Parametric and nonparametric measures of phenotypic stability in linseed (Linum usitatissimum L.)

A combined analysis with three parametric and two nonparametric measures to assess G × E interactions and stability analyses to identify stable genotypes of linseed across 18 environments in Ethiopia were undertaken. The combined analysis of variance for environments (E), genotypes (G) and G × E interaction was highly significant (p<0.01), suggesting differential responses of the genotypes and the need for stability analysis. The parametric stability measures of coefficient of variability and the stability variance showed that R12-N10D was the most stable genotype, whereas cultivarsʹ superiority measure indicated Chilalo to be the most stable cultivar. Like most of the parametric methods, the non-parametric measures revealed that R12-N10D had the smallest changes in ranks and thus was the most stable genotype in contrast to R12-D24C, which was unstable and the lowest yielder. A comparison of the five stability measures showed that the coefficient of variability, stability variance and variance of ranks were similar in assessing the relative stability of the genotypes, whereas cultivarsʹ superiority measure deviated from the others. The stability variance and variance of ranks were significantly rank correlated, and were the best in determining the comparative stability of linseed genotypes. The coefficient of variability was also relatively better than the cultivarʹs superiority measure. Further studies of repeatability tests are, however, needed to determine the best methods. The stability statistics generally identified R12-N10D, followed by Chilalo, as the most stable varieties, whereas R12-D24C and R11-M20G were the least stable varieties.