The demographic inferences summarized in Table 1 suggest strong bottlenecks with little Ginsenoside-F5 
subsequent recovery of size in the non-German populations. Moreover, although most loci show strong genetic structure, differentiation is lower with the German population. Pairwise comparisons also reveal a high proportion of shared variants and few fixed differences between Germany and other populations. Even populations as different genetically and geographically as Canada and Russia each possess extensive shared variation with Germany, suggesting that the nonGerman populations sampled represent subsets of the diversity in Germany. Consistent with this, all of our pairwise comparisons show a higher proportion of unique variants in Germany. Both FST and Bayesian cluster analyses reveal unusually strong population structure for an outcrossing herbaceous species, providing little evidence for recent admixture or gene flow, but suggesting long-term persistence of isolated populations. This finding is supported by analysis of an alternate demographic model that explicitly estimated low pairwise migration between Germany and other populations. It is possible, of course, that migration from unsampled populations or species contributes to observed patterns of diversity. One would expect such migration to increase both diversity and LD, but our data show higher LD only in non-German populations with lower levels of diversity. Although the data to explicitly test this hypothesis are not currently available, our sequence data provide no compelling evidence that migration from unsampled populations has strongly affected our sampled populations. Although our demographic model does not aim to infer a definitive history, it is important to consider how inclusion of nonequilibrium processes may affect estimation of divergence times. Our estimates are much lower than calculations based solely on median pairwise FST values, which yields divergence times ranging from,90,000 years between Germany and Butenafine hydrochloride Iceland to,170,000 years between Germany and Russia. However, our estimates are considerably older than the end of the most recent Ice Age, when Northern Europe was most likely re-colonized by A. lyrata. We note, however, that the 95% credible intervals of our estimates generally include times as recent as 10,000 years ago, and that because tS estimates in years are proportional to the mutation rate.