Re‐evaluation of a method used to study nutritional effects on bumble bees

Bee declines are an increasingly recognised problem globally. Nutritional stress due to agricultural intensification is one of the drivers of bee declines. Therefore, understanding the nutritional requirements of bees is crucial to mitigate the effects of food scarcity on bee populations. Laboratory studies evaluating nutritional effects on bumble bees often use microcolonies as a model system for effects on queenright colonies. Microcolonies consist of workers, whereby one worker will lay male-destined unfertilized eggs. Consequently, microcolonies exclusively produce males, while queenright colonies produce (female) workers and reproductives (males and queens). A comparison between Bombus terrestris (L.) microcolonies and queenright colonies was made by feeding three diets of varying nutritional quality. The experimental diets affected most fitness parameters of microcolonies differently than those of queenright colonies. Low-protein, largely monofloral, pollen affected queenright colony fitness by reducing colony size and increasing larval mortality, while these fitness parameters were not affected in microcolonies. High-protein polyfloral pollen reduced larval mortality in queenright colonies but did not affect larval mortality in microcolonies. For both colony types, a similar result was obtained when assessing offspring body mass, which was reduced by the low-quality diet and increased by the high-quality diet. In conclusion, our results demonstrate that microcolonies are inaccurate model systems and their use can lead to erroneous conclusions in terms of effects of nutritional stress or pollen quality on bumble bees. Our results further highlight the importance of high-quality (here high-protein polyfloral) pollen for sustaining colony development of bumble bees.