Projecting maize yield under local‐scale climate change scenarios using crop models: Sensitivity to sowing dates, cultivar, and nitrogen fertilizer rates

Abstract The APSIM‐Maize and CERES‐Maize models are widely used in impact studies to analyze the effect of climate change on future maize yield. The study objectives were to develop climate scenarios, assess crop model's sensitivity, and predict the impact of climate change on rainfed maize yield based on five global climate models under two RCP (RCP4.5 and RCP8.5) scenarios. The scenarios were based on the mid‐century and tested for sowing dates (SDs), maize cultivars, and nitrogen fertilizer rates (N). For field calibration and validation, the split‐split‐plot experimental design with three replicates was set up at Mount Makulu, Zambia. The treatments were SD, cultivar, and N‐rate were the main plot, subplot, and subsub plot, respectively. The APSIM‐Maize and CERES‐Maize models were used to run simulations using seasonal analysis. The impacts of climate change on maize yield were simulated for the future 2040–2069/1980–2010 using the AgMIP Protocols. The ensemble means from the simulation result in precipitation decrease and temperature increase. Days after planting to anthesis and maturity would reduce in 2050 (2040–2069). The % change in grain yield would range from 2.78% to 9.94%, −3.81% to −8.88%, and −2.33% to 10.63% under N1 (55.2 N kg/ha), N2 (110.4 N kg/ha), and N3 (165.6 N kg/ha) as affected by SDs, respectively. The simulation showed evidence of climate change and hence affect maize growth and yield. Therefore, there is a need to put in place strategies for alleviating the impact of climate change in maize production in Zambia.