Smallholder dairy production in Kenya

L. Reynolds, T. Metz and J. Kiptarus

L. Reynolds is an agricultural research and development consultant. His address is: 70 Springfield Crescent, Kibworth Beauchamp, Leicester LE8 0LH, UK.
T. Metz and J. Kiptarus can be contacted at the National Dairy Development Project, Kenya Ministry of Agriculture, Livestock Development and Marketing, Hill Plaza, PO Box 34188, Nairobi, Kenya.

PRODUCTION LAITI�RE EN PETITE EXPLOITATION AU KENYA

Le pr�sent article d�crit l'�tat actuel de la production laiti�re en petite exploitation au Kenya et �voque la n�cessit� d'une expansion du secteur, de fa�on � r�pondre � la demande d'une population de plus en plus nombreuse. Il attire l'attention sur certains perfectionnements techniques prometteurs qui devraient aider les producteurs � faire face � l'�volution anticip�e des besoins.

PRODUCCION LECHERA DE PEQUE�OS PROPIETARIOS EN KENYA

En el presente art�culo se describen los logros conseguidos en la producci�n lechera por parte de peque�os propietarios en Kenya y se examina la necesidad de ampliar la industria para satisfacer la demanda futura de una poblaci�n en aumento, prestando especial atenci�n a las novedades tecnol�gicas con posibilidades que permitan a los productores satisfacer las necesidades previstas.

Before Kenya gained its independence in 1964, commercial dairy production was the sole preserve of white farmers and small-scale farmers were not allowed to own dairy cattle. Twenty-five years later, Kenya is broadly self-sufficient in milk and dairy products, with over 400 000 smallholder dairy farmers producing more than 70 percent of the marketed milk in the country. This article examines the dairy sector and the developments needed to meet future demand.

NUMBERS AND DISTRIBUTION OF DAIRY CATTLE

There are slightly over three million dairy cattle in Kenya, mainly black and white Friesians, Ayrshires, Channel Island breeds and various crosses, concentrated in high-potential areas near to centres of population (Table 1). The Central Province has the highest concentration of dairy cattle (106/km²), although by far the largest number of animals (1.7 million) are found in Rift Valley Province. These two provinces also have the highest numbers of dairy cattle per head of population and are the only areas fully capable of meeting local demand and providing surplus produce for other parts of the country. The high-potential highland areas, where temperature is moderated by altitude, receive a greater and more reliable rainfall than medium-potential areas that are predominantly found at lower altitudes. These factors largely explain the current distribution of dairy farming in Kenya, as forage production is related to rainfall, disease risk is reduced at higher altitudes and market demand arose from centres of population located in the highlands.

1
Numbers and distribution of dairy cattle in Kenya, 1991
Nombre et r�partition des bovins d'�levages laitiers au Kenya, en 1991
N�mero y distribuci�n de los vacunos lecheros en Kenya, 1991

Source: Jaetzold and Schmidt (1982); Kenya Ministry of Livestock Development, 1991

The expansion in the distribution of dairy farms after independence was initially possible because of the availability of heifers and cows from the commercial farming sector and the subsequent use of imported and locally produced semen for artificial insemination. Initially, dairy cattle on small-scale farms were grazed, but as farm size decreased and grazing was less readily obtained, cultivated forages were adopted.

