Understanding the environmental effects shaping plant distributions is crucial for predicting future ecosystems under climate change. The effects of different environmental factors may vary in their importance in determining plant distributions at different spatial and taxonomic scales, which affects our understanding of plant–environment relationships. However, this has not yet been systematically explored.
Here we combined global distribution data of 205 widely distributed plant families and environmental data from multiple global databases. We then used the random forest algorithm to quantify the relative importance of environmental factors (including climate, soil, and topography) on the distribution of plants at three taxonomic levels (family, genus, and species) and multiple spatial scales (10 spatial extents from 1° × 1° to 10° × 10° randomly located across the globe). Mixed-effect models were used to assess the significance of spatial and taxonomic scales on relative environmental effects across the globe.
We found that climate factors had increasing importance on plant distributions at higher taxonomic scales and larger spatial scales (yet stochastic effects at spatial extents finer than 4° × 4°). Edaphic factors congruously decreased their importance on plant distributions as spatial and taxonomic scales increased. Topographic factors had a relatively larger influence at higher taxonomic levels (i.e., family>genus>species), but with a relatively slow rise with the increase in spatial scale.
Our findings are generally aligned with current knowledge but have also indicated the potential complexity underlying the scale-dependence of relative environmental effects on plant distributions. Overall, we highlight a multi-scale insight into ecological patterns and underlying mechanistic processes.