Résumé
Heterotic patterns among European maize (Zea mays L.) populations are strongly affected by genotype X environment (GE) interactions and no single heterotic pattern has been identified so far that is not influenced by GE interaction. The objectives of this work were to study (i) the mean performance and stability of the heterotic patterns 'Humid Spain X Southern France' and 'Dry Spain X Humid Spain' and (ii) the influence of some environmental and genotypic covariates on G and E main effects and their interaction. We studied the GE interaction for grain yield in eight environments using Sites Regression (SREG) and factorial regression models. The biplot obtained from the SREG model allowed visual cultivar evaluation. The factorial regression model incorporated genotypic and environmental covariates that enhanced biological interpretation of GE interaction. The heterotic patterns Humid Spain x Southern France and Dry Spain X Humid Spain had similar mean performance across environments, but the former, represented by the cross Lazcano x Millette du Lauragais, was more stable. Effects of G, E, and GE for grain yield were mainly due to earliness, vigor effects, and/or environmental factors related to cold stress. An adequately long vegetative cycle along with early vigor had a great influence on the mean grain yield performance of Lazcano × Millette du Lauragais. Additionally, its intermediate number of days to silking and tolerance to temperature stresses could be related to its stability. Breeding for tolerance to temperature stresses could render more stable maize genotypes.