Removing both large- and small-bodied animals from an ecosystem has potentially wide-reaching impacts on the functionality of that system. By analyzing the degree of ecosystem coupling (the overall strength of correlation-based associations between above- and belowground plant, animal, and microbial communities) and of communities with their surrounding physicochemical environment, we can gain a better understanding of the consequences of losing interactions between multispecies communities and their environment. Visually and analytically, ecosystem coupling can be represented as a network in which individual species are substituted with multispecies communities (e.g., microbes, plants, and nematodes). Under undisturbed conditions we would expect that the communities are more strongly connected with one another and their abiotic environment than under disturbed conditions.
We carried out a 5-year field experiment in subalpine grasslands in which we progressively excluded large, medium, and small mammalian vertebrates, and, ultimately, all aboveground dwelling invertebrates with size-selective fences. We predicted that the loss of large animals would reduce ecosystem coupling more than a subsequent loss of smaller animals.
The field experiment encompassed 18 size-selective exclosure setups distributed across two clearly differentiated vegetation types in the Swiss National Park (nine in short-grass, nine in tall-grass vegetation). Each exclosure setup contained four treatment plots that progressively excluded large, medium, and small-sized mammals and aboveground dwelling invertebrates. Next to each exclosure (>5m away) a reference plot provided access to all animals. We considered six ecosystem functions and process rates: soil net nitrogen (N) mineralization, soil respiration, plant tissue N content, plant species richness, root biomass, and microbial biomass carbon, and calculated ecosystem multifunctionality based on the multiple threshold approach.