Harry Siviter’s article, Quantifying the impact of pesticides on learning and memory, was published yesterday (Wednesday 11th July 2018). Here he explains why policy makers need to consider the effects of agrochemicals on pollinators.

Foraging bees have a tough time of it. In social bees, such as bumblebees and honeybees, workers have to forage across large landscapes, collecting nectar and pollen from flowers as they go. They then return to their nests, where they feed the collected food to the colonies developing brood. This process involves remembering the location of thousands of different food resources, while continuously tracking the profitability of each resource. Bees that aren’t able to effectively learn and memorise the location of rewarding flowers are likely to be at a disadvantage, which could ultimately influence bee colony health.

I, as a PhD student at Royal Holloway, University of London, where I am supervised by Elli Leadbeater and Mark Brown, was tasked with investigating whether agricultural pesticides influence bee cognition. Bee cognition, or animal cognition, involves investigating how animals perceive, learn and memorise information from the surrounding environment.

Siviter’s research demonstrates how pesticides and chemicals, commonly used in agriculture, affect bees’ knowledge of their environment

To protect agricultural crops from unwanted pests and diseases, we typically use a whole host of different insecticides and pesticides within agricultural environments. Insecticides are often insect neurotoxins, which means they act on the nervous system of invertebrates, which can potentially influence their cognitive ability. When I first began to evaluate the literature on the impacts of pesticides on bee cognition, it quickly became apparent that a number of studies had investigated this topic, the majority of which had used a classical learning paradigm called the proboscis extension response (from here after PER). In the same way that Pavlov’s dogs famously learnt to associate the sound of a bell with food, bees in a PER experiment learn to associate a rewarding odour with sucrose solution. That odour can then be presented to bees at a later date to assess their memory. By sheer coincidence during the time I was familiarising myself with this literature, I attended a lecture by our future collaborator Julia Koricheva (also from Royal Holloway, University of London), which first brought meta-analysis to my attention.

Meta-analysis is a technique used by researchers to provide robust conclusions to ecological questions by systematically combing the data from the literature and conducting statistical analysis. Given the rising number of studies investigating the impacts of pesticides on bee cognition it seemed like a meta-analysis would provide an essential tool for researchers and policy makers alike.

We decided to conduct a systematic review and meta-analysis that assessed the impacts of oral pesticide exposure on bee learning and memory in PER based paradigms. While we would have liked to have included the results from experiments that had used other non-PER experimental designs, the majority of studies had used a PER design and was the obvious choice for our study.

One of the most contentious issues in pesticide research is the classification of ‘field realistic’ dosages. We re-classified each experiment used in our meta-analysis based on up to date field realistic concentrations of pesticides that had been extracted from the pollen, nectar and honey of bee colonies. Our results combined data from 23 previously published papers and confirmed that pesticides appear to influence bee learning and memory at field realistic dosages and that further these effects still occurred when bees were only exposed once. This means that foraging bees could potentially become cognitively impaired whist performing a single foraging trip, which could influence foraging efficiently and their ability to navigate back to the nest.

With the commonly used neonicotinoid-insecticides set to be banned from the agricultural landscape in Europe in December this year, we thought it was important to consider how other non-neonicotinoid pesticides influence bee learning and memory. Our results showed that both neonicotinoids and non-neonicotinoids have similar negative impacts on the learning and memory of bees, suggesting that replacing the soon to be banned neonicotinoids with other pesticides should be done with caution.

There are still lots of unanswered questions. Does it matter if pesticides influence bee cognition? Does impaired cognition influence bee foraging and navigation, or do pesticides simply make bees feel ‘unwell’ or unmotivated? We still don’t have a definitive answer to these questions. Our results do however confirm that pesticides have a significant impact on bee learning and memory. These results and other similar studies stress the need for policy makers and regulators to consider the sub-lethal impacts of insecticides, prior to their registration for agricultural use. Failure to consider these important implications will result in bees, and potentially other important pollinators, being subject to further negative effects of agrochemicals.

The full article, Quantifying the impact of pesticides on learning and memory is free to read for a limited time in Journal of Applied Ecology.