Enhancing sorghum yield and risk management via optimizing crop design

Abstract

Globally, the need to intensify food production and accelerate crop yield gains requires new strategies for crop improvement. Agricultural production outcomes, such as grain yield and crop failure risk, are complex and emerge from interactions that occur between genotype (G), crop management (M), and the environment (E) during crop growth and development. With no feasible means to assess all possibilities, these G × M × E interactions complicate crop improvement decision-making and limit our ability to enhance production over diverse environments and conditions. Further complicating this problem are productivity-risk trade-offs, which make simultaneous improvements in multiple production criteria difficult. This study introduces the CropGen platform, which offers a simulation-based approach to explore crop-adaptation landscapes to identify optimal G × M strategies, called crop designs, for target E. By connecting the Agricultural Production Systems Simulator sorghum model with an evolutionary optimization algorithm, the CropGen platform enables the exploration of crop-adaptation landscapes and the generation of optimized crop designs allowing for the trade-offs among production criteria. This study details the testing and development of the CropGen platform, including its application to a sorghum crop improvement case study in situations varying in yield potential. Findings indicate the CropGen platform is capable of generating physiologically sensible sets of Pareto-optimal solutions that represent a range of trade-offs between yield and crop failure risk. The potential for CropGen to help guide and focus research and breeding efforts for the adaptation of crop production and the advancement of crop improvement is highlighted.