Using the traits of submersed macrophytes to predict the effects of insecticides in aquatic communities

Background/Question/Methods

In aquatic communities, insecticides can directly influence the abundance and behavior of sensitive species, resulting in indirect effects on less sensitive taxa. For example, insecticides can reduce cladoceran zooplankton, resulting in phytoplankton blooms that can shade periphyton and cause negative effects on the growth and survival of periphyton consumers. Previous work has demonstrated that the submersed macrophyte, Elodea canadensis, mitigates insecticide effects on aquatic communities. Because the generalizability of this function remains unclear, we performed an outdoor mesocosm experiment in which we crossed two insecticide treatments (control, 36 ug/L malathion applied every 2 weeks) and seven macrophyte treatments (no macrophytes, and monocultures of E. canadensis, Ceratophyllum demersum, Myriophyllum spicatum, Potamogeton crispus, Utricularia vulgaris, Vallisneria americana). We hypothesized that macrophyte species would differ in mitigation success because of differences in their ability to remove insecticides from the water column and competitively suppress phytoplankton.

Results/Conclusions

We found strong significant effects of insecticides on cladoceran abundance (ANOVA, F_1,42 = 73.5, p < 0.001), but no macrophyte main effects and no macrophyte x insecticide interaction, suggesting that macrophyte species did not differentially influence malathion’s direct impact. In phytoplankton, however, there was a significant macrophyte x insecticide interaction (ANOVA, F_5,36 = 3.2, p = 0.018), suggesting that the indirect effects of insecticides were strongly influenced by macrophyte species identity. These data suggest that insecticides caused strong enough cladoceran declines in each treatment to initiate trophic cascades, but whether those cascades actually translated into greater phytoplankton abundance depended on the identity of the macrophyte species in the community. Further, the data largely supported our predictions that E. canadensis and C. demersum would effectively suppress phytoplankton from blooming as well as our predictions that other species would be poor competitors against phytoplankton (P. crispus, V. americana, U. vulgaris). Though the mitigation ability of M. spicatum was lower than we predicted, this species generally did not survive well in our tanks. These results have important implications for predictive models aimed at predicting the effects of chemicals in different environments.