Selection of Superior Traits and Diverse Parents in Maize (Zea mays L.) Based on Multivariate Analysis

Abstract

An experiment comprising 40 maize (Zea mays L.) genotypes was conducted in a randomized complete block design with three replications at research farm of Faculty of Agriculture Science and Technology, AKS University, Satna (M.P.) India. Highly significant (p ≤ 0.05) differences among genotypes were observed for all the traits studied. The genotypic and phenotypic coefficients of variation ranged from 2.40-32.34% and 2.89-32.68%, respectively, with relatively higher estimates for seed yield per plant, tassel length and seed index, indicating substantial genetic dispersion for these traits. Broad-sense heritability was high to very high for most characters (65.00–99.20%), and genetic advance as percent of mean ranged from 4.12 to 65.91%. Correlation analysis revealed that seed yield per plant was positively and highly significantly associated with cob length, cob diameter, cob weight, number of seeds per cob, number of cobs per plant and days to maturity, whereas days to 50% silking was negatively associated with yield, indicating the importance of sink traits and flowering synchrony in yield expression. Path analysis showed high explanatory power at both genotypic and phenotypic levels (R² = 0.9102 and 0.9018), with days to maturity, cob length/cob diameter and number of seeds per cob exhibiting positive direct effects on yield, confirming these as robust indirect selection criteria. Mahalanobis D2 analysis grouped genotypes into six clusters with maximum divergence between Cluster IV and Cluster VI, indicating scope for exploiting heterosis by selecting parents from highly divergent clusters.