Yield stability analysis and multi-trait selection of maize (Zea mays L.) hybrids in multi-environment trials using integrated AMMI, BLUP, WAASB, and MTSI models

Abstract The methods utilized to analyze genotype-by-environment interaction (GEI) and assess the yield stability of crop varieties are constantly evolving, favoring integrated multi-method analytical frameworks. In this study, an integrated BLUP-AMMI-WAASB-MTSI framework was deployed to evaluate kernel yield (KY) stability and multi-trait performance in maize. Ten genotypes (nine elite experimental hybrids and one commercial check, CHK) were evaluated in a randomized complete block design across 18 environments (nine locations over two cropping seasons). Combined analysis of variance revealed significant effects for location, genotype, and spatio-temporal interactions (LG and YLG) on KY. AMMI analysis decomposed the multiplicative interaction into four significant interaction principal components (IPCs), explaining 81.70% of the total GEI variation. Under the mixed-model (REML/BLUP) framework, high genotypic mean heritability (0.69) and selective accuracy (0.83) were achieved. According to the WAASB quadrant biplot and WAASBY superiority index (50:50 yield: stability weighting), hybrids H2, H3, and H1 combined top-tier predicted BLUP yields (13.35, 12.95, and 12.80 t ha-1) with broad environmental stability, outperforming the check cultivar (CHK: 12.65 t ha-1). For multi-trait selection across 10 traits, exploratory factor analysis (FA) partitioned the traits into four dimensions. Based on the multi-trait stability index (MTSI) at a 30% selection intensity, hybrids H2, H6, and H3 were selected as ideal ideotypes, achieving positive genetic gains for KY (+3.91%) and grain-filling duration (+2.20%), alongside desirable reductions in plant height (-1.55%) and ear height (-0.47%) for lodging resistance. Overall, hybrids H2 and H3 were identified as consensus elite germplasm for commercial deployment.