Kernel-FA Unlocks Key Environmental Drivers of GEI

This study investigated genotype-by-environment interaction (GEI) in temperate and tropical maize using multi-environment trial (MET) data and factor-analytic (FA) mixed models. Likelihood ratio tests confirmed significant genotypic variation for grain yield (GY), supporting the use of FA structures to model GEI. The FA5 model best described temperate maize, explaining ~79% of the genetic variance, whereas FA4 captured ~84% in tropical maize, indicating that few latent dimensions summarize the major GEI patterns in both groups. To interpret these dimensions, FA loadings were correlated with environmental covariates processed by EnvRtype. The primary factors were strongly associated with thermal-hydric gradients, while secondary factors captured radiative, evaporative, and soil-related variation. These relationships revealed distinct environmental drivers shaping GEI in each adaptation group and helped identify conditions favoring superior performance. The integrative approach demonstrates the value of combining FA modeling with environmental descriptors to contextualize GEI and support predictive breeding across diverse environments.