Agent-based modeling of the long-term effects of pyriproxyfen on honey bee population
Résumé
Honey bees are beneficial insects playing a key role in pollinating wild plants and crop plants. Unfortunately, during their foraging activity they can be exposed to pesticides and other xenobiotics. The members of the colony can also be poisoned indirectly by contaminated food and water brought back to the hive by foragers. While the assessment of the potential adverse effects of a xenobiotic on individual bees can be easily done by using normalized laboratory tests, toxicity estimation at the population level suffers from technical difficulties and is costly. Fortunately, population modeling is increasingly used to overcome this problem of transition from the individual to the population level. In this context, an agent-based model, called SimBeePop, was designed for predicting the long-term effects of man-made chemicals on honey bee populations. After a detailed description of the structure of the model, its performances were estimated on pyriproxyfen, a juvenile hormone mimic. To do so, the physicochemical properties, environmental fate, use, and toxicity of this insecticide to honey bees were reviewed in order to elaborate realistic scenarios of contamination. Application of these scenarios in SimBeePop reveal that while pyriproxyfen is classified as moderately toxic or not toxic to bees under current regulations, it can induce significant adverse effects on bee populations at sublethal concentrations