Nonparametric stability analysis in multi-environment trial of canola

Abstract

Rapeseed is the world s second most important source of vegetable oils. Development of genotypes having highseed yield with stable performance is of paramount importance. In the present investigation seventeengenotypes were grown at seven locations during two growing seasons in semi-cold regions of Iran. Datarecorded on seed yield were subjected to different nonparametric measures which do not requiredistributional assumptions. Results of nonparametric tests of G, E and GE interaction and a combinedANOVA across environments showed there w ere both cross over and non-cross over interactions for G and Eand only non-cross over type for GE interaction. In this study, high values of Top (proportion of environmentsin which a genotype ranked in the top third) and mean of rank were associated with high mean yield. HoweverRank-sum measure was successful to detect genotypes showing simultaneous high yield and yield stability.Cluster analysis and principal component (PC) analysis help to group genotypes and measures and theyrevealed association among different statistics. Finally, among nonparametric measures, Top, Si(1) and Rank-sum statistics of nonparametric procedures were found to be useful in detecting the stability of the genotypes .According to these parameters Geronimo was found as stable and high yield canola genotype.