Impacts of intra-seasonal agricultural decision-making and forecast information on maize production in Zambia

Abstract:

Maize is the most important food staple in sub-Saharan Africa. Climate change and rainfall variability pose great risks on maize production in this region. Intra-seasonal adaptive management combined with more skillful weather forecasts has the potential to improve the resilience of agricultural systems. Our aim is to understand the extent to which within-season agricultural management decisions can mitigate weather risks to maize production, and the degree to which this mitigation varies as a function of when the decision is made and the trajectory of weather. Using Zambia as a test case, we conducted crop-modeling experiments to determine which crop and water management decisions (typical of smallholder farmers) are most effective in mitigating rainfall-driven yield reductions under three precipitation scenarios (below normal, normal, and above normal). Yields were simulated using the DSSAT CERES-Maize model driven by an ensemble of historical weather data. Potential maize yields under different management options were simulated from different forecast points during the growing season, starting at planting and then in successive two-week intervals through the grain-filling period. The yield distributions were constructed as a function of the weather conditions and the management options, with results indicating which decision options provide the most mitigation in relation to a) the particular point in the growing season at which they are made, and b) the potential rainfall scenario. This study will help to understand how smallholder farmers in semi-arid systems may increase their resilience to highly variable weather by using typical within-season management options, and which decisions are most robust to forecast uncertainty.