Chronic oral lethal and sub‐lethal toxicities of different binary mixtures of pesticides and contaminants in bees (Apis mellifera, Osmia bicornis and Bombus terrestris)

Abstract

Chemical hazard assessment for bees generally starts with laboratory testing of acute effects in the honeybee, Apis mellifera. Whether acute effects observed in this model species translate to different species and longer exposure periods are key issues for robust hazard assessment. The aim of this study was to develop and trial a series of toxicity tests for testing the prolonged (up to 240 h) exposure of A. mellifera, Bombus terrestris and Osmia bicornis to a range of insecticides (clothianidin, dimethoate, tau-fluvalinate), other pesticides (propiconazole, 2,4-D) and trace metals (cadmium and arsenic) and selected mixtures. Oral toxicity tests for the species were developed from standardised procedures that were modified to account for species ecologies and behaviours. Tests with A. mellifera identified that toxicity decreased in the order clothianidin > dimethoate > cadmium > arsenic > tau-fluvalinate > propiconazole >= 2,4-D. This order of sensitivity was broadly consistent for the other two species. The only chemical showing any trend for interspecies variation in sensitivity was tau-fluvalinate. While not toxic to A. mellifera at the maximum tested concentration, mortality effects were seen in the other species. Patterns of toxicity showed that LC 50 s decreased with time. Extension of tests to 240 h and prediction of exposure concentration effects up to a theoretical Apis worker bee lifetime (720 h) suggest that long-term toxicity may exceed predictions based on short-term tests by an order of magnitude. Mixture tests showed that most commonly tested combinations were additive and non-interactive. Studies with clothianidin and propiconazole did point to a slightly increased toxicity for the neonicotinoid in the presence of the fungicide. For dimethoate and clothianidin in B. terrestris and to an extent O. bicornis, weak antagonistic interactions were found. These finding suggest that, at least for initial assessment, current mixture models frequently provide a relevant indication of likely joint effect.