Soybean (Glycine max) is a photoperiod-sensitive and self-pollinated species. Days to flowering (DTF) and maturity (DTM), duration of flowering-to-maturity (DFTM) and plant height (PH) are crucial for soybean adaptability and yield. To dissect the genetic architecture of these agronomically important traits, a population consisting of 309 early maturity soybean germplasm accessions was genotyped with the Illumina Infinium SoySNP50K BeadChip and phenotyped in multiple environments. A genome-wide association study (GWAS) was conducted using a mixed linear model that involves both relative kinship and population structure. The linkage disequilibrium (LD) decayed slowly in soybean, and a substantial difference in LD pattern was observed between euchromatic and heterochromatic regions. A total of 27, 6, 18 and 27 loci for DTF, DTM, DFTM and PH were detected via GWAS, respectively. The Dt1 gene was identified in the locus strongly associated with both DTM and PH. Ten candidate genes homologous to Arabidopsis flowering genes were identified near the peak single nucleotide polymorphisms (SNPs) associated with DTF. Four of them encode MADS-domain containing proteins. Additionally, a pectin lyase-like gene was also identified in a major-effect locus for PH where LD decayed rapidly. This study identified multiple new loci and refined chromosomal regions of known loci associated with DTF, DTM, DFTM and/or PH in soybean. It demonstrates that GWAS is powerful in dissecting complex traits and identifying candidate genes although LD decayed slowly in soybean. The loci and trait-associated SNPs identified in this study can be used for soybean genetic improvement, especially the major-effect loci associated with PH could be used to improve soybean yield potential. The candidate genes may serve as promising targets for studies of molecular mechanisms underlying the related traits in soybean.

Key message Twenty-two loci for soybean SW and candidate genes conditioning seed development were identified; and prediction accuracies of GS and MAS were estimated through cross-validation and validation with unrelated populations. Abstract Soybean (Glycine max) is a major crop for plant protein and oil production, and seed weight (SW) is important for yield and quality in food/vegetable uses of soybean. However, our knowledge of genes controlling SW remains limited. To better understand the molecular mechanism underlying the trait and explore marker-based breeding approaches, we conducted a genome-wide association study in a population of 309 soybean germplasm accessions using 31,045 single nucleotide polymorphisms (SNPs), and estimated the prediction accuracy of genomic selection (GS) and marker-assisted selection (MAS) for SW. Twenty-two loci of minor effect associated with SW were identified, including hotspots on Gm04 and Gm19. The mixed model containing these loci explained 83.4 % of phenotypic variation. Candidate genes with Arabidopsis orthologs conditioning SW were also proposed. The prediction accuracies of GS and MAS by cross-validation were 0.75–0.87 and 0.62–0.75, respectively, depending on the number of SNPs used and the size of training population. GS also outperformed MAS when the validation was performed using unrelated panels across a wide range of maturities, with an average prediction accuracy of 0.74 versus 0.53. This study convincingly demonstrated that soybean SW is controlled by numerous minor-effect loci. It greatly enhances our understanding of the genetic basis of SW in soybean and facilitates the identification of genes controlling the trait. It also suggests that GS holds promise for accelerating soybean breeding progress. The results are helpful for genetic improvement and genomic prediction of yield in soybean.

Soybean seed weight is not only a yield component, but also a critical trait for various soybean food products such as sprouts, edamame, soy nuts, natto and miso. Linkage analysis and genome-wide association study (GWAS) are two complementary and powerful tools to connect phenotypic differences to the underlying contributing loci. Linkage analysis is based on progeny derived from two parents, given sufficient sample size and biological replication, it usually has high statistical power to map alleles with relatively small effect on phenotype, however, linkage analysis of the bi-parental population can’t detect quantitative trait loci (QTL) that are fixed in the two parents. Because of the small seed weight difference between the two parents in most families of previous studies, these populations are not suitable to detect QTL that have considerable effects on seed weight. GWAS is based on unrelated individuals to detect alleles associated with the trait under investigation. The ability of GWAS to capture major seed weight QTL depends on the frequency of the accessions with small and large seed weight in the population being investigated. Our objective was to identify QTL that had a pronounced effect on seed weight using a selective population of soybean germplasm accessions and the approach of GWAS and fixation index analysis. We selected 166 accessions from the USDA Soybean Germplasm Collection with either large or small seed weight and could typically grow in the same location. The accessions were evaluated for seed weight in the field for two years and genotyped with the SoySNP50K BeadChip containing >42,000 SNPs. Of the 17 SNPs on six chromosomes that were significantly associated with seed weight in two years based on a GWAS of the selective population, eight on chromosome 4 or chromosome 17 had significant Fst values between the large and small seed weight sub-populations. The seed weight difference of the two alleles of these eight significant SNPs varied from 8.1 g to 11.7 g/100 seeds in two years. We also identified haplotypes in three haplotype blocks with significant effects on seed weight. These findings were validated in a panel with 3753 accessions from the USDA Soybean Germplasm Collection. This study highlighted the usefulness of selective genotyping populations coupled with GWAS and fixation index analysis for the identification of QTL with substantial effects on seed weight in soybean. This approach may help geneticists and breeders to more efficiently identify major QTL controlling other traits. The major regions and haplotypes we have identified that control seed weight differences in soybean will facilitate the identification of genes regulating this important trait.

Achieving appropriate maturity in a target environment is essential to maximizing crop yield potential. In soybean [Glycine max (L.) Merr.], the time to maturity is largely dependent on developmental response to dark periods. Once the critical photoperiod is reached, flowering is initiated and reproductive development proceeds. Therefore, soybean adaptation has been attributed to genetic changes and natural or artificial selection to optimize plant development in specific, narrow latitudinal ranges. In North America, these regions have been classified into twelve maturity groups (MG), with lower MG being shorter season than higher MG. Growing soybean lines not adapted to a particular environment typically results in poor growth and significant yield reductions. The objective of this study was to develop a molecular model for soybean maturity based on the alleles underlying the major maturity loci: E1, E2, and E3. We determined the allelic variation and diversity of the E maturity genes in a large collection of soybean landraces, North American ancestors, Chinese cultivars, North American cultivars or expired Plant Variety Protection lines, and private-company lines. The E gene status of accessions in the USDA Soybean Germplasm Collection with SoySNP50K Beadchip data was also predicted. We determined the E allelic combinations needed to adapt soybean to different MGs in the United States (US) and discovered a strong signal of selection for E genotypes released in North America, particularly the US and Canada. The E gene maturity model proposed will enable plant breeders to more effectively transfer traits into different MGs and increase the overall efficiency of soybean breeding in the US and Canada. The powerful yet simple selection strategy for increasing soybean breeding efficiency can be used alone or to directly enhance genomic prediction/selection schemes. The results also revealed previously unrecognized aspects of artificial selection in soybean imposed by soybean breeders based on geography that highlights the need for plant breeding that is optimized for specific environments.