Plant breeding has been concerned with genotype by environment (GE) interaction and high yielding genotypes with stable performance are desirable while this target is difficult to achieve due to high environmental variations and unpredictable GE interaction. Stability of grain yield performance of 18 barley genotypes was evaluated at 5 locations for 3 years in the rainfed conditions and it was studied through 25 nonparametric stability methods. Four nonparametric tests indicated highly significant GE interaction due to differential performance of genotypes across fifteen environments. Regarding mean yield and six Hühn’s statistics, genotype G12 (1946 kg ha-1) was the most favorable genotype while based on the RN1, G4, G10 and G11 were the most stable genotypes. Genotypes G4, G8 and G10 were the most favorable genotypes according to rank-sum while genotypes G2, G13 and G18 were the most favorable genotypes based on nonparametric superiority. The relative interactivity index identified G4, G16 and G8 as the most stable genotypes while the genotypic classification identified G1, G2, G13 and G18 as the most stable genotypes. Clustering of the nonparametric stability methods indicated that there were two groups with different static and dynamic characteristics. In this study, five nonparametric stability methods as GC, FM, PA, RN2 and KR2 were associated with high grain yield and reflected the dynamic concept of stability, but the other twenty nonparametric stability methods were not positively correlated with yield and characterized a static concept of stability. Finally, genotypes genotype G13 (2114.13 kg ha-1) and G18 (2062.69 kg ha-1) were found to be the most favorable genotypes and are thus recommended for commercial release in semiarid areas of Iran.
