Bees represent the most important pollinator taxon in nature, so it is of great interest to understand the factors that influence the way they behave and interact with other organisms in their environments. Body size is one factor that has been shown to have a significant relationship with a plethora of different bee traits including: maximal foraging distance, pollen load, pollination efficiency, and thermoregulatory capability. Previous attempts to classify the functional variation of bee body size have used a categorical approach which places bees of similar size into discrete groups; however, we believe the use of continuous models to be a better option when trying to identify patterns in bee size. To test this view, we measured the body length of five bee species, two social: Bombus impatiens, characterized as a large species, and Halictus ligatus, characterized as a small), and three solitary (Anthidium manicatum, characterized as a medium, Agapostemon virescens, characterized as a small, and Ptilothrix bombiformis, characterized as large) that were sampled from urban community garden sites across St. Louis, MO. We compared body size models generated with the package ‘pollimetry’ in the R computational environment.
Results/Conclusions
Body size range was significantly greater in the social bees (B. impatiens and H. ligatus), likely due to the caste system of eusocial species. Most of the variation was attributed to phenology effects, but some to interannual variation. The introduced European wool carder bee, A. manicatum, was next in exhibiting the most variation along the growing season, but also significant sexual dimorphism. Ptilothrix bombiformis also exhibited a fair amount of variation, attributed mostly to sexual dimorphism. Finally, the green metallic sweat bee, A. virescens, had the least variation and was attributed mostly to sexual dimorphism. Continuous models accounted for greater variation than discrete models in four of the five species. A. virescens was the only species for which there was no significant difference between discrete and continuous models. The European wool carder bee could have been classified as a medium or a large bee depending on the time of year. Our results suggest that the use of discrete body size categories when determining the functional diversity of bee communities may mask significant variation and could potentially lead to incorrect results.