PRODUCTION SYSTEMS

A variety of production systems are employed by smallholder dairy farmers, ranging from stall-fed cut-and-carry systems, supplemented with purchased concentrate feed in areas of high population density where extensive systems are not possible, to free grazing on unimproved natural pasture in the more marginal areas. Exotic dairy breeds tend to be kept in stall-feeding units, while free-grazing dairy animals are more likely to be cross-bred cattle.
In order to meet market demand for milk from an expanding and increasingly urbanized population, the National Dairy Development Project (NDDP), under the Kenya Ministry of Agriculture, Livestock Development and Marketing, has been promoting intensive, stall-feeding units for more than a decade. Farmers are encouraged to establish Napier grass (Pennisetum purpureum), construct a stall-feeding unit and draw up a farm budget. Where appropriate, farmers are assisted in approaching the Agricultural Finance Corporation for loans to buy pure- or cross-bred in-calf dairy heifers. Regular visits from specialized dairy extension workers reinforce extension messages to assist farmers in the management of the enterprise. The NDDP, with over 4 500 registered dairy farmers, is active in the high-potential areas of Central, Eastern, Western, Nyanza and Rift Valley Provinces, and in the medium-potential areas of the Coast Province, adjacent to Mombasa. It is estimated that 420 km² of cultivated grass has been established, with two-thirds of this in the Central Province, all of which is used for cut-and-carry feeding (Kenya Ministry of Agriculture, 1991). In the high-potential areas only 10 percent of the agricultural land remains under natural pasture, compared with 50 percent in medium-potential areas, indicating that there is limited room for expanding agriculture in the high-potential zone.
A direct comparison of independent dairy and NDDP farms was made in Kilifi District of Coast Province between 1985 and 1987 by Leegwater, Ngolo and Hoorweg (1991). Kilifi is representative of the medium-potential areas into which dairying is expected to expand in the future. Rainfall and soil fertility are lower in Coast Province, and disease risk is higher than in the highland areas. The average size across all dairy farms in Leegwater, Ngolo and Hoorweg's survey was 11 ha, significantly larger than the NDDP farm average of 4.4 ha in Kilifi (van der Valk, 1992). This in turn was larger than the national average for NDDP farms of 2.7 ha (NDDP, 1993). Leegwater, Ngolo and Hoorweg (1991) found that farmers associated with the NDDP have greater access to resources (land and capital) than independent dairy farmers.

FEEDS AND FEEDING

Present estimates, based on farm size, land allocation and ecological potential, indicate that smallholder dairy farmers can only produce 70 percent of the feed required from their own resources. Nationally, within a survey group of 2 716 NDDP farmers, 21 percent of farmland was planted to Napier grass. While almost all NDDP farms have planted Napier grass, 41 percent also have natural pasture on-farm (0.89 ha). A small proportion of farms (11 percent) have Napier grass-legume mixtures and 30 percent have planted fodder trees (usually Leucaena leucocephala), but the areas are small. Other roughage sources commonly used include maize stover, banana stems and roadside grass (a generic term for natural pasture from a variety of sources) on 76, 62 and 34 percent of farms, respectively. In addition, 29 percent of farmers report that they regularly purchase fodder. The NDDP estimates that on-farm Napier grass production ranges from 5 to 22 tonnes/ha, but extension visits indicate that only two-thirds of the farms maintain adequate amounts of fodder in the feed trough, preventing ad libitum access to roughage for lactating cows. Less attention is paid to feed for young stock and dry cows, who are often tethered outside the stall unit on natural pasture.
There is clearly a need to improve forage availability. Farmers are reluctant to allocate more land to forage production, but increased yields per unit area could be achieved with greater fertilizer input. Most farmers use manure for more than one type of crop, but emphasis is placed on forage crops. Over 75 percent of farmers return at least half the manure from the stall unit to Napier grass plots. There appears to be little scope for improving the level of manure use, but qualitative changes to increase the efficiency of storage and application may be feasible. At present, less than 40 percent of farmers use inorganic fertilizer on forage plots, even though the financial returns are attractive.
Sales of commercial dairy feed in 1991 totalled around 60 000 tonnes, equivalent to only 0.2 kg/cow for each day of lactation. Other supplements are also used, such as brans and brewer's wastes which increase the total additional feed offered to dairy animals. NDDP farmers are reported to offer about 3 kg of all daily supplements to their lactating cows, sufficient to increase milk production by 50 percent over the level expected from a purely grass diet. Even though other supplements make up part of the total, NDDP farmers make more use of commercially prepared dairy feeds than do independent small-scale dairy farmers.
The survival rate of calves is affected by the value placed on them by farmers. Farmers are advised to feed 415 kg of milk plus 70 kg of early weaner pellets to calves before weaning. As can be seen in Figure 1, female calves are offered more milk than are male calves. Female calves are valuable for herd replacements and for sale to other farmers. The proportion of male calves receiving milk is surprisingly high, considering that only 7 percent of NDDP farms contained a bull over 15 months of age. The market price for meat will not justify rearing a male calf on milk for slaughter, which suggests that farmers may be raising males as part of a risk-avoidance breeding strategy.

