Previous Page Table of Contents Next Page

Smallholder dairy in Ethiopia

Tesfaye Kumsa

Bako Agricultural Research Centre
P.O. Box 3, Bako, Ethiopia

The project area
The experimental animals and feeds
Selection of participating farmers
Obligations of participating farmers
Obligations of the research centre
Data recording
Project achievements



Ethiopia's population, is estimated at 49.3 million with an annual growth rate of 2.9 percent Over 80% of this population is engaged in agriculture, which contributes 50% to the gross domestic product and 35% to the country's export revenue. Ethiopia has among the highest livestock numbers in Africa. The role played by livestock resources in the economy of the country is varied. The development of dairy programmes in the last 2-3 decades had limited impact on modifying traditional livestock management systems. In an attempt to overcome this problem the institute of Agricultural Research launched an on-farm small-scale dairy research project in 1988. Although various political and technical constraints have affected the implementation of the project, studies have been conducted showing that the F1 Friesian and Jersey crossbred cows with appropriate management could perform reasonably on-farm in respect to milk production and reproduction and also maintain good health.


Ethiopia is a country with a human population currently estimated at 49.3 million and growing at an annual rate of 2.9%. The country covers a total land area of 1.2 million km². Agriculture dominates the economy, accounting for about 50% of the gross domestic product and 35% of the export revenue. Over 80% of the human population is engaged in this economic activity (Goshu et al, 1989; Central Statistical Authority (CSA), 1990).

The country ranks top of the list of African countries with large livestock populations. Although statistical data for livestock in Ethiopia have never been consistent, the latest estimates indicate that there are 27 million cattle, 24 million sheep, 18 million goats, 7 million equines, 1 million camels and 52 million poultry (World Bank 1984; Gryseels and de Boodt, 1986). Over 60% of the cattle and sheep are found in the highlands, while goats and camels are predominantly found in the lowlands.

The role played by livestock in the economy of Ethiopia, as in many other developing countries, is varied but substantial. Livestock contribute to the production of food (milk, meat, eggs and blood), industrial raw materials (wool, hair, hides and skins), inputs for crop production (draft power and manure) and export earnings (live animals, skins and hides). They also generate cash income which can be used to purchase food grains, seeds, fertiliser and farm implements and for financing miscellaneous social obligations. Equines provide an important means of transporting goods and people in the majority of the areas in the mid- and high-altitude parts of the country, while camels have a high premium as food and transport animals in the pastoralist dry lowlands of Ethiopia. Jahnke (1982) estimated that livestock contributed about 3% of Ethiopia's agricultural GDP in 1980 apart from its contribution to draft power, transport and manure.

Dairy development research programmes have been undertaken in the past two to three decades in Ethiopia by various organizations. A review of these programmes by Beyene Kebede (1987) indicated that they made minimal impact on the dairy industry. This was after taking into consideration the huge livestock population, the high domestic demand for dairy products and the climatic conditions conducive to high productivity in most parts of the country. Besides the technical and policy factors that have contributed to this failure, the main limitation has been that most programmes involved in on-station activities lacked on-farm components. To fill this gap, the Institute of Agricultural Research (JAR) of Ethiopia launched a small-scale on-farm dairy production research project in 1988 with financial assistance from the International Development Research Centre of Canada (IDRC). The project was based at two locations, one in the central highland zone and the other in the subhumid zone of the western region of the country. The overall objectives of this study were to evaluate the on-farm performance of F1 cows and to introduce improved feeding packages, livestock management practices and the marketing of livestock products. The acceptability of these technologies was also an important component.

The project area

This project was undertaken in two different ecological zones, the central highlands and the western subhumid area, where IAR had been involved in livestock, crop, forage and pasture research since the mid-1960s.

The central highlands (Holetta)

In this area the rainfall pattern is bimodal, with a short rainy period from February to April and a long rainy season from mid- June to September. The annual temperature ranges between 18°C and 24°C and rainfall is recorded between 1000 and 1100 mm. The predominant soil type is Vertisol. Mixed farming is a common practice the traditional sector and major food crops include wheat, barley, teff, pulses and oilseeds. The central zone is the most densely populated region in the country with 9.5 million inhabitants and over 60% of the total cattle and sheep populations. Consequently, grazing areas are not only largely limited but also overgrazed. The Dairy Development Enterprise of the Ministry of State Farms, Coffee and Tea Development has most of its dairy farms in this zone and collects milk form farmers within a radius of 120 km of Addis Ababa.

