Updated 8/9/2013 Two files of microsatellite data are provided (A-B, below). Each file contains the entire data set, but subject to different data processing functions as necessary for appropriate analyses. All are tab-delimited text files suitable for importing into Excel or other spreadsheet programs. The remaining files contain additional information used to create the figures. A. File Taylor-Hoffman-2012-data-A_origenotypes.txt, contains genotypes in the form of repeat numbers, i.e., with primer sizes subtracted the result divided by the repeat size. This data was used for analyses in which association with original genotypes was necessary, as described in the paper. B. In the file Taylor-Hoffman-2012-data-B_nullcorr.txt, the data from "-A_origenotypes.txt" has been corrected for the presence of null alleles, as described in the publication. Because of the techniques used for null allele correction, data lines represent the collected alleles found within a population, but associations between different loci are random and do not necessarily represent the genotypes of specific individuals. These data were used primarily for calculating summary statistics representing populations or differences between or among populations. C. File "Taylor-Hoffman-2012-data-C_clustering.txt contains the aggregate clustering percentages used to create Figure 2. This data was generated in Structure and aggregated using Clumpp, as described in the publication. D. File Taylor-Hoffman-2012-data-D_Geodistances.txt contains geographical distance matrices used in production of Figure 3, and for Mantel tests.