The Kilimanjaro dairy feeding systems: an attempt at evaluating the impact of on-farm livestock research

N.S.Y. Mdoe and G.I. Mlay
Department of Rural Economy
Sokoine University of Agriculture
P.O. Box 3007, Morogoro, Tanzania

Abstract
References

Abstract

Introduction
2. Methodology
3. Survey Findings
4. Implementations of the evaluation results

The paper presents results of an evaluation survey conducted in the Kilimanjaro highlands to determine the impact of phase I of the Dairy Feeding Systems (DFS) project. Deviations from the conventional experimental design is advanced as the major reason for employing expost evaluation instead of quantitative analysis based on statistical method in assessing the impact of on-farm trials conducted in phase I of the DFS project.

The results of the evaluation survey reveal that the project has had a positive impact on the performance of dairy cattle. Farmers felt that the DFS project has been instrumental for the current closer link between farmers and extension staff. However, extension staff would have wanted to be more involved in future project activities. Farmers considered shortage of veterinary drugs, improved dairy heifers and scarcity of feeds as the main factors constraining dairy development.

Introduction

Traditional approach to on-farm livestock research include steps such as farm surveys, research at experimental station(s), implementation, extension and interpretation of results (Davendra, 1987). These steps have more or less been followed in on-farm livestock research conducted in the highlands of Kilimanjaro by an inter-disciplinary research team of the Dairy Feeding Systems (DFS) project. Preliminary farm surveys were conducted in 1984 (Urio and Mlay, 1984, Mdoe, 1985). Experimentation on use of crop residues, molasses-urea mixture and pasture agronomy have been going on at the Sokoine University of Agriculture. The implementation stage involved the selection of project participating farmers among the farmers who were willing to participate. On-farm trials conducted during phase I of the project included the incorporation of bean haulms, maize stover and molasses-urea mixture in livestock basal rations. Preliminary work on the incorporation of legumes on farmer plots as a means to improve nutritive value of home grown pastures was also carried out.

Except for the pasture agronomical studies started towards the end of phase I of the project, the design of most of the on-farm trials deviated from the conventional experimental design. It was decided to ignore blocking for non-experimental variations within and across households. Sampling was to be purposive rather than random and the trials were to be farmer managed. This decision was necessary after considering the following factors:

i. The herd size per household was small, averaging 4 cows. This factor restricts the number of treatments possible within a single household.

ii. Large variation within and between households exists in terms of breeds kept, age of animals, number and stage of lactation of individual cows and feed management including composition and feeding levels. Thus, under conventional methods of experimentation, a design used need to separate out these sources of variation from the treatment effects.

iii. Farmers attach high value to their animals and short of providing them with animals for the trials, an experimental design that requires differential treatment of the animals or transfer of daily management to "outsiders" is bound to fail.

The deviations from the conventional experimental design imply that quantitative analysis based on statistical methods cannot be employed in assessing the impact of the series of on-farm trials conducted in phase I of the Dairy Feeding Systems project. This paper employs expost evaluation in assessing the impact of the on-farm trials on dairy production in Kilimanjaro highlands. Expost evaluation is the analysis after the completion of a project or of a distinct phase of it. According to FAO (1981), the primary purpose of expost evaluation is to review the overall relevance, efficiency, and effectiveness of a project with a view to generating empirical lessons for planning, designing and implementation of similar future activities. An expost evaluation of projects requires considerable data and information entailing measurement and assessment of effects and impact. This calls for a systematic collection and compilation of information on the selected key indicators (e.g. increased production, benefit-cost ratio etc.) through project monitoring and well designed periodic surveys. It should be emphasized that the evaluation of the phase I of DFS project is just an attempt to assess the impact of the farmer managed on-farm trials. The evaluation is based on a single visit survey and its scope and vigour fall short of a complete expost evaluation.

The objectives of the evaluation were to determine:

- the impact of the DFS project on dairy cattle management
- the impact of the series of on-farm innovations on milk production
- the impact of the project on research-extension-farmers linkage
- ways and means of improving future project activities

2. Methodology

The evaluation was based on three target groups: project participating farmers, non-project participating farmers and extension staff at village, division and district levels. Data for the evaluation were obtained by administering structured questionnaires to the three target groups. It was planned to interview all twenty project participating farmers but at the time of the survey only 9, 4 and 5 farmers were available from Ng'uni, Wandri and Mowo Njamu villages respectively. In case of extension staff, the intention was to interview all of them. However, at the time the survey was conducted only 9 extension staff were available.

3. Survey Findings

3.1 Project participating farmers

3.1.1. Contribution of the Dairy Feeding Systems Project to Dairy Production

All the interviewed project participating farmers in the three villages indicated that they were not incorporating legumes in fodder production before the project. 55% of the farmers interviewed in Ng'uni village reported to have been incorporating molasses in the basal ration of dairy cattle before the project. All farmers in Mowo Njamu village indicated to have not been incorporating molasses in the basal ration while only 50% of the farmers interviewed in Uandri village reported to have been using molasses even before the project. The information given during the survey clearly indicates that all the project participating farmers in the three villages were feeding dairy cows with bean straw and maize stover.

