of markets, availability of inputs),
and existing research and extension recommendations and activities. Extension
staff are usually able to provide valuable information about local cropping
patterns and, since they are in close contact with smallholders,information
about smallholder complaints and the ways in which smallholder practices differ
On-farm interviews-and observations are an essential activity. These
visits to smallholder farms and talks with smallholders permit the research
team to appreciate the multiple objectives of smallholder farming and the resul-
tant complexity. Interviewing farmers about their cropping patterns and deci-
sions while actually standing in their fields allows the team to check verbal
information with actual observations. The interaction of research staff and
smallholders is a consistent feature of farming systems research. Smallholder
farmers are at the same time the ultimate clients of this research and actual
participants and partners in the research process. Research staff have one
type of expertise, and smallholders have another. They have a great deal of
experience with local conditions. They know what they are trying to accomplish
with their multiple enterprises, although they may not be able to clearly
explain their goals verbally. The research team must actively encourage and
support smallholder participation in describing, analyzing, prescribing, testing,
and evaluating technologies and systems.
The incorporation of farmers and extension staff into the research is
part of the general method employed in farming systems research. Another part
of the method is the use of multidisciplinary research teams. Since the unit
that is being investigated (the farming system) is complex and includes a wide
variety of factors, the research team includes production and socioeconomic
staff from various disciplines.
The initial descriptive and diagnostic stage ends with the identification
of some high priority targets for adaptive research. Farming systems work is
action-oriented. The team must constantly remind itself that the goal is rapid
development of appropriate technology that smallholders can and will use. The
single most important criterion for evaluating the success of the farming systems
approach is the extent to which smallholders adopt technologies developed by
the approach. Innovative technologies that are not adopted are failures. Non-
adoption of the innovation probably means it is inappropriate for smallholders.
The second step after diagnosis is the design of alternative technolo-
gies. These alternatives are intended to improve the smallholder's exploitation
of the biological potential of his or her environment and enhance the farmer's
overall utility/satisfaction (Collinson 1980). Based on the diagnosis of high
priorities for research and on an understanding of the resource capabilities
of farmers in the recommendation domain, the proposed alternatives are intended
to modify the existing system rather than dramatically change it. The reason
for this modest aim (gradual modification rather than radical change) is the
recognition that smallholders are reluctant to undertake radical changes which
entail a lot of uncertainty. Existing technologies may involve some biological
inefficiencies but they are tested and well-understood by the local farmers.
Innovations by their very nature mean the farmer must try something that he or
she does not know from experience.
The third step is testing the proposed alternatives to see how they
perform. Although some testing might occur on research stations, the preferred
form of testing is on-farm and farmer-managed. When on-station trials are
needed to evaluate some relationship under close controlled management, they
are always supposed to be followed by on-farm trials to test the adaptability
to farm conditions. Since we are discussing a complex adaptation to natural
and socioeconomic conditions, and a need to integrate any proposed technology
into an ongoing complex system, there is a need to have farmers manage the
trials. Simply placing the trial on a farmer's land only tests adaptation to
natural conditions. It is when smallholders actually manage the trials that
they and the research team are best able to measure and evaluate the systemic
adaptability of the proposals. Farmer management does entail some changes in
trial design. Random bloc designs and designs with multiple repetitions are
difficult for smallholder farmers to understand and operate. A more suitable
design is when each treatment occurs only once per farm, and each farm is then
considered a repetition of the trial.
The fourth step occurs when the tested innovation proves to be a good
and acceptable modification of the system. At that time the proposal is handed
over to the extension service. Note that extension is receiving a site-tested
adapted innovation that has been tested and approved by local farmers as well
as by research. If the testing (third step) reveals problems, then the fourth
step does not occur until the innovation is finally cleared. The fourth step
is facilitated when extension staff are involved with the process of on-farm
testing so that they are aware of the reasoning behind the process and the
innovation. This accentuates once again the need to continually include exten-
sion staff and smallholders in a collaborative farming systems research program.
Extension is an essential step in the farming systems research process
since adoption of the proposed alternative technologies is the single most
important criterion by which the program will be evaluated. This is sometimes
signified by adding extension to the program description and calling it farming
systems research/extension. The best way to evaluate the success of the program
is to resurvey the locality two or three years after starting the extension step
to test how many smallholders are adopting the alternatives and to analyze any
reasons for delays in adoption.
