W. S. BoitumeloDivision of Animal Production and Range Research
Department of Agricultural Research
P.B. 0033, Gaborone, Botswana
Introduction
Development of the research programme
Introduction of the technical package
Summary of research results
Evaluation of the research programme
Impacts of the research programme
References
|
Summary The increase in milk production by peri-urban small-scale farmers in Botswana became a major objective of the Ministry of Agriculture during the National Development Plan 6 (NDP6) (198591). it was envisaged that increased production in this sector would not only increase local supplies for home consumption but would also lead to surpluses available for sale. To support the need for increased milk supplies and to examine the potential for small farm production the Animal Production Research Unit (APRU) and the Department of Agricultural Research established a project to investigate the problems of milk production by small-scale communal cattle owners around urban areas. Lack of adequate nutrition, especially during the dry-season, was identified as a major constraint to sustainable milk production in small-scale farm conditions (1983-84). Increased milk production in a small-scale setup requires investigations into potential feeding strategies using locally available feed resources. Considerable quantities of crop residues are produced and can be harvested for dry season feeding of lactating and in-calf cows. On small farms, the nutritional quality of crop residue-based diets can be improved by either providing a high quality legume fodder (high protein) and/ or combining with other available milling byproducts such as sorghum bran (moroko). There. fore, the research emphasis has been to encourage farmers to harvest and conserve all their crop residues and to integrate fodder crops into the farming systems and establish a practical feeding programme based on planted fodder and crop residues. Problems which have been isolated are mainly related to feed procurement. Despite this problems, fodder production and conservation of crop residues such as sorghum bran for dry-season feeding, are a major activity of the 46 farmers participating in the dairy production project. The feeding system for in-calf and lactating crossbred and Tswana cows are based on local home-grown feed resources. The dry season feeding strategy using high protein farmer-grown lablab hay in association with crop residues and by-products has been a significant intervention in small scale dairy farms. Despite animal health problems, Simmental crossbred cows are promising as a dairy breed for small-scale farms. The management system is less intensive than for pure bred dairy cows and retains flexibility in terms of increased milk yield and sale of animals for beef. |
Botswana is situated in the centre of the southern Africa plateau It is a land-locked country of 570,000 km² with a mean altitude of about 1000 metres above sea level. It has a semiarid, continental climate with an average rainfall of 450 mm which is both erratic and unevenly distributed. The rains occur during the summer season as rain storms. The weather ranges from tropical in the north to semitemperate in the south, where winters are dry and temperatures vary from day to day with frequent frost. The maximum temperature for the hottest month (December) is about 35°C and the minimum temperature for the coldest month (July) is 3°C; occasionally temperatures can fall below 0°C.
More than half of the country is covered with shrub and tree savannah which deteriorates into a low shrub savannah in the Kgalagadi Desert. The national rangeland is mainly suitable for extensive grazing of ruminant livestock and wildlife.
The main resource of Botswana apart from its people, has been its cattle of which there were three million in 1982 (Ministry of Agriculture, 1986). The cattle population is now estimated at 2.6 million (Ministry of Finance and Development Planning, 1991). About 84% of the national herd is kept on communal land, whilst 16% is kept on commercial ranches. The cattle industry is a major contributor to exports; beef export earnings more than cover the cost of basic cereal imports. Although the contribution of the agricultural sector has been declining since Independence, the sector still remains an important source of food, income, employment and capital for the majority of the population.
The common breed used for both milk and meat production in Botswana is the indigenous Tswana breed. However, there are few commercial dairy farmers who keep exotic breeds such as the Holstein-Friesians, Jerseys, Simmental, Brown Swiss and some crossbred animals. In 1988, it was estimated that commercial dairy farms had a total of 2750 milk cows of which 1340 produced an average of 4.1 million litres milk per year (Animal Production Division, 1989). This only met 26% of the local demand. The deficit is met by imports from South Africa and sometimes from Zimbabwe.
A research programme was initiated on dairy production to develop appropriate dairy production systems for both small and large scale producers. Traditional small-scale milking entails overnight separation of cows and calves and morning milking using suckling to stimulate milk let down. It was found necessary therefore, to determine the potential of such a system for increasing milk production and to develop appropriate management packages for the system.
The major objective of the programme was to increase milk production by small-scale communal pert-urban cattle owners.
