Predicting the long-term effects of metal pollutants on the honey bee colony: Integrating experimental data in a colony population model

Honey bee colonies have long served as biomonitors of metal pollution in the environment, but the effects of metals on the health of colonies is relatively understudied. Unlike pesticides, metals never break down, allowing them to persist in the environment and potentially accumulate in colonies over time. In contrast, our understanding of the effects of metals in honey bees is based on relatively brief laboratory assays. As a consequence, there’s interest in modeling approaches that can predict the cumulative effects of metals to honey bee colonies over longer periods of exposure. For the present study, I conducted chronic (10-day) toxicity trials with honey bees fed several metals (As, Cd, LiCl, and Zn). Using a simple colony population model, I used the data to predict the cumulative effects of these metals to honey bee colonies over 150 days of exposure. I compare two different modeling approaches for extrapolating toxicity data from the laboratory to the colony level: a “traditional” approach based on dose-response curves and a recently published mechanistic modeling approach (the General Unified Thresholds Model of Survival, GUTS). This work demonstrates how different approaches to modeling toxic effects can be incorporated into honey bee population models for chemical risk assessment.