Graphical Analysis of Grain Yield Stability for Selection of Superior Faba Bean (Vicia faba L.) Genotypes

To investigate stability, pattern of G×E interaction, 10 faba bean genotypes (G1-G10) as well as four control cultivars; Barekat, Saraziri, Baloochi and Zereshki were evaluated using randomized complete block design with three replications in four agricultural research field stations of Gorgan, Dezful, Brojerd, and Iranshahr, for two cropping seasons (2014-15 and 2015-16). The results of composite variance analysis showed that the main effect of year, the interaction of year × location and the triple effect of year × location × genotype on grain yield were significant at the probability level of 1%. GGE -biplot analysis (genotype effect+ genotype×environment interaction) revealed that the two first and second principal explained 84.7% and 6.6% of total variation, respectively, and 91.3% of the grain yield variation. Based on a hypothetical ideal genotype biplot, the genotype G9 was better than the other genotypes across environments for stability and grain yield. View of polygon graph revealed superior mega-environments and the compatible genotypes were determined for each mega-environment; Gorgan- Brojerd (G9) and Iranshahr (G14). G9, G1, and G4 genotypes with average grain yield of 3479, 2808 and 2739 kgha-1, respectively, had the highest grain yield. Based on GEI and GGE-biplot analysis, Gorgan and Brojerd experimental environments had good differentiation ability. Finally, genotypes G1, G5, G9, and G10 were the most stable genotypes and can be used in breeding programs.