The increase in milk production by peri-urban small-scale farmers in Botswana has been the major objective of the Ministry of Agriculture in the past six-year NDP 6 (Ministry of Finance Development Planning, 1985). It was thought that increased production in this sector would not only increase local supplies for home consumption but could also lead to surpluses available for sale to the rapidly increasing urban population. With increased urbanisation, the demand for milk and milk products increased by 15% during the NDP 6. In 1985, Botswana imported 80% of its fresh milk needs and the entire requirement for other milk by-products. Today the country spends about 16 million Pula (an equivalent of US$ 8 million) annually on these imports (Ministry of Finance Development Planning, 1991). This is a food security issue and cannot be left unchecked. In response, investigation programmes into milk production systems were initiated in 1979 (APRU, 1982).
The development of a milk industry in Botswana is affected by both technical and non-technical constraints. Cattle production depends on the availability of adequate feed supplies and good management. Recurring drought, unreliable rainfall and poor soil fertility cause feed supplies to fluctuate both in quantity and quality. Protein and mineral content of the natural pastures is generally low, especially during the dry season. Conservation and storage of feed from the time of its maximum availability to the time of its use could be a useful strategy. Though crop residues are available after grain harvest, the use of these plant materials as livestock feed by the small-scale farmers is constrained by lack of transport, labour and the long distance between homestead and fields.
Improvement of the Tswana breed for milk production has been done through crossbreeding by using performance tested Simmental bulls. The major problems affecting the provision of this service to livestock producers are: inadequate artificial insemination facilities; insufficient numbers of selected bulls for interested farmers; and the shortage of qualified personnel at both technical and professional levels.
Livestock diseases have been controlled quite effectively through strategies developed by the Department of Animal Health and Production. Despite this, problems such as the lack of veterinary requisites, capital, transport and qualified staff to conduct research and field investigations pose problems in the control of livestock diseases.
Surplus milk produced in the traditional sector during the wet season is fermented to make madila (sour milk). The long distances involved, lack of transport and of proper cooling facilities preclude the sale of fresh surplus milk in urban areas.
A small-scale dairy research programme was conceived and implemented to diagnose the major constraints contributing to low milk production, followed by on-farm testing of alternative technologies. An on-farm dairy research project was started and located in the traditional farming area in the Gaborone Region. The aims of the project were (1) to examine the potential for increasing production of milk and milk products; (2) to test management technologies developed on-station to improve milk production; and (3) to evaluate the economic viability of these interventions. This project is jointly financed by the Government of Botswana and the International Development Research Centre (IDRC) of Canada. To date, there are 46 farmers participating in the project and all of whom practise mixed crop-livestock production within the Gaborone pert-urban communal areas.
Implementation of the programme
The major objectives of the project were to:
· increase milk production by small-scale farmers through improved breeding, feeding and management· identify the socio-economic factors and constraints limiting production, marketing and consumption of milk in the rural areas
· evaluate the technical and economic performance of new interventions
· strengthen the local research capacity through training.
Research methodology
Project implementation started in February 1985. Initially the approach was to review the objectives and proposed methodology and to establish a work plan for research activities. This included identification and designation of small-scale dairy development localities and the areas of collaboration within and between the various research disciplines in the Animal Production Research Unit (APRU), the Department of Agricultural Research, the Department of Animal Health and Production and District and Regional Agricultural Extension staff.
In the meantime, the ground work for the project implementation was completed. This included the selection of Simmental-Tswana crossbred cattle for exchange with farmers' cows; the identification and selection of participating farmers based on the agreed criteria; the baseline survey; measurement of each farmer's crop area and yield of crop residues and planted fodder; construction of basic dairy infrastructure; and purchase of project vehicles, laboratory equipment, veterinary supplies, dairy utensils, fodder seeds and fertiliser.
Finally a "Small-Scale Dairy Technical and Management Package for on-farm Research" was prepared and agreed upon in August 1985. The introduction of this technical and management package marked the actual field implementation of the project.
Simmental-Tswana crossbred cattle
Simmental-Tswana heifers (in-calf to a Simmental bull) were selected from APRU ranches and aggregated from about August 1985, for one month, at a temporary assembly and distribution point at the central research station. During this period they were adapted to the recommended cereal stover-based feeding system, trained and conditioned by the project staff.