1
Proportion of calves receiving milk, by age
Proportion de veaux au pis, par mois d'�ge
Proporci�n de terneros que reciben leche, por meses de edad

REPRODUCTIVE PERFORMANCE AND ARTIFICIAL INSEMINATION

The NDDP survey in 1992 found that the median calving interval was 426 days. Artificial insemination (AI) was used in 54 percent of services, with just over half resulting in a pregnancy. The efficiency of AI on zero-grazing farms is constrained by poor heat detection and the unreliability of delivery of the service. In the face of budgetary difficulties, with staff costs accounting for over 80 percent of the total extension budget, charges were introduced for services with the eventual aim of privatization. Initially, charges were set at K Sh 4 per insemination, although private veterinarians were charging K Sh 30 (de Haan and Bekure, 1991). However, the real cost varied from K Sh 20 to K Sh 250 depending on location, and there was little improvement in the efficiency of the government service after the introduction of charges. In Machakos District, where dairy cattle numbers increased by 11 percent annually during the 1980s while cattle numbers overall remained static, farmers increasingly turned to natural service (Akelo-Ogutu, 1991). Hence, access to good-quality bulls remains a critical factor for the future of the dairy industry, even in high-potential areas where the provision of AI services is most apparent.
There is evidence from Coast Province (Leegwater, Ngolo and Hoorweg, 1991) that the mortality rate of calves on NDDP farms is lower than on non-NDDP smallholder dairy farms (12 versus 28 percent). In the NDDP survey, between 40 and 50 percent of heifers over 15 months of age were pregnant, but the median age for a first pregnancy was 30 months. Hence, farmers raising their own heifers face a lengthy unproductive period before cash can be obtained from milk sales.

DISEASE CONTROL

Exotic and cross-bred dairy cattle are the source of most of the marketed milk in Kenya, but they are more susceptible to disease than local zebu animals. In particular, East Coast Fever, a tick-borne disease, can have a severe impact on dairy cattle. Legislation requires cattle in high-potential areas to be regularly dipped to control ticks. In 1990, there were a total of 4 800 functioning dips operated by the Department of Veterinary Services. Charges for dipping were steadily raised to reduce the burden on the Treasury. Eventually, a policy of local management and full cost recovery was implemented so that, by mid-1992, almost 3 000 dips were supposed to be under the management of local dip committees, with initial short-term support from the government and the Department of Veterinary Service. It is estimated that a regular throughput of over 200 animals is needed for a dip to be economic, but numbers have fallen short even where dipping is required by law. While farmers were willing to pay for cross-bred and exotic dairy cattle to be dipped, they were less convinced of the cost effectiveness of treatment for zebu animals. The effectiveness of dipping under local management was also questionable, and farmers increasingly turned to hand-spraying at home, as evidenced by the increase in sales of small packs of acaricide.

MILK PRODUCTION, CONSUMPTION AND MARKETING

Milk production and offtake from different classes of livestock are shown in Table 2. Marketed milk and dairy products arise exclusively from dairy cattle, although zebu cattle and camels are important sources of milk for home consumption.

2
Milk production and offtake in Kenya, 1991
Production laiti�re et pr�l�vement enregistr�s au Kenya, en 1991
Producci�n y extracci�n de leche en Kenya, 1991

Source: Adapted from FAO (1985); Kenya Ministry of Livestock Development (1991).