The subhumid zone (Bako)

This is a zone covering a total area of about 750 km with an altitude range of 1600 to 2000 m. The area receives annual rainfall of 1200 mm, 89% of which falls between May and September. The rainfall pattern here is also bimodal, with short rains coming in March and April. There is a distinct dry season extending usually from November to February. Temperatures average at 27°C with a range of 22-31°C. The dominant soil types are Nitosols with fertile alluvial soils in valley bottoms and depressions. Major crops in the area, in order of importance, are maize, teff, noug, pepper, sorghum, millet and pulses. Although 70% of the farmers own cattle, there is an acute shortage of milk in most urban centres of the region, mainly because of low production of the indigenous cattle and the lack of an established collection and marketing system for milk.

The experimental animals and feeds

Earlier research work at the Holetta and Bako centres indicated that Friesian and Jersey crossbreds were superior in dairy production and reproductive traits at both locations. It has also been observed that oats and vetch at Holetta and Rhodes grass and Leucaena at Bako performed better than other evaluated forage crops. These constituted the livestock feed packages in the two areas. For the programme F1 crosses at Barka, an indigenous zebu breed with Friesian and Jersey were used. In this study 20 of these cows were used on-farm at Holetta and 10 were used at Bako. Experimental cows grazed on unimproved natural pasture during the day and on communal land in the highlands and in an enclosed area in the subhumid zone. They were enclosed at night and fed on supplemental oat/vetch or Rhodes hay depending on the site; concentrates were given whenever available. Cows were hand milked in the morning and evening after first giving allowance for a few minutes of calf suckling.

Selection of participating farmers

When the project started in 1988, priority was given to farmers' cooperatives, due to the government policy of the time. Later with a change in economic policy, cooperatives were dissolved and the orientation switched to individual farm holdings. In both situations, a team of researchers in the disciplines of animal production, nutrition and feeds, health and agricultural economics supported by other technical staff were responsible for selecting farmers eligible for the project. The selection criteria were site suitability for dairy farming, accessibility and the commitment of farmers to buy a pregnant crossbred cow or heifer and to adhere to project conditions.

Obligations of participating farmers

Project farmers had to sign a memorandum of understanding with IAR concerning the following points:

· that they establish an improved pasture of a recommended forage variety, enough to sustain project animals all year-round

· that they handle project animals according to given recommendations and refrain from either using them for purposes unrelated to the project or selling them until after the project had ended

· that they provide all labour inputs required on-farm by the project

· that they construct a simple separate barn for project animals to enable overnight supplementary feeding.

Obligations of the research centre

According to the memorandum of understanding, IAR's obligations were to:

· sell a pregnant cow or heifer to the farmer at a subsidised price

· make seeds of improved forage crops available to project farmers

· provide farmers with technical support regarding improved dairy management and forage production

· help farmers get veterinary drugs and vaccines.

Data recording

On-farm data were collected on the following:

· daily milk yield of each cow

· monthly animal weight

· insemination date

· calving date

· calf birth weight

· animal health.


Milk yield

Average lactation and daily milk yields of project animals at the two project sites are summarised in Table 1. The political instability that prevailed in the country in 1991 made regular visits to farms impossible, which in turn affected data collection at various levels. The interruption was greater at Holetta (highlands site), where milk yields were taken for six months only for some farmers and less for others. At Bako (mid-altitude site) farmers continued milk recording without interruption and a full lactation record was obtained (Table 1).

Table 1. Average milk yield and lactation length of cows on-farm at Bako and Holetta.


Breed type

No. Cows

Mean milk yield (kg)

Mean lactation length (day)

Mean daily (kg)































a.= Subhumid mid-altitude climate
b.= Cows owned by individual farmers.
c.= Highland climate.
d.= Milk yield for 6 months.
e.= Cows owned by farmers cooperatives.
BaF = Barka x Friesian.
BaJ = Barka x Jersey.