Each of the 46 farmers was provided with one of these in-calf heifers in exchange for a cow from his herd. The crossbred heifers were allocated on a strictly random lottery basis.
The Simmental crosses were considered for free replacement by the project if they failed to produce milk satisfactorily or if the calf died within the first two months of exchange or if they died or were injured on-farm due to project related activities (e.g. heartwater vaccination).
Crop and fodder production
All participating farmers were provided with 20 kg of Dolichos lablab (Lablab purpureus) seed, enough to plant one hectare of land. All cultivation, planting and harvesting was done by the farmers. The farmers were also responsible for obtaining their normal subsistence crop seed and fertiliser for routine seasonal planting. Allocation of land for various crops was left to the farmer.
At the end of the cropping season farmers harvested all legume fodder hay (lablab) cereal stovers and all crop residues. These were then stored for winter or dry-season feeding to in-calf cows, milking cows and calves.
Feeding system
A flexible, practical feeding system was recommended which was also dependent on the availability of local feed resources and seasonal variations. The feeding system was divided into summer (wet) season and winter (dry) season.
Summer season - December to April
· natural pasture grazing;
· sorghum bran (moroko);
· mineral supplements - bonemeal and salt (1:1) or;
· dicalcium phosphate (pop) and salt (1:2) ad lib.
Winter season - May to November
Four alternative rations were recommended and compared in on-farm trials:
· natural pasture grazing;
· lablab hay;
· sorghum bran (moroko);
· mineral supplements as above; and
· sorghum stover.
The project started during the drought of the mid-eighties. Due to prevailing climatic conditions and to assist farmers, the project purchased and supplied additional cereal crop stovers, legume hay and sorghum bran (moroko) to enable a minimum of one and a maximum of four lactating cows to be fed during the first two winter seasons (120 days each season). The maximum crop stover provided by the project was up to 100% of the stover harvested by each farmer. The provision of this feed subsidy was only intended to assist farmers in the first two years of operation.
Breeding system
Artificial insemination was chosen as the most appropriate method for breeding to enable Simmental semen to be used on the farmers' existing Tswana cows and to provide a "grading-up" mechanism for those crossbred cows provided by the project and kept on-farm.
Provision was made for the farmers' Tswana cows to be sent to the nearest AI centre for insemination between November and March. The Simmental crossbred cows owned by the farmers were inseminated on-farm. The semen and the insemination service were provided by the project. Farmers assisted in recording the incidence of first heat so that the next cycle could be anticipated and prepared for. Local bulls did not have access to these cows. initial cost of this AI programme was borne by the project.
Milking
It was recommended that cows be milked twice a day, in the morning and evening. However, milking frequency depended on the farmer's management decision. To initiate milk let-down, calves were allowed limited suckling at the time of milking.
Calf rearing
Calves were given access to initial milk (colostrum) during the first three days of lactation to ensure an adequate supply of colostrum. Subsequently, calves suckled for a minimum of 30 minutes in the morning and evening after milking. The condition and weight of calves were monitored and adjustments to milking made accordingly. Calves were kraaled separately and, depending on their age, were provided with good quality stover, legume hay, sorghum bran and mineral/vitamin supplements and adequate health and sanitation care.
Feeding
Lactating cows were stall-fed in winter and grazed in summer in accordance with the recommended seasonal feeding system. Cattle were trekked to the nearest water point since it was difficult to provide water in the kraals.
Health and sanitation
The project provided routine veterinary assistance through the services of the Botswana College of Agriculture and the Animal Health and Production Department of the Ministry of Agriculture. The assistance included:
Tick control. The project supplied a knapsack sprayer and acaricides to be shared by three participating farmers. Acaricides were provided free of charge in the first two years of participation. Cattle were sprayed by the project staff once a during the winter and twice a month during the summer.
Vaccinations. Routine vaccinations against anthrax, pasteurella, calf paratyphoid, brucellosis and heart water were provided free of charge by the Department of Animal Health and Production of the Ministry of Agriculture.
General health and veterinary services. Parasitic control and deworming practices were based on recommendations made by the Veterinary Department. Dosing was done twice a year, at the start and the end of summer. The drugs were provided by the project for the first two years of participation.
Milk recording equipment
Each farmer was provided with the following milk recording equipment:
· milk recording forms
· two calibrated 10-litre plastic milk buckets for recording milk yields
· one 10-litre milk can for storage
· one seamless milking bucket.
