YIELD POTENTIAL AND STABILITY OF GERMOPLASM COMBINATIONS DEVELOPED AMONG MAIZE GROUPS

Abstract

Germplasm knowledge and classification are directly associated to possibilities of success in any plant breeding program. In order to assess the behavior of combinations among five contrasting gene pools adapted to El Bajío area, and to identify those with highest yields and stability, several linear models whit a combination of fixed and mixed effects in three contrasting environments. were compared during the Summer 2002. Grouping tendencies for combinations and environments were simultaneously evaluated along with the interaction among these factors, by using biplot graphics generated by the “additive main effects and multiplicative interaction” (AMMI or GE graphing), and “sites regression” (SREG or GGE graphing) models. The results indicated that the mixed lineal models provide more precise estimators and predictors than those based on fixed effects, by achieving wider inference spaces. The biplots obtained from the GE and GGE programs show the following advantages: the classification and clustering of main effects allows an easier interpretation; they estimate the magnitude and the sense of the genotype x environment interaction; and they allow visualization of the best performing combinations for a particular environment, thus allowing a better selection. The heterotic combinations with the highest yielding potential for the El Bajío region were Dwarf x Tropical maize and Dwarf x QPM, and the most versatile heterotic group for heterosis was Dwarf corn, which produces excellent combinations when crossed with all the other heterotic groups.