An analysis of agricultural livestock and traditional crop production statistics as a function of total annual and early, mid and late rainy season rainfall in Botswana

Abstract

Total rainfall amounts are usually not considered as useful indicators for agricultural season qualities because they do not reflect other parameters such as rainfall distribution, crop water requirements and soil moisture-holding capacities. Most methods that also take into account the above parameters have not been validated for Botswana because of the country's restricted meteorological network and because of the non-availability of the necessary soil data. Rainfall alone is, therefore, one of the most frequently used parameters in drought monitoring and quality assessments of agricultural seasons and provides essential, but rough information several months before the real agricultural production data have become available. In this paper, Botswana's national cattle herd size between 1965 and 1985, regional cattle death ratios and regional traditional crop production between 1979 and 1985 have been analysed as a function of total annual and early, mid and late rainy season rainfall as defined by Bhalotra (1985a). It was found that in Botswana, with marginal rainy season conditions both for cattle and crops, such an analysis can be used as a useful indicator for the agricultural quality of a rainy season. To obtain an indicator of the season quality for livestock performance, a weighted average for the past three seasons and for October December (early), January February (mid) and March-April (late) rainfall has to be used. A good indicator for the crop season quality can be obtained from a weighted average of early, mid and late rainfall of the actual season. These indicators cannot, however, be used as indices to forecast livestock condition or crop production, but only to carry out acceptable assessments of rainy season conditions on a subnational or areal scale.