Study of Genotype × Environment Interaction and Yield Stability of Tobacco Genotypes Using AMMI and Tai analysis

To determine the yield stability and adaptability of the genotype of tobacco, 5genotypes of flue-cured tobacco were evaluated in experiment using a randomizedcompletely block design (RCBD) with three replications at two locations includingRasht and Tirtash tobacco Research Centers (IRAN), during the growing season of2008-2010 (four environment). The interaction of genotype × environment in AMMIand Tai analysis method was significant (p 0.01). In Tai regression analysis method ofgenotype 1, 3, 5 were in part stability. But genotype (G3) had a high stability comparingthe others because of having near to 1 and near to zero and also because of a higheraverage yield than the total average yield. In AMMI1 model, genotype 3 was introducedas a stable genotype, and also in AMMI2 model, the first two principle component axes(PCA 1 and 2) were significant (p 0.01) and the first Interaction Principal ComponentAxis (IPCA1) of the interaction captured 87% of the interaction sum of squares. TheSecond Interaction Principal Component Axis (IPCA2) explained a further 12% of GEIsum of squares. And cumulatively contributed 99% of the total GEI. Thus, in this modelAMMI2, the first two principle components were considered. And the other PCA wereconsidered as noise. The biplot technique was used to identify appropriate genotype forspecial locations. Result show that genotype 3 (Caspian z1) with the highest interactionwere the most stable genotype. Furthermore, genotype 3 was known as more suitablefor Tirtash (S2) and Rasht (S1) locations. In general, in this study, genotype 3 wasknown as the most suitable genotype on the base of Tai regression analysis and AMMImethod.