The intensification of milk production is greater on NDDP farms than on other dairy farms. NDDP farmers' dairy herds are one-third the size of those of other dairy farmers, but they produce twice as much milk. Non-NDDP farmers had only one-third of their cows in milk, compared with 85 percent in NDDP herds. National daily milk yield averages 7.9 kg/day on NDDP farms, almost twice the estimated national average on non-NDDP farms (van der Valk, 1992; NDDP, 1993). Peak yields of 11.3 kg/day occur in the second month post-partum and remain relatively stable at 5 to 6 kg/day even in extended lactations after 12 months post-partum (Figure 2).

2
Average milk yield according to the stage of lactation of NDDP farm cows
Production laiti�re moyenne selon le stade de lactation des vaches dans les exploitations relevant du NDDP
Producci�n media de leche seg�n la fase de lactaci�n de las vacas en las granjas del NDDP

Although NDDP farmers are recommended to dry cows off after a ten-month lactation, a significant proportion of animals are milked for extended periods (Figure 3). Total milk yield on NDDP farms reaches 2 510 kg in ten months and 2 695 kg in a 12-month lactation. NDDP farmers retain a lower proportion of the total production for home consumption than do independent dairy farmers but both groups retained similar amounts, i.e. 1 to 2 kg/day (Huss-Ashmore, 1992).

3
Distribution of NDDP lactating cows according to stage of lactation
R�partition des vaches allaitantes selon le stade de lactation dans les exploitations relevant du NDDP
Distribuci�n de las vacas lactantes por fases de lactaci�n en las granjas del NDDP

Arguably, market demand for milk and the ability of farmers to earn an attractive return from the sale of dairy products are the greatest stimuli for expansion of the dairy sector. Demand for dairy products is related to income, and hence the collection, processing and distribution system is largely organized to supply urban consumers who have greater purchasing power than rural communities. As shown in Table 3, during 1991 urban consumption of milk and dairy products was estimated to be 70 kg per caput/year, taking 52 percent of total marketed output of dairy produce. Unmarketed milk by definition is consumed by rural households with dairy cattle, zebu cattle and camels. The mean per caput annual consumption on dairy farms was 59 kg but consumption across all other rural households averaged only 38 kg per caput.

3
Annual consumption of milk and dairy products across different sections of population in Kenya, 1991
Consommation annuelle de lait et de produits laitiers dans diff�rentes cat�gories de la population au Kenya, en 1991
Consumo anual de leche y productos l�cteos en los distintos sectores de la poblaci�n en Kenya, 1991

Source: Adapted from FAO (1993).

Urban areas, where many international tourist hotels are located, are particularly important markets for butter, cheese and other dairy products. About 80 percent of processed dairy products (other than liquid milk) are consumed in the Nairobi and Mombasa areas, although they account for less than 15 percent of the national population.
A guaranteed market for producers was offered by the Kenya Cooperative Creameries (KCC) which held a monopoly for many years. Prices were set annually by government decree, so producers knew how much they would get from milk sales. In 1992 a more open market was permitted, and prices for milk and dairy products were decontrolled outside the main markets of Nairobi and Mombasa. Where possible, milk is sold locally at a higher net price than that received from the KCC, but in high-potential areas there is a surplus of milk over local demand. In high-potential areas, dairy cooperatives play an important role in marketing and in organizing collection and delivery to the KCC. Producers have to meet the cost of transporting milk to the KCC. Currently, there are about 700 dairy cooperatives handling 80 percent of the milk that passes through the KCC.

