Evaluation of genotype‐by‐environment interaction for grain yield in elite soybean ( Glycine max ) germplasm by AMMI and GGE modeling

Soybean ( Glycine max ) grain is among the world's most important agricultural commodities. Globally, genotype‐by‐environment interaction (GEI) significantly affects soybean production, reducing selection accuracy and limiting genetic gain. The objective of this work included the examination of elite, commercially active soybean cultivar performance across seven production regions within the state of Michigan for a preliminary characterization of GEI patterns and mega‐environment structure. Nine genotypes were tested for grain yield during the 2025 season. Additive‐main‐effect‐and‐multiplicative‐interaction (AMMI) and genotype‐plus‐genotype‐by‐environment (GGE) biplots were used to inspect genotype stability performance and classify environments. The AMMI analysis of variance revealed that grain yield was significantly ( p < 0.001) affected by genotype, environment, and GEI, each explaining approximately 16.39%, 44.47%, and 13.76% of observed variation, respectively. Based on the AMMI and GGE models, genotypes DF_EX25E2.3 and DF_2311E3 exhibited minimal GEI response paired with exceptional yield performance across locations. Genotype DF_214N also showed favorable grain yield, though it was relatively less stable across environments. Among test environments, Lenawee and St. Joseph County sites were shown to be most favorable for yield response, while the Hillsdale County site provided significant downward phenotypic pressure. AMMI‐II analysis showed St. Joseph, Hillsdale, and Allegan to represent the most discriminative test environments, while Ingham was found to be the least. The AMMI model also showed strong response correlations between St. Joseph and Lenawee, as well as between Sanilac and Allegan sites. These results serve as preliminary evidence for near‐term commercial cultivar recommendations and the potential refinement of Michigan‐based soybean variety trial design.