STABILITY ANALYSIS BY RANKS AND SIMULTANEOUS ESTIMATION OF YIELD AND YIELD STABILITY IN BREAD WHEAT

Abstract

Genotype x environment (GE) interactions, in particular leading to different rankings of the genotypes in variable environments, are a major challenge to plant breeders. Numerous methods of characterizing stability of genotypes across environments are available. The objectives of this investigation were to study non-parametric measures of stability based on the ranks of genotypes in different environments and simultaneous evaluation of yield and yield stability by means of Hühn's S_i⁽¹⁾, S_i⁽²⁾ ,S_i⁽³⁾, S_i⁽⁶⁾. Jensen's H_i⁽⁵⁾ statistics and two indices. The indices were derived from the sum of the two ranks; X_i. rank + W_i⁽²⁾   rank, and r_i + S_i⁽²⁾, rank, respectively. The data used in estimating these statistics were obtained from the Official Winter Bread Wheat Registration Trials with 15 genotypes tested at 13 environments (year-location combinations). The S_i⁽¹⁾, and S_i⁽²⁾ statistics selected higher yielding genotypes. The S_i⁽³⁾and S_i⁽⁶⁾ statistics, however, appeared to give more weight to stability than did S_i⁽¹⁾, and S_i⁽²⁾ . Rank correlation coefficients indicated that low S_i⁽³⁾, S_i⁽⁶⁾, and W_i⁽²⁾were generally associated with low yield. Jensen 's H_i⁽⁵⁾, statistic and Index 2 (sum of mean rank (r_i) and S_i⁽²⁾, rank) showed a high degree of association with yield. Repeatability of stability statistics were studied by measuring rank correlations between high-yield and low-yield environments. The repeatability of Index 2 between high-yield and low-yield environments was relatively high and significant. Repeatability of the other statistics between the two groups of environments was negligible.