Yield stability analysis of winter canola (brassica napus l.) Genotypes

Presence of genotype x environment interaction necessitates evaluation of genotypes in a wide range of environments to find desirable genotypes. Twenty four winter canola (Brassica napus L.) genotypes were evaluated in nine cold and temperate regions in 1999-2000 and 2000-2001 cropping seasons using a randomized complete block design with four replications. After homogeneity test for error variances, combined analysis of variance was performed assuming genotype effect as fixed, and years and locations effects as random. Effect of genotype was highly significant. Except the genotype x year interaction effect which was non-significant, other interaction effects were highly significant. The statistical methods that used for measuring yield stability were Eberhart and Russell's deviation from regression parameter, coefficient of variation (C.V.), Wricke's ecovalence , Shokla's stability variance and Finlay and Wilkinson regression coefficient. Analysis of the grain yield by the method of Eberhart and Russell showed significant difference for the genotype x environment (linear) interaction. According to the classification of genotypes based on the mean of grain yield, coefficient of regression and deviation from regression, genotypes SLM046 and Parade with 3.43 and 3.34 tha-1 of grain yield, respectively and with well adaptability in all environments, were selected as desirable genotypes. Meanwhile, using other stability analysis methods (coefficient of variation, Shokla's stability variance and Wricke 's ecovalence) introduced genotypes such as Parade, SLM046, Fornax, Licord and Colvert as more stable and adaptable genotypes which already confirmed the results obtained from Eberhart and Russell's methods