Evaluation of different crop management options to adapt maize to current and future climate using a cropping systems model

Smallholder food security is vulnerable to climate change, particularly with regard to maize production, which is affected by rising temperatures and changing rainfall in Ethiopia. The objective of this study was to assess the integrated crop management options under current and future climate scenario conditions on the combined impacts on productivity and resilience of maize production using a global climate change cropping systems model. The model DSSAT-CERES-Maize was calibrated and validated at three sites (Ambo, Bako, Melkassa) with the field data (nRMSE < 10% for phenology and yield; d-index > 0.90). Various combinations of three cultivars (MH140, BH546, BH661), planting dates (May 15–June 15), and densities (4.4–7.14 plants m− 2 ) were tested, along with various rates of Nitrogen/Phosphorus fertilizers. Future scenarios were derived from five ensemble members of the GCMs for RCP4.5 and RCP8.5 conditions for the 2040s and 2070s, respectively, with the inclusion of CO 2 fertilization effects. Due to heat stress and loss of (Leaf area index), yields were decreased by up to 22% at Bako, and biomass by 20-60% at Melkassa without adaptation. Impacts were well mitigated by optimized packages (site-specific early-mid planting, higher density/fertilization, and resilient cultivars). Future climate showed changes between −22% and +12% baseline management (climate effect alone). Under the same future climate, the improved yields by applying adapted packages ranged from +10% to +121% when compared to the baseline yields under the same future climate (adaptation effect), while the total change ranged from +68% to +121%. This shows that climate change imposes negative pressure in some areas, but that in other areas, where biophysical limits are not exceeded, agronomic adaptation can significantly mitigate the outcome of climate change. Gains were observed at Bako (+12%) and Melkassa continued to experience losses (maximum 60% loss in biomass), suggesting biophysical constraints. Genotype × environment × management interactions were detected and the direction and magnitude of any changes in yield were dependent upon site and management package. While interactions coefficients could not be formally estimated due to the management factors being combined together rather than in a fully crossed factorial design, these results illustrate the necessity to consider site-specific integrated management instead of the imposition of general recommendations. The products of integrated agronomic packages may sustain or even improve the productivity of maize at the site-specific level under future climate conditions. Findings enable adaptation strategies to be developed for long-term food security in Ethiopian maize systems based on location and justify on-farm testing.