A modeling System for Identification of Maize Ideotypes, optimal sowing dates and nitrogen fertilization under climate change – PREPCLIM-v1

Abstract. The impact of climate change on crops and agricultural yield is an actual threat while being a challenging issue due to the high complexity of factors that intervene at the local scale of the crop. Assessing it, requires the use of coupled models climate-phenology, meanwhile methods to identify management and genotypes suitable for local future conditions, in order to sustain adaptation strategies. We present the implementation and use of a new integrated climate-phenology adaptation support modeling system based on regional CORDEX climate models and the CERES Maize model from DSSAT platform, with new modules for optimal management and genotype identification using a hybrid method: deterministic modeling and -ML/ genetic algorithms. It was run as a regional pilot over Romania, operating in real-time in interaction with users, performing agro-climate projections (combination of fertilization, sowing date, soil) and providing best crop management simulated under climate change projections. Multi-model ensemble simulations for two climate scenarios RCP4.5 and RCP8.5 and twelve management scenarios show new results for the region.

For the actual genotype we find a projected mean decrease in yield in both climate scenarios for all sowing dates and fertilization levels tested, response shown to be sensitive to initial soil parameters. This response was linked to two factors: a shorter growing season by up to 10 % and a loss of fertilization efficiency in a warmer climate. A warning points to results showing a narrowing of agro-management opportunities for crop yield but in opposite it is shown a significant role of optimal genotype-range identification that may provide crop solutions under climate change even in extreme years. Identifying best genotype under warmer climate along sets of six cross-parameter simulations show systematic lower values of the maximal yields, but emphasizes genotype windows of increases in the intermediate yield values in scenarios compared to actual climate. The highest harvest sensitivity to genotype is shown to be to changes in the thermal time to juvenil respectively to maturity stage under warmer climate. The results sustain using a deterministic coupled modeling system combined with data-driven modeling for identifying optimal adaptation including fertilization paths that contribute to climate change mitigation.