Management scale drives bee and forb biodiversity patterns in suburban green spaces

Land management can occur at very small scales, such as in individual home gardens, and at very large scales, such as at the municipal scale or beyond. The scale of management can affect environmental heterogeneity and thus can affect biodiversity at alpha, beta, and gamma scales. Urban areas provide a unique opportunity to evaluate the effects of large- versus small-scale land management on different scales of biodiversity. We surveyed floral resources and bees in residential backyards and powerline corridors in the Chicago suburbs, with yards representing small-scale management and powerline corridors representing large-scale management. We calculated alpha diversity, beta diversity, and gamma diversity for both floral and bee species. We also determined absolute nestedness for corridors and neighborhoods to gain further insight into the effects of management on species composition. We found that beta diversity for both floral resources and bees was highest in response to small-scale yard management and its creation of heterogeneous landscapes, while alpha and gamma diversity tended to be lower in yards compared to the powerline corridors. We also found that the floral resource and bee communities demonstrated absolute nestedness in both powerline corridors and residential neighborhoods, respectively. Our results indicate that taxa may respond to the effects of management scale differently owing to variations in their mobility or resource requirements. Our study results demonstrate that suburban residential areas may be prime targets for biodiversity conservation due to their environmental heterogeneity.

Methods
Data were collected in 5m X 5m sampling modules in powerline corridors and residential backyards. Floral resource surveys were conducted by a single observer who counted all open/blooming floral units visible within the sampling module and identified them to the lowest possible taxonomic level. Timed pollinator observations were conducted by a single observer who spent 5 minutes observing each half of the sampling module, counting any visible bees in the module during that time and identifying them to the lowest possible taxonomic level. Bee bowls were set out for ~24 hours in arrays of 6 bowls laid in a diagonal transect within each sampling module. Each bee bowl transect included 2 white, blue, and yellow bowls, and no two bowls of the same color were placed immediately adjacent to one another in the transect. Bowls were filled with a dilute soap solution. Bees captured in bee bowls were stored in 70% ethanol until lab processing, at which point they were washed, rinsed, dried, and identified to the lowest possible taxonomic level.