This discrepancy in data collection precluded any reasonable comparison between the performance of breeds or sites. Nevertheless, Friesian crosses appeared to perform better in both milk yield and lactation length than did the Jersey crosses. Cows at Holetta tended to follow the same trend although the differences seemed quite low (Table 1). However, at both sites, the number of cows involved was too small to make a statistically sound inference about the two test groups.

Earlier on-station evaluation results indicated that there was no significant difference in the fat-uncorrected annual milk yield between Friesian and Jersey crosses (Sendros et al, 1987a). Friesian and Jersey F1 crossbred cows at Bako were able to produce only 81% and 56%, respectively, of the milk they yielded on-station. The yield achieved on-farm, however, did not take account of the portion of milk suckled by calves before each milking. Survey results for Holetta (Zinashi and Seyoum, 1991) and Bako (Tesfaye, 1991) revealed that indigenous cattle breeds at these sites did not yield on average more than one kg of milk per cow a day. The limited available data in this small-scale dairy study showed that with reasonable management practices, Jersey and Friesian F1 cows could, on average, yield four to five times the milk obtainable from indigenous breeds on-farm. However, at the higher altitudes and more temperate climate of sub-Saharan Africa, crossbred cattle can out-yield indigenous stock fourfold, provided that modest improvements are made in their management and nutrition.

The extreme variability in milk yield in the study, between and within breeds, could be attributed to both genetic and management factors. As test animals were F1 progeny from unselected populations and managed differently (especially feeding) by farmers, variability in milk production was not unexpected. Farmers tended to feed their cows with whatever feed resources they had at their disposal. Thus the feed varied in nutrient content, which would have influenced animal performance. This remains one of the key problems contributing to the difficulties of on-farm livestock research. A comparison between individually and co-operative owned and managed cows shows that individually owned cattle were more productive (Table 1) probably because they receive more attention.

Milk marketing

An efficient milk collection and distribution system to distribute milk from the farmer to the consumer is a critical factor in dairy development. Most milk produced in the rural areas of sub-Saharan Africa is consumed at home or bartered, either as fresh or sour milk. Only in the vicinity of urban markets are milk surpluses processed into dairy products with longer shelf-life (O'Mahony and Peters, 1987). Brumby and Gryseels (1984) reported that in dairy development projects where milk collection and cooling centres had been established to eater for groups of consumers, smallholder participation in dairy development flourished. There is some evidence that traditional marketing systems tend to be low-cost operations and that they are in a position to pay higher prices to producers. Experience has shown that, modern marketing often cannot compete with the traditional systems for raw milk supplies, especially if the official prices are fixed.

In this study marketing and prices of milk have not posed any problem at either the highland or the subhumid sites. Although the highland site is within the radius of the government milk collection scheme, farmers preferred the traditional marketing system in which they sold the product directly to consumers at a negotiable price. This price was generally higher than that offered by the Ethiopian Dairy Development Enterprise. The subhumid site, a distance from the collection scheme, also relied on a traditional marketing system. Studies showed that at both sites milk was sold fresh and supply fell far short of demand resulting in a continuous price rise.

Reproductive performance

Originally it had been planned to use artificial insemination (AI) at the highland project site and natural service at the subhumid zone site. The extension division of the Ministry of Agriculture provided AI services in the highland site area, however, as the project progressed it was found that farmers were not efficient in either heat detection or early reporting. As a result, the timely synchronisation of the insemination of cows was not achieved. The project team, therefore, decided to use natural mating. In both sites, the breeding scheme has not been attractive as farmers have had to rely on borrowed bulls from the research centres. An attempt was made to adopt seasonal breeding and farmers were told to report to the research centre well ahead of a planned mating season in order to be able to borrow breeding bulls on time. This arrangement did not work out smoothly due to farmers' unfamiliarity with heat detection and their not being able to appreciate the importance of timely mating to facilitate conception.

