Indian Journal of Genetics and Plant Breeding Genetics
Genotype × Environment Interaction and Yield Stability of Maize Hybrids Across the North-Western Himalayas Using AMMI and GGE Biplot Analyses
Abstract
Multi-environment testing is needed to distinguish high yield from repeatable performance in the environmentally heterogeneous Kashmir Valley. We evaluated 20 coded maize hybrids across 12 location-year environments (six locations in 2024 and 2025) in randomized complete block trials with three replications. Plot-level grain yield was analyzed using fixed-effects ANOVA, AMMI and environment-centered GGE biplots. Environment, genotype and genotype × environment interaction were highly significant ( p < 0.001) and accounted for 41.96%, 35.75% and 21.39% of the corrected total sum of squares, respectively. AMMI IPCA1 and IPCA2 explained 39.51% and 23.35% of interaction variation. G5 had the highest marginal mean yield (7.58 t ha⁻¹) but ranked 13th for ASV, giving YSI = 14. G11 combined high yield (6.79 t ha⁻¹) with ASV = 0.332 and had the lowest YSI (7); G19, G14 and G15 each had YSI = 11. Independently calculated GGE PC1 and PC2 explained 78.55% of G + GE variation. All environment markers fell in the G5 sector in the two-PC which-won-where display, whereas observed means identified G6 at ENV7–ENV8 and G19 at ENV11, showing the limits of the two-dimensional approximation. ENV12 had the longest GGE vector (1.420) but a 44.1° angle to the average-environment axis; ENV1 and ENV5 were more representative by angle. Thus, G11 provided the strongest yield–stability compromise, whereas G5 was the highest-yielding candidate across the evaluated environments. Independent multilocation and multiyear validation is required before cultivation recommendations or mega-environment delineation.