Application of Additive Main Effects and Multiplicative Interaction and Biplot Graphical Analysis Multivariate Methods to Study of Genotype- Environment Interaction on Safflower Genotypes Grain Yield

To evaluate the effect of different environments on grain yield, 25 genotypes of safflower along with one control namely Kouse were studied in three locations (Tehran, Birjand and Isfahan) and three years (from 2015 to 2017), in each using a completely randomized block design with three replications. The results obtained from AMMI analysis demonstrated that the main effects of genotype, environment and genotype × environment interaction were highly significant. The first four principal components justified around 83% of the sum of squares of the interactions, and explained 40%, 19%, 13% and 9% of variances, respectively. Drawing the biplot of the first principal component and the average yield for genotypes and environments suggested that genotypes including Kouse, E2417, PI-250537 with higher than average yields and desired stability were selected. Biplot of the first two principal components showed that the interaction between genotypes Kouse, Hartman and PI-250537 were identified as genotypes with good compatibility.  Based on GGE Biplot method, two main components of the spatial regression models explained about 96% of grain yield variability. Biplot of relevance between environments confirmed the existence of almost identical reaction between genotypes in both locations including Isfahan and Tehran, but genotypes in Birjand showed a different reaction than these two environments. Based on the polygonal figure, genotypes including A2 in Birjand and Hartman in Isfahan as well as Tehran had a higher degree of narrative compatibility and performance stability. Genotype of Hartman and Isfahan location were the most ideal genotype and environment compared to other genotypes and environments.