THE MALAWI PROGRAM
Although the University of Florida/USAID agricultural research project
with its farming systems research component was planned and accepted in 1979,
the first actual demonstration of the farming systems approach in Malawi was
carried out by Dr. Mike Collinson of the Eastern African office of the Inter-
national Maize and Wheat Improvement Center (CIMMYT). Dr. Collinson conducted
a diagnostic survey (the first step) of Ntcheu in February 1980. This involve-
ment reflects the interest in farming systems research expressed in the various
international agricultural research centers (see Technical Advisory Committee
1978; Gilbert, et.al. 1980; Collinson 1980). Another expression of the wide-
spread interest in this approach is the fact that farming systems research
programs are already in various stages of development in the countries neigh-
boring Malawi (Zambia, Zimbabwe, Tanzania, Kenya, etc.) as well as in Malawi.
The program in Malawi started effectively with my arrival earlier this
year. My responsibilities as farming systems analyst with the Department of
Agricultural Research include developing such a research program for Malawi,
collecting and analyzing data on actual farming systems in various regions and
areas, and helping train Malawian research (and other Ministry) staff in the
methods and intent of farming systems research ,so that they will continue the
program after my departure.
As has been noted earlier, this is a multidisciplinary program. It
focuses the attention of numerous scientific disciplines on ways to effectively
improve existing systems of smallholder farming. Many of these disciplines are
already present within the Department. Instead of isolating a few scientists
in a separate farming systems unit, the procedure that will be followed in
Malawi is that this sort of adaptive technology research will be considered an
important aspect of almost everyone's research. All of the different specialists
within research will be involved at one time or another in surveying or in plan-
ning or testing alternative technologies. Agronomists will be most involved,
and eight agronomists from the Chitedze, Bvumbwe, Makoka and Baka Research
Stations have already participated to varying degrees this year. Other
specialists in plant breeding, pathology, soil science, etc., have also been
involved, primarily in planning trials.
Only two new disciplines are being added to the Department in order
to complete the multidisciplinary coverage needed for more effective small-
holder research. These two are agricultural economics and applied anthropology,
which are being added to the Department as sections of production economics
and farming systems analysis, respectively. The latter is called the farming
systems analysis section because I am spearheading the introduction of the
program from that section, while simultaneously selecting and training staff
in my own discipline of applied anthropology. In actuality, the farming
systems program will be part of almost everyone's research, not the specialty
of any one section.
The farming systems analysis program has been initiated this year in
three areas of Malawi: the Lilongwe plain of Lilongwe Agricultural Development
Division (ADD), the Phalombe area of Blantyre ADD, and the Bulambia plain of
Chitipa District of Karonga ADD. These areas were chosen in consultation with
Ministry of Agriculture staff. Diagnostic surveys have been planned and
carried out in each place; problems and constraints have been identified; and
adaptive research trials have been planned in two of the areas.
In each place the process began with planning meetings with the ADD
program manager and with other management and technical staff (primarily
evaluation and extension) and agricultural research staff. From these plan-
ning meetings and a review of secondary materials, a relatively homogeneous
area (in terms of rainfall, soils, and cropping patterns) was chosen in each
ADD to be surveyed. Since the purpose of the survey was to rapidly identify
systemic characteristics, problems and constraints, survey team members were
chosen from the professional and higher and more experienced technical levels.
This survey work cannot be left to less skilled or experienced staff.
Each survey team included approximately eight people who were drawn
from research (agronomists and myself), evaluation, extension, and other ADD
sections. Each person received a copy of an interview guide prepared by
CIMMYT that outlined the variety of topics to be covered during the survey.
The purpose and methodology of the survey was explained, and it was emphasized
that the team was going to learn from the smallholders, not lecture to them.
The actual on-farm surveying lasted three days in each area, with the
team members being split each day into three or four interviewing groups (two
to four people in each group). After the first half day of interviewing, most
people felt reasonably comfortable in their new role of listening and observing.
At the end of each day the groups met together to exchange information and to
synthesize as a research team their impressions and beliefs about local farming
systems and constraints. At the end of the three days of surveying the teams
spent from half a day to two days discussing what they had discovered about
the locality and what recommendations they had for further research or for
In each area the three day survey was sufficient time for the team to
identify the basic characteristics of the local farming systems and to identify
a number of ways in which local farmers were compromising the yield potentials
of specific crops. The teams were also able to identify some of the reasons
why smallholders were unable or unwilling to follow research recommendations.
These surveys were action-oriented; they were intended as rapid ways for skilled
staff to assimilate the outlines of the local systems, constraints and oppor-
tunities. The surveys were successful in that each did identify what they set
out to identify. Each survey resulted in new insights into the local situations
and good ideas about targets for adaptive trials.
Farming systems programs in different countries have evolved different
ways of verifying the truth of what is discovered in the exploratory or rapid
surveys. Some country programs in fact have elected to spend one or more
years in exhaustive background surveying before even attempting to identify
targets for adaptive research and extension. In Malawi the program is more
action-oriented. Instead of waiting until the local systems are fully under-
stood (a process that could take years of complex study), we will set up adap-
tive trials on the basis of the rapid survey as supported by existing data from
evaluation, agro-economic surveys, extension, and ADMARC. My original intent
was to conduct formal verification surveys of selected random samples of local
smallholders to validate the information and impressions from the first sur-
veys. As it has turned out, all of the areas are project areas with evaluation
sections, and all of the areas have a lot of available data on farming outputs.
