Date of Award

Spring 5-11-2019

Level of Access Assigned by Author

Open-Access Thesis

Degree Name

Doctor of Philosophy (PhD)


Ecology and Environmental Sciences


Cynthia S. Loftin

Second Committee Member

Francis A. Drummond

Third Committee Member

Samuel Hanes

Additional Committee Members

Dana Marie Bauer, Alison C. Dibble


Commercial production of lowbush blueberry (Vaccinium angustifolium Aiton) in Maine relies primarily on managed honeybee hives; however, naturally occurring wild bees are more efficient pollinators of the crop. Wild bees have short foraging distances and must nest near crop fields to provide pollination services. After crop bloom, the surrounding landscape must provide sufficient forage to maintain wild bee populations for the remainder of the growing season. Lowbush blueberries in Maine are produced in a mixed-use landscape with two distinct landscape contexts. Here, we document bee communities and habitat resources (nesting and floral) in power line rights-of-way and eight land cover types including and surrounding lowbush blueberry fields. We assess landscape pattern surrounding crop fields in the two contrasting contexts and determine any effect of arrangement of habitat patches on wild bee abundance or diversity. Additionally, we use our field data to inform and validate predictions of wild bee abundance from a spatial model applied to the lowbush blueberry production landscape and assess any influence of landscape pattern on prediction accuracy. Finally, we describe a collaboration with lowbush blueberry growers to develop an interactive web mapping tool that provides maps of habitat resources and predicted wild bee abundance.

We documented 168 wild bee species across 72 study sites; three bee species had not been previously recorded in Maine. Power line rights-of-way had diverse and abundant bee communities owing to high habitat quality, especially within resource-poor landscapes near lowbush blueberry fields. We observed abundant floral resources in lowbush blueberry fields, forest edges, and small towns and found ample nesting resources in lowbush blueberry fields and shrubby wetlands. Bees were less abundant and diverse in a homogeneous landscape context; however, that homogeneity led to more accurate model predictions of bee abundance in crop fields. We improved prediction accuracy in a mixed-use landscape and produced accurate predictions in non-crop land cover types in a heterogeneous landscape context; however, we found that predictions of wild bee abundance in crop fields are influenced by landscape heterogeneity. The maps we share through the web tool aid growers and other stakeholders in developing pollination management and conservation plans.

Files over 10MB may be slow to open. For best results, right-click and select "save as..."