An assessment showed that feed shortages in the dry season resulted in poor fertility. Although grass/legume hay was produced and given as supplementary feed to cows, no farmer produced enough surplus to take the animals through to the dry season. Another factor that negatively affected reproductive performance was the incidence of abortion. A few cows aborted and that seemed to discourage the cooperating farmers; causes of this phenomenon have yet to be determined.

Animal health

Major health problems encountered were liver fluke and ticks at the highland site and bacterial infection diseases and ticks at the subhumid site. Four cows and three calves died at the highland site while no deaths were recorded at the latter site. Causes of mortality were not determined as farmers reported cases too late for laboratory investigation. In addition, a shortage of drugs and acaricides coupled with the lack of fuel for the vehicles to visit the farms affected the smooth progress of the project.

Disease and parasite control remains a major concern, especially for dairy cattle and upgraded cows as they are usually much more susceptible to health problems than indigenous animals. A competent veterinary service is essential to overcome these problems, however, shortage of veterinary staff and the lack of cheap prophylactic and treatment drugs remains significant to effective disease control measures.

Calf growth and cow body condition

Calves were allowed to suckle colostrum for the first 4-5 days after parturition. Thereafter, their access to dams was limited to a few minutes before and after each milking. They grazed around the homestead on natural pasture at Holetta and on Rhodes grass established pasture at Bako and their feed was supplemented with hay and concentrates whenever available.

Growth patterns of the calves born at Bako are indicated in Figure 1. Friesian cross calves consistently grew at a faster rate than the Jersey crosses. Their average on-farm weights at three months (89 kg) and six months (165 kg) were even better than on-station weights where the means of 73 and 114 kg, respectively, were reported (Sendros et al, 1987). This better on-farm performance was achieved probably because calves grazed on improved pasture throughout the dry season as opposed to the calves on-station which were kept on natural pasture.

Figure 1. On-farm calf growth performance at Bako (subhumid mid-altitude climate) up to six months of age.

It is a common phenomenon in a tropical environment for animals to deteriorate in condition during the dry period and regain condition in the wet season. Cow weight fluctuations for six months at Holetta and nine months at Bako are indicated in Figures 2 and 3, respectively. Data was lacking to show the full seasonal fluctuation trends over a one-year period. However, cows tended to continue losing weight until March.

Figure 2. Liveweight patterns of Friesian cows on-farm at Holetta.

Figure 3. Liveweight patterns of on-farm Barka Friesian and Barka Jersey F1 crossbred cows over nine months at Bako (subhumid climate).

Project achievements

Various shortcomings encountered during project implementation hindered achievement of the objectives. Nevertheless, the following were some indicative observations:

· Improved technology transfer: This project revealed that the transfer livestock production technology (breed, health and nutrition) has a significant effect on augmenting the income of farmers under highland and subhumid zone conditions of Ethiopia.

· Farmers' assessment: All farmers were invariably of the opinion that one F1 crossbred cow was superior to indigenous cows and was equivalent in milk production to at least four indigenous milking cows. The number of project and non-project farmers willing to acquire crossbred cows was far in excess of supply.

· Sustained performance: It has been possible to show the possibility of maintaining reasonable performance levels on-farm comparable to on-station levels.

· Multidisciplinary approach: The IAR scientists gained experience in the organisation of multidisciplinary research teams and on-farm research approach.

· Support for on-station research: The procurement of vehicles, farm machinery and laboratory equipment with financial assistance from the project greatly contributed to strengthening on-station livestock research.


The following are among the major problems that hampered project implementation:

· Late arrival of project vehicle and farm implements: This was a major hindrance and to a large extent delayed the start of the project.

· Change in economic policy: A shift in the country from socialism to a mixed economic system of production brought about the demise of farmer cooperatives. Cooperative project activities were totally disrupted. Arrangements had to be made afresh with individual farmers half way through the projects.

· Political unrest: The political turmoil that culminated in a change in government in May 1991 compounded the problem by causing a temporary closure of the research centres that were responsible for the project implementation. This played a key role in reducing the period of the project.

· Fuel shortage: Lack of fuel for a long period in the wake of the ebbing strength of the former government further curtailed the implementation period.

· Shortage of veterinary drugs and acaricides: The inadequate supplies of these in the country affected the health care of project animals.

