Assessment of Yield Stability in Common Bean Cultivars Based on Univariate and Multivariate Methods

Stability and adaptability of ten common bean cultivars were studied in water stressed and normal conditions, three planting dates (including 5 May, 20 May and 5 June), and two years (2014 and 2016) in Shahrekord, Iran. Combined analysis of variance for yield indicated that the effect of genotype by environment interaction was significant; therefore, more accurate statistical analyses were required to evaluate yield stability of each cultivar. Stability analysis was performed using some univariate methods and a multivariate method (AMMI model). Comparison of univariate and multivariate stability parameters showed that AMMI model is more effective and practical than univariate methods. In AMMI model, graphical discrimination of the genotypes and the environments facilitated the identification of stable and high yielding genotypes. In this regard, Chiti cardinal genotype had high yield and wide adaptability to diverse environments. This genotype had the minimum amount of the first and the second interaction principal components scores (IPC1 and IPC2) in the AMMI model as well as the minimum value of AMMI stability parameters including EV4 and SIPC4; hence this genotype is introduced as a genotype having general stability. In contrast, genotype Akhtar had a high yield potential; however, in biplot graph this genotype was located close to IPC1 and away from IPC2, which implies that Akhtar has specific adaptability to some environments. Biplot graph also showed that among the different environments which were involved in this experiment, year had not fundamental effect on various performance of genotypes. In the second planting date, the genotypes performances were more consistent than the first planting date; moreover, the genotypes responses under moisture stressed conditions were different and unpredictable. In conclusion, sowing of Cardinal genotype in second planting date (20 May) under well-irrigated conditions is recommended for the Shahrekord in central west of Iran.