9.8 Implementation of trials with farmer involvement

Contents - Previous - Next

9.8.1 Regular Farm Visits

It has been said that the best 'fertilizer' a farmer can apply to the crop is foot prints in the field. In other words, regular field visits are extremely important for ensuring a good crop or a productive animal. The same applies to good trials.

Well trained technicians can be very helpful in implementing trials and collecting data, but their performance will be only as good as their supervision. Regular visits are important for several reasons:

• Only through regular visits will researchers understand what has happened in the trial throughout the season and be able to correctly interpret the results,
• Only through these visits can the researchers observe and guide the performance of the technicians and farmers -- in the case of researcher-managed trials -- and be sure that the trials are implemented properly and the data are collected properly.
• For trials with researcher management, it is particularly important that researchers are present when treatments are being applied and critical events occur in the implementation of trials (e.g., when crops from trials are harvested). Mistakes during such activities could greatly affect trial results and lead to either misinformation being produced or the loss of valid results from a trial.
• Lastly, in researcher-managed trials, only through regular field visits can researchers identity problems in a timely manner (e.g., developing weed infestations, developing insect problems, or disease problems in livestock) and deal with them appropriately.

Thus, particularly in researcher-managed trials, there is no substitute for regular visits by senior researchers. Probably visiting once a week is a minimum. Even with FMFI trials, senior researchers must visit each trial occasionally to ensure that the trials are valid for analysis and to understand what has occurred so as to interpret the results correctly.

9.8.2 Interacting with Farmers

Interactions between research and extension personnel and farmers sometimes can be poor because of shyness, a lack of respect, or not knowing how to communicate with the farmers In order to improve the quality of such interaction, it is important for FSD and extension staff to accept the following:

• Researchers should realize that they are there to serve the farmers, not the other way around
• Researchers must understand that the farmers are not there only to do trials and demonstrations. They have other more important objectives, like working to teed their families. Therefore, farmers cannot always give high priority to trials and/or demonstrations. Therefore, it is important to be reasonable about the level of cooperation and collaboration requested of the farmer. For example, efforts should be made to avoid becoming over-committed and putting him/her into a high risk situation.
• Efforts should be made to ensure that both farmers and researchers know exactly what the agreement is. In other words, a type of contract -- which may simply be verbal in nature -- needs to be agreed upon. Also, care should be taken by FSD team members to deliver on any promises of help that they make and to be very wary about building any reward systems into the contract terms. In general, reward or insurance terms, such as guaranteeing a certain level of return, should be avoided, particularly in trials at the FMFI level,
• Sometimes FSD and extension staff feel they must give answers to all questions, whether they are sure of their facts or not. Instead, they should realize that giving wrong information to farmers is much worse than saying 'I don't know', because eventually they will lose the farmer's confidence. In this case, it is best to answer: 'I don't know, but I will find out', Researchers also can check with other farmers to see if they have any good solutions.
• The concept of the trial and the procedures to be used must be explained fully and carefully, several times, especially where farmers are conducting experiments for the first time, because they sometimes have difficulty understanding the concept of 'comparisons,, Any disappointments that result from farmers not collaborating to the degree anticipated should be transmitted to the farmers in a constructive rather than a destructive manner. It is important to try and learn from such experiences and plan in the future, if possible, to make changes that will avoid these happening again,
• Because no technology works in all places, at all times, it is dangerous to tell farmers 'this works', 'this doesn't work,. Instead it is better to say, 'in our experience, this works more often than that', then suggest the farmers try it and form their own opinions,
• It is also important to be well prepared, Farmers lose confidence when they come to a demonstration and the equipment doesn't work, or they are promised seed and it is not delivered. FSD workers should prepare thoroughly, in advance, before meeting farmers and work hard to fulfill their promises.
• It is important to listen to and, where possible, desirable, or relevant, attend to farmers' problems quickly.

