GWAS Fine-Mapping for Soybean First Pod Insertion Height Identifies Candidate Genomic Regions Involved in Plant Development

Background/Objectives: Soybean first pod insertion height (FPIH) is a key trait established during plant development, but its genetic architecture in Eurasian germplasm remains largely unknown. Methods: We performed a GWAS and fine-mapping for FPIH using 180 Eurasian varieties (SoySNP50K array, imputed to ~4M SNPs) phenotyped in four Russian environments (2021–2022), using two models: a residual-based and a covariate-adjusted. SuSiE fine-mapping was applied to refine candidate loci. Results: The covariate-adjusted model demonstrated better control of genomic inflation (λ = 1.08 vs. 1.16) and higher SNP heritability (0.381 vs. 0.014); the lower heritability in the residual-based model was expected, as this model removes environmental main effects and the genetic variance associated with them. Thus, the models are complementary: one captures stable genetic effects, the other highlights environment-dependent signals. SuSiE refined three loci. On chromosome 13, two stable independent signals (Gm13_30553403, Gm13_30909346; PIP ≥ 0.99) were identified near MYB83 and BEN1, consistent across both models. On chromosome 4, the signal shifted to Gm04_36933515 (PIP = 0.9999) near COBL4/IRX6, although model dependency was observed, and this locus is not currently recommended for marker development. On chromosome 5, the original GWAS SNP was not causal; two tightly linked SNPs (Gm05_38427309 and Gm05_38710046, r2 = 0.714, PIP ≥ 0.99) formed a haplotype block located near ARR1/ARR2 and a B-box/CCT domain gene. Crucially, this signal was absent in the residual-based model (max PIP = 0.33), suggesting that the chromosome 5 locus may modulate developmental plasticity rather than exerting a direct main effect on FPIH. However, as we did not perform formal G × E testing, we present this as a hypothesis requiring further validation. The environment-dependent behavior of this locus indicates that it should be used with caution in breeding programs and validated under specific target environments. Conclusions: Chromosome 13 SNPs provide stable genetic associations across environments, whereas the chromosome 5 block requires environment-specific validation. This work provides the first fine-mapped GWAS for FPIH in Russian soybean germplasm and highlights how model choice can uncover or obscure environment-dependent genetic loci affecting plant development.