Jump to Main Content
Temporal scale dependent interactions between multiple environmental disturbances in microcosm ecosystems
- Garnier, Aurélie, Pennekamp, Frank, Lemoine, Mélissa, Petchey, Owen L.
- Global change biology 2017 v.23 no.12 pp. 5237-5248
- aquatic ecosystems, carbon, dissolved oxygen, global change, models, oxygen, shade, synergism, temperature
- Global environmental change has negative impacts on ecological systems, impacting the stable provision of functions, goods, and services. Whereas effects of individual environmental changes (e.g. temperature change or change in resource availability) are reasonably well understood, we lack information about if and how multiple changes interact. We examined interactions among four types of environmental disturbance (temperature, nutrient ratio, carbon enrichment, and light) in a fully factorial design using a microbial aquatic ecosystem and observed responses of dissolved oxygen saturation at three temporal scales (resistance, resilience, and return time). We tested whether multiple disturbances combine in a dominant, additive, or interactive fashion, and compared the predictability of dissolved oxygen across scales. Carbon enrichment and shading reduced oxygen concentration in the short term (i.e. resistance); although no other effects or interactions were statistically significant, resistance decreased as the number of disturbances increased. In the medium term, only enrichment accelerated recovery, but none of the other effects (including interactions) were significant. In the long term, enrichment and shading lengthened return times, and we found significant two‐way synergistic interactions between disturbances. The best performing model (dominant, additive, or interactive) depended on the temporal scale of response. In the short term (i.e. for resistance), the dominance model predicted resistance of dissolved oxygen best, due to a large effect of carbon enrichment, whereas none of the models could predict the medium term (i.e. resilience). The long‐term response was best predicted by models including interactions among disturbances. Our results indicate the importance of accounting for the temporal scale of responses when researching the effects of environmental disturbances on ecosystems.