In addition, all farmers' cattle were eartagged and weighed every two months. Project staff and participating farmers kept performance records and additional farm related data records.
Rural milk marketing and processing
Milk collection and cooling centres are currently being established in each village to facilitate marketing of the farmers' milk. Already one of these centres is being equipped with solar powered cooling facilities. Farmers will deliver their fresh milk in cans to the centres for sale. When milk supply exceeds the demand in the villages the surplus milk will be sent to Gaborone Dairy Cooperative for processing and marketing.
During 1987/88, studies to investigate traditional milk processing methods by farmers were initiated. A survey of milk processing techniques (madila), chemical composition and microbiological studies of fresh milk and processed products have been undertaken (APRU, 1988).
Au participating farmers were located within the communal areas of the Gaborone region and practiced mixed crop-livestock farming. Some of these farmers already had experiences in harvesting and feeding crop residues. The project encouraged farmers to collect and conserve crop residues for feeding to livestock in the dry season; the advice was generally well appreciated.
Research has shown that the controlled feeding of crop residues to beef animals can improve animal performance (APRU, 1986). The dairy project in a parallel investigation is testing a package based on locally available feeds and improved cow breeds which is proving particularly attractive to a wide range of farmers.
Milk represents an important element in the diet of most farm families and the high milk yields of the Simmental crossbreds has increased the availability of milk for family consumption. Some participating farmers already produce surplus milk. The balance is sold fresh or in a semi-processed form as sour milk (madila). The availability of milk for home consumption and the generation of income through the sale of surplus have been the major benefits to the participating farmers.
Dairy production based on dual purpose cattle using the Simmental crossbred is promising. The management system is less intensive than for purebred dairy cows and retains flexibility in terms of increased milk yield and the sale of animals for beef. Participating farmers showed enthusiasm for the cow-for-cow exchange of the crossbred and also sent 194 open Tswana cows for artificial insemination to the government centre between 1985 and 1990 (DPR, 1989). Therefore the multiplication of the Simmental crosses is picking up well.
Each participating farmer made his own decision in terms of land preparation, planting, weeding, harvesting and allocation of land for various crops. The predominant crops were sorghum and millet intercropped with either cowpeas or water melons. The project succeeded in convincing participating farmers to allocate one hectare of land for fodder (Lablab purpureus) production. Here fodder production has become a major component of the cropping system (Boitumelo et al, 1991).
The feeding system is practically oriented and hence based on local feed resources i.e. winter (dry season) feeding using conserved crop stovers, legume fodders and sorghum bran (moroko) and summer (wet-season) feeding on natural grazing. In a comparative winter-feeding trial the dry-matter intake values were not different from the range recommended for dual-purpose cattle.
The dry-season feeding strategy using high protein lablab legume fodder, planted and harvested by farmers, fed in association with crop by-products and post-harvest residues has been a significant intervention in the small dairy farms. All farmers harvested and stored on the average 34,63; 28, 70; 113, 32; and 35, 26 tonnes of dry matter of lablab and sorghum stover for the 1985/86, 1986/87, 1987/88 and 1988/89 seasons, respectively (Boitumelo et al, 1991).
In terms of management, all farmers were responsible for feeding and watering their cattle. Some farmers milked the cows twice a day but the majority milked once a day. Those participating were provided with standard forms to record the daily milk yield of each cow. Surplus fresh milk or sour milk was sold to the nearby villages.
Animal performance data indicated that under similar nutritional and management conditions Simmental crossbred cows produced about twice more milk per lactation than Tswana cows. Average milk produced per lactation (average 250 days and 200 days) by breed for the 1985/86, 1986/87, 1987/88 and 1988/89 seasons were 760, 338; 890, 319; 818, 474, and 988, 420 litres for Simmental crossbred and Tswana cows, respectively (DPR, 1989). These quantities do not include the portions suckled by the calves.
Average birth weights for Simmental crossbred calves (34 kg) were higher than for Tswana calves (28 kg). At 12 months of age the average liveweights for Simmental crossbred calves and Tswana calves were 147 kg and 124 kg, respectively (DPR, 1987). This shows that the crossbreds are also good beef producers.
