Accurately estimating the risk of sensitive populations subjected to toxicant disturbances is central to our ability to protect ecosystem services. While the gold standard for assessing risk historically involves static measures such as the LD50 or LC50, more sophisticated approaches have been developed in an attempt to capture more nuanced outcomes. In the 1980s the International Organization for Biological Control (IOBC) developed a tiered approach to determine the compatibility of pesticides and natural enemies in the context of integrated pest management (IPM). We analyzed the IOBC approach using stage-based matrix models to project population outcomes for four parasitoid species, Diaeretiella rapae, McIntosh, Fopius arisanus, (Sonan), Diachasmimorpha longicaudata Ashmead, Psyttalia fletcheri (Silvestri) and the predator, Coccinella septempunctata L. By imposing mortality levels in matrix models equivalent to those outlined in the IOBC Tier 1 Class 1 (29 %) (harmless) and Class 2 (79 %) (slightly harmful) mortality classes, we explored discrepancies between the IOBC approach and population outcomes generated by these models. Our results highlight that the IOBC Class 1 and 2 levels of mortality are too high to protect many natural enemies from pesticides, setting the stage for unrealistically optimistic views of pesticide compatibility in many cases. Furthermore, a one size fits all approach to protect natural enemies from pesticides does not work because of differences in demographic rates among species which will be less negatively affected by Tier 1 levels of mortality and those that do not reproduce quickly and will thus be more vulnerable to pesticides. Therefore, the IOBC method should be used cautiously if at all, and results should be interpreted with the caveats and pitfalls highlighted here. Results of this study indicate that it is time to reevaluate how we estimate pesticide compatibility with natural enemies and adjust the methods and mortality thresholds based on more realistic measures of toxicity.
Read full abstract