RESUMO
Risk assessment for chemicals in the United States relies upon the honey bee (Apis meliffera L. [Hymenoptera: Apidae]) as a surrogate for other bee species. There is uncertainty in extrapolating honey bee toxicity data to bumble bees due to differences in life history strategies, food consumption, and nest structure. Here we evaluated the design of a queenless bumble bee microcolony test that could be considered for generating larval toxicity data. Three microcolony studies were conducted with Bombus impatiens to evaluate the effects of exposure to 1) diflubenzuron in pollen, 2) dimethoate in pollen, and 3) dimethoate in sucrose. Immature drone bee emergence, worker survival, pollen, and sucrose utilization were measured throughout the study duration. For dimethoate, a 10-d chronic adult bumble bee study was also conducted to compare microcolony endpoints to toxicity endpoints on individual adults. Microcolonies exposed to 10 mg diflubenzuron/kg pollen produced fewer adult drones despite no effects on worker survival. Microcolonies treated with dimethoate at ≥3 mg a.i./kg pollen and ≥0.1 mg a.i./kg sucrose solution produced fewer drones. Exposure to dimethoate in the 10-d chronic adult study resulted in direct mortality to the adult workers at ≥0.1 mg a.i./kg diet. Results from the 10-d study suggest direct effects of dimethoate on workers in the microcolony will alter provisioning of diet to the brood, resulting in lower drone production in the microcolony. Our data suggest that the microcolony study is only appropriate to assess brood effects to bumble bees for substances with low toxicity to adults, as demonstrated with diflubenzuron.
