Study of statistic stability to select high-yielding and stable peach genotypes

Abstract

In peach breeding, suitable identification methods for performance stability studies as well as the associations between stability parameters are poorly understood. Therefore, the aims of this work were to compare parametric (S2 ij , biS2 xi, Wi, i and Ii) and non-parametric (S(1) i , S(2) i , S(3) i and Pi) stability measures, evaluate the level of association among them, select superior accessions and identify major environmental variables as causes of yield variation among years. Fruit yield stability was studied using data of fruit yield from 25 peach genotypes under three environments, arranged in a completely randomized design with three replications. Frosts, chilling, heat, rainfall and the interactions among them were considered as explanatory variables of yield variation through years. Crossover was the main effect of genotype-byenvironment interaction indicating that the selection of high-yielding and stable peach genotype would be a laborious task for breeders. The interaction between rainfall and heat accumulation during fruit development period explained 96.7% of yield variation among years. Yield (Yi) exhibited negative correlation with Wi and i, while Wi showed negative association with Pi. S(1) i , S(2) i , and S(3) i were positively associated with each other, showing that just one of these three statistics would be sufficient to select stable accessions although they were not correlated with Yi. Fruit yield was positively correlated with Pi and Ii, these two measures were also positively associated with each other, and therefore, only one of them would be enough for selection of superior peach accessions. Both, Pi and Ii statistics revealed that accessions Sunprince, Flameprince, María Aurelia, Vega, Starlite and Flavorcrest were the most stable and high-yielding genotypes across environments