A foundational paradigm in biological and earth sciences is that our planet is divided into distinct ecoregions and biomes, demarking unique assemblages of species. These high-level, categorical classifications have greatly shaped scientific research, global modeling efforts, and policy decisions. Given recent advances in available technologies and the advent of global biodiversity monitoring programs, however, we are now poised to ask whether historical descriptions of ecoregions meaningfully delimit biological communities, or whether ecoregions represent an artificial categorization of continuous gradations in species composition. We tested this question using over 175 million point-occurrences [plants, arthropods, birds, reptiles and amphibians, and fungi], the most commonly used, conservation-relevant map of global ecoregions, and 10,000 random walk transects distributed around the globe. We developed two models to test strength of ecoregion borders, one which compared species accumulation rates at and away from ecoregion borders and the other that measured community similarity within and across ecoregions. We then tested for significance by using permutation tests that compared models fit with ‘true’ ecoregion borders to randomly generated ecoregion borders and asking whether models incorporating the actual location of boundaries performed better than 95% of models fit with randomly generated borders.
Results/Conclusions
To interpret the results, we juxtaposed two hypotheses; the sharp-transition hypothesis and the gradual-transition hypothesis. The sharp-transition hypothesis posits that ecoregion borders should coincide with abrupt discoveries of novel species upon crossing into them as well as distinct changes in community makeup. In contrast, the gradual-transition hypothesis holds that species accumulation and turnover should be independent on ecoregion borders, and instead rely only on geographic distance a transect has traversed. In considering either species accumulation or community turnover we find strong support for the sharp-transition hypothesis. We can evaluate significance as the number of transects where known ecoregion boundaries better explain the rate of species discovery than 95% of randomly drawn boundaries. We found that across reptiles and amphibians (52.39%), mammals (36.97%), birds (29.99%), plants (27.37%), arthropods (25.3%), and fungi (22.50%) of the transects met a significance threshold of p < 0.05, when considering spices accumulation rates. A similar pattern was true when we examined community composition [mammals – 35.76%, plants – 25.02%, reptiles and amphibians – 23.18%, birds – 22.22%, arthropods 19.43%, and fungi 13.35%]. Our results here indicate that knowing ecoregion boundaries can be an important, but by no means the only factor, in developing effective large-scale conservation plans for biotic communities.