Review of approaches to the pesticide regulatory risk assessment for bees and other pollinators

An inventory and review of current pesticide regulatory risk assessment procedures for bees and other pollinators is presented. For North America, the European Union, Brazil, China, and Australia, the work is based on a review of the published literature, covering pesticide regulatory risk assessment approaches of high-income countries. For low- and middle income countries in the African, Pacific and Caribbean region, the work is based on an online questionnaire, conducted during the summer of 2021, followed by ad-hoc questions for further details and clarifications. In hindsight, the focus on the above-mentioned high-income countries is justified, as the expectations were met that it is a widespread practice for low- and middle-income countries to consult assessments by APVMA (Australia), USEPA (USA) and EFSA (European Union). Protection goals in the published approaches focus on bees, especially the honey bee as a prevailing model organism. Approaches incorporating further taxa, in particular with regard to local pollination services, may be adopted in the future, but are currently prevented by knowledge gaps. All established risk assessment procedures follow a tiered approach, and establishing a lower-tier system in regions currently lacking risk assessment procedures is a much more important step toward improved pollinator protection compared to risk refinement at higher tiers. Several established risk assessment approaches include toxicity classification, which are similar to each other in general, offering an opportunity to assess risks at a screening level without data on exposure. However, basic toxicity data on pesticides are currently not required at all in many low- and middle-income countries. Thus, establishment of a general requirement of data on acute toxicity for honey bees would markedly improve risk assessment in these countries, given that obtaining and assessing such data is a comparatively standardized endeavour and not particular resource demanding. Exposure is handled in different ways by the established risk assessment procedures, with similarities in considered application methods, but marked differences between further parameters of exposure scenarios, e.g. regarding spatial and temporal relation between pesticide application and exposure event, contaminated matrices considered etc. There is a general consensus in the established risk assessment procedures with regard to adopted guidelines and guidance documents used for effect assessment. Approaches established by EPPO, EFSA and USEPA all compare exposure-toxicity quotients to predefined levels of concern for identification of unresolved unacceptable risks to bees, while the calculative basis for these quotients and/or threshold values differs between these approaches, reflecting particular application or situation specific factors or models, accounting inter alia for background mortality, interspecific extrapolation, uncertainty, measured residue data. The other established risk assessment procedures largely follow the approaches of EPPO, EFSA or USEPA. Risk enveloping might offer a resource saving strategy to assess risks of pesticides with similar use patterns in similar cropping systems, and adopting an approach similar to a working example set by IBAMA in Brazil might be feasible, at least for similar geographic regions, given that there are some commonalities in pollinator depend crops among the surveyed low- and middle-income countries. Risk management and uncertainty analysis should always be linked to the risk assessment at all stages, and resulting risk mitigation measures should be practicable in the local agricultural context and comprehensible, while not compromising product efficacy.