Master of Science (MS), Bowling Green State University, 2017, Biological Sciences

This study took place within the Oak Openings Region, a biodiversity hotspot in northwestern Ohio. The Oak Openings was created during the Cenozoic period, a time where glaciers continuously moved across Ohio, creating valleys and riverbeds. When the Wisconsin glaciers melted back from Ohio around 14,000 years ago, waters were released into multiple lakes with sandy beaches. Overtime, these sandy beaches started to became dunes inundated by rainwater that could not drain past the clayey till and bedrock ground layer. Water that would stand between the clay layer and sandy dunes provided moisture to eventually support oak savanna habitat. Between the sand ridges, rainwater would saturate the land, which created open areas of wet prairie (Higgins, 2003; Grigore, 2004). The combination of oak savanna habitat and open wet prairie gave this region the name of “Oak Openings” (Grigore, 2004). Today, the Oak Openings looks quite different as a result of the economic development and agricultural pursuits along the Toledo-Detroit corridor. This area is highly fragmented by roads, agricultural fields and urban/residential development. Roughly 45% of the Oak Openings Region contains urban and suburban development and roughly a quarter of the region has been converted to areas of agricultural production (Schetter and Root, 2011). However, there are still remnants of wild lands that exist west and south of the city of Toledo. This area is Ohio's largest single surficial sand covering which is 1-12 meters thick and consists of wet lowland and sand ridge terrain with elevations up to 210 meters above sea level. This region is approximately 8 kilometers wide and 32 kilometers long (Higgins, 2003; Figure 1) and contains oak savanna, oak woodland and wet prairie habitats on post glacial beach ridges and swales and covers 478 km2 (The Nature Conservancy, 2001; Grigore, 2004; Schetter and Root, 2011). The Oak Openings Region contains one third of all of Ohio's endangered plant commu (open full item for complete abstract)

Committee: Karen Root (Advisor); Raymond Larsen (Committee Member); Andrew Gregory (Committee Member) Subjects: Animal Sciences; Animals; Biology; Conservation; Ecology; Environmental Studies; Statistics; Wildlife Conservation; Wildlife Management; Zoology

Doctor of Philosophy (Ph.D.), Bowling Green State University, 2024, Biological Sciences

Conflict with wildlife continues to escalate as human population increases and development expands. Understanding how vertebrates interact with the environment is a critical component to conservation ecology. Movement patterns reflect spatial and temporal changes associated with resource availability, life history stages, and habitat use. This study explored how vertebrate mortality could be used to understand the critical factors impacting the consequences of permeability, i.e., ability to move between patches on the landscape, in a mixed disturbance landscape. We assessed how spatial and temporal heterogeneity influenced terrestrial vertebrate mortality. In the biodiversity hotspot of the Oak Openings Region (OOR) of northwest Ohio, we surveyed repeatedly, across three years, approximately 50 kilometers of road segments. Vertebrate mortality locations (N=654) were related to road (e.g., traffic, road width), structural (e.g., canopy cover, soil average water capacity), compositional (e.g., landcover) and productivity (e.g., NDVI) measurements. We found vertebrate mortality locations were positively related to traffic, road width, canopy cover, and normalized difference vegetation index (NDVI) but negatively related to landcover as it becomes more altered (i.e., natural to agricultural). Our consistent findings across years suggest that the spatial components were influencing mortality differences more than temporal differences, and intra-year differences do not impact mortality in a way that would steer long term mitigation of permeability issues. We developed spatially explicit models for predicting current vertebrate mortality probabilities across the entire OOR. Proportion of residential/mixed landcover area was the most influential variables of mortality occurrence probability. We found mortality was well predicted and the results of the same key variables were robust across taxa and years. The models developed can serve as an assessment tool for evaluating co (open full item for complete abstract)

Committee: Karen Root Ph.D. (Committee Chair); Timothy Schetter Ph.D. (Committee Member); Raymond Larsen Ph.D. (Committee Member); Jeffrey Miner Ph.D. (Committee Member); Erin Labbie Ph.D. (Other) Subjects: Biology; Conservation; Ecology; Wildlife Conservation; Wildlife Management

Master of Science (MS), Bowling Green State University, 2020, Biological Sciences

With the expansion of the human population, new roads are continually being built, leading to an increase in fragmentation or loss of natural habitats. These roads can act as corridors for vertebrates connecting natural habitats or aiding in migration, but they can also act as barriers or boundaries leading to fragmentation, isolation, and/or mortality on roadways. Research has focused mainly on larger highways, outside of the U.S., and away from larger urban areas. The Oak Openings Region, a biodiversity hot spot in northwestern Ohio, is a matrix of human dominated land use and remnant natural patches. This research focused on identifying vertebrate mortality hot spots and the spatial and temporal variables associated with vertebrate road mortality. From mid-April to the end of September 2019, we surveyed 38 road transects within and around three protected (natural) areas to record diversity, abundance, and distribution of roadkilled vertebrates and to identify influencing factors (e.g., land use, canopy cover, environmental data, road characteristics). We found an uneven spread of roadkill, with 45% of roadkill found on nine out of the 38 (24%) surveyed transects, highlighting potential areas to prioritize for mitigation. During the spring months (April 14 - June 20), nine of the transects had no roadkill, but in the summer months (June 21 - September 18), only one transect had no roadkill. This suggests that seasonality influenced the abundance and distribution of roadkill. As transects moved northward, there was a positive trend of more roadkill with increased developed land use. We found a total of 297 roadkill on or around the roads, with all vertebrate taxa included. Mammals made up 49.8% of the roadkill found. Mammal roadkill showed a significant positive trend with average canopy cover (p < 0.0001). Amphibians made up the second largest roadkill group, but showed a seasonal peak in August. These results can help land managers predict where animal dispersal (open full item for complete abstract)

