Botswana, with nearly 80% of its total area in natural rangeland, supports around 3.3 million cattle in addition to a sheep and goat population of approximately 1.8 million (1977 estimates) 1. This yields a 5 to 1 ratio between LUs and human population, the highest in Africa. The main emphasis in livestock research and development in the past has been on commercially oriented beef cattle production systems. As a result, exports of animal products, mainly in the form of fresh meat, account for 20 to 25% of the country's total exports. Before the recent expansion of the mining industry, the livestock sector was even more important to the economy 2.
1 Livestock constitute the principal agricultural resource, contributing 80% of the value added by agriculture (ILCA, 1979).
2 Before 1974, livestock exports accounted for over 50% of export earnings (ILCA, 1979).
Despite its high livestock population. Botswana imports relatively large quantities of milk products. Dairy imports in value terms increased from P 1.3 million in 1975 to P 7.6 million in 1979, reflecting a substantial increase in both the price and the volume of imports 3. The main sources of dairy products are the Republic of South Africa and Zimbabwe. The prices of imports from both sources are expected to increase rapidly.
3 One P equals approximately US$ 1.20.
The high increase in the imports of dairy products has resulted mainly from rapid urbanization 4, the attractiveness of imported products to urban consumers (cost, convenient packaging, hygienic products) and the difficulties (cost, perish-ability, product presentation) of getting surplus milk (if any) from rural areas to urban markets. The demand for milk in the towns is growing at a rate of some 20 to 25% per year.
4 The average annual population growth rate in the main towns is about 8% and that of the major villages about 6%. For Gaborone alone the growth rate is 15% (Fielding, 1978a).
Milk consumption in the rural areas is usually seasonal. Partial milking of indigenous cows is traditionally practiced, providing more than adequate milk supplies for household consumption during the rainy season. Most of the milk sold in village areas is in the form of local milk products (madila). Another important activity has been cream production for both domestic and export markets. However, the importance of this occupation has declined in most areas, probably due to the more favourable price of beef, the recognition that milk deprivation is responsible for poorer calf growth and the difficulties of obtaining farm labour 5.
5 Cream exports from Botswana in 1976 were only 55% of their 1972 level in value terms, and in volume terms only 37.7% (Fielding, 1978a).
In view of this situation, there is considerable interest at the national level in the development of the dairy industry. Due to the vastness of the country, efforts so far have concentrated on the development of small-scale dairies in or near each major town to supply local needs. However, because of the considerable importance attached to the viability of the beef cattle industry, an important policy question has been the extent to which indigenous and crossbred cows can be used for dual-purpose production without impairing their performance as beef producers 6. It is frequently contended that increasing milk offtake prejudices calf growth and should be discouraged. However, provided the value of milk obtained exceeds the loss of value in calf growth, overall productivity may be in favour of milking at some level. Thus, with the existing and rapidly growing milk markets in Gaborone and other towns and in view of the present deficit in milk production, the formulation of an optimum milking policy based on economic considerations is necessary.
6 Recently, APRU has initiated field experimentation with a dual-purpose herd at Broadhurst Farm, near Gaborone. The purpose of this experimental herd, consisting of about 40 cows, is to investigate the technical and economic factors in the partial milking of T and ST cows.
Several studies have examined the economic feasibility of dairying and the technical and institutional constraints in the development of the dairy industry (Silichena, 1976; Fielding, 1978a, 1978b; Rose, 1978; APRU, 1980a). In addition to the substantial foreign exchange savings, expanded rural milk production would bring other even more important benefits, such as higher incomes to livestock producers and the creation of jobs in milk production, processing and distribution. However, these conclusions are reached from simple static cash-flow analyses based on certain assumptions on livestock productivity and potential milk offtake rates. The dynamic effects of implementing such alternative production regimes are not elaborated in these studies. ILCA (1978) provides an indication of these dynamic effects and an overall idea of the direction of short-and long-term costs and benefits. That study suggested that milking can result in short-term gains, however, at the expense of a longer term reduction in the capital value of the herd due to increased calf mortality, deferred female maturity and extension of intercalving intervals. The exact nature of the economic trade-offs between milk and meat production are not elaborated, nor is the sensitivity of the results at different milk off take rates, supplementation levels, and input and output prices.
The objectives of the present study are to evaluate the potential of two genotypes (T and ST crosses) as milk producers, to determine the viability of alternative production options within a dynamic system, and to establish the economic trade-offs between milk and meat production under various levels of feed supplementation. The study area is the Masiatilodi and Matlolakgang ranches, both west of Gaborone, a brief description of which is given in Section 2. The analysis employs a dynamic and stochastic cattle simulation model (Konandreas and Anderson, 1982). Section 3 provides an overview of this general model and its components, and gives complete details of the model's driving variables. The validation of the simulation model in the study area is presented in Section 4. Section 5 presents the response of key production variables to various interventions on the input and output sides of the herd enterprise and compares the simulated performance of the two genotypes on the basis of individual production traits. The overall comparison between the two genotypes, for the range of interventions and on the basis of the economic criteria, is presented in Section 6. Optimum input and output levels are estimated as well as the sensitivity of the results to different pricing structures. Finally, Section 7 presents overall conclusions and policy recommendations.
