Ecological niche models as hypothesis generators of functional genetic differentiation and potential local adaptation in a Mediterranean alpine ecosystem

Geographically disparate populations within a species range may show important differences including variation in ecological, demographic, genetic and phenotypic characteristics. Based on the Center-Periphery Hypothesis, it is often assumed that environmental conditions are optimal in the geographic center of the range and stressful or suboptimal at the periphery, implying ecological marginality is concordant with geographic periphery. But this assumption has been challenged as geographical and ecological gradients are not necessarily concordant. The conservation value of populations inhabiting environmentally marginal areas is still under debate and is closely related with their evolutionary potential. Strong selective pressures caused by stressful conditions may generate novel adaptations in marginal areas, conferring these populations distinct evolutionary potential. But populations inhabiting marginal areas may also show reductions in neutral and adaptive genetic diversity via drift and inbreeding. In this work we explore the potential of ecological niche models (ENMs) to identify environmentally optimal and marginal areas, as well as the principal putative selective pressures likely to act. To do so, we built a carefully parameterized ENM of Silene ciliata, a dominant plant species of Mediterranean alpine habitats. Complementarily, we selected wild populations inhabiting contrasting environmental conditions and carried out common garden experiments to detect genetic differentiation among populations associated with functional traits. With the resulting information, we tested whether environmentally marginal populations defined by the ENM had genetically differentiated phenotypes that are potentially adaptive and, thus, of conservation value. We found genetically based phenotypic differentiation of phenological traits between populations inhabiting areas identified by the ENM as marginal and optimal, as well as between populations with different habitat suitability values. Results supported ENMs as powerful tools for determining environmental marginality and identifying selection pressures, and thus also as hypothesis generators for divergent selection. Furthermore, genetically based phenotypic differentiation found underlines the potential adaptive value of populations inhabiting marginal areas. The approach developed here provides a theoretically justified and practical way to study adaptive processes and provide insights about the conservation value of marginal populations.