Leveraging synthetic environmental covariates to explore crop genotype by environment interactions
Environments are unpredictable. Genotype by environment interactions (GEI) further complicate any assessment of genotype performance or stability across diverse environments posing a challenge for plant breeders. Several approaches exist to address this challenge. However existing approaches present a trade-off between scale and interpretability often acting in a 'black box' manner. Here we extend the Finlay-Wilkinson regression on synthetic environmental covariates of Piepho and Blancon (2023) to dissect GEI at the level of individual developmental windows, resolving which growth stages drive GEI and how genotype stability varies across them. Yield data for 961 genotypes from a barley breeding program, tested across 12 environments, is used as a case study to illustrate how this approach can provide valuable insight into the GEI present. Exploring the dynamic impact of environmental variables across growth stages revealed that environmental factors during early grain fill contributed the most to yield GEI variance. Stability was found to be dynamic across development windows, presenting growth-stage-specific yield stability as a new potential breeding target. This approach yields new, growth-stage-resolved measures of yield stability and a new means of dissecting GEI in complex traits, while also opening a route to environmental clustering based on stage-specific GEI patterns.