Farmers, participation is critically important in FSD. The level and degree of participation of course varies according to the nature of the trial. In RMRI trials, the farmer provides the land and the researcher provides all the inputs and controls the trial. The farmers' contribution is in a sense minimal. On the other hand, in RMFI trials, the farmers' participation is much greater, because they are responsible for implementing the trial under the management of the researcher. Thus, in this type of trial, the quality of the interaction becomes critically important. Finally, in FMFI trials, the farmer is in complete charge and the researcher is dependent on good interaction in order to reap any benefits from the trials.

Although the intensity of interaction with the farmer will depend on the type of trial, a number of factors will influence the nature of the relationship. These are discussed below for each type of trial. Researcher-farmer relations, location of trials on the farm, on-farm trial designs, field data management, and standardization are a few things that need to be considered when conducting research in farmers, fields with their active participation. Although the tone of the following discussion implies emphasis on on-farm trials involving crops, many of the principles apply to on-farm livestock trials as well.

BOX 9.10: RMRI PLOT SIZES ARE DETERMINED BY ENVIRONMENT AND TOPIC

In semi-arid Botswana, the size of RMRI plots has generally been a minimum of 6m by 20m when less precise methods of planting and spreading fertilizer have been among the included treatments. These plot sizes could be smaller if more precise placement methods were used.

For soil moisture conservation/tillage trials using animal draught, plot sizes of 40m to 50m in length and 10m to 15m wide have been satisfactory. Using tractors in such trials might require plots up to 0.5 hectares in size. For these types of trials, large plots are necessary to ensure that treatment effects are separated within the experiment. the researcher's supervision? members of the host family can participate when invited or when they have time and are interested in learning more about the technology. The farmer also can be hired to do some of the work, for example, weeding, bird scaring, or threshing, and may provide the draught power. The farmer is paid for this work, and in most cases, the grain yields are given to the farmer in lieu of rent.

9.8.3 Implementation of RMRI Trials

Relevant points with reference to RMRI trials are as follows:

• Selection of Farmers. In this type of trial, the selection of farmers depends less on their resources -- because these are provided by research -- and more on the technical environment (e.g., are soils representative for the area?), Nevertheless, significant secondary criteria in the selection process are the interest and potential for cooperation individual farmers have shown, Also, they need to possess a large enough area of suitable land to allocate to a replicated trial. Each farmer is approached individually. Researchers meet the farmer before the planting season to discuss the proposed trial and to find out if the farmer is willing to host a trial. The farmer is encouraged to take an interest in the implementation of the trial and, to the extent possible, to comment on and evaluate both the proposed treatments and the results. The more the farmer understands about what is going-on, the more capable he/she will be in explaining it to visitors and friends, so that more individuals can learn from the trial.
• Site Selection. The site should be representative of the target environment and should include the type of soil, topography, etc. suitable for research work. It is important to make clear to the farmer that such conditions are needed to prevent bias in the results. A trial site should not be located where there are paths, ditches, large trees and other conditions that are not normally part of the environment. To ensure good researcher-farmer relations, it is very important that the site be satisfactory for both the farmer and the researcher. Plot sizes obviously are determined by the content of the trial and usually can be the same as used on-station (Box 9.10).
• Implementation. Farmers must understand the importance of the trial to the researchers. The risk of not completing trials is higher on farms than on experimental stations. All the arrangements and the implementation are carried out by the researcher, but the farmer, as a partner, should be informed of the whole procedure so that he/she will not destroy it through independent action or by accident. Although all the work is carried out by the researcher or under
• Data Collection. Less information usually is recorded in on-farm trials than in trials conducted on experiment stations, Researchers should not collect data beyond what is appropriate for the area and the trial, Farmers develop high expectations when something is done on their farms, and their curiosity should be satisfied as quickly as possible if their support and assistance are to continue, Consequently, it is important to ensure that as little time as possible lapses between the completion of trials (i.e., recording data) and informing the farmer of the results,
• Yield Estimation and Result Assessment. For most trials, the final stage is assessing and evaluating a given technology, with yields from the different plots representing the differences between treatments, A suggested procedure is as follows:

