Pre-breeding in alfalfa germplasm develops highly differentiated populations, as revealed by genome-wide microhaplotype markers

Plant genebanks contain large numbers of accessions that likely harbor useful alleles or genes absent in commercial plant breeding programs. Broadening the genetic base of commercial alfalfa germplasm with these useful genetic variations can be accomplished by screening the extensive genetic diversity in germplasm collections and enabling maximal recombination among selected genotypes. In this study we surveyed the genetic diversity and differentiation of germplasm pools selected in northern US latitudes (USDA Plant Hardiness Zone seven or less) originating from Eurasian germplasm. The germplasm evaluated here included four BASE populations (C0) from different geographical origins (CASIA, EURO, OTTM, SYBR), 20 cycle-one populations (C1) generated from each of the four BASE populations selected in five locations in the USA and Canada, and four commercial cultivars. A panel of 3,000 SNP Diversity Array Technologies (DArTag) markers harboring ~ 12,000 microhaplotypes were used to quantify genetic diversity and population structure. Principal Component Analysis and Discriminant Analysis of Principal Components identified substantial population structure among the alfalfa populations based on their geographical origin, while the check cultivars formed a central cluster. Inbreeding coefficients (F _IS ) ranged from − 0.1 to 0.006, and 27 out of 28 populations had negative F _IS values, indicating an excess of heterozygotes. Interpopulation genetic distances were calculated using Rho and analysis of molecular variance (AMOVA) parameters. Pairwise population Rho values ranged from 0.007 to 0.336. All BASE populations had the lowest Rho values compared to C1 populations and check cultivars. AMOVA found high variance among individuals within populations and low variance between populations. Variation among population was highest among check cultivars and lowest in BASE populations at 10.6% and 7.3% of total molecular variation, respectively. This study shows that BASE populations have high gene diversity, low interpopulation distances, and minimal inbreeding which is required for base-broadening selection.