Patterns of genetic variability in central and peripheral populations of <Emphasis Type="Italic">Dipteryx alata</Emphasis> (Fabaceae) in the Brazilian Cerrado

This study tested whether genetic parameters in Dipteryx alata populations, estimated from genomic and chloroplastidial microsatellite markers, were distributed according to a central-peripheral model, inferring which factors drive this spatial distribution of genetic variability within populations. For each of the 23 populations sampled throughout the species’ range, the mean number of alleles per locus, expected heterozygosity and intrapopulation fixation indices were calculated using a rarefaction approach based on 54 alleles from 8 nuclear microsatellites. Explanatory variables were grouped into three subsets: the ecological suitability estimated by combining different techniques of ecological niche modeling, variables expressing human occupation, and a historical variable represented by the first eigenvector from the pairwise F _ST matrix based on cpDNA microsatellites. Each response variable was modeled using first (linear) and second (quadratic) order trend surface analysis (TSA). Multiple regressions were then used to evaluate the relative effects of the explanatory variables, based on AIC multi-model selection. In general, the genetic parameters did not follow a classical central-periphery model. Ecological suitability had a significance influence in all genetic parameters, so more suitable regions have higher genetic diversity and low endogamy. There was also a relationship between fixation indices and human impacts. The high genetic diversity in the southwestern region of Cerrado suggested that recent range expansion (after the Last Glacial Maximum) may also influenced the observed intrapopulation genetic patterns. Thus, complex combinations of both historical and ecological drivers, as well as contemporary human occupation, seem to drive current genetic composition within D. alata populations throughout its geographic range.