EFFECTS OF DAIRYING ON THE RURAL COMMUNITY

Two socio-economic studies have been undertaken in Coast Province, where small-scale dairy production is less well established than it is in the central highlands. The standard of living of the population of Coast Province is lower than the Kenyan average, and the introduction of dairying by NDDP was expected to bring both nutritional and economic benefits for dairy farming households. Huss-Ashmore (1992) found that livestock farmers (NDDP and independent) in Coast Province had considerably more land, larger households and higher incomes than the general coastal population. The customers of NDDP farmers also had relatively high incomes, primarily owing to off-farm employment. Adoption of intensive dairying was associated with households that had a steady cash income, either from off-farm employment or from cash crops. Two divergent strategies were employed in the NDDP households: one group used family labour for the dairy enterprise and relied on cash cropping for a substantial proportion of farm income; the other group substituted cheaper hired labour to work with the cattle, and directed family labour to off-farm employment, thereby spreading risk. NDDP farms employing labourers had more cows than those relying on family labour. They also had a higher total income and almost as much income arising from off-farm employment as the overall income on family-only dairy farm households (Table 4). Thus, the wealthier households, comprising a large proportion of the small-scale dairy sector, created rural employment opportunities for others.

4
Characteristics of NDDP households in Coast Province, with and without employed labour
Caract�ristiques des m�nages relevant du NDDP situ�s dans la Province c�ti�re, avec et sans main-d'œuvre embauch�e
Caracter�sticas de los hogares del NDDP en la provincia costera, con y sin empleo de mano de obra

¹ K Sh/year (1 US$ = K Sh 28 in 1991).
Source: Leegwater, Ngolo and Hoorweg (1991).

Studies in the Kenyan highlands have indicated that, with increasing commercialization, the control of income from milk sales shifts from women to men, in a similar manner to income from horticulture and other traditional female enterprises (Huss-Ashmore and Curry, 1992). Income accruing to women is more likely to be used to provide food for the households. While undernourishment is prevalent in children throughout Coast Province, better nutritional status was associated with smaller household size, higher income and steady sources of income such as wage employment or dairying. Nutritional status was sharply higher for families of labourers employed on dairy farms compared with the general population.

GOVERNMENT POLICY

Government policy, as expressed in the National Food Policy (Kenya Ministry of Agriculture, Livestock Development and Marketing, 1993a) and the Kenya Dairy Development Policy (Kenya Ministry of Agriculture, Livestock Development and Marketing, 1993b) is to maintain a position of broad self-sufficiency, establish reserve stocks to ensure security of food supply and ensure an equitable distribution of foodstuffs to each section of the community. The dairy sector is expected to become more intensive and its output to rise through facilitated access to appropriate production technologies and inputs.
The processing and marketing system will become more efficient and self-sustaining through policies designed to encourage competition. In the veterinary field, AI and clinical services are being privatized and government financial support for the running of dips is being withdrawn. Dairy cooperatives are being encouraged to increase the range of services they offer to their members, and to work more closely with private veterinarians.
To assist in this changeover, the government is to emphasize livestock extension within the dairy sector, improve access of small-scale producers to credit facilities, lower taxes on a range of imported inputs and stimulate oilseed production within Kenya to increase availability of supplementary feed. Price structures have been liberalized and the KCC is to be reorganized so as to respond better to the needs of the dairy industry. Other locally based processing and marketing organizations will be encouraged to compete with the KCC. The importance of an adequate all-weather rural road network is recognized, in order to enable new areas in high-potential zones to deliver their perishable dairy produce to market.

ABILITY OF THE DAIRY SECTOR TO MEET PREDICTED DEMAND

The population of Kenya is 26 million, 23 percent of whom live in urban areas. By 2003 the population is predicted to rise to more than 36 million, with urban numbers increasing faster than the rural population, partly because of migration in search of employment (Winrock International, 1992). Elasticity of demand for livestock products is three to five times higher than for cereals, so if real incomes rise, the demand for livestock products increases faster than demand for cereals. The projected demand for marketed liquid milk in 2003 is shown in Table 5 based on three growth scenarios.

5
Demand estimates for marketed liquid milk in Kenya
Demande estim�e de lait liquide commercialis� au Kenya
Estimaci�n de la demanda de leche l�quida comercializada en Kenya

Source: FAO (1993).

