Integrating science and policy: examining effects of neonicotinoid insecticides on honey bee (Apis mellifera L.) and bumble bee (Bombus impatiens Cresson) queens, worker bees, and colony development

Abstract

Exposure to neonicotinoid insecticides has been highly debated as a major contributor to the decline of bee health. In response to concerns over the potential effects of neonicotinoids on bees, the United States Environmental Protection Agency (US EPA) has made improvements to the regulatory risk assessments process for systemic insecticides. These improvements refine toxicity testing procedures to better establish appropriate uses (acceptable targets) and inform labeling (application rates and timing) to minimize non-target, or unintended, effects on bees. The main objective of these improvements is to increase field realism to obtain more accurate assessments on risks to bees in a natural setting. Each of the following chapters of this dissertation addresses a different aspect of the new risk assessment approach and provides data or suggestions to improve the procedure to better assess non-target effects of systemic neonicotinoid insecticides on bee pollinators. In the first chapter, experiments are described whereby honey bees (Apis mellifera L.) and bumble bees (Bombus impatiens Cresson) were fed imidacloprid syrup in acute toxicity tests at different concentrations, either at the standard quantity (10µl) or at a more field-realistic quantity (50µl) that resembled the amount of nectar foraging bees may hold in their crop to bring back to the hive, or nectar carrying capacity. Results indicated that bumble bees were more sensitive to imidacloprid than honey bees. Bumble bees were particularly more susceptible to becoming unresponsive after exposure rendering them nonfunctional and thus “ecologically dead.” In addition, the “lethal” concentration was considerably lower when measured endpoints included bees that were either physiologically dead or ecologically dead, suggesting higher toxicity to bees when toxicological consequences were considered in an ecological context. In chapter two, honey bee (A. mellifera) colonies consisting of various worker-population sizes (1500, 3000, and 7000 bees) were exposed chronically to imidacloprid syrup (0, 10, 20, 50, and 100ppb) for three weeks to assess the potential effects on queen bee fecundity and colony development. My results are the first to show dose-dependent effects of imidacloprid exposure on queen egg laying and locomotor activity, as well as on worker bee behavior, brood production, and pollen stores in all treated colonies. In addition, such negative effects lessened as colony size increased, suggesting that population size may act as a buffer to pesticide exposure through foraging efforts and increased nestmate interactions that might dilute exposure levels. In chapter three, I fed individual bumble bee (B. impatiens) queens imidacloprid chronically for 18-days. Imidacloprid was administered in syrup (1, 5, 10, 25 ppb) and pollen (0.3, 1.7, 3.3, 8.3ppb) in cage studies. Bumble bee queens exhibited reduced survival at all doses, even as low as 1 ppb. All treated queens that survived their initial exposure exhibited a significant dose-dependent delay in nest initiation or egg-laying, which led to the delayed emergence of worker brood. However, treated queens eventually commenced nest initiation after exposure, suggesting some recovery potential. The final chapter reviews the existing literature on effects of neonicotinoids on bees and organizes the findings using the Adverse Outcome Pathway approach to create a conceptual framework from which to better understand causal linkages among adverse effects across studies. In doing so I created a mechanistic explanation or pathway relating the mode of action of neonicotinoids to the specified adverse outcome: inhibition of colony development in honey bees and bumble bees by neonicotinoid exposure.