Recent bee declines have ignited interest in understanding how humans affect invertebrates. To test mechanisms driving these declines and predict future fluctuations, large-scale monitoring of populations is required.Citizen science projects, such as BeeSpotter and iNaturalist, invite non-scientists to upload photographs with proper metadata to help monitor bees. Photo surveys that control for collection effort potentially provide quantitative measures of diversity and relative abundance that approximate traditional methods. The purpose of this research was to compare citizen scientist photo survey techniques with standard netting techniques. In 2018 and 2019, we conducted netting surveys within three hours of photo surveys conducted by trained citizen scientists. After controlling for collection effort, we tested for differences in diversity metrics. In 2018, six experienced undergraduate ambassadors trained small groups (4-6 individuals) of citizen scientists in identifying target taxa (i.e., bumble bees and honey bees), conducting high-quality photo surveys, and submitting bee observations to the online portal. In 2019, citizens were trained and performed two styles of photo surveys to include 5 locations monthly. We compared diversity as recorded by three data collection methods (netting, stationary photo surveys, and walking photo surveys).
Results/Conclusions
Over 60 citizen scientists participated in 6 beeblitz surveys. One third of the participants conducted more than one survey (many returning for up to 5), indicating continued interested and commitment to the program. Photo surveys provided similar patterns to traditional aerial netting surveys. Five bumble bee species and honey bee were recorded regardless of data collection method. Relative abundances fluctuated throughout the year, with Bombus auricomus and B. pensylvanicus dominating early in the year when host plants, such as Monarda species were available. Simpsons’s diversity was equivalent across survey methods. However, rarefaction curves suggest that photo surveys may accumulate rarer species more slowly, suggesting 10-15 spottings per species in the potential species pool. Our study illustrates that, given adequate training, citizen scientists can contribute high-quality data that approximate traditional estimates of bumble bee diversity. Citizen science programs have the potential to increase the geographic scale, temporal scale, and ecological depth of data while enhancing science communication with nonexperts. However, further exploration of the potential biases of these methods versus traditional methods are required to appropriately interpret the resulting data.