Predictions of the number and distribution of dairy cattle in 2003 are shown in Table 6, based on a 1.5 percent annual increase in animal numbers. Assuming small improvements in reproductive performance and milk yield for the national dairy herd, milk offtake could be sufficient to meet the high growth rate and demand scenario. However, to reach this output, an intensification of production will be needed with greater reliance on supplementary feeding, over and above feedstuffs currently produced on-farm. As population increases and demand for land rises, farm size will shrink. To increase output, farmers will be forced to produce more forage from less land and to use more purchased supplements. As indicated earlier, greater use of manure and/or inorganic fertilizers will increase per hectare forage yields. Increased supplementary feeding will depend on the reliability and availability of supply as well as on the attractiveness of profit margins. Kenya could be self-sufficient in oilseeds, of which protein concentrates are a by-product after the extraction of oil for human consumption but, at present, cooking oil is derived from imported palm oil and very little oilseed is grown in the country. The continued importation of oilseed cake for animal feed depends on the availability of foreign exchange - local sourcing of supplies would be less dependent on external factors.

6
Predicted numbers and distribution of dairy cattle in Kenya for 2003
Pr�visions concernant le nombre et la r�partition des bovins d'�levages laitiers au Kenya, en 2003
N�mero y distribuci�n previstos de los vacunos lecheros en Kenya para el a�o 2003

Source: Adapted from Kenya Ministry of Livestock Development (1991).

PROMISING RESEARCH ACTIVITIES

When the Kenya Agricultural Research Institute (KARI) listed its research priorities, dairying came top of the list because of the contribution to national GDP, national food security, likely improvement in output, the probability of research success and of adoption by farmers, employment opportunities and an equitable distribution of benefits to broad sections of the community (KARI, 1991).
The needs of dairy farmers are dependent on location. In high-potential areas, solutions are needed for problems of intensification. In medium-potential areas, such as in Coast Province, attention is focused on reducing risks - disease risk, especially of East Coast Fever and trypanosomiasis; risk to forage supplies, especially from cultivated forages with poor rainfall - and on optimizing breeding systems where farms are too dispersed for AI services to be economically viable. Particular attention is given to adaptive research to ensure that promising technology is suited to local circumstances. KARI, in collaboration with the International Livestock Research Institute (ILRI) and the extension service within the Kenya Ministry of Agriculture, Livestock Development and Marketing are cooperating with farmers to conduct on-farm research in parallel with regional on-station research.
Forage from grass-legume combinations for cut-and-carry feeding have shown particular promise in terms of dry matter yield and quality and in raising milk production on-station. They have also proved acceptable to farmers (Murithi et al., in press; Muinga, Thorpe and Topps, 1992). These agronomic and feeding studies have demonstrated a sustainable production system, appropriate to smallholder resources for both high- and medium-potential areas of the country. However, while the Napier grass species is productive under good rainfall conditions, it is less persistent under frequent harvesting when rainfall is low and sporadic. Further research is needed to produce adoptable and sustainable, intensive forage production for dairy farms in medium-potential areas.
In Africa, Kenya is at the forefront of the move towards privatization and market liberalization for the dairy industry. This will be a fruitful area of research, providing valuable information for other countries considering similar action.

CONCLUSIONS

Farmers will accept a degree of risk in adopting production systems when they are assured of a profitable market outlet. More affluent farmers, which in the small-scale sector means those with regular off-farm income, can accept higher levels of risk. Research to reduce risk, rather than to maximize production, will widen the spread of benefits in the community. Small-scale dairy production offers a route to increase rural employment and improve household welfare. Technology to intensify production in high-potential areas, where the majority of dairy farms are located, has been developed by KARI and the NDDP. Collaborative on-farm research, involving researchers, extension staff and farmers, to reduce risk in parallel with on-station studies will provide technology options to expand dairying in medium-potential areas.
Organization and infrastructural constraints are also important. Privatization and market liberalization are designed to alleviate problems caused by limitations in public finances. However, expansion of the dairy industry into medium-potential areas must be supported by government-provided services until there are sufficient producers for private sector services to be financially viable.
Kenya should be able to meet projected demand for milk and dairy products, and the relative success of the strategies employed will provide lessons for other countries wishing to develop their own dairy sector. 

