12. Decisions on whether to undertake tsetse control in the Zambezi Valley in support of the introduction of cattle should be based upon economic benefit-cost analysis, using techniques which have already been applied elsewhere in Africa (e.g. Jahnke, 1974; Habtemariam, 1983; Shaw, 1987; Brandl, 1988). The direct costs of tsetse and trypanosomiasis control operations can be quantified far more easily than potential benefits and disbenefits. An adequate basis exists for estimating the effect of disease control on animal productivity and the economics of alternative land uses. But the difficulty arises in projecting over a ten to twenty year period the pattern and intensity of economic activity that is likely to develop in a project area with and without tsetse control being undertaken (Straw, 1987). Standard techniques of sensitivity analysis provide a methodology for appraising the implications of uncertainty providing there are plausible limits to such uncertainty.
13. Implementation of tsetse control can be justified only if there is confidence that subsequent rural development will be in accordance with the land use plans and projections on which benefit-cost analysis has been based. In many respects the level of land use planning taking place in the Zambezi Valley is high. But there may be a real danger that land use plans will not be implementable in the long run. Planners themselves appear unconvinced that they know how the marginal, semi-arid lands in this part of Zimbabwe can be used in a sustainable and economically viable way. Plans are being prepared on the basis of experience in more establised farming areas of the country, experience which may prove inappropriate in the Zambezi Valley.
