Non parametric measures to estimate GxE interaction of dual purpose barley genotypes for grain yield under multi-location trials

Abstract

GxE interaction of seventeen dual purpose barley genotypes evaluated at ten major barley locations of the country by non parametric methods. Non parametric measures had been well established and expressed ad-vantages over their counter parts i.e. parametric measures. Simple descriptive measures based on the ranks of gen-otypes i.e. Mean of ranks (MR) pointed towards RD2925 and BH1008 and standard deviation of ranks (SD) for KB1401 and UPB1054 whereas Coefficient of variation (CV) for JB322 and RD2925 as stable genotypes. Nonpara-metric measures based on original values (Si1, Si2, Si3, Si4, Si5, Si6, Si7) indicated the stable performance of NDB1650, JB322 and UPB1054 while UPB1053, RD2715, RD2927 and RD2035 were observed of unstable nature. CSi1, CSi2, CSi3, CSi4, CSi5, CSi6 and CSi7 measures based on the ranks of corrected grain yield identified JB322, RD2552, RD2925 and NDB1650 as stable genotypes. Spearman’s rank correlation established highly significant positive correlation of yield with SD (0.67), Si1(0.65), Si2(0.59), Si5(0.68), Si7(0.67) whereas negative association observed for CMR (Mean of corrected ranks) (-0.62), CMed (Median of corrected ranks) (-0.60). NPi(2) expressed negative correlation with CV(-0.32), Si6 (-0.30), CMR(-0.34) and CMed(-0.48). More over NPi(3) maintained negative correlation with most of the measures though the magnitude was of low magnitude.