Comparing the Potential of Indices and Multivariate Statistical Techniques to Select Drought Tolerant Genotypes in Barley (hordeum Vulgare L.)

A multi-trial research was performed to compare some multivariate statistical methods and various stress tolerance indices with the aim of introducing an optimal method for selecting drought-tolerant genotypes of barley. Ten diverse varieties of barley were assayed during four cropping years under irrigated and rain-fed conditions. Drought tolerance and susceptibility indices were calculated. These indices included the stress susceptibility index, mean productivity, tolerance, stress tolerance index, geometric mean productivity, harmonic average productivity, yield (YI), yield stability and linear regression coefficient (b). The studied varieties showed significant differences (p ≤ 0.01) in terms of grain yield and its components. Multivariate statistical techniques including discriminant function analysis and factors analysis along with stress tolerance score (STS) as a function of all conventional indices were applied in order to select high-yield and drought-tolerant varieties. Based on the results of discriminant function analysis, factor analysis and STS, the ‘Kavir’ variety was more tolerant to drought stress and generated the highest yield, compared to other varieties during each year of the four cropping years. The similarity of obtained results from the various methods revealed that STS index due to easier calculation and more accurate than other statistical analyses and indices can be considered as an integrated criterion to identify drought-tolerant genotypes in barley and a broad spectrum of grain crops over all the world.