Committee: Karen Root PhD (Advisor); Enrique Gomezdelcampo PhD (Committee Member); Shannon Pelini PhD (Committee Member) Subjects: Biology; Conservation; Ecology; Wildlife Conservation

Master of Science (MS), Ohio University, 2018, Biological Sciences (Arts and Sciences)

As bobcat (Lynx rufus) populations recover in Ohio, an accurate evaluation of demographic and population metrics such as mortality rates (natural or human-induced) and occupancy are critical to understanding past and future population trajectories. To identify predictors of road mortality, we used logistic regression to analyze landscape and local variables using a long-term data set (1978-2016) of georeferenced road kills. To explore whether bobcats exhibit road avoidance behaviors, we used telemetry data from 18 individuals to compare road crossings along actual trajectory paths with random road crossings using correlated random walks. We then computed the population-level mortality rate in a framework that combines traffic, behavioral, and landscape variables. Results show that bobcats are being killed at higher rates on interstates regardless of surrounding landscape variables, and that landscape variables are more useful at predicting mortality on smaller roads. We also found that bobcats exhibit avoidance of certain road types (predominantly county, municipal, and US routes), and that 6-30% (mean = 18%) of individuals are lost to vehicle strikes annually in Ohio. We then tested the use of a multi-method monitoring system (camera traps and hair snares) to understand occupancy of bobcats in southeast Ohio, their interactions with the eastern coyote (Canis latrans), and to evaluate bobcat abundance. We found that the probability of bobcat occupancy across our surveyed sites in southeast Ohio was 0.40 (0.290 - 0.533), and that baited hair snares had limited success in detecting bobcats (~0.1 detection probability). Coyotes and bobcats co-occurred throughout the landscape, though their spatial dynamics are complicated; both species are likely partitioning activity centers, though more research is needed. Our results add important information for evaluating the long-term population viability in Ohio, identify potential areas for mitigation of vehicle-strikes, and em (open full item for complete abstract)

Committee: Viorel Popescu PhD (Advisor); Nancy Stevens PhD (Committee Member); Johnson Joseph PhD (Committee Member) Subjects: Ecology; Wildlife Conservation; Wildlife Management

Master of Science (MS), Ohio University, 2017, Environmental Studies (Voinovich)

Madagascar is home to a broad array of intriguing, endemic, and increasingly endangered species. The fosa (Cryptoprocta ferox) is the largest living (non-introduced) mammalian carnivore on the island and is considered a keystone species for maintaining ecosystem complexity in a broad range of Madagascar's forested habitats. Sadly, the fosa is threatened, with viable populations remaining in only two protected areas. In this context, complex interactions among fosa, prey, and myriad introduced species, reveal a dynamic that is increasingly sensitive to human pressures (e.g., hunting, deforestation for agriculture and fuel wood). This project assembles detailed geographic information to augment long-term data collection and help inform the fosa conservation effort. Research was conducted in Ankarafantsika National Park, a dry deciduous forest in the northwest region of the country, and one of the two places where viable fosa populations still exist. Species encounters and trap rates, roadkill patterns, and the spatial and activity patterns of GPS-collared fosas were analyzed to gather a comprehensive assessment on habitat pressures experienced by the fosa. Species encounter and trap rates documented a shift in prey item encounters over time. Roadkill surveys, together with the roadkill death of one of the GPS-collared study animals, demonstrated failure of current mitigation efforts in addressing roadkill dangers. Fosa space use and activity patterns clearly reveal that they rely almost exclusively upon forested habitats, and that they avoid human settlements. Interestingly, although fosa do rely heavily on forest habitat, their ability to use forest edges and narrow forested strips suggests that corridors may be a viable method to enhance habitat connectivity and promote positive conservation outcomes.

Committee: Nancy Stevens (Advisor); Geoffrey Dabelko (Committee Member); Viorel Popescu (Committee Member) Subjects: Environmental Studies; Wildlife Conservation

Master of Arts, University of Toledo, 2016, Geography

Roadkill is a widespread occurrence in the United States. Particularly, deer-vehicle collisions (DVCs) pose societal, economic, and traffic safety concerns. DVC data was combined with landscape and roadway variables in a geographic information system (GIS) to examine factors associated with deer roadkill in Lucas County, Ohio. To explore patterns of DVCs, the data was analyzed for distribution, space-time trends, and hot spot clustering. Two regression methods were developed from 13 candidate variables to find contributing factors of landscape and roads experiencing statistically significant DVC incidents. In Ordinary Least Squares (OLS) regression, distance to parks & water, forest & habitat areas, traffic volume, time of year & day, and road contour were variables that significantly contributed to DVCs. In another method, Geographically Weighted Regression (GWR), it was found forest & building area, population, length of street networks, and distance to parks & water influenced DVCs. It is recommended to implement enhanced deer-crossing signage on the identified, critically high DVC routes.

Committee: M. Beth Schlemper Ph.D. (Committee Chair); Bhuiyan Alam Ph.D. (Committee Member); Patrick Lawrence Ph.D. (Committee Member) Subjects: Geography; Statistics; Transportation Planning; Wildlife Conservation; Wildlife Management

Master of Arts, Miami University, 2011, English

The Animal Remains is a collection containing three sections: straddling the shoulder, poems that utilize mixed media and lyric forms to interrogate the boundaries of gender; Telling Stories in the Window, poems that explore domestic spaces with autobiographical narratives, both linear and nonlinear; and Twins Talking, narrative prose poems concerning the feminine body in adolescence and the forming of identity through the disjointed retrieval of memory.

Committee: Keith Tuma Dr (Committee Chair); Christopher Cheek Dr (Committee Member); Catherine Wagner Dr (Other) Subjects: Animals