Data from: Species turnover and invasion of dominant freshwater invertebrates alter biodiversity-ecosystem function relationship

Freshwater ecosystems rely on allochthonous resources. Integration of these subsidies depends on diversity of both terrestrial resources and aquatic shredder and decomposer communities, but the diversity effects on leaf litter breakdown and decomposition are less clear in aquatic than terrestrial ecosystems. We need a better understanding of this relationship because aquatic communities are rapidly changing with species invasions and anthropogenic impacts. Here, we experimentally disentangled the effects of leaf and shredder richness on leaf litter breakdown by macroinvertebrates in mesocosm experiments using three species of amphipods, a dominant guild of crustaceans in European freshwater ecosystems. Increased leaf richness led to lower-than-predicted leaf consumption by native shredders, with mixed evidence of resource-switching or prioritization of preferred food items within a leaf mix. Higher shredder species richness never promoted leaf consumption rates compared to predictions from relevant single-species experiments, and instead sometimes substantially decreased leaf consumption. We then conducted a meta-analysis of leaf litter consumption rates by seven widely distributed amphipod species (the three used in the experiments and four additional species). As expected based on our own experiments, non-native amphipod species generally had lower biomass-adjusted leaf litter consumption rates, although their larger body size led to higher per-individual leaf consumption rates. Contamination of the water by metals, pesticides and other chemicals additionally significantly decreased leaf litter consumption by multiple native and non-native species compared to unpolluted systems. While the meta-analysis suggested that litter consumption, and thus breakdown, would decline if native shredders are replaced by non-native heterospecifics, complete species replacement is not the only outcome following immigration in a meta-community context. Our experiments suggest that breakdown rates could remain reasonably high where native species coexist with non-native arrivals. Experiments which neglect the ecological realism of species coexistence will necessarily mischaracterize effects on ecosystem functioning