GENETIC VARIABILITY, INTERRELATIONSHIP AND PATH ANALYSIS FOR YIELD IMPROVEMENT OF RICE GENOTYPES

Abstract

An investigation was carried out to study the genetic variability, association among the yield component traits, their direct and indirect effects on the yield. All the genotypes showed considerable amount of variation in their mean performances with respect to the characters studied, indicates presence of sufficient variability and scope for further selection and breeding superior and desirable genotypes. Genotypic coefficient of variation played a major role for the expression of the traits and ranged from 4.15 (days to maturity) to 27.58 (grain yield per plant). Heritability is ranged from 18.00 (flag leaf length) to 92.17 (number of spikelets per panicle). High genetic advance was observed for the trait, number of spikelets per panicle (99.83) coupled with high heritability (92.17), indicating the preponderance of the non-additive gene action; suggesting that hybridization breeding will be effective. Biological yield per plant (0.78. 0.72), harvest index (0.57, 0.58) and number of spikelets per panicle (0.18, 0.18) were positively associated with grain yield at both genotypic and phenotypic levels respectively, indicating improvement in these traits will increase the grain yield. Number of panicle per plant (4.33, 0.08), biological index per plant (1.04, 0.80), harvest index (0.90, 0.67), panicle length (0.23, 0.03) and test weight (0.02, 0.03) had positive direct effect on yield at both genotypic and phenotypic levels respectively, indicating importance of these characters and which can be strategically used to improve the yield of rice grains. The present study concluded that the genotypes CB 08-534 and CRK-27 were identified as best genotypes for grain yield.