• It is important to inform the farmers that production obtained from the trial will eventually be theirs to keep. One approach is to leave all the grain at the farmer's house for drying, threshing, and weighing. However, sometimes this is not possible if detailed information is required on seed weight, seed number. etc.
• Farmers should be given the opportunity to see the yield differences that occurred between the plots before the grain is combined for storage. This provides them with an opportunity to appreciate the treatment differences attained.
• After harvest estimation is completed, farmers should be thanked for their assistance, and plans for the coming year discussed with them. For example, whether the same trial will be repeated, what modifications will be made, whether the same or a different location should be used, etc.

9.8.4 Implementation of RMFI Trials

Relevant points with reference to RMFI trials are as follows:

• Selection of Farmers. Farmers for these trials are selected in the same way as in RMRI trials, They are contacted individually before the beginning of the planting season. Because in RMFI trials, farmers carry out the work under the guidance of the researcher. it is extremely important that the farmer understands the differences between the treatments, what is being compared, and what is required for the different treatments and appreciates the importance of proper experimental procedures. It is also essential that the farmer has a genuine interest in the trial, so that he/she is serious about properly implementing the procedures. The farmer also must have the necessary resources to conduct the trial. Proper selection of farmers in this case requires a good deal of researcher-farmer discussion. Proper selection of participating farmers can make all the difference between successful and unsuccessful RMFl trials.
• Site Selection. As with RMRI trials, a representative location should be chosen for the trial, so that the results can be generalized. Portions of the field with big trees or stumps, paths, ant hills, and ditches should be avoided. Again, the location must be agreed upon mutually by both parties. Because part of the purpose is to examine the effect of farmer implementation of the technology, farmers must have sufficient space to simulate a real operation. In other words, they must have a sufficiently large plot so that the implementation is not very much different from the way it would be done on the whole field. This enables collection of 'real' socio-economic data relating to labour and traction. Where animal draught is being used, plots need to be long enough so that the animals are not turning around constantly. Also, there must be sufficient space at the end of each plot for the animals to turn without entering other experimental plots. For example in trials where animal traction is used, a plot length of 40 to 50 metres and a width of 10 to 15 metres may be satisfactory. For crop variety trials where labour or animal traction inputs are not likely to vary substantially between treatments (i.e., except for differences in time required to harvest the different yields), plot sizes can be smaller. Finally, where larger plots are used, the total number of plots must be restricted severely, so that all plots can be planted within a reasonably short period (e.g., a day).
• Implementation. When conducting RMFI trials, the researcher must appreciate that the farmer its not simply someone to be given instructions on what to do. Rather, the work that the farmer is to do must be discussed and agreed upon mutually. Researchers must accommodate the other priorities of the farmer within the experimental design. Thus, although the farmer does all the implementation under the guidance of the researcher, there is a sharing of responsibilities, and the farmer plays a significant role so that he/she can be in a position to assess the potential value of the technology under practical farming conditions.
• Data Collection. The farmer should be informed during the planning of the trial what data will be collected. Before the trial work begins, it is important to decide on how to record the data collected. It is advisable to develop some standard procedures on how to collect the data. These procedures will help to protect the reliability of the information and to speed up the data processing and analysis.
• Yield Estimation and Result Assessment. It is important that the farmer understands the harvest procedures and that he/she is to compare the yields of the different plots. It is usually necessary, every year, to remind the farmers about this before harvest time and to check the procedures during and after harvest. Generally, the farmer will assess the value of the different treatments, not only in terms of yields, but also in terms of the inputs required to obtain those yields. In this type of trial, both farmer assessment and researcher judgement -- based on documented figures -- are used in the final assessment of the different treatments.

