COS 117-3 - Trees for bees: The effect of tree species composition on forest bee communities

Background/Question/Methods

Temperate forests in eastern North America harbor a unique assemblage of spring-flying bee species that do not persist in other landscapes. Thus, conservation of high-quality forest habitat is likely important for preservation of native bee diversity in this region. Yet, we lack forest bee management guidelines because it is not known what forest characteristics drive habitat quality for bees. Although data on bee scopal pollen shows that bees use large amounts of tree pollen, it is not clear whether tree composition influences the bee community present at a site. Here, we investigate how tree species composition is associated with bee species composition across 31 forest patches in New Jersey, USA. We pan-trapped wild bee communities and quantified the surrounding tree community at a 4 hectare scale. Specifically, we used distance-based tree sampling to estimate the density and basal area of 12 tree species that previous studies suggest provide important floral resources to bees. For a subset of 13 forest-associated and 6 habitat generalist bee species, we then used negative binomial mixed models to test for associations between tree species densities and bee species abundances.

Results/Conclusions

Our datasets as analyzed consisted of 5488 bee specimens of 19 species, and 2974 measurements of individual trees of 12 species. The preliminary results of our mixed models show a significant positive effect of forest patch size for most forest-associated bee species, in contrast with negative or non-significant effects of patch size for all habitat generalist species. Acer rubrum is positively associated with the abundance 5/13 forest-associated and 2/6 habitat generalist bee species, and Acer saccharum is positively associated with the abundance of 6/13 forest-associated and 2/6 habitat generalist bee species. Over half of the species analyzed had significant negative associations with Quercus spp. These results indicate that forest size and tree species composition influence habitat suitability for forest-associated and habitat generalist bees. Maple (Acer) species, which are among the first trees to bloom in our study region, may provide particularly important spring resources for both forest-associated and habitat generalist bee species. Despite contrasting responses to forest size, both species groups appear to utilize tree resources. This research represents an important step toward assessing drivers of habitat quality for forest bees and will contribute to informing forest management practices that support both native bee diversity and spring crop pollination services.