A risk based pollination network for non-Apis bees demonstrates the importance of understory plant contamination

Understanding the distribution of pesticides in the floral landscape is critical for land managers and regulators, particularly since identifying where exposure occurs is critical to pesticide mitigation. In this study, we developed a bee-plant network for a commercial sweet cherry (Prunus avium L.) system and the surrounding unmanaged floral habitat. We estimated the pesticide contamination of flowering plants in this network by trapping pollen from honey bee colonies, identifying the plant species of origin of the pollen, and relating this to the non-Apis bee visitation and toxicity of pesticide detections. Over 90 plant-bee interactions from non-Apis species were matched with honey bee collected pollen. By combining bee visitation and pollen data, we attributed the pesticide hazard to 33 plant genera. Unlike previous studies, we observed the greatest hazard to non-Apis bees did not come from visits to the crop or from pesticide drift off the orchard, but from contamination of an orchard understory plant (genus Taraxacum). The importance of this plant in pesticide exposure was related to both the hazard of the pollen and the frequency of visitation by non-Apis bees. Our findings caution against generalizing how non-Apis bee species become exposed to pesticides.