Generalization of a density‐dependent ecosystem function in dominant aquatic macroinvertebrates

The processing of organic matter is a central ecosystem function in freshwater ecosystems that allows the integration of terrestrial plant material into aquatic heterotrophic food webs. From an energetic perspective, many temperate aquatic systems fundamentally depend on such allochthonous inputs. Current research established that this central ecosystem function may be linked to population density of few yet dominant invertebrate detritivores, showing a negative density‐dependent relationship. Here, we extended the current knowledge and experimentally assessed the processing of leaf litter by freshwater amphipods, using broad density gradients, interspecific competition, and laboratory and field experiments. We used two species of dominant amphipods, namely the native Gammarus fossarum and the non‐native Gammarus roeselii, varying their density by two orders of magnitude, both in single‐species and two‐species treatments. Results from 252 mesocosms in the lab and 97 mesocosms in a natural stream show that per capita leaf litter processing rates are strongly negatively dependant on population density in monocultures. Interspecific competition in two‐species treatments corroborated the negative density‐dependent ecosystem function and highlighted the functional redundancy of the two detritivore species. We identified a flattening in the processing rates at previously unreported but well‐defined breakpoints. These breakpoints may reflect minimal metabolic requirements at which survival takes precedence over any other process, such as interference competition. The breakpoints were consistent across both species, indicating that they must process leaf litter equalling one fifth to one fourth of their dry body weight per day as a minimal nutritional threshold. Our results suggest the need for integrating nonlinear density‐dependencies and breakpoint population densities in key ecosystem functions in aquatic ecosystem modelling alongside classical biodiversity–ecosystem functioning relationships. Depending on the population density the corresponding ecosystem functioning could differ significantly. This corroborates the need to better understand biodiversity–abundance relationships when protecting aquatic ecosystems.