Toxicity of binary mixtures of metals and pyrethroid insecticides to Daphnia magna Straus. Implications for multi-substance risks assessment.

Two different concepts, termed concentration addition (CA) and independent action (IA), describe general relationships between the effects of single substances and their corresponding mixtures allowing calculation of an expected mixture toxicity on the basis of known toxicities of the mixture components. Both concepts are limited to cases in which all substances in a mixture influence the same experimental endpoint, and are usually tested against a "fixed ratio design" where the mixture ratio is kept constant throughout the studies and the overall concentration of the mixture is systematically varied. With this design, interaction among toxic components across different mixture ratios and endpoints (i.e. lethal versus sublethal) is not assessed. In this study lethal and sublethal (feeding) responses of Daphnia magna individuals to single and binary combinations of similarly and dissimilarly acting chemicals including the metals (cadmium, copper) and the pyrethroid insecticides (lambda-cyhalothrin and deltamethrin) were assayed using a composite experimental design to test for interactions among toxic components across mixture effect levels, mixture ratios, lethal and sublethal toxic effects. To account for inter-experiment response variability, in each binary mixture toxicity assay the toxicity of the individual mixture constituents was also assessed. Model adequacy was then evaluated comparing the slopes and elevations of predicted versus observed mixture toxicity curves with those estimated for the individual components. Model predictive abilities changed across endpoints. The IA concept was able to predict accurately mixture toxicities of dissimilarly acting chemicals for lethal responses, whereas the CA concept did so in three out of four pairings for feeding response, irrespective of the chemical mode of action. Interaction effects across mixture effect levels, evidenced by crossing slopes, were only observed for the binary mixture Cd and Cu for lethal effects. The analysis of regression residuals showed that interaction effects across mixture ratios were restricted to feeding responses in binary mixtures that included Cu. These results indicate that the ability of the CA and IA concept to predict mixture toxicity effects varies from lethal to sublethal endpoints irrespective of their primary mode of action. This suggests that when considering complex responses, the pharmacological notion of mode of action should be extended to encompass an ecotoxicological mode of action, based on the concentration at which various toxicological effects become operative in the biological system under consideration.