The neuroethology of insecticide toxicity: Effects on locust visual motion detection and collision avoidance behaviour

Abstract

Agrochemicals are paramount to supporting current agricultural practices, despite the costs to ecosystems. However, sublethal effects of agrochemicals on non-target organisms are poorly understood. Additionally, novel insecticides are being developed continuously, and often can be found in complex pesticide mixtures applied as seed treatments. One of the most controversial of these are the class of insecticides categorized as neonicotinoids, which are nicotinic acetylcholine receptor agonists. These insecticides, lauded for their specificity for the insect receptor, are known to affect many aspects of insect behaviour and physiology. Wild and domestic bees are especially sensitive to these insecticides, which are thought to affect their flight and navigation ability and contribute to colony collapse disorder. Here, I explore whether a common neonicotinoid, imidacloprid, affects visual motion detection and collision avoidance behaviours in the locust, Locusta migratoria. These behaviours and neural circuits are well conserved among species, but are especially well described in the locust, making this a highly tractable system for exploring these effects in vivo. Through a series of three projects I uncovered how imidacloprid affects the responses of important descending visual interneurons to an ecologically-important visual stimulus: the image of an object on a direct collision course (looming). This stimulus elicits robust escape behaviours in the locust, either while in flight or while standing. I show that low, sublethal exposure to imidacloprid resulted in reduced visual motion processing in multiple descending neurons, and that these effects were present between 1 and 24 hours after treatment. I correlated these effects with reduced escape behaviours - animals treated with a single dose do not steer or jump to avoid an impending collision. I show that these effects also resulted from treatment with metabolites of imidacloprid. This is significant as these metabolites exist both in the environment and within insects for a longer time and sometimes at a higher concentration than the parent compound, suggesting an additional source of exposure. Finally, using a comparative analysis I show that another agonist of the nicotinic acetylcholine receptor, the novel insecticide sulfoxaflor, did not produce the same effects as an equal dose to that used with imidacloprid. I argue for the utility of using neuroethological assays to answer questions in toxicology, as these assays link neural and behavioural effects thus offering a more complete picture than single endpoint assays often employed by toxicologists. My results show effects of imidacloprid on visual motion detection and escape behaviours, suggesting that similar effects may occur in non-target insects, including bees, when exposed to these insecticides.