Pervasive sublethal effects of agrochemicals as contributing factors to insect decline

Abstract

Insect biomass is declining across the globe at an alarming rate. Climate change and the widespread use of pesticides have been hypothesized as two underlying drivers. However, the lack of systematic experimental studies across chemicals and species limits our causal understanding of this problem. Here, we employed a chemical library encompassing 1024 different molecules (including insecticides, herbicides, fungicides, and plant growth inhibitors) to investigate how insect populations are affected by varying concentrations of pesticides, focusing on sublethal doses. Using a controlled laboratory pipeline for Drosophila melanogaster, we found that 57% of these chemicals affect the behavior of larvae at sublethal concentrations, and an even higher proportion compromises long-term survivability after acute exposure. Consistent with these results, we observed that exposure to chemicals at doses orders of magnitude below lethality induced widespread phosphorylation changes across the larval proteome. The effects of agrochemicals were amplified when the ambient temperature was increased by four degrees. We also tested the synergistic effects of multiple chemicals at doses found widely in nature and observed fitness-reducing changes in larval developmental time, behavior, and reproduction. Finally, we expanded our investigation to additional fly species, mosquitos, and butterflies and detected similar behavioral alterations triggered by pesticides at sublethal concentrations. Our results provide experimental evidence that strongly suggests sublethal doses of agrochemicals coupled with changes in environmental temperatures are contributing to the global decline in insect populations. We anticipate that our assays can contribute to improving chemical safety assessment, better protect the environment, secure food supplies, and safeguard animal and human health, as well as understand our rapidly changing world.

Statistical signature of subtle behavioural changes in large-scale behavioural assays

“…Tracking data were initially generated using MWT [40] as described above. We post-processed them using the pipeline introduced in [9], limiting the representation of a larva to 5 points along its anteroposterior axis [9] (tail, lower neck, neck, upper neck, May 3, 2024 9/31 head) to allow the representation to be used with very low-resolution imaging of larvae (as in [53]).…”