Evaluation of adaptability and seed yield stability of soybean (Glycine max L. Merril) promising lines using GGE biplot analysis

Selection of adapted genotypes with high seed yield and yield stability is the goal of soybean breeding programs. To evaluate the adaptability and seed yield stability of soybean promising lines, 19 promising lines and cv. Williams as check were evaluated using randomized complete block design with four replications in four locations: Karaj, Gorgan, Moghan and Khoramabad in Iran during two growing seasons (2013 and 2014). GGE biplot analysis was employed to evaluate the adaptability and seed yield stability. Combined analysis of variance showed thatyear, location, genotype, year × location, year × genotype, location × genotype and genotype × location × year interaction effects were significant on studied traits. The contribution of year, location and genotype variance to total variance was 0.01, 0.60 and 0.02, respectively, indicating considerable contribution of location variance. The first two components of PC1 and PC2 explained overall 58% of total observed variation of genotype and genotype × environment (G + GE). In this study, three mega-environments were identified. The first mega-environment included: E2 (Karaj 2014), E5 (Moghan 2013) and E8 (Gorgan 2014) and G16 was the superior genotype in this mega-environment. The second mega-environment included: E3 (Khorramabad 2013) and E4 (Khorramabad 2014) and G8 was the superior genotype in this mega-environment. Third mega-environment consisted: E1 (Karaj 2013) and E7 (Gorgan 2013) and G17 was the superior genotype in this mega- environment. Biplot analysis showed that genotypes: G17 (L85-3059) with 2702 kg.ha-1 and G16 (L12/Chaleston × Mustang) with 2750 kg.ha-1 were highly adapted genotypes with high seed yield and yield stability. The E7 environment (Gorgan, 2013) was the most desirable environment in respect to its discriminating ability among soybean genotypes and the best representative of the target environments.