Main Article Content
Assessment of yield stability in sorghum
Abstract
genotypes in eight environments (Locations x years combination). There were significant differences among the genotypes, the environments and GxE interactions. Thus, the three types of univariate stability models: Type-1 (CVi and S2i), Type-2 (W2i, s2i, and bi) and Type-3 (Sd2i) were compared for ranking of the genotypes. The
parameters of W2i and si2 had perfect positive correlation (r=1.0) and strong positive correlation with bi (r=0.80),but either weak or no correlation with the rest of the parameters. Similarly, CVi and S2
i had strong rank correlation(r=0.97) but both had either very weak or no rank correlation with the rest of the parameters tested. The Sd2
i had very weak negative correlation with the remaining parameters. Based on the three stability statistics, the different genotypes were classified as stable. To compliment and verify findings of this univariate approach, the GxE which uses a mulivariate approach was used. The multivariate approach (AMMI model) gives a broader inference. Based on the AMMI model, genotypes 2 and 5 were the most stable, although genotypes 1 and 3 had satisfactory levels of yield performance as well as stability. Therefore, these four genotypes with wider adaptation
are recommended for sorghum growing dry lowlands of the country.