Whether warming magnifies the toxicity of a pesticide is strongly dependent on the concentration and the null model

How global warming changes the toxicity of contaminants is a research priority at the intersection of global change biology and ecotoxicology. While many pesticides are more toxic at higher temperatures this is not always detected. We studied whether deviations from this general pattern can be explained by concentration-dependent interaction effects and by testing the interaction against the inappropriate null model. We exposed larvae of the mosquito Culex pipiens to three concentrations of the pesticide chlorpyrifos (absence, low and high) in the absence and presence of 4 °C warming. Both the low and high chlorpyrifos concentration were lethal and generated negative sublethal effects: activity of acetylcholinesterase (AChE) and total fat content decreased, and oxidative damage to lipids increased, yet growth rate increased. Warming was slightly lethal, yet had positive sublethal effects: growth rate, total fat content and metabolic rate increased, and oxidative damage decreased. For four out of seven response variables the independent action model identified the expected synergistic interaction between chlorpyrifos and warming. Notably, for three variables (survival, AChE and fat content) this was strongly dependent on the chlorpyrifos concentration, and for two of these (AChE and fat content) not associated with a significant interaction in the general(ized) linear models. For survival and fat content, warming only potentiated chlorpyrifos (CPF) toxicity at the low CPF concentration, while the opposite was true for AChE. Our results highlight that taking into account concentration-dependence and appropriate null model testing is crucial to improve our understanding of the toxicity of contaminants in a warming world.status: publishe