Because of the available data, and because up to now studies of the available
data tend to confirm the results of the rapid surveys, we shall continue to
set up trials without carrying out formal verification studies.
It is important to note here that the farming systems method is iterative.
That means that there is a continual cycling back of information into planning.
As trials are conducted, they are monitored for smallholder reactions and to
gain more data on the costs and benefits (yields, labor, capital inputs, cash
and food outputs). This new information is fed back into the planning process
to modify our understanding of local systems and to modify our recommendations.
This feedback process is especially important when we are dealing with such a
complex unit as a farming system. Thus, instead of delaying any trial research
until all information on local systems is collected and analyzed, we will move
rapidly into on-farm trials and intensive monitoring to discover more about the
systems as they respond to our proposed innovations.
Another basic feature of the approach in Malawi has been the continual
interaction among research, ADD management and staff. After each survey there
has been a meeting in which the -assembled ADD staff had an opportunity to hear
the results of the survey and discuss its significance. In both Blantyre and
Karonga ADDs the meeting combined the presentation of survey and evaluation
data, and these joint presentations were especially effective in permitting
ADD management and staff to put together this overlapping data and interpre-
tation of data. Joint meetings of agricultural research and ADD staff are
essential to achieve a successful turnover of research recommendations (the
fourth step in the general method). ADD staff usually know a great deal about
local conditions and farming patterns, and this needs to be included in planning
adaptive trials. Extension staff, in particular, need to be involved in setting
up and monitoring on-farm trials so that they appreciate the results and are
able to correctly interpret them to local smallholders.
This is the present status of the fanning systems research program in
Malawi. We have surveyed three areas; joint meetings of research and ADD staff
have evaluated the survey findings; adaptive trials have been planned in two
of the areas; and trials will commence with the rains of 1981 (for the 1981-82
cropping season). It is too early to point to recommendations that have emerged
from the surveys and been tested on farms. At this stage we have generated
hypotheses about systems and'recommendations but not yet tested them. Although
it is too early to evaluate the results of trials, let me present some of the
issues that have been raised and the hypotheses generated, so that you gain
more understanding of the sorts of results and the orientation of the program.
ISSUES AND EXAMPLES
I will start by presenting an example of one of the areas surveyed -
Phalombe since it is from the study of actual conditions and systems that
issues and hypotheses arise. Phalombe is a drought-prone area bounded on the
north by Lake Chilwa, on the east by the Malawi/Mozambique border, on the south
by Mount Mulanje, and on the west by the Phalombe River and Traditional Autho-
rity (chiefly) boundaries. The area lies between 1900 and 2400 feet above
sea level; rainfall is unreliable, the average rainy season lasting three to
four months but frequently being too short for a good maize yield; and the
distribution of rainfall across the area varies from being highest near Mount
Mulanje and to the east of the mountain and being lowest near Lake Chilwa,
where it averages below 30 inches a year.
There is intense population pressure on arable land, and the average
amount of cultivated land per household is slightly more than one hectare
(1.02 ha.). The pressure on land and the resultant small landholdings were
noted in the survey and corroborated in evaluation and agro-economic surveys
for the 1978/79 and 1979/80 cropping years. More than 6C~0 of all households
cultivate less than one hectare, and approximately one quarter of the house-
holds actually cultivate less than half a hectare apiece.
People have responded to the dual constraints of short and unreliable
rainfall and the shortage of arable land with two obvious strategies: intense
intercropping and off-farm employment. Every farmer interviewed during the
May survey mixes cowpea and maize seed before planting so that the two seeds
are planted together. Fields were frequently observed where three, four and
even five crops were mixed. Maize, sorghum or cassava are present in almost
every upland (munda) field, since only the minority of farmers have enough land
to devote a field to a non-nsima crop. In the areas of higher rainfall, small-
holders with very little land even intercrop sorghum and maize, sorghum
serving as an insurance nsima crop in years of drought, although smallholders
with slightly more land prefer to separate sorghum and maize. In the drier
areas, cassava replaces sorghum as the major insurance nsima crop. In addition
to the crops already mentioned, sunflower, pigeon peas, groundnuts, various
varieties of beans, chickpeas, and grams are also intercropped in various
combinations, primarily with maize. Rice is grown as an alternative nsima
crop, but usually where there is dambo land available.
