Initial pesticide residue concentrations in nectar and pollen after spray application on flowering plants for assessing impacts on pollination

Abstract Background Pesticide application is one of the factors reported to contribute to the decline of pollinators, such as honey bees and bumble bees, and guidelines for risk assessment have been established recently. We collected a total of 1720 data sets of pesticide residue levels in pollen and nectar from EFSA documents and from the scientific literature, covering 71 different chemicals and 20 different crops. Out of this data set, 164 events for nectar with 27 different chemicals and 161 events for pollen with 32 different chemicals were chosen that describe the initial ( t = 0–1 day) residue per unit dose (RUD) after spray application. Median RUD values were calculated to describe the typical realistic initial applied dose. Results Overall, we can confirm earlier findings found in the frame of risk assessment of plant protection products on bees. RUD pollen is higher than RUD nectar, and there is a wide variability of results. Polar pesticides had higher median RUD values, in particular in nectar, but the explained variability of substance properties is generally small (≤10%). There are clear indications that the setup and the conditions during the experiment override substance-specific differences. Factors contributing to the observed differences in RUD were identified: spray direction (up/sideward > downward), sampling method of nectar and pollen (from flower > by traps > from bees), experimental setting (greenhouse > field or tunnel) and geometry of flowers (Fabaceae had highest RUD nectar). Conclusions A goal of the study was to find a prediction method for initial pesticide residues in nectar and pollen after spray application to flowers and plants. Thus, median RUD values for the different conditions were provided. However, nothing can be said about how combinations of the mentioned factors affect RUD values.