Red clover (Trifolium pratense) is one of the most important fodder crops worldwide. The knowledge of genetic diversity among red clover populations, however, is under development. This study provides insights into its genetic diversity, using single nucleotide polymorphism (SNP) markers to define population structure in wild and cultivated red clover. Twenty-nine accessions representing the genetic resources available at NordGen (the Nordic gene bank) and Lantmännen (a Swedish agricultural company with a red clover breeding program) were used for this study. Genotyping was performed via SeqSNP, a targeted genotype by sequencing method that offers the capability to target specific SNP loci and enables de novo discovery of new SNPs. The SNPs were identified through a SNP mining approach based on coding sequences of red clover genes known for their involvement in development and stress responses. After filtering the genotypic data using various criteria, 623 bi-allelic SNPs, including 327 originally targeted and 296 de novo discovered SNPs were used for population genetics analyses. Seventy-one of the SNP loci were under selection considering both Hardy-Weinberg equilibrium and pairwise FST distributions. The average observed heterozygosity (HO), within population diversity (HS) and overall diversity (HT) were 0.22, 0.21 and 0.22, respectively. The tetraploids had higher average HO (0.35) than diploids (0.21). The analysis of molecular variance (AMOVA) showed low but significant variation among accessions (5.4%; P < 0.001), and among diploids and tetraploids (1.08%; P = 0.02). This study revealed a low mean inbreeding coefficient (FIS = −0.04) exhibiting the strict outcrossing nature of red clover. As per cluster, principal coordinate and discriminant analyses, most wild populations were grouped together and were clearly differentiated from the cultivated types. The cultivated types of red clover had a similar level of genetic diversity, suggesting that modern red clover breeding programs did not negatively affect genetic diversity or population structure. Hence, the breeding material used by Lantmännen represents the major genetic resources in Scandinavia. This knowledge of how different types of red clover accessions relate to each other and the level of outcrossing and heterozygosity will be useful for future red clover breeding.

Red clover is a highly valuable crop for the ruminant industry in the temperate regions worldwide. It also provides multiple environmental services, such as contribution to increased soil fertility and reduced soil erosion. This study used 661 single nucleotide polymorphism (SNP) markers via targeted sequencing using seqSNP, to describe genetic diversity and population structure in 382 red clover accessions. The accessions were selected from NordGen representing red clover germplasm from Norway, Sweden, Finland and Denmark as well as from Lantmännen, a Swedish seed company. Each accession was represented by 10 individuals, which was sequenced as a pool. The mean Nei’s standard genetic distance between the accessions and genetic variation within accessions were 0.032 and 0.18, respectively. The majority of the accessions had negative Tajima’s D, suggesting that they contain significant proportions of rare alleles. A pairwise FST revealed high genetic similarity between the different cultivated types, while the wild populations were divergent. Unlike wild populations, which exhibited genetic differentiation, there was no clear differentiation among all cultivated types. A principal coordinate analysis revealed that the first principal coordinate, distinguished most of the wild populations from the cultivated types, in agreement with the results obtained using a discriminant analysis of principal components and cluster analysis. Accessions of wild populations and landraces collected from southern and central Scandinavia showed a higher genetic similarity to Lantmännen accessios. It is therefore possible to link the diversity of the environments where wild populations were collected to the genetic diversity of the cultivated and wild gene pools. Additionally, least absolute shrinkage and selection operator (LASSO) models revealed associations between variation in temperature and precipitation and SNPs within genes controlling stomatal opening. Temperature was also related to kinase proteins, which are known to regulate plant response to temperature stress. Furthermore, the variation between wild populations and cultivars was correlated with SNPs within genes regulating root development. Overall, this study comprehensively investigated Nordic European red clover germplasm, and the results provide forage breeders with valuable information for further selection and development of red clover cultivars.