Simulation model of millet and cowpea intercrop

The model is developed within the generic GROWIT framework (which calculates daily plant growth based on temperature, rainfall, and soil characteristics) and uses a spreadsheet. A spreadsheet template is also provided for making stochastic dominance comparisons. The millet-cowpea model and the stochastic dominance spreadsheet are applied to the evaluation of millet and cowpea production strategies in Niger. Use of the GROWIT-based simulation model to evaluate production strategies confirms the utility of using generic models for analyzing crop innovations. The model can be adapted to new situations quickly. Over a period of several months, the GROWIT model was modified for intercropped annuals by two researchers working part-time, who took about four person-weeks to carry out preliminary calibrations and validations. The validations suggest that the model captures the main tendencies of the yield distribution, but that accuracy, in terms of simulated yields as a percentage of observed yields, is poor, especially in low rainfall years. In the economic analysis, simulation-derived distributions were used to rank crop production strategies according to risk characteristics using the stochastic dominance template. The conclusions obtained with the help of simulation complemented information from other sources. The model is expected to help biological and economic researchers working on the development of new intercrop technologies. Since the millet-cowpea crop association is grown widely in West Africa, the model can be used directly in areas other than Niger. Finally, the generic nature of the model makes it possible to adapt it to other grain/legume intercrop systems. (Author abstract)