Examining the effects of habitat fragmentation on a southern California native bee assemblage through historic pitfall trap samples

Background/Question/Methods

Bees (Hymenoptera: Anthophila) are an ecologically essential insect group currently thought to be undergoing global declines in diversity and abundance, with habitat fragmentation being a key driver of these declines. Here, we examine the response of native bee assemblages to urbanization-induced habitat fragmentation in a coastal sage scrub ecosystem in San Diego County, CA, USA. Using bees collected in an array of ca. 635 pitfall traps deployed in 2004, we examined bee assemblages in eight scrub habitat fragments embedded in an urban matrix and also in eight study sites across two large natural reserves. We identified all bees to species or morphospecies, and for each species, obtained data on their body size, diet breadth, nesting habits, and degree of sociality. Because of numeric differences in bee abundance across sites, we used rarefaction in calculating site-specific species richness and diversity indices. Our general hypothesis was that compared to natural reserves, habitat fragments would harbor bee assemblages that are depauperate in species and dominated by bee groups more tolerant of disturbance (e.g. honey bees and other dietary generalists).

Results/Conclusions

We collected 2,579 native bees belonging to 71 species in 29 genera across six families, and an additional 1,217 honey bees. Although our sampling method did not target bees specifically and thus yielded relatively few species, several species present in the samples are unknown from the well-sampled adjacent Riverside County. As predicted, native bee species richness was significantly higher in reserve sites (mean = 23.0 species) than in fragment sites (mean = 20.3 species), but only slightly so. Comparisons of species composition and distribution also suggested that fragmentation caused shifts in the community composition of native bees. For example, the halictid bee Conanthalictus bakeri was more common in reserve interiors than in fragments, while the small carpenter bee Ceratina arizonensis was more common in small fragments than in other habitat types. However, we surprisingly detected no effect of fragmentation on the abundance, evenness, and dominance of native bees; even when bees grouped by natural history (e.g. large vs. small, generalist vs. specialist, social vs. solitary) were analyzed separately. Moreover, opposite to our prediction, average honey bee abundance in reserves was almost twice that in fragments. Our results suggest that explaining variation in native bee assemblages across study sites may require detailed measures of the diversity and availability of local floral resources. Research is ongoing to account for the patterns detected in this study.