Abstract
We explored if beta diversity and the distance decay of community similarity depend on spatial resolution. We studied if patterns of beta diversity are related to variability in climate, land cover or geographic distance and how the effects of these variables depend on the spatial grain of the data.
We used data on European birds, plants, butterflies, amphibians and reptiles, British plants, Catalonian birds and Finnish butterflies. We fitted nested grids of varying resolutions to each of these datasets.
For each grid we calculated differences in climate, differences in land-cover composition and beta diversity between all pairs of grid cells. In a separate analysis we looked specifically at pairs of adjacent grid cells (the first distance class).
We then identified the magnitude of independent statistical associations of climate, land cover and geographic distance with spatial patterns of beta diversity. Beta diversity at any given distance decreased with increasing grain.
Geographic distance was always the best predictor of beta diversity for all pairwise comparisons at the extent of Europe. Climate and land cover had weaker and grain-dependent effects.
Climate was more important at coarse grains, land-cover effects were stronger at fine grains. In the country-wide analyses, climate and land cover were more important than geographic distance.
Climatic and land-cover models performed poorly and showed no systematic grain dependence for beta diversity between adjacent grid cells. Relationships between geographic distance and beta diversity and the environmental correlates of beta diversity are systematically grain-dependent.
The strong effect of distance indicates that a substantial fraction of species is missing from areas with a suitable environment. The effects of geographic distance (at continental extents) and land cover (at fine grains) indicate that species distribution models should take both environment and dispersal limitation into account.