Date of Graduation


Document Type


Degree Name

Master of Science in Biology (MS)

Degree Level



Biological Sciences


Brett DeGregorio

Committee Member

John D. Anderson

Second Committee Member

Eunchun Park


Armadillo;Density;Mesocarnivore;Occupancy;Random Encounter Model;Urban to rural gradient


Human development is a rapidly expanding phenomenon impacting wildlife communities globally. As the human footprint expands, natural greenspace and nature reserves become fragmented and isolated. Wildlife species and communities respond to these changes in different ways. Urban exploiters persist in more urban areas and are able to take advantage of anthropogenic subsidies (food, shelter, etc). Urban avoiders, however, are species that may be pushed out of developed areas or occur in lower densities relative to undisturbed areas. Northwest Arkansas is an optimal location to study the impacts of urbanization on wildlife communities, as the current human population is expected to double in the next 20 years and human development is continually encroaching into natural areas. In Chapter I, I evaluated mammalian mesocarnivore density along an urban to rural gradient in the Ozark Ecoregion, Northwest Arkansas, USA. I deployed 287 trail cameras across 12 natural areas ranging from urban parks within the city of Fayetteville to more rural sites in the Ozarks. To estimate density, I used the Random Encounter Model (REM), developed to estimate density of unmarked individuals via trail camera data. I focused my analyses on the five most commonly detected mesocarnivore species: bobcat, coyote, raccoon, red fox, and Virginia opossum. I found that both raccoons and coyotes densities had a positive correlation with anthropogenic noise, indicating an affinity for areas with human activity. Red fox and opossum densities both had a positive correlation with developed open space, representing novel, human-maintained habitat within urban areas such as golf courses, cemeteries, and city parks. Red fox density also had a negative correlation to distance to water, and opossum density had a positive though insignificant trend with distance to water. Bobcat density was not correlated with any variables selected for analysis and may need to be evaluated on a larger spatial scale. My results support past research showing coyote, red fox, opossum and raccoon to be urban exploiters. They also add support for the REM as a reliable method to estimate density based on trail camera data. My density estimates for all five species were comparable to density estimates from the literature that used traditional density estimation techniques and allowed us to evaluate trends in density based on environmental and anthropogenic resources. In Chapter II, I evaluated the density and occupancy probability of the nine-banded armadillo (Dasypus novemcinctus) along an urban to rural gradient. I used a total of 352 camera locations from the 2020-2021 winter season and the 2021-2022 winter season. I found that detection probability increased for the armadillo as temperature increased but decreased as precipitation increased. Occupancy probability had a negative correlation with the angle of the slope, and a positive correlation with elevation. Armadillo density was not influenced by any anthropogenic or environmental variables, but anecdotally was higher at sites with higher levels of human development. Collectively, these results shed insight into how 6 different mammal species respond to human development across an urbanization gradient in a quickly growing part of the country. Understanding impacts of urbanization on wildlife allows managers to predict populations of wildlife and potential conflicts that may arise in response to changes in their natural environments, as well as make well informed management decisions.