Publications

Community composition modifies direct and indirect effects of pesticides in freshwater food webs

Zhao, Qinghua; De Laender, Frederik; Van den Brink, Paul J.

Résumé

For environmental risk assessment, the effects of pesticides on aquatic ecosystems are often assessed based on single species tests, disregarding the potential influence of community composition. We, therefore, studied the influence of changing the horizontal (the number of species within trophic levels) and vertical composition (number of trophic levels) on the ecological effects of the herbicide linuron and the insecticide chlorpyrifos, targeting producers and herbivores, respectively. We tested how adding, to a single primary producer, 4 selected competing producer species, 0–1-4 selected herbivore species, and one selected predator species resulting in 1, 2 and 3 trophic levels, changes the effects of the two pesticides. Linuron decreased producer biovolume less (17%) when the 4 producers were added, because insensitive producers compensated for the loss of sensitive producers. However, linuron decreased producer biovolume 42% and 32% more as we increased the number of herbivore species from 0 to 4 and as we increased trophic levels from 1 to 3, respectively. The indirect negative effect of linuron on herbivore biovolume was 11% and 15% lower when more producer and herbivores were added, respectively. Adding a predator increased this indirect negative effect by 22%. Chlorpyrifos decreased herbivore biovolume about 10% less when adding multiple herbivore or producer species. However, adding a predator magnified the direct negative impact on herbivores (13%). Increasing the number of producer, herbivore species and adding trophic levels increased the indirect positive impact on producer biovolume (between 10% and 35%). Our study shows that changing horizontal composition can both increase and decrease the effects of the selected pesticides, while changing vertical composition by adding number of trophic levels always increased these effects. Therefore, single species sensitivity will not always represent a worst case estimate of ecological effects. Protecting the most sensitive species may not ensure protection of ecosystems.