Adaptation and Stability Analysis in Some Bread Wheat Genotypes

Genotype x environment interaction is a challenging issue to plant breeders in improving high-yielding, stable genotypes for variable environments. Several methods of measuring stability of genotypes tested across a range of environments have been proposed. The grain yields of fifteen bread wheat ( Triticum aestivumL.) genotypes grown at eight different locations were used to assess adaptation and stability characteristics of the genotypes and to study associations among these parameters. According to most of the parameters used in the present study, Kiraç 66, Es-kbvd-15 and Ank-92-1 appeared to have a good level of general adaptation to all environments. Mean yield was highly correlated with S i 2 , b i , b i , and a i . The regression coefficients b i , b i , and a i were perfectly rank correlated (r=1.00) with each other as well as with S i 2 and CV i . Eberhart and Russell's Sd i 2 statistic was highly associated (r=0.93**) with Tai's l i . The stability statistics S i (2) and S i (3) of Hühn showed a good rank correlation with CV i , W i 2 , and Sd i 2 . The statistics based on the ranks of genotypes over environments, such as S i (2) and S i (3) , would be a potential alternative to the parametric approaches currently used.

Adaptation and Stability Analysis in Some Bread Wheat Genotypes

Genotype x environment interaction is a challenging issue to plant breeders in improving high-yielding, stable genotypes for variable environments. Several methods of measuring stability of genotypes tested across a range of environments have been proposed. The grain yields of fifteen bread wheat ( Triticum aestivumL.) genotypes grown at eight different locations were used to assess adaptation and stability characteristics of the genotypes and to study associations among these parameters. According to most of the parameters used in the present study, Kiraç 66, Es-kbvd-15 and Ank-92-1 appeared to have a good level of general adaptation to all environments. Mean yield was highly correlated with S i 2 , b i , b i , and a i . The regression coefficients b i , b i , and a i were perfectly rank correlated (r=1.00) with each other as well as with S i 2 and CV i . Eberhart and Russell's Sd i 2 statistic was highly associated (r=0.93**) with Tai's l i . The stability statistics S i (2) and S i (3) of Hühn showed a good rank correlation with CV i , W i 2 , and Sd i 2 . The statistics based on the ranks of genotypes over environments, such as S i (2) and S i (3) , would be a potential alternative to the parametric approaches currently used.