Fewer bowl traps and more hand netting can increase effective number of bee species and reduce excessive captures

Abstract

Abstract Reports increasingly point to substantial declines in wild bee abundance and diversity, yet there is uncertainty about how best to measure these attributes in wild bee populations. Two commonly used methods are passive trapping with bee bowls or active netting of bees on flowers, but each of these has drawbacks. Comparing the outcomes of the two methods is complicated by their uncomparable units of effort. The abundance distribution of bee species is also typically highly skewed, making it difficult to accurately assess diversity when rarer species are unlikely to be caught. The effective number of species, or Hill numbers, provides a way forward by basing the response metric on the number of equally abundant species. Our goal is to compare the effective number of bee species captured between hand netting and bowl trapping in wheatgrass prairie in South Dakota and tallgrass prairie in Minnesota, USA. Species overlap between the two methods ranged from ~40% to ~60%. Emphasis placed on rare species was important, so that 95% confidence limits overlapped between the two methods for species richness but netting exceeded trapping for Shannon's and Simpson's diversities. Netting always captured more bee species with fewer bee individuals than trapping. In most cases, the number of bees captured in bowl traps indicated substantial over‐sampling, with little increase in bee species detected. Correlations between bee and floral abundance, richness, and diversity differed between netted and trapped samples. We conclude that netting and trapping together produce a more complete account of species richness, but shifting sampling emphasis from trapping to netting will result in fewer bees, but more bee species captured. Due to the different relationships between bee and floral diversities that depended on sampling method, it is unwise to compare habitat associations determined by netting with those determined by trapping.

Keywords

• bee survey methods
• bee‐habitat associations
• Hill numbers
• sampling effort