COS 116-6 - Wild bee communities in non-crop land cover in Maine’s (USA) wild blueberry production landscape

Wednesday, August 9, 2017: 3:20 PM
C125-126, Oregon Convention Center
Brianne Du Clos, Wildlife, Fisheries, and Conservation Biology, University of Maine, Orono, ME, Cynthia S. Loftin, U.S. Geological Survey, Maine Cooperative Fish and Wildlife Research Unit, Orono, ME and Frank A. Drummond, Biology and Ecology, University of Maine, Orono, ME
Background/Question/Methods

Wild blueberry (Vaccinium angustifolium) production relies heavily on insect pollination to set fruit. Wild bees provide up to 30% of this pollination service. The Maine (USA) wild blueberry production landscape consists of irregularly-shaped crop fields that range widely in size. There are two wild blueberry growing regions that differ in landscape structure and composition. Large crop fields, forested wetlands, and coniferous forest dominate the Downeast region, whereas the Midcoast region has a more complex arrangement with many small patches of urban and agricultural land in a deciduous/mixed forest matrix. Crop growers are motivated to pursue wild bee pollination resources owing to rising costs and pathogen pressure of rented honey bee hives, however, we lack baseline knowledge on wild bee communities in non-crop cover types and availability of forage beyond crop bloom. We surveyed wild bee communities and habitat resources in seven non-crop land cover types throughout the growing season in both regions to determine differences in wild bee communities across land cover types and landscape patterns. Surveys consisted of bowl trapping and live netting wild bees, identifying and measuring patch area for all blooming plant species, and assessing nesting substrate availability at each site.

Results/Conclusions

We collected 2,094 bee specimens representing 135 species over two growing seasons. Wild bee communities differed across growing regions and among land cover types. Bee abundance and species richness responded strongly to land cover type in both regions, decreasing in forested land cover while remaining similar among non-forested land cover types. Though wild bee communities were more diverse and abundant in the more complex Midcoast growing region, analyses of landscape pattern, such as patch size, spatial distribution, and shape complexity, were inconclusive, suggesting that spatial resolution finer than the 10m resolution used in our analyses is necessary to realistically model wild bee perception of complex landscape patterns. Wild bee communities in wetlands were more diverse than expected. This is encouraging for crop pollination and wild bee conservation as wetland land cover comprises 15-40% of the landscapes surrounding our survey sites. We are sharing our results directly to crop growers through BeeMapper, an interactive, collaboratively developed online tool. Understanding how wild bees use the landscape surrounding Maine wild blueberry will aid crop growers and natural resource managers in decision-making regarding pollination management, including habitat restoration and honey bee hive rental, and inform wild bee conservation.