GENETIC VARIABILITY IN BREAD WHEAT (Triticum aestivum L.) GENOTYPES FOR YIELD, YIELD RELATED TRAITS AND PROTEIN CONTENT AT HARAMAYA, ETHIOPIA

Abstract

The study of genetic variability is the critical step in crop improvement programs either to use the natural variation that already exists or produce genetic variation. This study was conducted to assess the genetic variability of 20 bread wheat genotypes and determine associations among yield, yield related traits and protein content. The experiment was laid out in randomized complete block design with three replications during 2017 at Raare research site of Haramaya University. The results from analysis of variance revealed the presence of significant difference (P<0.01) among 20 bread wheat genotypes for 14 quantitative traits but not for days to grain filling, number of productive tillers and number of kernels per spike. Moreover, the genotypes had mean grain yield in the range between 5564.2 kg/ha for (ETBW7790) and 7945.7 kg/ha for (variety Jalanne) with overall mean of 6582.73 kg/ha; and in the range of 9.20 for (Qulqulluu) to 12.87for (ETBW7790) for grain protein content. The estimates of phenotypic (PCV) and genotypic (GCV) coefficient variations ranged from 1.96 to 20.05% and 1.37 to 13.91%, respectively, in which the lowest and the highest estimates were computed for hectoliter weight and peduncle length, respectively. A high and moderate values of PCV and GCV were computed for peduncle length, whereas moderate PCV and GCV values were for spike length, thousand kernel weight and harvest index. The estimated values of broad sense heritability (H 2 B) and genetic advance as percent of mean (GAM) ranged from 48.14 to 92.74 and 1.98 to 20.69%, respectively. High H 2 B and GAM values obtained for spike length while high H 2 B coupled with moderate GAM values for number of spikelet per spike, thousands kernel weight, biomass yield, harvest index, grain protein content and wet gluten content suggesting improvement of the traits through selection could be fairly easy. Grain yield had positive and significant association with harvest index but negative and significant correlations with spike length both at genotypic and phenotypic levels. Harvest index and spike length also exerted positive and negative direct effects on grain yield, respectively, both at genotypic and phenotypic levels. Both spike length and harvest index had positive indirect effects on grain yield through thousands kernel weight and hectoliter weight at genotypic level. This suggested harvest index but not spike length could be considered for indirect selection of genotypes for high yield. The first five principal components explained cumulatively 82.76 % of the total variation observed among bread wheat genotypes. The Euclidean distance for all possible pairs of 20 bread wheat genotypes ranged from 1.54 to 8.60 and genotypes were grouped into eight distinct clusters. Cluster III consisted of five genotypes (25%), Cluster I, IV and VII consisted of each three genotypes, Cluster II and VIII constructed each by two genotypes and Cluster V and VI consisted of only G10 and Jalanne variety, respectively. The selection to be made from genotypes grouped under Cluster I, VI and VIII, consisted genotypes that had high yield and protein content. In conclusion, it was observed the presence of considerable variations and genetic divergence among bread wheat genotypes that could be a good opportunity for breeders to develop variety (ies) either by further evaluation of high performing genotypes or through Heterosis breeding by crossing of genotypes with varied desirable traits of different clusters.