Genotype × Environment Interaction in Maize

Genotype × environment (GE) interactions are a challenge to plant breeders because they cause difficulties in selecting genotypes evaluated in diverse environments. When GE interaction is significant, its cause, nature, and implications must be carefully considered. No information is available on the contribution of weather variables and environmental index ( Ȳ . j or mean yield of all cultivars in j th location minus Ȳ .. or overall mean yield for all cultivars and all locations) to GE interaction for yield in maize ( Zea mays L.). The objective of this study was to determine the effect of an environmental index, maximum and minimum temperatures, rainfall for the growing season, preseason rainfall, and relative humidity (covariates) on GE interaction for yield. Seventeen hybrids grown in 12 environments (4 locations × 3 yr) were studied. The GE interaction was significiant and was partitioned into σ 2 i (stability variance) components assignable to each genotype (hybrid). Heterogeneity (non‐additivity) due to each covariate was removed from the GE interaction, and the remainder of the GE interaction variance was partitioned into s 2 i components assignable to each genotype. Environmental index accounted for 9.61% of the GE interaction sum of squares ( P = 0.10). This was the largest amount of heterogeneity removed from the GE interaction by any single covariate. Rainfall during the growing season removed 1.4% of the GE interaction sum of squares and preseason rainfall removed 1.1% of the GE interaction sum of squares as heterogeneity. Minimum temperature, maximum temperature, and relative humidity removed a negligible amount of heterogeneity from the GE interaction. The 17 hybrids evaluated in this study were differentially affected more by differential fertility and/or cultural practices (environmental index) than by weather factors. Seven of the 17 hybrids had unstable performance across the 12 environments.