Bibliography

Akelo-Ogutu, A.C. 1991. Livestock production. In M. Tiffen, ed. Environmental change and dryland management in Machakos District, Kenya, 1930-1990, p. 45-94. Overseas Development Institute Working Paper No. 55. London, ODI.
de Haan, C. & Bekure, S. 1991. Animal health services in sub-Saharan Africa. Initial experiences and alternative approaches. Technical Paper No. 134. Washington, DC, World Bank.
FAO. 1985. Population supporting capacity study for Kenya - Animal production study. A proposed model for estimating the human food supply produced by animals in Kenya, by P.H. Sloane. Rome. 23 pp.
FAO. 1993. Dairy Development Project. Report No. 34.93 CP-KEN 33. Rome.
Huss-Ashmore, R. 1992. Nutritional impact of intensified dairy production: an assessment in Coast Province, Kenya. ILRAD Technical Report No. 1. Nairobi, ILRAD. 35 pp.
Huss-Ashmore, R. & Curry, J.J. 1992. Impact of improved livestock disease control on household diet and welfare: a study in Uasin Gishu District, Kenya. ILRAD Technical Report No. 2. Nairobi, ILRAD. 97 pp.
Jaetzold, R. & Schmidt, H. 1982. Farm management handbook of Kenya. Vol. II, Natural conditions and farm management information, Parts A, B and C. Nairobi, Kenya Ministry of Agriculture.
KARI. 1991. Kenya's agricultural research priorities to the year 2000. Nairobi, KARI. 110 pp.
Kenya Ministry of Agriculture. 1991. Annual Report. Nairobi.
Kenya Ministry of Agriculture, Livestock Development and Marketing. 1993a. Sessional Paper No. 3 of 1993 on National Food Policy. Nairobi. 53 pp.
Kenya Ministry of Agriculture, Livestock Development and Marketing. 1993b. Kenya Dairy Development Policy: a strategy towards the development of a self-sustaining dairy sector. Nairobi. 25 pp.
Kenya Ministry of Livestock Development. 1991. Annual Report. Nairobi.
Leegwater, P., Ngolo, J. & Hoorweg, J. 1991. Dairy development and nutrition in Kilifi District, Kenya. Report
No. 35/1991. Nairobi, Food and Nutrition Planning Unit, Kenya Ministry of Planning and National Development and Leiden, the Netherlands, African Studies Centre. 134 pp.
Muinga, R.W., Thorpe, W. & Topps, J.H. 1992. Voluntary feed intake, liveweight and lactation performance of crossbred dairy cows given ad libitum Pennisetum purpureum (Napier grass var. Bana) supplemented with leucaena forage in the lowland semi-humid tropics. Anim. Prod., 55: 331-337.
Murithi, J.G., Reynolds, L., Thorpe, W. & Tayler, R.S. Fodder production based on Pennisetum purpureum and Leucaena leucocephala in an alley farming system. In Proceedings of the 2nd International Alley Farming Conference. Ibadan, Nigeria, IITA. (in press)
NDDP. 1993. Results of the farm survey in all districts. NDDP/ME/93/047. Nairobi, Kenya Ministry of Agriculture, Livestock Development and Marketing. 23 pp.
van der Valk, Y.S. 1992. Review report of the surveys with the Dairy Evaluation and Advice Form during 1991. NDDP/M-46/246. Nairobi, Kenya Ministry of Agriculture, Livestock Development and Marketing. 46 pp.
Winrock International. 1992. Assessment of animal agriculture in sub-Saharan Africa. Morrilton, Ark., USA. 125 pp.