There are thousands of pesticides used worldwide.  Many pesticides have similar modes of action, and thus different pesticides within a class of pesticides might have very similar effects on ecosystems.  If so, we might be able to predict the community- and ecosystem-level effects of new or relatively unstudied pesticides simply by knowing the class or mode of action to which they belong.  We tested this hypothesis by applying 14 pesticide treatments, all at expected environmental concentrations, to 56 freshwater mesocosms.   There were two pesticide types (insecticides and herbicides), two classes within each pesticide type (triazine herbicide, chloroactenilide herbicide, carbamate insecticide, organophosphate insecticide), and three different pesticides in each of the four classes, as well as water and solvent controls.  We recorded several ecosystem-level variables and the abundance of over 20 species.

Results/Conclusions

For ecosystem-level variables, there was substantial consistency within the pesticide types.  Insecticides increased phytoplankton abundance (presumably by reducing caldocerans), which in turn elevated dissolved oxygen and pH and reduced water clarity and periphyton.  Herbicides had exactly the opposite effect, reducing phytoplankton and shifting the community to periphyton dominance.  For community-level variables, there was substantial consistency in effects within pesticide classes but variability among classes.   Insecticides reduced the abundance of insects, cladocerans, and salamanders and increased tadpole and snail survival.  The triazine herbicides reduced anuran survival but increased snail survival.  These results suggest that ecosystem-level responses might be predictable based strictly on pesticide type, whereas community-level responses might be predictable based on pesticide class.  Consequently, reducing pesticides to classes based on modes of action might help to simplify the daunting task of understanding the impacts of the thousands of pesticides on the environment.