Is the USDA core collection of common bean representative of genetic diversity of the species, as assessed by SNP diversity?

Core collections are envisioned to be a representative subset of larger germplasm collections. They were introduced to facilitate the characterization and use of these germplasm collections. The common bean (Phaseolus vulgaris L.) core collection of the USDA Western Regional Plant Introduction Station was one of the first collections to be established in the early 1990s. Here, we evaluate the representativity of this common bean collection in light of the availability of a single nucleotide polymorphism (SNP) platform and new information about genetic diversity of the species, including phaseolin and seed type data. The SNP diversity was studied with a combination of STRUCTURE, principal coordinate analysis (PCoA), and neighbor-joining analysis (NJA). STRUCTURE analyses were conducted for K (number of subpopulations) = 3 and K = 7, based on the ad hoc statistic Delta K. The K = 3 analysis recognized the split between Andean and Mesoamerican domesticates and the subdivision of the Mesoamerican domesticates into high- (Durango/Jalisco) and low-altitude (Mesoamerica) ecogeographic races. The K = 7 analysis further subdivided the Andean group identified for K = 3, as well as the high-altitude group from the Mesoamerican gene pool. It also identified smaller groups consisting of Mesoamerican wild beans. The PCoA and NJA confirmed the STRUCTURE results and highlighted the existence of presumed hybridization among groups. Our results suggest that this core collection should be updated by adding domesticated categories, developing a separate wild common bean core collection, and developing cores for specific purposes.