CropGen: A novel tool for optimising sorghum crop design

Abstract

Sorghum is an important food and feed crop globally. Strategies to optimise its yield performance are needed to help meet rising demands in changing future environments. Typically, sorghum is grown as a dryland crop, with water availability being a major limitation for yield. Here we present a new tool, CropGen, for optimising sorghum design and yield performance. CropGen connects the APSIM sorghum model with an evolutionary optimisation algorithm to enable iterative exploration of genotype × agronomy combinations to identify optimal strategies based on a number of objectives, including grain yield, water use, and crop failure risk over diverse production environments. Initial model testing was undertaken using single year simulations both with and without irrigation to explore tillering × maturity combinations and quantify their yield and water use. In the irrigated scenario, optimal yields were achieved with high tillering and long maturity sorghum designs, whereas lower tillering and shorter maturity designs were identified for the non-irrigated scenario. These findings are consistent with known effects of water availability, crop growth, and yield interactions. CropGen was then applied to optimise tillering × maturity designs for Dalby, a key sorghum production region in Australia, by quantifying the yield, water use, and crop failure risks. The nuanced understanding of crop performance over diverse environments generated by CropGen is key in identifying optimal strategies for guiding research and plant breeding efforts to maximise crop improvement.