Non-parametric stability analyses of protein content in multi-environment trials of wheat (T. aestivum L.)
According to literature, a detailed paper has not been published yet on using non-parametric stability statistics for evaluating genotypic stability in protein content (PC) of wheat. Thus, this study aimed to investigate the stability for PC of wheat using sixteen non-parametric stability measures (YSD-PC standard deviation, RM-Rank mean, RSD-Rank?s standard deviation, RS-Rank Sum stability statistic, PA-Percentage of adaptability, R1 and R2-Range indexes, TOP-Ranking, Si(1), Si(2), Si(3), Si(6), NPi(1), NPi(2) NPi(3)and NPi(4) rank statistics, together with Y-PC mean). The study included 13 wheat genotypes, consisting of 5 registered cultivars and 8 breeding lines, selected from National Wheat Breeding Program of Turkey. The genotypes were grown in ten rain-fed environments, representative of major rain-fed wheat-growing areas of Turkey, during 2011-2013 cropping seasons. The ANOVA showed that the effects due to environments (E), genotypes (G) and GE interaction (GEI) were significant (P < 0.01). Spearman?s rank correlation and principal component analyses (PCA) also revealed that two types of associations were found between the stability parameters: the first type included Si(1), Si(2), Si(3), Si(6), NPi(1), NPi(2) NPi(3), NPi(4), RSD and YSD parameters which were related to static stability, whereas the second type consisted of the Y, RM, TOP, PA, RS, R1 and R2 parameters which are related to dynamic concept of stability. Among the 8 breeding lines, G7 and G8 were the best genotypes in terms of both high PC and stability. In conclusion it could be suggested that dynamic non-parametric stability statistics should be used for selecting genotypes with high PC and stable when tested across a wide range of environments.