Deciphering genetic variability and trait associations in Nepalese finger millet (Eleusine coracana L. Gaertn) landraces for yield improvement

Abstract Knowledge of genetic variability and the degree of association between yield and yield-attributing traits is fundamental for effective selection and yield improvement in breeding programs. Thus, eighteen finger millet landraces and one released cultivar (Kabre Kodo 2) were evaluated in a randomized complete block design with three replications. Ten key agromorphological traits were assessed to study genetic variability and characterize trait associations. Analysis of variance (ANOVA) revealed significant differences (p ≤ 0.05) for all the studied traits among the landraces. The phenotypic coefficient of variation (PCV) was greater than the genotypic coefficient of variation (GCV) for all the studied traits. Higher heritability (h2 b), coupled with genetic advance as a percentage of the mean (GAM), was observed for ear head length, thousand-seed weight, and grain yield. The grain yield was positively correlated with finger width, thousand-seed weight, and ear head weight at both the genotypic and phenotypic levels (p ≤ 0.05). Path analysis at both the genotypic and phenotypic levels revealed that the thousand-seed weight and number of fingers had the greatest positive direct effects on grain yield. Thus, future breeding programs should employ selections based on ear head weight, finger width, number of fingers, and thousand-seed weight to improve yield in finger millet.