Date of Award
2024
Degree Type
Open Access Thesis
Degree Name
Master of Science (MS)
Department
Biological Sciences
First Advisor
Valerie E. Peters
Second Advisor
David Hays
Third Advisor
Jennifer M. Koslow
Abstract
Both the productivity of agricultural systems and functioning of terrestrial ecosystems are reliant on animal pollinators to facilitate plant reproduction. Yet the integrity of these pollination services is threatened by the declines of pollinators. Mounting evidence indicates landscape homogenization in agricultural, urban, and most notably in natural systems have corresponded with substantial changes in bumblebee distribution, abundance, composition, and forage resource availability. Naturally occurring forest openings, barrens, and grasslands have historically provided important habitat for native bees across the otherwise forested landscape in the eastern United States. Fire suppression and the loss of these important habitats are well-documented, but it is unknown if bumblebee composition or forage plant availability changes between high-quality and low-quality forest openings. To determine effective strategies for ecological stewardship and resource extraction, a fundamental understanding of how different quality ecosystems support bees and their host plants is essential. Despite the increasing number of studies evaluating threats to pollinators and quantifying the effect of their decline on ecosystem services, few studies have concurrently studied the extent to which the outcome of modern land management and concomitant changes plant communities affect the variation in bee composition and their dietary breath. A lack of reliable data also hinders assessing long-term population trends for many bumblebee species, precludes the reconstruction of historic plant-pollinator network structure, and elucidation of species-specific habitat requirements and dietary niches. To investigate this, bumblebees and their pollen were collected within high quality forest openings, actively managed by the Office of Kentucky Native Preserves or private landowners, and compared to those collected from adjacent low quality, unmanaged forest openings. The data presented in this thesis address the hypotheses that (Chapter 1) the abundance, species richness, and composition of bumblebee assemblages vary in response to forest quality, management practices, and surrounding landscape variables, and (Chapter 2) the composition of plants detected in corbicular pollen loads and plant species topological roles in bumblebee-plant networks differ systematically between high quality, managed habitats and low quality, unmanaged habitats. The results from Chapter 1 demonstrate bumblebee abundance, species richness, diversity, and evenness indices are not directly influenced by management type or percentage of surrounding landcover type but are associated with increased floral abundance and plant richness. Increased floral abundance and plant richness were significantly higher in high quality, managed transects than in low quality, unmanaged transects. The zoo-centric bumblebee-plant networks constructed using DNA metabarcoding described in Chapter 2 demonstrate that bumblebees collect pollen from a wide range of plant species with considerable variation in pollen breadth among different species. The composition of plants detected in bumblebee pollen collected in managed sites and unmanaged sites were significantly different, with management type explaining approximately 48% of the variation in diet breadth. Both managed and unmanaged bumblebee-plant networks in spring-early summer and managed late summer-fall network were not significantly modular, possessing 4 modules each. The late-summer fall network was significantly modular also with 4 modules. Network level properties were relatively similar between bumblebee-plant networks constructed from managed and unmanaged transect across both seasons (spring-early summer and late-summer-fall). Of the 79 plants in the managed, spring-early summer bumblebee-plant network, six plants were identified as connectors, seven were identified as modular hubs, and no plants were identified as network hubs. Of the 94 plants in the unmanaged, spring-early summer bumblebee-plant network, three plants were identified as network hubs, three plants were identified as connectors, and no plants were identified as a modular hub. Of the 170 plants in the managed, late summer-fall bumblebee-plant network, nine plants were identified as connectors, seven were identified as modular hubs, and one plant was identified as a network hub. Of the 155 plants in the unmanaged, late summer-fall bumblebee-plant network, ten plants were identified as connectors, two plants were identified as network hubs, and three plants were identified as modular hubs. There was no plants that performed the same topological role in the managed and unmanaged spring-early summer bumblebee-plant networks. However, in the late summer-fall bumblebee-plant networks, Solidago canadensis was identified as modular hubs in both the managed and unmanaged networks. Securigera varia was the only plant identified as having a core topological role in all four networks in this study.
Recommended Citation
Carter, Kassandra, "Influence of Forest Management on Bumblebee-Plant Networks and Community Compositions in Eastern Kentucky Forest Openings" (2024). Online Theses and Dissertations. 820.
https://encompass.eku.edu/etd/820
