Analysing patterns of spatial and niche overlap among species at multiple resolutions

Abstract

Aim

Analyses of spatial overlap in species distributions are frequently used to test a range of ecological and evolutionary hypotheses, from the role of competition in community assembly to the geography of speciation. Most studies quantify overlap at one spatial resolution. Here we explore the effects of measuring spatial and niche overlap patterns for the same clade (Banksia) at multiple resolutions.

Location

Australia.

Methods

We quantify overlap among species using broad overlap of species range polygons, proximity of occurrence points and co‐occurrence within small survey plots. We compare overlap patterns with null models using age–range correlations and the frequency of sympatric sister species. We then use similar methods to examine patterns of overlap in environmental niche dimensions.

Results

Banksia species show a wide range of overlap values based on range polygons and point proximities, but very low levels of co‐occurrence at the local scale. Intercepts of age–range correlations point to higher levels of overlap among recently diverged species than expected. However, comparing the frequency of sympatric sister species with an evolutionary null model supports a prevailing allopatric mode of speciation. In many cases, niche overlap between species exceeds that expected from phylogenetic relatedness or spatial overlap alone.

Main conclusions

Patterns of broad geographical overlap among Banksia species support a predominantly allopatric mode of speciation, combined with post‐speciation range drift. There is more evidence for niche conservatism than for rapid niche divergence among closely related species. This pattern is consistent with broad‐scale geomorphic and landscape complexity as a driver of plant speciation in south‐west Australia. It is less consistent with finer‐scale mechanisms of species divergence such as fire mosaics, or with ecological divergence in sympatry. Analysis of species overlap patterns at different resolutions is a useful approach for revealing the multiple ecological and historical factors that influence species distributions.