BOX 9.11: FARMER GROUPS FACILITATE TESTING A WIDE VARIETY OF TECHNOLOGICAL OPTIONS

FSD teams in Botswana have had extensive experience in using farmer groups for organizing FMFI trials. A wide range of options has been tested in such a format including:

• Tillage/water conservation techniques
• Planting method options
• Crop varieties
• Manure and fertilizer options
• Forage and fodder production options
• Evaluation of implements
• Seed protection options

9.8.5 Implementation of FMFI Trials

FMFI trials can be undertaken very conveniently through farmer groups (see Section 9.8,6). The object is to test a broad range of technologies under a wide range of conditions. As well as helping to determine what types of innovations are most appealing to farmers with a wide range of resource conditions, they provide a good way of estimating the robustness of different technologies under real farm conditions Additionally, FMFI trials provide an opportunity for involving farmers directly in the technology generation and assessment process, Points to note about a suitable procedure are as follows:

• Selection of Farmers. Prior to the onset of the experimental period, research staff may attend a village meeting at which they make a formal report describing the results of the previous year's research, They also describe the plans for farmer group activities for the coming year and invite anyone wishing to participate in the farmer group to attend the first group meeting. At the first group meeting, a wide range of technology options is discussed (Box 9.11). Farmers then are asked to select any technology they wish to test. They also are invited to add to the list any items that have not been mentioned. Those farmers wanting to undertake a trial form the farmer group that meets on a regular basis (e.g., monthly), Local extension staff also are invited to attend these meetings, where progress and problems are discussed,
• Site Selection. Farmers themselves choose the sites for their trials, However, they are asked to select sites for the test and control plots that are as similar as possible. Factors to consider in determining plot size are the same as discussed for RMFI trials (see Section 9.8.4).
• Implementation. Trial procedures are discussed at the first group meetings and are repeated in later meetings. Additionally, where necessary, village staff visit farmers to assist in the implementation. Items that are to be tested are supplied to the individuals in the groups. Any equipment being tested is given out on a loan basis. After selecting the technologies they would like to test, farmers cut pegs for marking the plots. FSD field staff visit each farmer to help in pegging each trial and to make sure that the farmers understand what they are supposed to do in terms of implementing each trial. Staff may help supervise implementation, particularly in the case of unfamiliar equipment. Field days often are held in each village, At these field clays, a selection of the participating farmers is given the opportunity to explain and show their trials to the rest of the group, as well as to farmers from outside the group, to FSD staff, to station-based researchers, and to extension staff,
• Data Collection. Because of the FMFI format of the trials, emphasis is given to obtaining farmer opinions rather than on gathering quantitative data, The farmers assess the technologies and discuss them during the regular meetings As a result, the farmers provide feedback on problems raised and, where possible, help find alternative solutions to them. Further development of PRA techniques (see Section 8.4.4) should increasingly enable a degree of quantification of this qualitative type data.
• Yield Estimation and Result Assessment. At the end of the season, the farmers harvest the plots separately to compare the yields and are helped by the field staff to weigh the produce. As in other types of trials, the farmers are allowed to keep any production from the testing work they conduct, Implements are assessed through informal discussions and by conducting a forma] end-of-season survey, through which farmers express their views on the particular item they have tested, All other trials are assessed by yield comparisons, informal discussions, and formal end-of-season surveys,

9.8.6 Farmer Groups

Because FSD teams in Botswana had a major role in developing the approach for using farmer groups in on-farm experimentation, which is summarised in a recent paper by Heinrich [1993], some time is spent here discussing its characteristics, The term 'farmer group' obviously can refer to any group of farmers who come together for any purpose, However, in the context of on-farm research in Botswana, the term farmer group has been used to signify a group of farmers who come together to test and adapt new agricultural technology options, to discuss the results of those tests, and to identify on-farm needs for other technology options, These groups are composed of farmer participants, researchers, and local extension personnel and generally meet at regular intervals throughout the cropping season. The farmers in these groups select the technologies they are interested in and test them on an individual basis under the FMFI format. They constitute a group in the sense that they meet to discuss together on a regular basis. These group meetings are supplemented with targeted problem identification/verification visits to specific fields by the researchers themselves during the trial implementation stage.

