Moving beyond honey bees as a surrogate: modeling approaches to expand the pesticide risk assessment landscape
On this page:
Honey bees (Apis mellifera) and native North American bees provide crucial pollination services that support both agricultural crops production and wild plant diversity. Yet they are threatened by a multitude of stressors including exposure to pesticides, and as a result many species are facing range declines and extinction risk. As part of the pesticide risk assessment process, the US EPA considers risk to pollinators using individual bee toxicity tests and semi-field colony feeding studies. But because these methods are poorly developed for native bee species, honey bees are used as a surrogate species for native bees in most cases. However, the diverse life histories of native bees may lead to key differences in pesticide exposures and impacts at the individual, colony, and population levels. To bridge this gap, EPA’s Office of Research and Development is to developing modeling methods that allow for extrapolation from honey bee experimental data to other bee species in ecologically relevant scenarios and real habitat areas. Here, we review past, current, and future efforts at EPA to expand the pesticide risk assessment landscape using modeling methods such as: agent-based models that simulate bee colony and/or population dynamics in response to stressors, Bayesian model inference methods for fitting models to data, species sensitivity distributions, and spatiotemporal modeling of pesticide exposures through geospatial data. Collectively, this research seeks to advance the Agency’s ability to protect a wide range of critical pollinator species and the ecosystem services they provide.