Race structure within the Mesoamerican gene pool of common bean (Phaseolus vulgaris L.) as determined by microsatellite markers

Common bean (Phaseolus vulgaris L.) cultivars are distinguished morphologically, agronomically and ecologically into specific races within each of the two gene pools found for the species (Andean and Mesoamerican). The objective of this study was to describe the race structure of the Mesoamerican gene pool using microsatellite markers. A total of 60 genotypes previously described as pertaining to specific Mesoamerican races as well as two Andean control genotypes were analyzed with 52 markers. A total of 267 bands were generated with an average of 5.1 alleles per marker and 0.297 heterozygosity across all microsatellites. Correspondence analysis identified two major groups equivalent to the Mesoamerica race and a group containing both Durango and Jalisco race genotypes. Two outlying individuals were classified as potentially of the Guatemala race although this race does not have a defined structure and previously classified members of this race were classified with other races. Population structure analysis with K = 1-4 agreed with this classification. The genetic diversity based on Nei's index for the entire set of genotypes was 0.468 while this was highest for the Durango-Jalisco group (0.414), intermediate for race Mesoamerica (0.340) and low for race Guatemala (0.262). Genetic differentiation (G ST) between the Mesoamerican races was 0.27 while genetic distance and identity showed race Durango and Jalisco individuals to be closely related with high gene flow (N m) both between these two races (1.67) and between races Durango and Mesoamerica (1.58). Observed heterozygosity was low in all the races as would be expected for an inbreeding species. The analysis with microsatellite markers identified subgroups, which agreed well with commercial class divisions, and seed size was the main distinguishing factor between the two major groups identified.