An Incremental, Farmer-Participatory
Approach to the Development of Aquaculture Technology in Malawi1

Randall E. Brummett and
F. J.K. Chikafumbwa
ICLARM African Aquaculture Project
Zomba, Malawi



Increasing population pressure in SubSaharan Africa has led to over-utilization of land and a subsequent decline in actual and potential agricultural productivity. Putting more land under cultivation will only exacerbate the decline of environmental quality which is already occurring (Brummett, 1994). Increasing intensity of production systems is therefore essential.

In Malawi, over 80 percent of the population is composed of small scale farmers. The farms which this group operates average about 1.5 ha and produce as many as twenty crops, in various sorts of rotation and integration, over the course of the year. In addition, the farms are highly variable from place to place, depending upon a wide variety of social, economic and environmental circumstances. This group of farmers has, in general, failed to benefit from "green revolution" agricultural innovations which were developed by research teams largely focused on commercial monocropping systems.

New methods for increasing the efficiency, productivity and sustainability of small scale farming systems must be found, and a logical first step is the development of improved methods for problem analysis and technology development aimed specifically at small scale farmers. A realistic alternative to traditional technology development and transfer has been utilized by the International Center for Living Aquatic Resources Management (ICLARM) to integrate pond fish culture into low-input farming systems in Malawi.

Integrated Resource Management

Integrated resource management (IRM) is the use of waste products from one agricultural enterprise to fuel another, thus creating economic, environmental and productivity synergisms which improve overall farm efficiency and profitability. IRM systems have been shown to operate effectively in

test cases in Malawi (Brummett and Noble, 1995a), Ghana (Prein, 1995) and the Philippines (Lightfoot and Pullin, 1995). In addition to immediate improvements in farm function, longer term ecological benefits such as reduced soil erosion and increased tree cover are predicted (Lightfoot and Pullin, 1995).

To create the maximum impact of these systems, however, requires that farmers have a holistic vision of their farm which permits the efficient management of resources. Unfortunately, many small scale farmers rely more on tradition than on any systematic approach to farming in making resource allocation decisions. Brummett and Noble (1995a) found, for example, that many farm enterprises are carried from year to year despite continually losing, or earning only negligible amounts of money.

Both the complexity of small farms and the conservatism of small scale farmers serve to dramatically increase the number of factors which must be weighed in decision-making. The simple reliance upon production and net return calculations which most agriculture researchers use to promote their technologies hence often falls short of convincing small scale farmers. The tools and methods of IRM (c.f., Lightfoot and Noble, 1993; Brummett and Noble, 1995b; Prein, 1995) are designed to take the complex reality of small scale farms into consideration when designing improved farming systems.

The concept of IRM goes beyond simple integration and technology transfer. The adoption by farmers of improved techniques may, in the short term, increase farm profitability and food production. However, unless these farmers can adopt also a spirit of innovation


1 Presented to the Association for Farming Systems Research and Extension, Pretoria, Republic of South Africa, 30 Nov – 4 Dec 1998.


and a general attitude of openness to improved farming methods, the gain will be lost over time as population continues to grow. Small scale farms must therefore evolve if they are to continuously improve their performance and become economically and ecologically viable and sustainable.


Poorly educated small scale farmers, operating from within the perspective of a rural ecosystem which incorporates a large number of unquantifiable social and environmental factors, often have difficulty explaining their situation clearly to researchers. At the same time, researchers too often use their advanced schooling to focus on the details of maximizing the output from their special crop. A first step in designing new, more appropriate technologies and in giving farmers the mental tools they need to adopt a more progressive approach to farming is the development of mechanisms for the interactive exchange of information and ideas between farmers and researchers.

Resource Flow Diagramming

In designing research projects aimed at the development of more appropriate technologies for small scale farmers, ICLARM uses resource flow diagramming (Lightfoot and Noble, 1993) to provide the basis for communication between farmers and researchers. The diagramming exercise begins with an interview during which the proposed activities are presented to farmers and their feedback is solicited. Over the course of the discussions, farmers who are truly interested in participating in collaborative research can be separated from those who wish only to be part of a development activity (Harrison, 1995). Following these initial discussions, a walking tour of the farm allows researchers to roughly characterize the farm for inclusion or exclusion from particular studies.

Once farmers who are both interested and fit the experimental design are identified, resource flow diagramming is used to characterize the farms in terms of their resource base. It also provides a means for giving the farmers a systematic perspective on their farming system which they may have never had before and which might arguably be a prerequisite for efficient farmer experimentation (Lightfoot and Noble, 1993).

After the main features of the farming system are noted, farmers are asked to describe how the various enterprises depicted relate to each other. This is done by connecting the various enterprises with arrows to show the direction of resource movement. If, for example, stovers from a maize crop are used to mulch a vegetable garden, a flow arrow is drawn on the map to connect the maize and vegetable plots with an arrowhead indicating the direction of the resource movement (in this case towards the vegetables). Materials which are eaten by the families would be

connected to the house with an arrow. Materials which are sold would be connected with an external market. Cash (generally in the form of payments for labor or money gained through sale of produce), fertilizers and animal feeds are included as resource flows which connect the farm to the local economy. For greater detail, values can be attached to flows to give them a quantitative dimension (Lightfoot and Noble, 1993).