There are three main purposes for the formation of farmer groups in on-farm research:

• To Expand the Range of Technologies That Can Be Examined in an' On - farm Research Programme. Trials that involve researcher management and supervision require the researchers to spend a good deal of time on field visits. This limits the number of technologies that a research programme can test and evaluate. However, many technology options that may be useful to farmers and others have been developed on the research station and will need field testing before being released to extension. As indicated earlier (see Section 9.5.2 and Box 9.6), by using farmer groups and including farmers in the testing process, research programmes can greatly expand the number of technology options being evaluated in the target area.
• To Include Farmers in the Technology Development Process. One of the main reasons for conducting on-farm research is to include farmers in the technology development process. However, simply having researchers conducting trials on farmers' fields is not the most effective way to undertake this. In the farmer groups, the participants select the types of technologies they wish to test, they perform the tests for themselves, and they discuss and evaluate the various technology options with other farmers as well as with researchers and extension personnel. The two most important criteria used in judging new technology options are: how the technology performs under farmer management and the farmers' joint evaluation of the various technology options. Additionally, at the regular meetings, farmers have the opportunity to present their observations and opinions to research and extension personnel. Thus, through the farmer groups, the farmers become full partners in the technology development process.
• To Create o Forum for Direct Interaction Between Farmers. Researchers and Development/Extension Personnel. One of the primary objectives of FSD is to create linkages between research, development/extension, and farmers at the field level. Farmer groups bring together farmers, researchers, and extension personnel at the field level, in order to work together to develop applied technologies to increase farm productivity.

There are several ways for farmer groups to be organized. There are research-oriented and extension-oriented farmer groups [Norman et al, 1988]. Although discussion here is confined mainly to research-oriented farmer groups, many of the points apply to extension-oriented farmer groups as well.

The groups are composed of FSD researchers, extension personnel (i.e., from village, district, and regional levels), and farmers, The groups are open to any village farmer who is interested in participating, At the first meeting of the group, researchers (i.e., both station-based and on - farm) discuss a wide range of technology options, addressing as many production constraints as possible, Farmers also are requested to raise questions about problems that they feel have not been addressed adequately. From the list of technology options that is thus developed, individual farmers select specific technology options they wish to test, However, across farms, trials of specific technology options are conducted according to a mutually agreed upon trial design. For example, it is suggested, a cowpea variety trial, if selected for testing by 10 farmers, is implemented using the same set of varieties, on the same size plots, and with the same seeding rate at all ten locations. This may allow some subsequent statistical analysis, However, if some farmers choose to deviate from that agreed plan, then that is of course, their prerogative. New equipment, small amounts of seed, and required chemical inputs are, if necessary, provided by researchers. Additionally, the researchers visit each trial once during the season to verify proper implementation, Throughout the season, the farmers, researchers, and extension agents meet as a group on a monthly basis. At these meetings, farmers discuss their progress with the trials, their observations, and any problems encountered, Possible solutions to the problems are discussed by the group.

Researchers collect data on the dates of field operations and the crop and variety used -known because of the standard trial design and the use of researcher-provided seed. The type of equipment used also is also recorded. At the end of the season, farmers harvest and sometimes research staff weigh grain yields and conduct end-of-season farmer assessment interviews. PRA interview techniques (Section 8.4.4) also could be used. These interviews are used to quantify farmer's opinions and perceptions of specific technology options (e.g., crop genotypes). Results are reported both to farmers and to other interested researchers.

