EMPIRICALLY MODELING THE INTERACTION BETWEEN INTENSIFICATION AND CLIMATIC RISK IN SEMIARID REGIONS

In many developing countries, yield increases obtained through intensi fication of agriculture seems necessary to face the high population gr owth rate. In the case of rainfed agriculture of semiarid regions the effect on yield of some intensive farming practices is highly variable , however, due to interactions with rainfall distribution and amount. This paper presents a model (SARRA) to perform regional comparisons of the risk associated with intensive practices in semiarid environments . The model, built for this specific and restricted objective, is suit able for crops and regions where few data are available. The intensifi cation level of a given crop situation is assumed to be the potential yield that would be obtained without any water constraint. The model w as calibrated and validated for millet in Senegal and then run for thr ee locations in that country, using long-term sequences of climatic da ta from wet and dry periods. Three potential yield levels and two runo ff loss levels were tested for each location and period. At one site, the study revealed that using management to increase potential yield i s highly risky. In a location only 200 km south of the former, the ris k associated with intensive practices is much less, provided that all rainfall infiltrates. Considering 25% rainfall loss by runoff, which w as reported to be common in this region, however, the risk associated with high yield potential increases. This case study highlighted the e xtent to which the risk faced by farmers in semiarid regions can vary spatially and with management practices. The approach proposed in this paper appears suitable for use in designing agricultural development strategies in semiarid regions. (C) 1997 Elsevier Science B.V.