Bees are fundamentally important for pollination services and declines in populations could have significant economic and environmental implications. Pesticide exposure and pathogen infection are recognised as potential stressors impacting upon bee populations and recently there has been a surge in research on pesticide-disease interactions to reflect environmentally realistic scenarios better. We critically analyse the findings on pesticide-disease interactions, including effects on the survival, pathogen loads and immunity of bees, and assess the suitability of various endpoints to inform our mechanistic understanding of these interactions. We show that pesticide exposure and pathogen infection have not yet been found to interact to affect worker survival under field-realistic scenarios. Colony-level implications of pesticide effects on Nosema infections, viral loads and honey bee immunity remain unclear as these effects have been observed in a laboratory setting only using a small range of pesticide exposures, generally exceeding those likely to occur in the natural environment, and assessing a highly selected series of immune-related endpoints. Future research priorities include the need for a better understanding of pesticide effects on the antimicrobial peptide (AMP) component of an individual's immune response and on social defence behaviours. Interactions between pesticide exposure and bacterial and fungal infections have yet to be addressed. The paucity of studies in non-Apis bee species is a further major knowledge gap.

Currently, there is a growing interest in developing biopesticides and increasing their share in the plant protection market as sustainable tools in integrated pest management (IPM). Therefore, it is important that regulatory requirements are consistent and thorough in consideration of biopesticides’ unique properties. While microbial pesticides generally have a lower risk profile, they present special challenges in non-target organism testing and risk assessment since, in contrast to chemical pesticides, their modes of action include infectivity and pathogenicity rather than toxicity alone. For this reason, non-target organism testing guidelines designed for conventional chemical pesticides are not necessarily directly applicable to microbial pesticides. Many stakeholders have recognised the need for improvements in the guidance available for testing microbial pesticides with honey bees, particularly given the increasing interest in development and registration of microbial pesticides and concerns over risks to pollinators. This paper provides an overview of the challenges with testing and assessment of the effects of microbial pesticides on honey bees (Apis mellifera), which have served as a surrogate for both Apis and non-Apis bees, and provides a foundation toward developing improved testing methods.

Bee declines have been associated with various stressors including pesticides and pathogens. We separately exposed immune-challenged adult worker honey bees (Apis mellifera L.) to two neonicotinoid pesticides, thiamethoxam (10 ppb) and imidacloprid (102 ppb), by dietary delivery. We found that whereas neonicotinoid exposure weakly affected transcriptional responses of antimicrobial genes, it did not detectably affect the physiological antimicrobial response as measured by a lytic clearance assay of haemolymph. Our findings add to the evidence that transcriptional responses in immune-related genes are not yet reliable indicators of pesticide impacts on bee health, which suggests caution in their future use as biomarkers in pesticide risk assessment.