Success of spatial statistics in determining underlying process in simulated plant communities

1. Spatial statistics are widely used in studies of ecological processes in plant communities, especially to provide evidence of neutral or non-neutral mechanisms that might support species coexistence. The contribution of such statistics has been substantial, but their ability to identify any links between underlying processes and emergent patterns is not certain. 2. We investigate the ability of a number of spatial statistics to distinguish theorized mechanisms of species coexistence (spatial and temporal niche differentiation, neutrality, the Janzen-Connell effect and heteromyopia) in a simulated plant community. 3. We find that individual statistics differ substantially in their sensitivity to these mechanisms, with those based on nearest neighbour species identities being the most sensitive. These differences are largely robust to changes in the strength of the modelled mechanisms when simulated independently and in combination. The spatial signal of niche differentiation is always distinct in simulations that combine mechanisms. 4. Synthesis. We describe full spatial signals of modelled coexistence mechanisms that are observed consistently across statistics and simulated strengths and combinations of mechanisms, and identify a set of spatial statistics that holds particular promise for empirical studies designed to investigate mechanisms of these kinds.