Many benefits are derived from using the farmer group approach. A few of the more Important ones are as follows:

• Efficiency. Contacting farmers one by one requires a lot of travelling by researchers. Also, when dealing with individual farmers, researchers must explain their work and objectives and the technology options they are interested in testing many times over. By addressing farmers as a group, researchers not only save a great deal of travel time, but they can explain their programme to many farmers at once, thus saving time on explanations as well. By holding regular group meetings with participating farmers throughout the season, farmers, researchers, and development/extension personnel can discuss progress, problems, and farmers' observations, without the researchers having to visit each farm every month. Thus by working with groups, instead of making a lot of visits to individual farmers, researchers save both time and money that would otherwise be spent on fuel.
• Expanded Range of Technologies Tested. When using the farmer group approach, researchers can work with many more farmers than when they work only with individuals. As a result of greater participation in the testing activities, many and varying technology options can be tested.
• Research-Extension Liaison. When researchers talk with one set of farmers and development/extension personnel talk with a another set of farmers' they may form different ideas about farmers' problems in a region and the best solutions for those problems. However, if research and extension personnel work together at the field level and meet together with large groups of farmers on a regular basis, this problem is reduced. Working together in the farmer group approach allows research, extension, and farmers to come to a common understanding of the problems. If the same groups work together to test various technologies that address those problems, then all are able to see which technologies are useful and practical and which are not. Furthermore, if extension personnel participate in technology development, they will be in a good position to later extend the same technologies to other farmers. Thus, the farmer group approach provides a forum where researchers, extension personnel, and farmers can meet on a regular basis to address problems in the field. This interaction can be of benefit to all participants.
• Group Dynamics and Immediate Feedback. An individual farmer may be reluctant to tell a researcher that a particular technology option is not good. It also may be difficult for an individual farmer to convince a researcher. However, if a group of farmers gets together before a meeting and share a common opinion of a new technology, then it is likely that someone will raise the point during the meeting. Furthermore, if one farmer tells a researcher that a particular technology is not working, the researcher may just assume that the farmer applied the technology incorrectly. This will not happen if 1() farmers at once explain that a technology option is not working. Thus, dealing with people in a group tends to help communications -- if the group is not too large -- and allows research and extension personnel to quickly obtain a practical assessment of any technology being tested. The groups also provide an opportunity for farmers to share opinions and observations among themselves and to develop their own solutions to problems. In societies that are not egalitarian, farmer groups probably will have to be formed for different 'classes' of people in the village, including, when necessary, different groups for men and women (Box 9.12). Also, as the group becomes bigger, productive interaction may be diminished. It is not clear at what size of group a breakdown in communication is likely to occur but, once this is known, the solution is easy -- break the group into two.
• Farmers Develop Their Own Production Packages. Farmers understand their own production problems and resource constraints better than researchers. Thus, when farmers are able to select technology options from a range of options and apply them where they wish, they often develop specific packages that are more practical than what researchers alone could have devised. Farmers also may find uses for technology items that researchers would not have considered. Thus, farmers can have a lot of input into developing improved production systems.
• Farmers Teaching Farmers. The farmer group approach provides good material for field days. Because the trials are FMFI, the farmer who has implemented a trial is able to present it and explain it to other farmers, Farmers learn best from other farmers, so these field days can be very effective when a technology option is ready for extension.
• Flexible Response to On-Station Research Needs. When large numbers of farmers are participating in the testing activities, it is very easy to introduce new technologies for field testing. For example, if the station-based cowpea scientists wish to get farmer evaluations of five new cowpea varieties, it is easy to introduce these to the groups and find farmers who want to try them. It would not be as easy if the researchers had to implement the tests themselves in several locations, Thus, the farmer group approach makes it easier for on-farm researchers to respond to the needs of on-station research programmes for field testing of their new technologies.
• Forum for Obtaining Other Information. These groups are effective forums for obtaining general information not only on the relevance and value of technologies being offered to farmers, but also on many other topics such as producer prices for products in an environment where there are no physical markets or market days; on aspects that affect all farmers (e.g., comments on development projects, extension, etc.); controversial issues (e.g., availability of inputs); trends (e.g., fertility, draught type, etc.): general opinions on issues; or information on variability (e.g., types of farms, farming systems, planting patterns).

