Jackson, J.K., R.J. Horwitz, and B.W. Sweeney. 2002. Transactions of the American Fisheries Society 131(5):910–930.
Six field experiments tested the effects of the microbial pesticide Bacillus thuringiensis israelensis (Bti) on pestiferous black flies (primarily the Simulium jenningsi group) as well as nontarget macroinvertebrates and fish in the Susquehanna River, Pennsylvania. The response of black fly drift after the Bti application was similar across all experiments. Black fly drift increased significantly within 0.5 h after application, peaked 2-3 h after application, and then remained high for several hours. Significant changes in drift were observed after Bti application for 31 of the 50 nontarget macroinvertebrate taxa examined, but none of these taxa exhibited an increase in drift density similar in magnitude and timing to that of black flies. Furthermore, none of the 31 taxa exhibited the same drift changes in all experiments, as was observed for black flies; most of the taxa exhibited a drift density change during only one posttreatment sample period within a single experiment. Most nontarget macroinvertebrates therefore appeared unaffected by Bti, including the chironomid midge Rheotanytarsus, which was previously reported as Bti sensitive. Fish species composition and abundance in riffles did not change following the Bti application. The importance of black flies in fish diets varied dramatically among individual fish, sampling periods, and years. Variation in the number of black flies in fish diets may reflect variation in discharge as well as in black fly and fish behavior. The number of black flies consumed did not increase immediately after Bti application (when black fly drift was high) or decrease later (when black fly densities were suppressed). Fish condition and short-term (7-10-d) growth rates did not change consistently across experiments. The absence of repeated responses that paralleled the magnitude of black fly responses suggests that the statistically significant changes observed for nontarget macroinvertebrates and fish were unrelated to the Btiapplication. Thus, we found no evidence that a single Bti application affected nontarget macroinvertebrates and fish to an ecologically significant extent.