β-diversity is defined as the complementarity in species composition between pairs of sites and plays a crucial role in biodiversity management and conservation, since it allows to detect those environmental gradients that functionally act in determining the variation of species composition. From an ecological viewpoint, the decay of β-similarity principally arises from: (i) a decrease in environmental similarity with distance, which inevitably creates a competition among species with different physiological abilities (niche difference model); (ii) a spatial configuration which creates spatial barriers and thus isolation among habitats, that influences species and gene movement across landscapes. To date, efforts were made to demonstrate the effect of distance on species similarity. However, few tests have been performed at local scale, i.e. at the community level. The aim of this study is to test the distance decay of β-similarity hypothesis at local scale (10x10 m plots) considering both (i) spatial distance between pairs of sites; (ii) the ecological distance between pairs of sites, measured as Euclidean distance in vector spaces defined by ecological variables (elevation, aspect, slope, assolation, spectral reflectance considering both the whole multispectral system of a Landsat ETM+ image and the Near Infrared Wavelength, band 4). β-similarity decay rates at different thresholds (τ) have been quantified by means of quantile regressions based on exponential models, focusing on the main factors influencing their trends.
