Abstract:
Okra is a traditional vegetable crop in northwestern Ethiopia but it has not given research attention and considered a minor crop. Thus this research was conducted to characterize and evaluate okra landraces and to estimate genotypic, phenotypic variability and genetic divergence and to estimate heritability and genetic advance under selection, degree of genotypic and phenotypic associations among yield and yield related traits. A total of 35 okra landraces from three districts (Guba, Mandura and Dangure) were evaluated for 23 agro-morphological and eight qualitative traits in 2017 at Pawe Agricultural Research Center in randomized complete block design. Results of analysis of variance showed significant differences among okra landraces for all traits and the landraces also distributed in different categories of qualitative traits. The landraces had number of tender fruits per plant ranged from 4.86 to 36.54 and 2.49 to 21.98 t ha-1 mean fruit yield per hectare. The estimates of genotypic (GCV) and phenotypic coefficients of variation (PCV) for 22 traits of 35 okra landraces were in the range between 2.95 and 54.92 and 4.96 and 55.22%, respectively. The broad sense heritability (H2 ) and genetic advance as percent of mean (GAM) estimated in the range between 35.36 and 99.6% and 3.62 and 112.66%, respectively. High GCV, PCV, H2 and GAM was estimated for plant height, number of primary branches/stem, internodes length, number of matured fruits per plant, weight of matured fruits per plant, dry weight of matured fruits/plant, number of seeds per fruit, number of tender fruits per plant and yield per hectare indicated the high heritability was due to the close correspondence between the genotypic and phenotypic variations as a result of relatively small contribution of the environment to the phenotype expression of the traits. Fruit yield per hectare had positive and significant genotypic and phenotypic correlation coefficients with weight of matured fruits per plant, dry weight of matured fruits per plant, hundred seed weight, number of tender fruits per plant and leaf length. Fruit yield per hectare had positive and significant phenotypic correlation coefficients with number of ridge and peduncle length. These traits also had positive direct effects on fruit yield at genotypic and phenotypic levels except leaf length exerted negative direct effect on yield at phenotypic level. This suggested direct and simultaneous selection of genotypes for yield and these traits is possible. Results of principal component analysis indicated the first four principal component axes (PCA1 to PCA4) accounted 65.59% of the total variation, of which PCA1 and PCA2 had larger contribution of 22.09 and 19.34%, respectively. The 35 okra landraces were grouped into 12 distinct clusters from Euclidean distances matrix using Unweighted Pair-group Methods with Arithmetic Means (UPGMA) of which Cluster IX consisted of 6 (17.14%), Cluster II, VI and X consisted each five and other clusters consisted of 1 and 3 landraces. The study results showed the presence of genetic variation among landraces for all traits suggested that selection could be effective to develop okra varieties for high fruit yield and other traits.