The importance of ecological networks in multiple-stressor research and management

Multiple stressors are increasingly affecting organisms and communities, thereby modifying ecosystems' state and functioning. Raising awareness about the threat from multiple stressors has increased the number of experimental and observational studies specifically addressing consequences of stressor interactions on biota. Most studies measure the direct effects of multiple stressors and their interactions on biological endpoints such as abundance, biomass, or diversity of target organisms. This yields invaluable information for the management and restoration of stressed ecosystems. However, as we argue in our perspective paper, this common approach ignores a fundamental characteristic of communities and ecosystems, i.e., that organisms in ecosystems are interlinked by biotic interactions in ecological networks. Examples from the literature show that biotic interactions can modify stressor effects, transfer stressor effects to distant groups of organisms, and create new stressor interactions. These examples also suggest that changes in biotic interactions can have effects of similar or greater magnitude than direct stressor effects. We provide a perspective on how to include network characteristics and biotic interactions into analyses of multiple-stressor effects on ecosystems. Our approach can also make use of biomonitoring data produced with established and intercalibrated methods, and can combine it with novel metrics used to describe the functioning of ecosystems, such as trait information or stable-isotope measurements. The insights on network-mediated effects gained via the approach we propose can substantially increase mechanistic understanding of multiple-stressor effects, and in turn, the efficiency of ecosystem management and restoration