Evaluation of genotype × environment interaction on yield and oil content of rainfed safflower genotypes by regression and GGE Biplot

To analyze genotype × environment interactions (GEI) on yield and seed oil content of 15 safflower genotypes (including 8 breeding lines, 6 cultivars and one landrace), multi-environment trials were conducted in three sowing times (fall, entezari and spring) at four dryland agricultural research stations including Kermanshah, Maragheh, Kurdistan and Zanjan (totally 21 environments) during 2011-2013. The two models combined index (regression analysis + yield) and GGE biplot were used for stability analysis. The combined ANOVA revealed that genotype, environment, and GEI were highly significant for the two traits (P < 0.01). The sum of squares of genotype, environment, and GEI were accounted 5, 62 and 24% for yield, and 25, 45 and 22% for oil content.The result of regression analysis indicated that the regression components were significant in both traits. According to combined index, G5, G1, and G6 for yield, and G5, G3 and G4 for oil content were superior genotypes. The polygon view of the GGE biplot showed that all test environments were divided into 3 environmental groups for both traits. Two environments of Maragheh for yield and six environments of Kurdistan for oil content were favorable. Based on GGE biplot, G10 and G5 for yield and G11, G4 and G5 for oil content had a high combination of yield (or oil content) and stability. High rank correlation coefficients were obtained between regression and GGE biplot models in both traits. Based on the results of two models, G5 (Padideh) is superior in terms of yield and oil content and is recommended for all rainfed environments.