Abstract
AIM: To infer the geological and climatic factors that have shaped the genetic diversity and structure of a willow species (Salix humboldtiana) in three basins of Central Mexico. LOCATION: Central Mexico. METHODS: We collected samples from 11 populations across two hydrological basins (Balsas and Lerma) and one population from another basin (Ameca) within the Mexican Central Plateau (MCP). Individuals were analyzed using sequences of two chloroplast DNA (cpDNA) regions and eight nuclear simple sequence repeats (nSSR). Population genetic diversity and structure were determined from these data. To evaluate whether genetic structure was associated with ecological niche differentiation, we determined whether there is niche equivalence, overlap, or divergence between the Balsas and Lerma basins. Also, we evaluated the relative contributions of geographic distribution and climatic variation on population genetic structuring through redundancy analysis (RDA) and partial RDA. RESULTS: Both cpDNA and nSSRs data indicated the presence of three highly differentiated genetic groups, mostly geographically congruent with the three main hydrological basins. According to nSSRs, the three genetic groups can be further subdivided into eight subgroups corresponding to different rivers within the main basins. The niche equivalency test showed that the niches of the species in the Balsas and Lerma basins are significantly nonequivalent. The RDA indicated a significant association of genetic variation among populations with climate variables (particularly those related to the precipitation regime), while controlling for geographic distribution. MAIN CONCLUSIONS: The genetic structure of S. humboldtiana is strongly associated with the historical and current geological configuration of the basins and the rivers within basins. The observed hierarchical genetic differentiation can be due to gene flow limitation resulting from physical barriers to the dispersal of S. humboldtiana, but also to some degree of isolation by environment, as suggested by the significant association between genetic variation among populations and precipitation regime.