SOYBEAN YIELD IN ARGENTINA IN RESPONSE TO WEATHER VARIATION AMONG CROPPING SEASONS

A relatively simple, mechanistic model describing soybean development and growth was used to analyze variation in yield among growing seasons and locations in Argentina. Comparison of simulation results with those obtained in a single year of intensive sampling of rainfed and irrigated treatments at Oliveros, showed good agreement. Particularly good agreement between observations and simulations was obtained in the rainfed treatment. In addition, the yield of the highest-yielding cultivar in each trial at three locations (Oliveros, Marcos Juarez and Pergamino) representing the main soybean-growing region in Argentina, and in three growing seasons was simulated well. The simulations highlighted the importance of inadequate soil water in February and March in all environments in constraining soybean yields. Even in a relatively wet year (1987/88), soil water decreased to sufficiently low levels in the model to inhibit crop growth. The possibility of extending the cropping season by delayed flowering to take advantage of March and April was examined and found to be beneficial only in the driest year (1988/89).