Once complete, the map shows the various farm enterprise systems and the movement of resources around the farm and into the surrounding economy (Figure 1). Depending upon the purpose to which it is to be put, details of soil type, slope and water resources can be easily added. Such mapping provides the researcher with a detailed picture of the diversity and distribution of land, soil and water resources from the perspective of the farmer. Farmers gain a perspective, often for the first time, of their own farm’s relationship to the surrounding rural community and its agroecological environment.

Farmer-Led Experimentation

Having established a map showing the enterprises and resource flows on the farm, the farmer is requested to imagine a scenario where a new or modified enterprise is incorporated into the farming system. In the case of ICLARM, researchers were trying to introduce the idea of IRM through the incorporation of a fish pond into the existing farming system. Once the new idea has been presented in general terms, the resource flow diagram is then re-drawn to show the theoretical relationships between the new activity and existing ones. The re-drawing of the map gives the researchers the opportunity to discuss the specific details of IRM.

An interesting point to note on the drawing is that only the fishpond has links with other resource systems and enterprises which do not pass through the household or an external market. Fish ponds have a particularly high capacity for using and transforming agricultural wastes without creating pest or human health problems (Lightfoot et al., 1993).

The theoretical farming system model created during the re-drawing session is used by farmers and researchers as a guide for conducting applied experiments. In the case of integrated agriculture-aquaculture, the farmer constructs the pond in a site selected in consultation with the researcher (to make sure that it will, for example, fill with water). Other than general advice, no inputs to pond construction are provided by the researcher. Once the pond is constructed and full of water, the researcher must provide the fingerlings for fish stocking. This is done to ensure that the quantity and quality of seed is known and controlled. The farmer then uses the resource flow diagram to manage the farm. The farmer records the amounts of materials which flow along the different pathways and notes any deviations from the design.


randall.JPG (56045 byte)


Replicated simulations are established on the experiment station to control (in the scientific sense) the experiments being conducted on-farm. Every week, samples of on-farm inputs are collected, dried and analyzed. This data is then loaded into PondSim, an ICLARM-developed spreadsheet which reduces the inputs used by farmers to dry matter, organic matter, nitrogen and phosphorus terms. Through a comparison of this data with observations made on-farm, inputs and management practices used by farmers can be mimicked on the experimental station.

This system provides the statistical comparisons needed to make general statements about farm management strategies. It also gives the researchers an opportunity to experience, at a personal level, the problems faced by farmers. This helps in providing insight to potential new technologies and, maybe more importantly, provides the shared experience necessary to the creation of a more positive mutual understanding between scientist-researcher and farmer-researcher. Such a relationship

is one of the keys to a more realistic and fruitful research and development methodology (Figure 2).

Technology Development

As Figure 2 illustrates, the conduct of farmer-led experimentation presents the opportunity to perform on-station research and development which is inherently appropriate to the farmer’s situation. In classical approaches, potential management strategies are compared to some sort of best case scenario. Researchers tend to choose systems for study which push the productivity of the fish pond without regard for whether or not farmers really have access to the inputs necessary to use the technologies developed on their farms.

In the Farmer-Scientist Research Partnership approach, the data gathered during the weekly visits, as mentioned above, is used to establish scientific


controls of on-farm studies conducted by farmers. These controls are also used by scientists on the experiment station to conduct studies of improved systems. For example, if a farmer elects to test the use of chopped maize stover as a pond input, the researchers establish a maize stover control which simulates the on-farm treatments. Then the researchers design other treatments which are also controlled by the same ponds which control the on-farm study. Such treatments might include the use of different presentation strategies for the maize stover such as grinding, composting etc.

When the growing season is over, all ponds both on-farm and on-station are harvested and the results analyzed and compared. The outcome is presented to the farmer for discussion. Sometimes, this requires the use of ingenious methods to clearly demonstrate to the farmers what has happened and care must be taken to avoid confusion (Hopkins, 1988). Group discussions among participating farmers are often useful (Lightfoot and Noble, 1993).

Based on the results of both the farmer’s and researcher’s studies, farmers are requested to re-draw their resource flow diagrams again to show how the system will be managed in the following year. Depending upon the objectives of the research

program, this cycle can be repeated and may, over time, help to develop in farmers a new ability to systematically analyze problems and empirically search for solutions.

