Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
4375681 | Ecological Modelling | 2015 | 10 Pages |
•We model 23 tree species local colonization and extinction rates using 46,596 plots.•We assess the role of climate, other species and dispersal in explaining the data.•Dispersal was the most important predictor of colonization.•Climate and dispersal (rescue effect) were equally important for extinction.•For the colonization and extinction the best model included all three drivers.
We studied the relative importance of climate, abundance of potentially competing species, and dispersal in explaining local colonization and extinction rates of tree species throughout mainland Spain. We used a Bayesian framework to parameterize a patch occupancy model to 23 species censused in 46,596 permanent plots in a 1 × 1 km grid across most Spanish forests. For most species, dispersal was the single best predictor of colonization, whereas climate and dispersal were equally important as predictors of extinction. Precipitation was positively correlated with the colonization rate of 12 out of 13 deciduous broad-leaved species, and negatively correlated with the extinction rate of nine of them. In contrast, precipitation equally decreased colonization and extinction of five out of eight of needle-leaved species (Juniperus and Pinus spp.). There was, however, marked variation among species in the magnitude of these effects, with some species exhibiting contrasting patterns for the colonization and the extinction process. Abundance of competing tree species (= summed plot basal area) was consistently correlated with decreased colonization of all needle-leaved species, and it increased the extinction rate of 6 out of 8 of these species. It had, nonetheless, weak facilitative effect on some broad-leaved species by promoting colonization (3 of 13 species) and decreasing extinction (7 of 13 species). With local colonization and extinction data, non-equilibrial and dynamic species distribution modelling can be improved by incorporating measures of biotic interactions and dispersal effects, along with traditional climate variables.