Análisis gráfico mediante biplot del comportamiento de híbridos de maíz

The Sites Regression (SREG) model is used for the multi-environment data analysis, especially when the environment (E) is the more relevant source of variation in relation to the contribution of the genotype (G) and the genotype-environment interaction (GE). This model that includes G+GE into bilinear terms provides a graphic analysis of the performance (yield and stability) of the genotypes called biplot GGE. This graph allows identifying highly productive genotypes in each environment and grouping genotypes and environments with similar response patterns. In order to evaluate the variability of yield maize grain and to determine the response patterns between hybrid and environment, 13 commercial hybrids were evaluate during 3 growing seasons at the Sampacho location, in the central region of Argentina. The pooled analysis of variance across environments showed differences statistically significant (P <0,01) among hybrid, environment and interaction hybrid-environment. The biplot allowed identify four superior hybrids AW190 MG, DK700, 32F07 and DK747, although DK700 genotype could be considered an ideal genotype in the area of Sampacho, because it had highest and stable yielding. The growing seasons 2002/03 and 2004/05 showed a similar pattern of response, while the evaluations carried out in the environments 2003/04 and 2004/05 showed a clear tendency to order the genotypes in a different way.The environment 2002/03, which was highly representative of the average environment, showed higher discriminating ability between genotypes. Differences in the sown dates, soil conditions and rainfall between growing seasons could explain the pattern of genotypic responses across environments.El modelo de regresión de sitios (SREG) se utiliza para el análisis de los datos provenientes de ensayos comparativos de rendimiento multiambientales, en especial cuando el ambiente (E) es la fuente de variación más importante en relación con la contribución del genotipo (G) y la interacción genotipo-ambiente (GE). Este modelo, que incluye en el termino bilineal G+GE, proporciona un análisis gráfico del comportamiento (rendimiento y estabilidad) de los genotipos denominado biplot GGE. Este gráfico permite identificar el genotipo de mayor potencial en cada ambiente y agrupar genotipos y ambientes con patrones similares de respuesta. Con el objeto de evaluar la variabilidad en el rendimiento de grano de maíz y determinar los patrones de respuesta entre híbridos y ambientes, se utilizaron los datos de producción provenientes de 13 híbridos comerciales probados durante 3 campañas agrícolas en Sampacho, Córdoba. El análisis de la varianza combinado a través de los campañas mostró diferencias estadísticamente significativas (P