Genotype by environment interactions, stability, and heritability of seed yield and certain agronomical traits in soybean [glycine max (l.) merr.]

This research was aimed to (1) evaluate genotypic yield performances of eight soybean genotypes, (2) determine their stability parameters and (3) estimate variance components and heritabilities of yield and yield-related traits. Eight soybean genotypes were evaluated in two locations (Mustafakemalpasa and Gorukle) of Southern Marmara Region, Turkey during two years (2000 and 2001), to identify patterns of genotype by environment interactions and their stabilities in terms of seed yield and yield-related components. Seed yield and its components are affected by plant genotype and environmental conditions. Plot data for plant height, first pod height, number of pods per plant, number of seeds per plant, number of seeds per pod, 1000 seed weight and seed yield were observed at all the four trials. There were significant differences between genotypes of one or two years at each location for all the traits. Significant differences among years or between locations were obtained in terms of all traits. Genotype x environment interactions at all the traits were highly significant. Thus the stabilities of eight soybean genotypes were different for all the traits. According to the stability parameters, none of the genotypes were stable for seed yield. Genotypes, Hodgson-78, A-3127, Ataem-1, Corsoy and SA-88 were considered as having high adaptability to good environments while Mitchell, Ataem-2 and Etae-8 adapted to poor conditions in terms of seed yield. Genetical components of variance at all the traits were highly significant. Genotype x location x year interaction variances were also found significant at all the traits except first pod height. The estimates of heritabilities with limited phenotypic variance definition were 0.14, 0.14, 0.21, 0.004, 0.13, 0.30 and 0.26 for plant height, number of seeds per pod, 1000 seed weight, first pod height, number of seeds per plant, number of pods per plant and seed yield, respectively. The heritabilities with complete phenotypic variance definition were 0.05, 0.05, 0.04, 0.003, 0.07, 0.19 and 0.20 for the same traits, respectively. Moderate or low heritabilities estimated for all the traits showed that family selection method could be used instead of individual selection in the breeding programs for improving seed yield and its components