No evidence of biodiversity effects on stream ecosystem functioning across green and brown food web pathways

Abstract

Abstract Biodiversity loss is known to affect the two fundamental and opposite processes controlling carbon and nutrient cycles globally, that is, primary production and decomposition, which are driven by green and brown food web compartments, respectively. However, biodiversity in these two food web compartments has been mostly studied independently, and potential reciprocal effects of biodiversity loss on ecosystem processes remain unclear. We conducted a 35‐day stream mesocosm experiment with two levels of algal diversity (natural and diluted periphyton communities) and three levels of litter diversity (no litter, monocultures of poplar, maple, and oak, and the three‐spp. mixture) to simulate changes in biodiversity in both the green and brown pathways of an aquatic food web. We then measured multiple ecosystem processes pertaining to carbon cycling. We predicted that algal diversity would enhance decomposition and sporulation of fungal decomposers, while litter diversity would enhance algal growth and net primary production, due to the more diverse algal exudates or litter nutrients being released from more diverse mixtures. In contrast to this hypothesis, we only found biodiversity effects on an ecosystem process within the green pathway: there was a relationship between algal diversity and biofilm carrying capacity. Nevertheless, we found that this relationship was affected by the presence or absence of litter (algal diversity increased the carrying capacity in presence of litter and decreased it in its absence), which also influenced the algal community structure. Our mesocosm experiment did not evidence relationships between biodiversity and ecosystem processes across different food web compartments, but further studies in more realistic conditions would be necessary to confirm this result. If supported, the lack of biodiversity–ecosystem functioning relationships across compartments would facilitate the prediction of the impacts of biodiversity loss on ecosystems.