BOX 9.12: FARMER GROUPS FACILITATE INTERACTION

Two very positive experiences have come from using farmer groups in Botswana:

• Members of groups have been together for some time, and, therefore, interactions between the farmers themselves and between the farmers and the 'outsiders' have become more relaxed, creating the potential for more productive discussions.
• The cultural setting is such that discussion is not inhibited between farmers in different recommendation domains, thus creating potential for greater variability in the views expressed. Also, the majority of the individuals attending the meetings have been women, some of whom came from male-headed households .

9.8.7 A Case Study: Evolution of Farmer Groups in the Lake Zone, Tanzania

This case study is based on the experiences of Roeleveld and Colleagues 11994] in the Lake Zone FSD team in Tanzania where, like many other countries, it is still called FSR. The material presented provides a good illustration of how farmer groups can evolve over time and develop a 'life of their own' and incorporate other functions as a result of the feeling of empowerment that members appear to develop. Undoubtedly, internal group dynamics can facilitate this feeling of empowerment.

Following a participatory informal survey conducted in November 1992 by the Lake Zone FSD team, an on-farm research programme was started and resulted in the development of farmer research groups (FRGs). The survey, which was carried out in collaboration with the extension service, focused on animal husbandry, a subject not covered adequately in an earlier survey. During the survey, three villages located in one of the three agroecological zones of Kwimba District (8000 square kilometres) were visited, each one for three days. Techniques such as transect walks, village mapping, kraal visits, and individual and group discussions were used. Male and female farmers from various socio-economic strata of the village communities participated. Within a month after the survey, village debriefing meetings took place to discuss survey findings and to identify an initial research topic. In the meetings, which were very well attended by both male and female farmers, proposals for research were selected. In all three villages, it was jointly decided that farmers would test an oxen-drawn weeder to alleviate the workload of men, and particularly of women, during the weeding season. A few weeks later, the first trial had started.

More farmers volunteered than could participate, because the number of weeders available were limited. Therefore, some selection was necessary. About 20 households per village could be included. Care was taken to include female-headed and non-oxen-owning households. Participants of this FMFI trial agreed to meet every two to four weeks to exchange experiences in the use of the weeder. The farmers liked this idea and decided that the 'group' should have a chairman and a secretary. This was, in fact, the start of the FRGs or in Swahili kikundi wa wakulima watafiti.

The village extension worker and research staff attended most of the meetings, which were organized regularly by the FRGs in two of the three villages. Attendance was high (i.e., 15 to 30 persons) with a surprisingly large number of women. Farmers were very positive about these exchanges, and discussion on subjects other than weeding quickly developed. Farmers who had not participated in the trial also started attending these one to three-hour meetings. In the third village, the development of the FRG initially stagnated because of village leadership problems, Once the villagers themselves had solved the problems, the FRG in this village also rapidly developed.

About two months after the start of the trials researchers suggested that the FRGs each organize field days to show farmers and extension workers from neighbouring villages, representatives of other FRGs, district extension personnel, and researchers the progress that had been made in ox-weeding. The results more than fulfilled expectations, The FRGs, in collaboration with their local extension workers, organized three extraordinary field days, which included talks by FRG members, field visits, and general discussions, and ended with songs and food, Transport of district staff and a few crates of soft drinks were the only contributions from the research side,

In that same season, FRG representatives visited farmers in a nearby district where ox-weeding had been introduced recently, Also, a male and a female farmer presented their experiences in a one-day workshop for extension and development staff at the district agricultural office, and all three villages participated in a presentation of the oxen-drawn weeder at the annual district agricultural show.

The first season of the ox-weeder testing ended in each village with an evaluation of the technology by male and female farmers (i.e., farmers' assessment). During these meetings, improvements in the design of next season's testing were discussed, and the meetings ended in big social gatherings with dance, music, and food. At the end of the first year, a second trial, feed supplementation of oxen, was implemented and a fourth research village added.