With regard to the impact of the project, most of the participating farmers interviewed in the three villages indicated that the project has had an impact on dairy management in terms of keeping records, use of molasses, better utilisation of maize stover by chopping, increased use of crop residues and improved pasture management practices (Table 1).

Table 1 shows that the greatest impact of the project on dairy management has been on increased use of molasses and keeping milk production records. The least impact has been on increased use of crop residues and pasture management.

Table 1: Impact of project on Dairy Management.

Source: Mlay and Mdoe (1987).

The resultant effect of changes in various management aspects presented in Table 1 was the improvement in performance of dairy cows in the research site. All project participating farmers interviewed in Ng'uni and Wandri villages reported performance of dairy cows to have improved as a result of the project. In Mowo Njamu, 80% of the participating farmers reported improvement in performance of dairy cows while 20% of the farmers indicated that the performance of dairy cows remained the same.

There has been an overall increase in milk yield per cow as reported by project participating farmers (Tables 2.1, 2.2 and 2.3). On the average, milk yield increased from 4.5 litres/cow/day to 5.5 litres/cow/day in Ng'uni village. This -as a rise of 22% over the project period. In Wandri village, milk yield increased by 21% from 5.25 litres/cow/day to 6.4 litres/cow/day. Average milk yield in Mowo Njamu increased by 23% from an average of 4.8 litres/cow/day to 5.9 litres/cow/day during the project period.

3.1.2. Contact with extension staff

It was revealed during the survey that the Dairy Feeding Systems project has had a very big impact on farmers contact with extension agents. All the projects participating farmers interviewed in the three villages reported that their contact with extension agents had improved since the commencement of the project. They strongly indicated that the project really opened way not only to frequent visits by extension staff but also to frequent visits by other government officials from district and regional levels.

3.1.3. Farmers' views on future project implementation

Most of the project participating farmers interviewed during the survey, had the view that the project should be implemented as it was being implemented in the first phase. This view was given by 44%, 50% and 80% of the farmers interviewed in Ng'uni, Wandri and Mowo Njamu villages respectively. Project participating farmers from Ng'uni village felt that the sample size should be increased while those from Wandri and Mowo Njamu villages felt that the frequency of visits by researchers should be increased while those from Wandri and Mowo Njamu villages felt that the frequency of visits by researchers should be increased. The project participating farmers interviewed reported a number of issues they would like to be addressed in future. The issue reported by majority of the farmers as one of the issues to be addressed in future was that of availability of veterinary drugs which are currently in short supply. Availability of minerals as one of the issues to be addressed was reported by 67% and 40% of the interviewed farmers in Ng'uni and Mowo Njamu villages respectively. The issue of incorporating legumes in pastures was reported by 60%, 22% and 25% of the project participating farmers interviewed in Mowo Njamu, Ng'uni and Wandri villages respectively. Other issues in the order of being reported by a large proportion of the farmers as issues to be addressed in future include research on improved pastures, availability of concentrates, selection of high yielding cows, availability of heifers and co-operative marketing of milk.

Table 2.1: Ng'uni village: milk yield before and during project period.

Source: Mlay and Mdoe (1987).

Table 2.2: Mowo Njamu village: milk yield before and during project period.

Source: Mlay and Mdoe (1987).

Table 2.3: Wandri village: milk yield before and during project period.

Source: Mlay and Mdoe (1987).

3.2. Non-project participating farmers

Majority of the non-project participating farmers interviewed in Ng'uni, Wandri and Mowo Njamu villages indicated that they were aware of the existence of the Diary Feeding Systems project in the area. The number of farmers who were aware of the project was relatively small in Ng'uni village whereas only 45% of the interviewed farmers in each of the remaining two villages indicated that they were aware of the existence of the project in their villages. The farmers indicated to have first learned about the project from other farmers, village leaders and extension staff.

3.2.1. Impact of the project on dairy production

The non-project participating farmers who reported to be aware of the Dairy Feeding Systems project were further asked to indicate if the project has made them change dairy management practices. In Ng'uni and Wandri villages' all these farmers reported that the project has had an impact on how they managed their dairy herd. However, only 71% of the farmers who reported to be aware of the existence of the project indicated to have changed dairy management practices as a result of the project.

Dairy management practices which have changed include feeding, pasture management, and record keeping. The non-project participating farmers were also of the opinion that the changes in dairy management have had an impact in the performance of their dairy herd and improved as a result of management changes. In Mowo Njamu village, only 86% of the farmers indicated improvement in performance of dairy herd due to changes in management practices.

