Abstract

The associations among yield components and their direct and indirect influence on the grain yield of finger millet were investigated. For this purpose, eighty-eight finger millet (Eleusine coracana (L.) Gaertn.) genotypes were tested using an augmented randomized complete block experimental design with two replications at Adet Agricultural Research Center in 2008. Accordingly, phenotypic and genotypic correlations among the traits and their path coefficients were estimated. Grain yield was significantly correlated with its component characters like plant height (r_p=0.446^** and r_g=0.574^**), number of ears per plant (r_p=0.364^** and r_g=0.443^**), number of fingers per ear (r_p=0.329^** and r_g=0.532^**), finger length (r_p=0.361^** and r_g=0.426^**), biomass yield (r_p and r_g=0.839^**), harvest index (r_p=0.336^** and r_g=0.476^**) and thousand kernel weight (r_p=0.225^* and r_g=0.267^*). Phenotypic path analysis showed biomass yield (0.835) and finger length (0.159), number of fingers per ear (0.016), and number of ears per plant (0.038) to exert positive direct effects on grain yield, while plant height, days to heading and days to maturity exhibited negative direct effects. Genotypic path analysis also revealed that biomass yield (2.240), number of tillers per plant (0.359) and finger length (0.242) exerted positive direct effects on grain yield. Thus, the correlation analysis showed plant height, number of ears per plant, number of fingers per ear, finger length, biomass yield, harvest index and 1000 kernel weight to be important yield components that can be used to improve the yield potential of finger millet genotypes.

Key words: Finger millet, path coefficient, yield related traits, correlation.