ICLARM’s approach utilizes the resource base and constraints faced by farmers to establish control conditions and works from there to modify the production system. Productivities of systems developed in this way are, of course, much lower than those designed by classical methods. There are also problems with communication, trust and misunderstanding motivations which must be overcome (Harrison, 1995; Noble, 1995). However, the results are much closer to those which farmers can actually expect to achieve. At the same time, doing the research with the farmer’s situation firmly in mind gives researchers a much clearer idea as to what might be possible within the context of the complete farming system than does the classical approach of focusing on the fish alone. Building new farming systems from the ground up in this way also gives the farmers a sense of propriety over new technology which facilitates its evolution into more sophisticated and productive forms (Chikafumbwa, 1995). The relationship established with the farming community as a result of this sort of exercise can also facilitate the collection of longer-term monitoring data on technology adoption and impact.


29.gif (47941 byte)



ICLARM uses the farmer-scientist research method to study small scale farming systems and devise innovative approaches to problems which constrain agricultural productivity. The farmers themselves use the outputs of the method for more than just one season of research. In the words of one farmer with whom ICLARM has worked for several seasons: "Participating in the research project has been helpful because I have learned how to keep records of activities and now know how my farm enterprises are being run in terms of labor cost and input costs. I now realize which enterprises are bringing more money to the family and which ones are not helpful."

Although not specifically designed to transfer technology, the

adoption rate of IRM-based aquaculture innovations among farmers who have participated in farmer-scientist research partnerships has varied between 65 and 100 percent. More remarkably, farmers involved in the development and testing of the methodology have expanded the fish farming components of their farms continually over the past three years despite continued droughts which have dramatically reduced the productivity of their fish ponds (Figure 3).

Whether the hoped for spirit of evolution has been instilled in these farmers remains to be seen, but preliminary indications are good. Lightfoot and Noble (1993) and Chikafumbwa (1995) noted that farmers with whom researchers had interacted had adopted, modified and spread to at least four other farmers each the technologies which they had helped to develop.


Figure 3. Pond productivity over time in FSRP Vs non-FSRP fish ponds in Southern Malawi.

fig3.JPG (13989 byte)

Note: Entry-level technology under the FSRP is, initially, much simpler and less productive than production-focused technologies but evolves on-farm as farmers who understand the technology are able to more efficiently manipulate it to suit their individual situation.





Brummett, R.E. 1994. The context of smallholding integrated aquaculture in Malawi. In R.E. Brummett (ed.) Aquaculture Policy

Options for Integrated Resource Management in SubSaharan Africa. ICLARM Conf. Proc. 46. International Center for Living Aquatic Resources Management, Manila, Philippines.

Brummett, R.E. & Chikafumbwa, F.J.K. 1995. Management of rainfed aquaculture on Malawian smallholdings. Paper presented at the PACON Conference on Sustainable Aquaculture, 11-14 June, Honolulu, Hawaii. Pacific Congress on Marine Science and Technology, Honolulu, Hawaii, USA.

Brummett, R.E. & Noble, R.P. 1995a. Aquaculture for African smallholders. ICLARM Tech. Rep. 46. International Center for Living Aquatic Resources Management, Manila, Philippines.

Brummett, R.E. & Noble, R.P. 1995b. Farmer-scientist research partnerships and smallholder integrated aquaculture in Malawi. In J-J. Symoens and J-C. Micha, eds. The Management of Integrated Freshwater Agro-Piscicultural Ecosystems in Tropical Areas. Technical Centre for Agricultural and Rural Cooperation, Wageningen, The Netherlands, and Royal Academy of Overseas Sciences, Brussels, Belgium.

Chikafumbwa, F.J.K. 1995. Farmer participation in technology development and transfer in Malawi. In R.E. Brummett, ed. Aquaculture Policy Options for Integrated Resource Management in SubSaharan Africa. ICLARM Conf. Proc. 46. International Center for Living Aquatic Resources Management, Manila, Philippines.

Harrison, E. 1995. Digging fishponds: perspectives on motivation. In R.E. Brummett, ed. Aquaculture Policy Options for Integrated Resource Management in SubSaharan Africa. ICLARM Conf. Proc. 46. International Center for Living Aquatic Resources Management, Manila, Philippines.

Hopkins, K.D. 1988. Reporting fishpond yield to farmers. Aquabyte 1(2):6.

Lightfoot, C. & Noble, R.P. 1993. A participatory experiment in sustainable agriculture. J. Farming Systems Rsch. Ext. 4(1):11-34.

Lightfoot, C. & Pullin, R.S.V. 1995. Why an integrated resource management approach? In R.E. Brummett, ed. Aquaculture Policy Options for Integrated Resource Management in SubSaharan Africa. ICLARM Conf. Proc. 46. International Center for Living Aquatic Resources Management, Manila, Philippines.

Lightfoot, C., Bimbao, M.A., Dalsgaard, J.P.T., & Pullin, R.S.V. 1993. Aquaculture and sustainability through integrated resources management. Outlook on Agriculture 22(3):143-150.

Noble, R.P. 1995. Research challenges in integrated resource management (IRM) in rural Africa. In R.E. Brummett, ed.

. Aquaculture Policy Options for Integrated Resource Management in SubSaharan Africa. ICLARM Conf. Proc. 46. International Center for Living Aquatic Resources Management, Manila, Philippines.
. .