GENETIC VARIABILITY AND ASSOCIATION OF TRAITS FOR YIELD AND YIELD RELATED TRAITS AMONG SORGHUM GENOTYPES (SORGHUM BICOLOR L. MOENCH) UNDER DROUGHT STRESS CONDITIONS

Ethiopia is the center of origin and domestication for sorghum with a wide range of sorghum collections for various agro-ecologies. However, there are many factors that hinder the production and productivity of sorghum. Drought is one of the most important factors that affect crop production worldwide and continues to be a challenge to plant breeders, despite many decades of research. Understanding the genetic variability and trait association is key to the development of improved sorghum cultivars. Therefore, the field experiment was conducted to estimate genetic variability, association of traits, and determine their direct and indirect effect on grain yield. A set of 72 sorghum genotypes advanced from a pedigree breeding approach was used in this study. The experiment was laid out using Row-Column design at Miesso Agricultural Research Station during 2021 main season. R software was used to analyze the data. The analysis of variance indicated that there were significant variations between the tested genotypes for all traits studied. Grain yield varied from 2.10 t/ha to 6.32 t/ha. Genotypic and phenotypic coefficient of variation ranged from 0.56% (stay green score) to 23.88% (harvest index) and 0.66% (stay green score) to 28.99% (harvest index) respectively. Broad sense heritability ranged from 25.56% (drought tolerance score) to 86.87% (grain filling rate) while genetic advance as percentage of mean ranged from 1.11% (drought tolerance score) to 43.40% (grain filling rate) respectively. Grain yield had positive and highly significant correlation with panicle weight, panicle yield, stand count and grain filling rate at both genotypic and phenotypic level. Grain yield had also negative and highly significantly correlation with days to flowering at both genotypic and phenotypic level and days to maturity at genotypic level. Path coefficient analysis indicated that grain filling rate exerted the highest positive direct effect on grain yield at both genotypic and phenotypic level. Principal component analysis (PCA) revealed that five principal components with Eigenvalue greater than unity accounted 74.1% of total variation. The first two PCs accounted for a cumulative of 39.5% of total variation indicating most of the important yield and yield attributing traits were present in these first two PCs. Cluster analysis grouped the test genotypes into five clusters. Cluster I, II, III, IV and V accounted for 41.667%, 6.944%, 26.389%, 16.667% and 8.333% of the tested genotypes in that order. The highest intra-cluster distance was observed for cluster V whereas the maximum inter-cluster distance, to be used as source for hybridization program was observed between cluster IV and V. The lowest intra cluster distance was observed for cluster III, whereas cluster I and III showed the lowest inter cluster distance. Overall, the variation observed among the tested genotypes indicates a potential to improve sorghum genotypes for drought tolerance and productivity but more tests are required to get a conclusive result.