· Scarcity of protein supplements: In Ethiopia oilcrop cakes are the common source of protein supplement during the dry season. The demand for oil cakes is generally high resulting in exorbitant prices making it difficult for the farmers to obtain these concentrates.

· Data: Farmers in Ethiopia do not have the tradition of keeping farm records. Thus when the research centres were closed due to political instability, they stopped collecting data on milk production and animal weight. This has generally undermined data analysis.

· Long periods of livestock research: One of the major drawbacks of livestock research as opposed to crop research is the relatively longer time span required to achieve results. All the above-stated constraints coupled with the nature of livestock study left much to be achieved in assessing the impact of this project. However, it is hoped that the project will continue for several years.


Beyene Kebede. 1987. The status of dairy research and development in Ethiopia. In: The status of livestock pasture and forage research and development in Ethiopia. IAR (Institute of Agricultural Research), Addis Ababa. pp. 24-33.

Brumby P J and Gryseels G. 1984. Stimulating milk production in milk-deficit countries of Africa and Asia. ILCA Bulletin 19:2-7. (ILCA (International Livestock Centre for Africa), Addis Ababa, Ethiopia.

Brumby P J and Scholtens R G. 1986. Management and health constraints for small-scale dairy production in Africa. ILCA Bulletin 25:9-12. (ILCA (International Livestock Centre for Africa), Addis Ababa, Ethiopia.

Central Statistical Authority. 1990. Area of population density by autonomous, administrative region and awraja. Statistical Bulletin 76. CSA, Addis Ababa.

Goshu Mekonnen. 1989. Central Statistical Authority (CSA), 1990.

Goshu Mekonnen, Teketel Forsido, Alemu G. Wold, Zeleke Dagnachew and Addis Anteneh 1991. The Ethiopian livestock industry: retrospect and prospects. In: Proceedings of the Third National Livestock Improvement Conference, 24-26 May 1989, Addis Ababa. IAR (Institute of Agricultural Research), Addis Ababa, Ethiopia. pp. 9-19.

Gryseels G and de Boodt K 1986. Integration of crossbred cows (Boran and Friesian) on smallholder farms in the Debre Zeit area of the Ethiopian highlands. ILCA Highlands Programme Report. ILCA (International Livestock Centre for Africa), Addis Ababa, Ethiopia.

Jahnke H E. 1982. Livestock production systems and livestock development in tropical Africa. Kieler Wissenchatsverlag yank, Kiel, Germany.

O'Mahony F and Peters K J. 1987. Options for smallholder milk processing in sub-Saharan Africa. ILCA Bulletin 27:2-17. (ILCA (International Livestock Centre for Africa), Addis Ababa, Ethiopia.

Sendros Demeke, Beyene Kebede, Tesfaye Kumsa, Taye Bekure and Hailu G Mariam. 1987. Preliminary result of cattle crossbreeding (European and zebu study. 1. Milk production performance of F1 cows. In: Proceedings of the First National Livestock Improvement Conference, 11-13 February 1987, Addis Ababa. JAR, Addis Ababa. pp. 61-65.

Sendros Demeke, Beyene Kebede, Taye Bekure, Mulugeta Kebede and Hailu Gebre Mariam. 1987. Preliminary result of cattle crossbreeding (European and zebu) study. 2. Growth performance of European and zebu crossbred calves. In: Proceedings of the First National Livestock Improvement Conference, 11-13 February 1987, Addis Ababa. JAR, Addis Ababa. pp. 73-75.

Tesfaye Kumsa. 1991. Livestock system of the western region of Ethiopia. IAR Research Report 12. JAR, Addis Ababa. pp. 10-12.

World Bank. 1984. Appraisal report on the Fourth Livestock Development Project. World Bank, Washington, DC, USA.

Zinash Sileshi and Seyoum Bediye. 1991. Utilization of feed resources and feeding systems in the central zone of Ethiopia. In: Proceedings of the Third National Livestock Conference, 24-26 May 1989, Addis Ababa. IAR (Institut of Agricultural Research), Addis Ababa, Ethiopia.

Previous Page Top of Page Next Page