Genotype-environment interaction in maize hybrids: an application of the AMMI model

The genotype-environment interaction is generally recognized when the same genotype is assessed in different environments, having a decisive influence in cultivar recommendation. Maize hybrids that are only adapted to particular environments can become a limiting factor to seed production in large scale. One of the main objectives of breeders is the obtainment of a hybrid with high mean yield and good adaptation to different environments. The present study assessed the grain yield stability in thirty-six maize genotypes in ten environments located in Central Brazil in the 1999/2000 growing season. The analysis of variance (ANOVA) detected significance (P<0.01) for the genotype-environment interaction (GE). The AMMI model (additive main effects and multiplicative interaction) was used to assess the additive and multiplicative effects of the interaction. The results indicated homogeneity of environments assessed and high correlation of the hybrid genetic constitution with the yield stability of the three types studied -single, triple and double crosses maize hybrids. The first multiplicative component of the interaction explained more than half (50.6%) of the original sum of squares (SS GE ). Most of the environments were neutral in their contribution to the interaction. The Uberlândia environment contributed positively, while the environment in Patos de Minas contributed negatively to the GE interaction. Most of the genotypes also presented a low contribution to interaction. Some single crosses hybrids presented greater mean yield (10182.0 kg.ha -1 ) while the double crosses hybrids presented greater stability to the