EXPLOITATION OF HETEROTIC POTENTIAL AND COMBINING ABILITY FOR GENETIC ENHANCEMENT OF GRAIN YIELD IN MAIZE

Abstract

The present investigation was undertaken to evaluate the performance, character association and stability of newly developed maize (Zea mays L.) genotypes across diverse environments. The experimental material was evaluated at three environments, viz., E₁ (Kolhapur), E₂ (Gurugram (Delhi NCR)) and E₃ (Jalna). Data were recorded for phenological, morphological and yield-contributing traits and subjected to analysis of variance, correlation analysis, path coefficient analysis, environmental index analysis and stability analysis following the Eberhart and Russell model. Considerable variability was observed among the genotypes for the characters studied. Pooled grain yield per plant ranged from 74.66 to 103.64 g, indicating substantial scope for selection. Among the environments, Kolhapur was comparatively favourable for grain yield, with an environmental index of 0.99, whereas Gurugram (Delhi NCR) was relatively unfavourable (−1.03) and Jalna was nearly average (0.04). Genotypic correlation analysis revealed that plant height had a consistent negative association with grain yield across Kolhapur, Gurugram (Delhi NCR) and Jalna, with correlation coefficients of −0.353, −0.293 and −0.312, respectively. Cob length and 100-seed weight also exhibited predominantly negative associations with grain yield. In contrast, number of kernels per row showed positive genotypic associations with grain yield at Kolhapur and Gurugram (Delhi NCR), although the relationship was negative at Jalna. Path coefficient analysis revealed considerable environmental variation in the contribution of yield components. Plant height showed a strong negative direct effect under Kolhapur and Jalna, whereas number of kernels per row exhibited a high positive direct effect (0.573) under Gurugram (Delhi NCR). Stability analysis demonstrated differential responses of genotypes to environmental changes. Genotypes possessing high mean performance, regression coefficient close to unity and low deviation from regression were considered broadly adapted, while genotypes with regression coefficients greater or less than unity were identified as relatively responsive to favourable or unfavourable environments, respectively. The study demonstrated that simultaneous consideration of mean performance, character association, direct effects and stability parameters is essential for identifying promising maize genotypes for wider and specific adaptation.