The majority of smallholders interviewed mentioned a consistent shortage
of nsima during the December through March period, and some started running
out of nsima even earlier. Most of them respond by working (ganyu) on local
farms or migrating farther away to work on estates. Few attempt to increase
their maize yields through applying manure because only a minority own cattle
(approximately 10% of households). Some of the smaller farmers had applied
fertilizer to their food (local) maize in the past but encountered two problems:
when the rains fail they lose their investment, and they need their available
money to buy food for this year, so instead of buying fertilizer in December
they buy maize to eat. The first priority for most smallholders appeared to
be ensuring a stable production of enough nsima for their household.
Even though enough nsima was the highest priority for many, that in
itself is not enough. All households have a need for some money income, and
some households that did not produce enough maize for themselves still sold some
of that maize after harvest in order to have some money. Others utilized off-
farm labor (ganyu) or businesses or the sale of crops such as sunflower, grams,
chickpeas (all of which are grown primarily as cash crops) or the surplus of
ndiwo crops. Only those households with enough land to devote some to non-
food production were growing and selling the cotton and tobacco.
The survey focused on crop production since the major livestock were
more of a minority concern: 101% of households owning cattle and 25% goats,
although we did recommend that the project supply Newcastle disease vaccine
for chickens. In terms of crops we focused on nsima production in an inter-
cropping context of scarce land and capital. The intercropping would have
to include ndiwo crops (since all farmers are trying to supply the basic diet
of nsima and ndiwo from their farms), although the ndiwo did not appear to
be a major problem, at least not in comparison to the nsima problem. Another
goal with the intercropping research would be the increased production of
money-generating crops, as long as this did not interfere with the primary
goal of ensuring a stable and increased production of nsima.
,What immediately becomes clear from this example is how the farming
systems approach to research starts from the existing systems and constraints
and attempts to deal with the highest priority problems that are identified
for the majority of local smallholders. In this instance, the approach means
that research should look at intercropped maize, small-scale farming, and
increasing production and stability with little money or land. How does this
obviously differ from the traditional research approaches? Monocropping,
single commodity, single discipline, and focusing on yield optimization with a
full package of inputs these are basic features of the existing national
But the two approaches are best seen as complementary. In this case,
farming systems research identifies existing problems and priorities for
research trials. Then specialists from the various other research programs
(in maize breeding and agronomy, minor legumes, grain legumes, sunflower,
and soil science in this instance) cooperate with the socioeconomic research
staff to plan, conduct, monitor, and evaluate research trials. These specialists
know a lot about pieces of the smallholder puzzle, and the farming systems work
helps focus that knowledge. The extent to which these two approaches assist
each other will determine the success of the Department in providing sound and
acceptable research recommendations for smallholders to extension.
BIBLIOGRAPHY ON FARMING SYSTEMS RESEARCH AND PHALOMBE PROJECT
1. Agro-Economic Survey, Malawi Government 1977 Phalombe (Summary Report).
2. 1980 Phalombe Summary Report.
3. Collinson, Michael 1980 A Farming Systems Contribution to Improved Rele-
vancy in Agricultural Research: Concepts and Procedures
and Their Promotion, by CIMMYT in Eastern Africa.
4. CIMMYT Eastern African Economics Programmne (no date) Detailed Guidelines
for Informal Survey Discussions with Farmers.
5. Douglas, Malcolm (no date).Geology and Geomorphology (of Phalombe).
6. Evaluation Unit, Blantyre ADD 1981 Phalombe Rural Development Project
Household Composition, 1980 (Working Paper 1/81).
7. 1981 Phalombe Rural Development Project
Garden Survey, 1980 (Working Paper 2/81).
8. 1981 Report of Preliminary Farm Systems
Survey Conducted in Phalombe, May 1981 (Working Paper 3/81).
9. 1981 Phalombe Rural Development Project
Yield Study Survey (Working Paper 4/81).
10. Gilbert, E.H., et.al. 1980 Farming Systems Research: A Critical Appraisal.
11. Hansen, Art 1981 Report on the Fanning Systems Survey Conducted in May
1981 in Phalombe Rural Development Project of Blantyre ADD.
12. et.al. 1981 Farming Systems of Alachua County, Florida: An
Overview with Special Attention to Low Resource Farmers.
13. Hildebrand, Peter E. 1979 Generating Technology for Traditional Farmers:
The Guatemalan Experience.
14. Malawi Government 1977 National Rural Development Programme: Phalombe
Rural Development Project Proposals.
15. Norman, David W. 1980 The Farming Systems Approach: Relevancy for the
16. Technical Advisory Committee, The Consultative Group on International
Agricultural Research 1978 Farming Systems Research at the
International Agricultural Research Centers.
The publications noted above as well as others 'concerning farming systems
research and smallholder farming systems are available for loan from my
office at Chitedze Agricultural Research Station.