Different effects of elevation, habitat fragmentation and grazing management on the functional, phylogenetic and taxonomic structure of mountain grasslands

Ecological theory suggests that several nested environmental filters, acting at different spatial scales, shape the assemblages of local plant species. However, it is unclear whether different biodiversity components, that is, taxonomy, function and phylogenetic, respond similarly to these filters. The integration of these different components into coherent and comprehensive analytical frameworks also remains unclear. In this study, we developed an approach to test the relative effects of elevation, habitat fragmentation and grazing management on the functional, phylogenetic, and taxonomic structures of mountain pastures. Thirty summer farms, distributed along an elevational gradient were surveyed in the Italian Alps. Within these farms (around 100 ha), we identified all areas as one of two types of management (intensive or extensive), and we sampled three plots for each area, a total of 180 plots. Using the same mathematical framework, we quantified the functional (FD), phylogenetic (PD) and taxonomic (TD) diversity of each plot. We tested the influences of three environmental filters (elevation, habitat fragmentation and grazing management), using a series of partial regression analyses within a univariate and multivariate framework, as well as specific permutation schemes that accounted for our nested design. We found that elevation, habitat fragmentation, and grazing management affected the community structure, but in different ways. This finding confirmed that these filters operate at different scales and, despite some similarities, have different effects on various biodiversity components. Interestingly, FD was the only component that responded to all three types of predictors. Regarding functional aspects, elevation, as a broad-scale environmental gradient, showed a greater influence on dominant trait values, whereas at finer scales, grazing management had a primary effect on both dominant trait values and diversity measures. Habitat fragmentation showed a primary influence on TD, probably because the effect on dispersal limitations concerned mostly species availability. The hierarchical, multi-faceted approach adopted in this study yielded insights into the factors influencing biodiversity and community assembly processes in mountain pastures, thereby highlighting the importance of considering multiple facets of diversity in assessing the role of environmental filtering in vegetation structure.