Integrative genome-wide association and haplotype-based analyses reveal genetic structure and local adaptation in Korean landrace soybeans

Soybean (Glycine max L.) is a major crop valued for both food and industrial applications. The genetic diversity preserved within landrace accessions serves as a critical resource for improving agronomic traits and enhancing adaptation to climatic variability. Korean landrace soybeans, shaped by long-term cultivation across diverse local environments, provide an ideal population for dissecting genetic variation underlying key traits and regional adaptation. A genome-wide association study (GWAS) was performed on 1,693 Korean landrace soybean accessions genotyped with 67,222 SNPs from the 180 K Axiom® Soya SNP array. Population structure and genetic diversity were assessed using model-based stratification (K = 15), principal component analysis (PCA), linkage disequilibrium (LD) decay analysis (r² = 0.2 at 309 kb), and FST-based differentiation. GWAS using the MLMM, FarmCPU, and BLINK models identified 38 significant SNPs associated with flowering date (FD), maturity date (MD), number of seeds per pod (NOSPP), and 100-seed weight (SW), including pleiotropic loci. Candidate genes within LD blocks were annotated and subjected to Gene Ontology (GO) enrichment, revealing biological processes such as signaling, transcriptional regulation, sterol metabolism, and cell division. Haplotype analysis supported phenotypic differentiation among allelic groups. For SNPs located outside of LD blocks, variation in allele frequencies and trait values across regions indicated evidence of local adaptation. These findings provide new insights into the genetic architecture of agronomic traits in Korean landrace soybeans. The identified loci, functionally annotated candidate genes, and region-specific alleles constitute a valuable genomic resource for breeding programs aiming to develop climate-resilient and regionally adapted soybean cultivars.