The second season started with village research planning meetings. Most of the participants had participated in the trials of the previous season. More trials, both agronomic and animal production related, were planned, Most experiments had been suggested by researchers but were based on problems identified during the surveys or expressed during FRG meetings. As a result of requests from female farmers, trials on sweet potato and cowpea varieties (i.e., 'women crops') were included, which mostly involved women, Despite suggestions from researchers to consider changes in the organization of the FRGs, because of the growing number of trials and participants, no changes occurred.

As a result of the increase in the research programme on the research side, little attention was paid to further development of the FRG approach during the second season. This was despite concerns regarding the 'openness' of the groups; little involvement of farmers in data collection; and some inefficiencies in the collaboration between researchers, extension workers, and FRG leaders and participants.

Despite this lack of 'active' guidance of the FRGs, they functioned well. FRG meetings continued, field days were organized for large groups of visitors, a village drama was performed during the annual agricultural show, and every trial was evaluated with farmers. Both extension staff and farmers (i.e., from the FRGs and from non-research villages) helped in testing the first extension leaflets.

Attendance at the different events was generally high. Activities for women often attracted 50 to almost a 100 persons. During the second year, the groups developed greater confidence and knowledge concerning the research done and appreciation of the actual role of 'research'. This was expressed very clearly during the many visits FRGs received from missions visiting the research station and from government officials. The FRGs organized the visitors' programme, and both men and women responded to almost all questions raised about the research undertaken.

The second year showed a number of interesting spin-offs. Four of them were as follows:

• An active women's group was established in one of the villages. Although its members individually participated in the trials, they participated as a group in the ox- weeder trial to generate income by cultivating a rented field. The group recently has applied for credit to increase their economic activities.
• Another activity undertaken by tanners, independently from the research programme, was the training of farmers in nearby villages in the use of oxen-drawn weeders. This training occurred following formal requests from neighbouring villages whose farmers had seen the trials on field days, had seen the equipment operating when passing by, or had heard about it from others. Both male and female farmers have visited the villages to help in the training, indicating that both male and female farmers were trained. Attendance at training was reported to be very high.
• As a result of a request for training in the use of oxen-drawn weeders, from a regional Integrated Pest Management Project, 40 subject matter specialists, extension workers, and farmers from three districts were trained during a two-day period by one of the FRGs.
• Finally, two of the FRGs have created a revolving fund that they want to use for income-generating activities such as the supply of inputs.

Currently, the FRGs are at the start of their third season. Attempts have been made by research to organize large, dynamic, research-planning meetings in the villages in order to enlarge the number of participants and to increase participation of different user groups. With this in mind, 'technology markets' were organized in which researchers demonstrated through posters materials, demonstrations etc. what they can offer, and farmers could choose to participate in one or more of the trials. Furthermore, FRGs have agreed on the delegation of tasks and responsibilities by appointing a farmer research coordinator for each type of trial. Participating farmers will also start collecting some experimental data, a task that will be monitored by the village extension workers. Training sessions are being organized to establish this new initiative with reference to the FRGs and research/extension collaboration at the village level.

In conclusion, a promising start has been made in developing a participatory on-farm research programme. Farmers are becoming increasingly involved in the different steps of the research process, Participation in most of the activities and events is generally high, and participants are becoming more self-confident and increasingly familiar with the 'why and how' of the trials. The FRGs have started developing activities of their own, particularly in the field of training and income generation.

Important components in the success of the FRG approach to date are believed to include the facts that the trials address major farmer-felt problems and that farmers are treated as real partners in the research process. The group approach fits well in the Sukuma culture. From the research side, emphasis currently is being placed on finding ways to ensure broad socioeconomic participation and on increasing the efficiency of the research process through better distribution of tasks among the partners, Also, increasing attention will be paid to medium-term research activities, while at the same time increasing the influence of farmers in defining the research agenda.

Contents - Previous - Next