Dryland Agricultural Research Institute, Kermanshah, Iran.
Evaluating cultivars in the presence of unpredictable genotype × environment interaction (GEI) is a major problem in crop breeding programs. This study was conducted to: (1) evaluate seed yield stability of 15 safflower genotypes grown in three consecutive seasons (2005-08) at three locations in Iran and (2) investigate the inter-relationships among eight agronomic traits using the genotype × trait (GT) biplot technique. The additive main effect and multiplicative interaction (AMMI) analysis, the stability parameters derived from the AMMI model, Shukla’s analysis, genotype plus GE (mean performance + stability) distance (GGED), yield-stability (YS) statistic, and several common stability procedures were used to analyze GEI. Results of the AMMI analysis showed that main effects due to genotype (G), environment (E), and GEI, as well as the first six interaction principal component axes (IPCA1 to 6), were significant (P0.01). The partitioning of GEI also revealed that both heterogeneity caused by the environmental index and heterogeneity caused by residual GEI were important sources of variation. According to rank correlation analysis, stability parameters can be categorized into three distinct sections related to static and dynamic concepts of stability. The parameters of the three groups ranked the genotypes in different ways. Genotypes G12, G9, G11 and G8 were superior in terms of both stability and high yield. The GT biplot revealed that seed yield was positively and significantly associated with seed weight and heads per plant, but oil content was not correlated with seed yield.