Insect communities with diverse life histories and morphologies, such as bees (Hymenoptera: Apoidea), are difficult to representatively sample. Methodological comparison studies have increased our knowledge of bee-sampling method biases. However, further understanding of how sampling methodologies are biased with respect to functional traits and the surrounding environment is needed. We examined the differences in taxonomic and functional trait representation of sampled bee communities between an active netting and passive bee bowl methodology. We also determined the influence of surrounding floral resources on each method's ability to capture bee abundance, richness, and diversity. We captured bees at 32 sites across North Dakota, USA using both methods. Netting and bee bowl methods reflected different bee communities defined by taxonomic families, genera, diet breadth, sociality, and body size. Bee bowls reflected more solitary and small-bodied bees while the netting method selected a greater representation of generalist, social, and large-bodied taxa. Bee bowls captured fewer bees as floral abundance increased, while bee abundance, richness, and diversity captured with the netting method showed the opposite trend with increasing floral resource availability. In addition, the netting method generally captured more varied samples than bee bowl surveys. Our results indicate the importance of using multiple types of sampling methods and of understanding how surrounding floral availability affects sampling of the target bee communities. Methodological biases influence how ecological studies interpret target communities, making it essential for future bee-sampling to incorporate multiple methods and to interpret their results carefully in the context of chosen methodologies and study systems.
Implications for insect conservationThe essential ecological services many bees provide, coupled with the concern over declines in bee populations and diversity, fuels the increased interest in bees and studies involving bee-sampling. Management and conservation policy for declining bee populations depend on data that is representative of the sampled bee communities. Therefore, it is important to understand systematic biases in our sampling methods as well as the effects of surrounding habitat on these biases.