Evaluation of Genetic Variability in Different Genotypes of Finger Millet (Eleusine coracan L.) through Path and Multivariate Analysis
Dinesh T.,Kangabam S.,Swapnil,Dash G. K.,Sil P.,Panda K. K.,Mitra, D.,Purkaystha S.
DOI: https://doi.org/10.9734/ajsspn/2024/v10i3321
2024-06-11
Asian Journal of Soil Science and Plant Nutrition
Abstract:The experiment was carried out using a Randomized Block Design (RBD) with three replicates. The trials were entirely composed of fifty-two genotypes for yield and their contributing characters for 13 different characters. For every attribute under study, an analysis of variance showed a highly significant difference. Among the genotypes studied, the evaluated characteristics showed varying degrees of heritability, genetic advancement, and variability. Both the genotypic and phenotypic coefficients of variation (PCV and GCV, respectively) ranged from low to high. The modest GCV and PCV values particular for the percentage of fingers per ear (17.47% and 17.50%), ear weight per plant (g) (13.42% and 16.28%), and 1000 seed weight (g) (11.79% and 14.09%), were observed respectively. Plant height (cm) (81.60%, 28.47) had the greatest wide sense heritability value, with the highest genetic advance, followed by ear weight per plant (g) (67.95%, 20.13). Thus, the findings of this study indicate that these genotypes have diversity in yield and other yield-related features, which should be exploited in subsequent breeding. The correlation and path coefficient analyses for yield and yield characteristics showed that grain yield per plant showed a strong positive relationship both at the genotypic and phenotypic levels, with days to 50% flowering, days to maturity, ear weight per plant, and harvest index. Path analysis exhibited that for genotypic and phenotypic coefficients, harvest index (0.33, 0.46) and ear weight/plant (0.86, 0.39) showed the most positive direct effects. These characteristics might thus be applied as selection criteria to identify finger millet genotypes that show promising results in future breeding. Multivariate approaches such as principal component analysis and cluster analysis are crucial statistical tools for examining genetic diversity in plant breeding programs, as are their significant quantitative features. Based on principal component analysis, the entire variance was provided by six main components, of which PC-1 and PC-2 contributed 21.97% and 18.32%, respectively, to the total variability. Euclidean distance was used to divide the fifty-two genotypes into five groups. The largest number of genotypes (twenty-two) in Cluster I was followed by Cluster III, with 16 genotypes. As a result, genotypes from these clusters may be used as parents for hybridization.