The effect of repeated, lethal sampling on wild bee abundance and diversity

1. Bee pollinators provide a critical ecosystem service to wild and agricultural plants but are reported to be declining world-wide due to anthropogenic change. Long-term data on bee abundance and diversity are scarce, and the need for additional quantitative sampling using repeatable methods has been emphasized. Recently, monitoring programmes have begun using a standardized method that employs a combination of pan traps and sweep netting, resulting in lethal sampling of bees. This standardized method can remove a large number of bees from sites during each sampling day, raising concern that the sampling itself could have a negative effect on bee populations. 2. We conducted an experiment to assess whether lethal sampling for bees using pan traps and netting affected bee abundance and diversity when done every two weeks throughout a season and over multiple years. We compared bee abundance, richness, evenness and functional group composition between sites that had been sampled every two weeks from 2009 to 2012 to similar sites not previously sampled. 3. We found that the standardized method for sampling bees, with specimens from 132 morphospecies, did not affect bee communities in terms of abundance, rarefied richness, evenness, or functional group composition. Thus, our results indicate that the bee communities we sampled are robust to such sampling efforts, despite removing an average of 2862 bees per season. 4. We discuss several explanations for why sampling did not affect bee abundance or community structure, including a density-dependent response to reduced competition for resources. 5. These results suggest that bee monitoring programmes sampling once every two weeks with pan traps and netting will not affect bee community structure. We urge researchers monitoring bees to utilize standardized protocols so that results can be compared across space and time.