Genotypey by environment interaction, genetic variability and path analysis for grain yield in elite soybean [Glycine max (L.) Merrill] lines.
MetadataShow full item record
Soybean [Glycine max (L.) Merrill] is the world’s leading source of protein and vegetable oil. However, its productivity is still low in the region due to limited availability of stable and high yielding cultivars. Therefore, the objectives of this study were: (1) to determine the magnitude of genotype by environment interaction and stability of elite soybean lines for seed yield, (2) to establish trait profiles of 25 soybean genotypes and to study the associations among characters, their direct and indirect effects on grain yield and (3) to estimate genetic parameters of traits related to seed yield and to analyse genetic diversity among elite soybean lines. To achieve these objectives, 25 genotypes (20 elite soybean lines and five commercial checks) were evaluated in multi-location trials conducted in the 2017/18 rainy season using six sites in four countries viz. Zambia, Malawi, Zimbabwe and Mozambique. Both AMMI and GGE biplot analyses indicated Lusaka West as the highest yielding and most informative environment and could be useful for selecting specifically adapted genotypes. Rattray Arnold Research Station was the most ideal environment as it was both informative and highly representative. The soybean lines TGx2002-17DM, TGx2001-10DM, TGx2001-18DM, TGx2014-24FM, TGx2001-6FM and TGx2002-3DM exhibited specific adaptation. Both GGE and AMMI models showed that TGx2014-5GM was more stable than the checks and was second to the highest yielding check. The genotype by trait (GT) and correlation coefficient analyses revealed that pod number per plant and hundred seed weight were the most positively correlated traits with grain yield, while days to 50% flowering had a negative association with grain yield. Sequential path analysis, showed that the number of pods per plant and hundred seed weight had the highest positive and significant direct effects on seed yield, implying that these two traits could be used as selection criteria for seed yield in soybean. The soybean lines TGx2014-5GM and TGx2002-23DM had good combinations of high yields with large seed size and high pod number. The analysis of genetic variability showed small differences between PCV and GCV values for all the traits except for pod clearance. This implied that there were minimal effects of the environment and high contribution of the genes in the phenotypic expression of the traits, except for pod clearance, which was more affected by the environment. Moderate GCV values of 13.45% and 13.49%, high heritability values of 70% and 69% and GAM values of 23.24% and 23.04% were recorded for grain yield and number of pods per plant, respectively. Only two principal components, PC1 and PC2 accounted for the variation, with a cumulative contribution of 68.25%. All the seven traits were useful in discriminating the genotypes as they had high eigenvalues in either PC1 or PC2. The 25 soybean genotypes were grouped into two main clusters, which were further sub-divided into eight sub-clusters based on the seven morphological characters. The genotypes TGx2014-5GM, checks SC Safari and SC Squire in sub-cluster 6 had the highest means of the most desirable traits (large seed size, high pod number per plant and seed yield). The three genotypes could be used in hybridisation programmes for improvement of grain yield, seed size and number of pods of the genotypes. Overall, the study identified soybean lines that could potentially be released as cultivars in the four southern African countries or used as parents in future soybean improvement programmes. It also revealed traits that could be used for indirect selection of seed yield and high genetic diversity among the genotypes for possible exploitation in soybean breeding programmes to increase seed yield.