The following thesis is submitted as two chapters. Chapter one contains background information, literature review, and predictions. Chapter two is formatted as a paper submission to Invasive Plant Science and Management. Note that this journal requires a combined results and discussion section and a summary of management implications written after the abstract. The appendix contains supplementary materials that are relevant to study hypotheses but do not necessarily align with the scope of the article submission. Oak (Quercus spp.) savannas are one of the most threatened ecological systems within the United States. Like many ecotones, these savannas are considered biodiversity hotspots due to their high environmental heterogeneity (EH). While select studies have assessed how EH in oak savannas relates to the success of individual species, less work has been completed on a community scale. To better inform management targets, quadrat-based understory vegetation surveys were taken alongside measurements of soil moisture, canopy cover, and elevation in a nested plot design. Two definitions of EH, horizontal heterogeneity (total variation over area) and spatial heterogeneity (intensity of clustering), were used to assess EH-vegetation quality relationships using generalized mixed linear models with nested covariates for study sites and plots. Metrics of vegetation diversity and quality included native richness, native cover, potential native vegetation (PNV), woody cover, exotic richness, and exotic cover. Species richness was intentionally omitted due to a strong link with woody cover and higher ratios of exotic to native abundance. Both horizontal and spatial metrics of canopy cover heterogeneity had only positive or neutral associations with vegetation quality and diversity metrics. More specifically, greater variation of canopy cover values was associated with higher percentages of savanna-associated vegetation (PNV) and lower exotic richness. Increases in native richness were observed alongside greater distinctiveness of shade and light patches, suggesting a significant role of niche partitioning in this environment. Soil moisture heterogeneity models had mixed effects on vegetation quality and diversity. Further experimentation is likely necessary to separate the roles of canopy shading and woody encroachment on resulting EH trends. Elevational heterogeneity had a moderate negative association with native species richness – an unexpected finding affirming previous results suggesting that steeper areas decrease native richness within this environment. These EH-vegetation quality trends may be partially explained by differences in how habit generalists and specialists respond to gradients in EH. More specifically, higher EH increased the presence of light specialist species and decreased shade specialist species, whereas high EH increased soil moisture generalists and decreased both moist and dry soil specialists. Model quality was consistently highest at the medium spatial extent (12 by 12 m), suggesting that environmental controls on the microclimatic scale have a strong impact on resulting savanna vegetation.


Matthew Kaproth

Committee Member

Ryan Kaproth

Committee Member

Ryan Wersal

Committee Member

Mriganka De

Date of Degree




Document Type



Master of Science (MS)

Program of Study



Biological Sciences


Science, Engineering and Technology

Creative Commons License

Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 License
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 License.

Included in

Biology Commons



Rights Statement

In Copyright