Modelling potential maize yield with climate and crop conditions around flowering

Understanding, and then modelling, the effects of sowing date and cultivar on maize yield is essential to develop appropriate climate change adaptation strategies. Here we test the WOFOST model and a hybrid model, based on physiological crop conditions around flowering, against observed data collected during 4 years of field experiments in a Mediterranean environment under fully irrigated conditions. We simulate sowing date and cultivar responses by using 45-year historical meteorological records from the experimental weather station and future climate conditions till 2060 as projected by a set of regional climate models. Both WOFOST and the hybrid approach reveal good performance in simulating average maize yield. However, the hybrid one outperforms WOFOST with respect to its responsiveness to changes in sowing date and cultivar. These findings, besides stressing the importance of crop conditions around flowering in determining maize yield, point to lower yields (14 %–17 %, average reduction) under future climate conditions. The estimated losses may only be partially offset by changes in phenology and sowing dates.