Ecosystem-level testing of a synthetic pyrethroid insecticide in aquatic mesocosms

Twelve 0.1-ha rectangular ponds (mesocosms) were constructed during 1986 and 1987 to test ecosystem-level responses to a pyrethroid insecticide containing esfenvalerate. Treatment consisted of 10 weekly drift and 5 biweekly runoff simulations of formulated esfenvalerate between July 9 and September 17, 1988. Control ponds received only untreated runoff, while low, medium, and high treatments received total loadings of 0.23, 4.1, and 23.3 g active ingredient (a.i.)/ pond. The top vertebrate predator in the mesocosms was bluegill (Lepomis macrochirus). Once adult blue-gill spawned and young fish were present, microcrustaceans declined in number. Following esfenvalerate applications, nauplii in medium- and high-rate ponds underwent further significant reductions, indicating a response to esfenvalerate not observed in control and low-rate ponds. Relative to other treatments, high-rate ponds were found to have significantly increased metabolism and reduced macroinvertebrates following application. Bluegill exhibited no significant differences in population structure among treatments, although 2-cm-size bluegill were missing from high-rate ponds at harvest. Patterns of response emerged in this whole ecosystem test of esfenvalerate. Variation measured among communities was high as is typical of natural ecosystems; however, we suggest that this experiment provided an acceptable way of detecting impacts at both the community and the ecosystem level.