Improved feeding, disease control and management has resulted in overall better condition of animals in the project area. For example, in the 1985/86 season five Simmental crossbred cows died, while in the 1986/87 season seven crossbred cows and calves died. In the 1987/88 no cow mortality was recorded. This trend shows a reduction in incidences of mortality.
Problems or failures
The problems that have been isolated are mainly related to feed procurement. Firstly, the major constraint observed during harvesting and conservation of lablab was the duration of time required to dry the stems. Lablab leaves dry and shatter within three days while it takes up to six weeks for the stems to dry completely (DPR, 1988). Secondly, the labour required to harvest sorghum stover is a constraint in that it affects the quantity that can be harvested and conserved by farmers (APRU, 1989).
Another problem relating to animal health concerns several reported deaths had no known causes. As a result it has not been easy to adequately address the problem. Lack of organised marketing outlets has also proved to be a problem.
The increase in milk production by small-scale pert-urban farmers has been a major objective of the Ministry of Agriculture during the six-year National Development Plan (Ministry of Finance and Development Planning, 1985). It was thought that increased production in this sector would not only increase local supplies for home consumption but could also lead to surpluses available for sale to the rapidly increasing urban population. Therefore, the government policy environment has been quite conducive to production.
The low performance of the dairy subsector has been largely attributed to the lack of technical expertise within the country. As a consequence there emerged a big gap between information and extension packages and the research and extension methodology needed for bringing about increased productivity. Due to this problem, technical support was offered by the co-sponsors of the project who provided an animal scientist to give expert advice. Provision was also made to strengthen research capacity by training a local young scientist and technician in various dairy related fields. In recognition of inadequate equipment, laboratory capacity, transport etc funds were provided the project to facilitate the procurement of these necessary inputs.
Dairy production based on dual-purpose cattle using the Simmental crossbred proved promising and appropriate because it required less intensive management skills than the pure bred dairy breed. It could also produce reasonable quantities of milk and be sold as a beef animal.
Lablab fodder legume has adapted quite well in the small farm conditions. It is drought tolerant, easy to manage, has grazing potential and yields quite reasonable amounts of dry matter (average of about 2.55 t/ha per farmer).
The target group used in this project seems to be suitable. However, constraints such as sufficient labour for harvesting and conserving the feed still need to be addressed. The on-farm milk production results are comparable with those found on-station (APRU, 1988), but the lablab yields have been higher on-station than on-farm (5.36 ton/ha against 2.55 ton/ha). It is thought that this difference was due to planting at the correct time on-station while usually farmers tended to plant legume fodder after planting their crops.
Availability of capital investment for adoption of the technologies does not seem to be a serious problem. There are various programmes within the Ministry of Agriculture that can provide funds to groups and individuals to construct dips and handling facilities (e.g. Arable Development Programme (ALDEP) and services to livestock owners in the communal areas (SLOCA)). Inputs like drugs are available in government-run livestock advisory centres. With the sale of milk, it is hoped that farmers will in time afford to purchase drugs. Although farmers are expected to produce the bulk of their feed requirements, feeds can also be purchased from LAC and other private companies.
A financial assessment of the prospects for small-scale dairy farming was carried out on the assumption that costs would be met by farmers. As it appeared that the labour shortage would prevent farmers from producing all supplementary fodder required by the improved package, it was therefore assumed that a certain proportion of the fodder would be purchased thus increasing the costs of the improved system (APRU, 1989). Therefore the production strategy is now aimed at reducing the costs associated with an improved system in order to increase the net benefit.
Socially the project has benefited the farmers by providing surplus milk which is sold to earn income. Some farmers have managed to purchase knapsack sprayers and acaricides with the money realised from surplus milk sales. Because of those spin-offs from milk sales it is also planned that the project should commence a socio-economic monitoring and evaluation.
A marketing strategy is being developed, involving the establishment of milk collection centres to be equipped with cooling and storage facilities. A survey on marketing and the consumption of fresh and sour milk was conducted to identify marketing constraints. In this survey over 60% of the farmers sold some of their milk either fresh or soured. The quantities sold were dependent on the amount of milk left after daily family requirements had been met (APRU, 1988).
It is not easy to quantify the adoption rate. However, out of the 46 farmers participating in the project, six other farmers have been provided with lablab seed to plant on up to one hectare of land. Also several farmers, both small-scale and medium-scale, have been enquiring about Simmental cows and establishment of lablab fodder. In one area where the construction work for a milk collection centre is complete, 20 farmers have formed into a milk marketing group. This includes 12 participating and eight non-project farmers.
