a study of yield stability in cold region ecotypes of alfalfa (medicago sativa l.) through non-parametric measures

selection of superior varieties of wide adaptability over diverse farming environments is of high importance, prior to any varietal recommendation. in order to study yield stability of alfalfa, 17 ecotypes were evaluated during 2004-2005 and 2005-2006 growing seasons at 10 research stations. each trial was carried out according to a completely randomized block design of 3 replications with fresh and dry yield traits being recorded. simple analysis of variance for each environment (combination of location×year) along with combined anova were carried out. the results of simple anova showed that there are significant differences among the studied ecotypes. the results of combined anova showed that year×location as well as ecotype×location× year interactions were significant at 1% probability level for both fresh and dry yield traits. in order to introduce stable ecotypes, several non-parametric measures including mean of rank, standard deviation of rank, fox et al. index, nonparametric methods proposed by tennarasue as well as nonparametric methods proposed by huehn were employed for both fresh and dry yield. overall, based on dry yield and according to standard deviation of rank and fox et al. index, ecotype 1 was found as the stable ecotype. in addition, the ecotypes 8 and 6 were the most stable ecotypes among the others according to huehn and tenarasue nonparametric measures. according to the results of all nonparametric methods and as based on fresh yield, ecotypes 10 and 11 were the most stable ecotypes. as a whole, and on the basis of different results of nonparametric methods and according to the dry mean yield, ecotype 8 was finally introduced as the most stable ecotype.