There had also been an overall increase in milk yield as a result of changes in management as reported by the non-project participating farmers. Tables 3.1 to 3.3 show the impact of changes in dairy management on milk production per cow per day. In Ng'uni village, average milk yield increased from 3.7 litres/cow/day to 4.3 litres/cow/day during the project period. Average milk yield in Mowo Njamu increased by 22% from 3.7 litres/cow/day to 4.5 litres/cow/day. In Wandri village average milk yield increased by 25% from 3.6 litres/cow/day to 4.5 litres/cow/day during the project period. This shows that impact was greatest for non-project participating farmers in Wandri village where milk yield increased by 25% as compared to 22% and 19% for Mowo Njamu and Ng'uni respectively.

Table 3.1: Ng'uni village: daily milk yield before and during project.

Source: Mlay and Mdoe (1987).

Table 3.2: Mowo Njamu village: daily milk yield before and during project

Source: Mlay and Mdoe (1987).

Table 3.3: Wandri village: Daily milk yield before and during project.

Source: Mlay and Mdoe (1987).

3.2.2. Views on future project implementation

The non-project participating farmers reported a number of issues that they would have liked to be addressed in future if the project was to be extended. These issues reported in their order of being reported by a large number of farmers include easy access to veterinary drugs, pasture establishment and improvement, easy access to concentrates and minerals, use of molasses, general dairy husbandry, upgrading of animals through artificial insemination and access to improved dairy heifers.

3.3. Extension staff

3.3.1. Contribution of the project

The main contributions of the Dairy Feeding Systems project in the research site as indicated by the extension staff interviewed during the survey include better use of crop residues through implementation, use of alternative feeds and improved pasture management. 56% of the extension staff indicated better use of crop residues through supplementation as the main contribution of the project in the research site. Improved pasture management practices was reported by 22% of the interviewed staff as the main contribution of the project. 11% of the extension staff reported that the main contribution was that farmers learned how to use alternative feeds for dairy cows. The remaining 11% had the opinion that the project had not had any significant contribution in the research site.

3.3.2. Views in future project implementation

About 44% of the extension staff interviewed gave the view that the project should have several focal points in Hai district. The views that researchers should frequently visit the project site, local extension staff should be fully involved in the project, and that the project should involve both poor and rich farmers were suggested by 33% of the extension staff. Other views given in the order of importance include intensification of follow up, more research on fodder production, use of visual aids to educate farmers on dairy husbandry, improving baling and transportation of crop residues and emphasizing feed supplementation.

4. Implementations of the evaluation results

The results of the survey indicate that the project has had a positive effect on the performance of dairy cattle mainly attributed to better utilisation of crop residues by choosing and supplementating with urea-molasses mixture. Work on pastures was just at its preliminary stage when phase I of the project ended and therefore it is not surprising that farmers' have not substantially changed their pasture management practices. Both farmers and extension staff were of the opinion that research on pastures should be given priority in future studies.

Farmers felt that the project has been instrumental for the current closer link between farmers and extension staff. However, extension staff felt that they would have wanted to be more involved in the project than it was made possible under the arrangement that prevailed in phase I of the project.

The main weaknesses related to the project include poor follow up as a result of remoteness of researchers from the project site, coverage (in terms of number of farmers and area) was considered too small and limited involvement of local extension staff at village level.

When proposals from farmers on problems to be addressed in future were examined, it was clear that they considered shortage of veterinary drugs and improved/exotic dairy cows as most important factors constraining dairy development in the study area. However, given that these can be available, feeds (particularly roughages) will continue to be a problem. This is supported by the fact that all the target groups gave emphasis to pasture research among proposed future activities.

On the basis of continuous monitoring of the on-farm experiments and the expost evaluation surveys, corrective measures have been undertaken to improve performance in phase II of the project. These measures include:

i. Location of a permanent technician at the project area to ensure a close follow-up on on-going project activities.

ii. significant part of phase II project activities to concentrate on on-farm forage research with both researcher and farmer managed trials.

iii. Incorporation of farmers field days in the programme of project activities.

iv. Making some project facilities available to extension staff as a deliberate effort to facilitate increased participation of extension staff in project work.

References

Davendra, C. 1987. The relevance of On-farm animal production research in Asia. paper presented at a workshop on On-Farm Research/Extension and its Economic Analysis held at the South-East Asian Regional Centre for Graduate Study and Research in Agriculture, Los Banos, Laguna, Phillipines, 19-23 January 1987.

FAO. 1981;. Monitoring Systems for Agricultural and Rural Development Projects. FAO Social Economic paper No. 12, Rome.

Mdoe, N.S.Y. 1985. An Economic Analysis of Alternative Dairy Feed Management Systems in the Highlands of Kilimanjaro, Tanzania. M.Sc. thesis, University of Guelph, Ontario, Canada.

Mlay, G.I. and Mdoe, N.S.Y. 1987. An evaluation report of Phase I of the Dairy Feeding Systems Project. IDRC project file No. 3-p-82-0085, Morogoro, Tanzania.