GENETIC VARIABILITY AND ASSOCIATION AMONG YIELD AND YIELD RELATED TRAITS IN LOWLAND SORGHUM [(Sorghum bicolor (L.) Moench)] GENOTYPE AT FEDIS, EASTERN ETHIOPIA

Sorghum is one of the major staple food crops in eastern Ethiopia where the area is known with an uncertainty of rainfall. Therefore, this study was conducted to assess genetic variability among sorghum genotypes, and to estimate associations among yield and yield-related traits of sorghum genotypes. Thirty-four sorghum genotypes and two improved varities in 6x6 simple lattice designs at Fedis Agricultural Research Center in 2018 cropping season. Results of analysis of variance revealed the presence of significant differences among genotypes for 12 quantitative traits and two qualitative traits. Grain yield varied from 1634.00 to 6522.00 kg ha-1with mean grain yield of 3390.4 kg ha-1and six genotypes gave maximum grain yield in the range between 5434 and 6522.00 kg ha-1 with the non-significance difference among them. A total of 58.3% and 63.8% had white seed color and straight panicle shape, respectively.Phenotypic(PCV%) and genotypic(GCV%)coefficient of variations ranged from 7.17to 37.65% and 7.24 to 43.74%, respectively, and the lowest and highest PCV and GCV estimates were registered for biomass yield and harvest index, respectively. Heritability (H2)in broad sense ranged between 74% (harvest index) and 99.8% (number of seed per panicle) and genetic advance as percent of the mean (GAM) ranged from 14.63% (biomass yield) and 71.42% (number of seed per panicle). High estimates of H2coupled with high to moderate GAM for grain yield, plant height, panicle length, panicle width, number of seed per panicle, panicle weight, biomass yield, and harvest index. These traits had a positive and significant genotypic and phenotypic correlation with grain yield. Moreover, leaf area, harvest index and biomass yield exerted high direct effects on yield both at genotypic and phenotypic levels. These traits also had positive genotypic and phenotypic indirect effects on yield via panicle weight and number of seeds per panicle. suggested the traits could be used for selection of genotypes for high yield. The first four principal component axis (PCA) accounted for 72.36% variation of among genotypes. The first principal component (PC1) had the largest contribution of 32.6% to the observed total variation among sorghum genotypes in which days to flowering, days to maturity, leaf area, number of seed per panicle, grain yield, and panicle weight had a higher load to the PC1. Thousand kernel weight, harvest index and biomass yield had the largest contribution to PC3. The genotypes were grouped into 13 distinct clusters of which Cluster II and IV consisted of 8(22.2%) and 7(19.4%)genotypes, respectively, Cluster IX, XI, and XII consisted of each two genotypes, Cluster XIII consisted of two released checks varieties (Teshale and Gubiye. It could be concluded that the observed genetic variations among genotypes for yield and yield-related traits potentially could be exploited in future breeding activities of sorghum in the study area.