Focus has been limited to the 46 farmers and the other six farmers around Gaborone. Expansion is awaiting the development of an extension package.
The project has been successful in both postgraduate and short courses training. Two scientists obtained MSc degrees from the University of Guelph while one scientist is just about to complete a PhD degree in Canada. Six project staff attended different specialised short courses in dairy related fields.
To assess whether the programme has changed or influenced production is not easy, this requires a step by step analysis of measurable and tangible changes. This paper will only highlight some areas where it is felt there has been some changes mainly due to the project.
Besides the collaborating farmers who have benefited directly, the results of the project have provided a framework for future development programmes in milk and beef production in the pert-urban areas of the country. This information from the project has been used in outlining the National Development Plan.
Although Botswana probably does not have a comparative advantage in milk production, the government's dairy development policy encourages this sector. This is evident from the assistance the government is providing to farmers through the Financial Assistance Policy (FAP) and the Dairy Industry Special Fund. The new government agricultural policy gives considerable attention to the cost of producing agricultural products. The overall policy is to provide sufficient income-generating opportunities to raise household incomes and improve access to food.
The low performance of the dairy subsector during the past NDP 6 was attributed largely to the lack of technical expertise within the farming community and the lack of organised marketing infrastructure necessary to run a viable dairy enterprise. During the NDP 7 (Ministry of Finance and Development Planning, 1991), the government therefore sought to give more emphasis to improvements in the milk marketing system and infrastructure and to further develop and disseminate dairy production technology. To this end, the government has allocated funds to be used mainly for developing the pilot milk collection centres and marketing systems in four project areas.
With more emphasis by government on improving the dairy industry, the dairy research programme needs to further develop, cost-effective packages that could be used by smallholders to increase production. The dairy development research project will mainly focus on the development of an efficient marketing system to stimulate dairy production in the peri-urban areas. Special attention will also be given to farmer training.
The general objective of the future dairy development research is to increase milk production at minimal costs and to develop small-scale processing technologies and marketing systems for milk and milk products in the rural areas. To achieve this objective, both on-station and on-farm research will be carried out. On-station work will be necessary to re-address the areas identified as problem areas and to further refine some technologies to be tested on-farm.
APRU (Animal Production Research Unit). 1979. Livestock and range research in Botswana.
Annual Report 1979. Government Printer, Gaborone, Botswana.
APRU (Animal Production Research Unit). 1982. Livestock and range research in Botswana. Annual Report 1981-82. Government Printer, Gaborone, Botswana.
APRU (Animal Production Research Unit). 1984. Livestock and range research in Botswana. Annual Report 1983-84. Government Printer, Gaborone, Botswana.
APRU (Animal Production Research Unit). 1986. Livestock and range research in Botswana. Annual Report 1986. Government Printer, Gaborone, Botswana.
APRU (Animal Production Research Unit). 1988. Livestock and range research in Botswana. Annual Report 1988. Government Printer, Gaborone, Botswana.
APRU. 1989. Livestock and range research in Botswana. Annual Report 1989. Government Printer, Gaborone, Botswana.
Animal Production Division. 1989. Annual Report. Division of Animal Production, Ministry of Agriculture. Government Printer, Gaborone, Botswana.
Boitumelo W S. Mahabile W and Madibela O R. 1991. Potential of crop residues as a feed resource for small-scale milk production: Dairy research experiences. A paper presented at Livestock Production Workshop, October 1991, Gaborone, Botswana. (In press).
DPR 1988 Dairy/beef production systems. Final report Phase I 1988. Government Printer, Gaborone, Botswana.
DPR 1989 Dairy/beef production systems. Annual reports 1985-89. Government Printer, Gaborone, Botswana.
DPR 1987 Dairy/beef production systems. Annual Report 1986-87. Government Printer, Gaborone, Botswana.
Ministry of agriculture. 1986. Agricultural statistics report. Division of Planning and Statistics. Ministry of Agriculture, Gaborone, Botswana
1985. National Development Plan NDP 6. 1985-91. Ministry of Finance and Development Planning. Government Printer, Gaborone, Botswana.
1991. National Development Plan NDP 7. 1991-97. Ministry of Finance and Development Planning. Government Printer, Gaborone.
