Can SPEcies At Risk of pesticides (SPEAR) indices detect effects of target stressors among multiple interacting stressors?

Pesticides are increasingly recognised as a threat to freshwater biodiversity, but their specific ecological effects remain difficult to distinguish from those of co-occurring stressors and environmental gradients. Using mesocosms we examined the effects of an organophosphate insecticide (malathion) on stream macroinvertebrate communities concurrently exposed to a suite of stressors typical of streams in agricultural catchments. We assessed the specificity of the SPEcies At Risk index designed to determine pesticide effects in mesocosm trials (SPEARmesocosm). This index determines the log abundance proportion of taxa that are considered physiologically sensitive to pesticides. Geographic variation in pesticide sensitivity within taxa, coupled with variation between pesticides and the effects of co-occurring stressors may decrease the accuracy of SPEARmesocosm. To examine this, we used local pesticide sensitivity assessments based on rapid toxicity tests to develop two new SPEAR versions to compare to the original SPEARmesocosms index using mesocosm results. We further compared these results to multivariate analyses and community indices (e.g. richness, abundance, Simpson's diversity) commonly used to assess stressor effects on biota. To assess the implications of misclassifying species sensitivity on SPEAR indices we used a series of simulations using artificial data. The impacts of malathion were detectable using SPEARmesocosm, and one of two new SPEAR indices. All three of the SPEAR indices also increased when exposed to other agricultural non-pesticide stressors, and this change increased with greater pesticide concentrations. Our results support that interactions between other non-pesticide stressors with pesticides can affect SPEAR performance. Multivariate analysis and the other indices used here identified a significant effect of malathion especially at high concentrations, with little or no evidence of effects from the other agricultural stressors. [Display omitted] •SPEARmesocosm is an index designed to respond selectively to pesticides in mesocosm experiments•We identify that SPEARmesocosm and two variants did not respond predictably to malathion.•All variants responded in a problematic manner to another stressor antagonism.•Mathematical simulation showed that ≈ 15% sensitivity misclassification can disrupt SPEAR.•Among multiple stressors SPEARmesocosm should be used with other metrics, indices and analysis.