Detectability of habitat influence on wild bee communities varies by collection technique

Effective and replicable monitoring is an important part of bee conservation.  Correctly interpreting influence of habitat characteristics on native bee communities is necessary to create strategies for bee conservation and provision of pollination service to agricultural crops and natural plant communities.  To evaluate influence of trapping method on detected bee community composition we used blue vane traps, colored pan traps, and an aerial net to assess bee communities at 32 sites in 28 fragments of Palouse Prairie. Differences in abundance, species richness, generic proportions and functional trait characteristics among the three sampling techniques are noted.  We also evaluated differences in relationships between bee community characteristics and habitat factors known to mediate bee communities: plant species richness at the collection site and the amount of suitable bee habitat at a distance.  Under our sampling regime blue vane traps had the highest abundance and species richness relative to colored pan traps and net collection. Net collected assemblages have a higher proportion of oligolectic bees. Colored pan traps and blue vane traps collect significantly more parasitic bees.  Community metrics for bees collected using blue vane traps were correlated with the amount of suitable habitat at a distance but not plant species richness at the collection site. Conversely, community metrics for bees collected using an aerial net were correlated with the plant species richness but not amount of suitable habitat at a distance. This suggests collection method can alter the detectability of habitat influence on bee communities, which has implications for conservation planning.