Semester

Fall

Date of Graduation

2013

Document Type

Dissertation

Degree Type

PhD

College

Davis College of Agriculture, Natural Resources and Design

Department

Wildlife and Fisheries Resources

Committtee Chair

John W. Edwards

Committee Co-Chair

Colin P. Carpenter

Committee Member

J. Todd Petty

Committee Member

Christopher W. Ryan

Committee Member

Michael P. Strager

Abstract

To date, no research studies in the eastern United States have addressed the spatial ecology of black bears (Ursus americanus) in urban and suburban habitats, and there is limited information regarding black bear space use, habitat selection, and harvest vulnerability. I assessed the harvest vulnerability, home range size, and spatial ecology of black bears in New Jersey, Pennsylvania, and West Virginia in collaboration with the New Jersey Division of Fish and Wildlife, Pennsylvania Game Commission, and the West Virginia Division of Natural Resources. The major objectives of my study were to identify and quantitatively assess: (1) if black bears shift home ranges seasonally with respect to urban/suburban habitats; (2) harvest vulnerability and cause-specific mortality of black bears in urban/suburban habitats; (3) if black bears captured as result of nuisance complaints are transient or reside in urban/suburban habitats; (4) habitat characteristics of movement corridors utilized by black bears in urban/suburban habitats; and (5) if corridors likely to be used by black bears accessing urban/suburban environments can be predicted by habitat modeling.;Over the course of the study (2010-2012), agency employees trapped, handled, and fit 119 bears with GPS-GSM collars. Individual study areas in each state were centered around West Milford, Stillwater/Branchville, and Vernon, NJ; Johnstown, Scranton/Wilkes-Barre, and State College, PA; Beckley, Charleston, and Morgantown, WV. A total of 57,816 bear locations were recorded in New Jersey, 114,451 locations were recorded in Pennsylvania, and 33,217 in West Virginia.;Black bears shifted spatial distribution on the landscape in response to resource availability, and consequently I expected bears in urban areas to shift their home ranges seasonally in urban environments. On average, bears were most often found near city limits (<5 km). Home range centers of male bears were twice as far from city limits as female bears (2.31 km vs. 0.91 km, respectively). Bear home range size did not differ among seasons, but did differ between sexes (male home ranges were 5.6 times larger than female bears) and among study areas. Bears did not shift their home ranges closer to urban areas during times of food shortage (spring or late fall). Urban bears lived near town and were resident to the edge of the urban area, but this distance varied with the study area in which they resided. As a result, managers seeking to understand where potential bear conflicts may occur should focus their efforts on the edge of urban and suburban areas (known as the exurban areas) in the Mid-Atlantic Region.;Regulated harvests have reduced mortality and allowed black bear populations to increase throughout the eastern United States over the past 30 years. This rapid and dramatic recovery in population size has led to increased human-bear interactions in New Jersey, Pennsylvania, and West Virginia. Harvest vulnerability of black bears is dependent on a variety of factors and therefore difficult to estimate. I measured harvest vulnerability by generating maximum entropy (Maxent) models of bear occurrence during the prehunting period and hunting season for each study area. I used Maxent to generate models of bear occurrence. In all but one study area, black bear occurrence was 5-75% less in the hunting season than in the prehunting period. Bear occurrence decreased from prehunting period to hunting season in both public hunting lands and urban areas. Bear occurrence probability shifted from public hunting areas to the periphery of the public hunting areas between the prehunting period and hunting season. Annual harvest rates of urban bears were highest in Pennsylvania (20.2%) and lower in New Jersey (5.9%) and West Virginia (17.3%). Despite the short timeframe (3 years) of my study, regulated hunting was effective in killing urban black bears Pennsylvania and West Virginia. Hunters in Pennsylvania harvested a similar proportion of urban bears to the long-term harvest rate of all bears in the state (20.2% vs. 20.0%, respectively). Probability of urban bear occurrence shifted from public hunting areas during the prehunting period to private lands on the periphery of urban areas during the hunting season. Average overall mortality rates of urban bears were highest in Pennsylvania (28.1%) and lower in West Virginia (17.5%) and New Jersey (15.1%). Despite the short timeframe of the study, regulated hunting was effective in killing a high number of urban black bears in New Jersey, Pennsylvania, and West Virginia, when all mortality sources were taken into account.;Black bear populations have increased nationwide over the past 4 decades due to reduction in direct mortality. Bear population sizes have increased over time and as a result, human-bear conflicts have increased. I sought to determine whether nuisance bears in urban/suburban areas are residents to the area or transient. I predicted that the majority of bears found in urban/suburban areas form resident populations on the urban perimeter, rather than transient individuals that leave the core forests and enter the urban areas periodically. There was considerable support for my prediction that bear populations in urban areas are resident and spend much of their time on the city's edge. Black bears used private lands on the periphery of urban areas. I posit that this may have been because these areas likely had abundant food and provided reduced risk of disturbance (e.g., hunting, human disturbance).;Given the recent explosive increase in urban bear populations, managers are charged to determine which areas of urban/suburban centers are likely to be used by bears. It is unknown whether black bears use travel corridors within urban/suburban matrix to travel between habitat patches. There exists a paucity of information on how black bears use urban and suburban habitats. We used boosted regression trees to create two predictive models of bear occurrence in urban and suburban habitats for (1) New Jersey and Pennsylvania, and (2) West Virginia. We separated West Virginia from New Jersey and Pennsylvania in the modeling process because West Virginia's topography is more rugged and the population density of people was the lowest of all three states. We randomly selected a subset of 40,000 bear locations in New Jersey and Pennsylvania, as well as, 30,000 bear locations in West Virginia from the full database of locations. We generated 40,000 random points within the study areas in New Jersey and Pennsylvania and 30,000 random points within the study areas of West Virginia. We built three models (1) for New Jersey and Scranton/Wilkes-Barre, PA, (2) State College and Johnstown, PA, and (3) West Virginia. We found that probability of bear occurrence was highest in New Jersey and Scranton/Wilkes-Barre study areas when bears were: (1) <1 km from edge forest, (2) <7.5 km from the nearest road, (3) <7.5 km from the nearest urban area, (4) land use/ land cover was forested, (5) <12 km from public land, and (6) NDVI < 0.3. We found that probability of bear occurrence was highest in Johnstown and State College study areas when bears were: (1) <1 km from edge forest, (2) <1 km from the nearest road, (3) <7 km from the nearest urban area, and (4) <7 km from public land. The highest probability of bear occurrence in West Virginia occurred when (1) NDVI was >0.6, (2) distance to public land was >22.0 km, (3) distance to urban areas was between 1-5 km, (3) topographic position index was >100 (steep, rugged terrain), (4) land use land cover was forested or "other", (5) distance to roads was >1.4 km, and (6) distance to core forest was >1.5 km. We found no support for our prediction that urban bears use corridors. Bears spent nearly 95% of their time on the edge of city limits and <5% of their time within city limits. We found no evidence that habitat quality on the edge of city limits was lower than that of "non-urban" bear habitat. There is likely not a physiological need for bears to traverse urban areas when they can remain in habitats where they would encounter less human disturbance. I found no support for my prediction that bears use corridors. Bears spent nearly 95% of their time on the edge of city limits and <5% of their time within city limits. I had no evidence that habitat quality on the edge of city limits was lower than that of non-urban bear habitat. There likely was no physiological need for bears to traverse urban areas because urban habitat patches are often safe from human disturbance and therefore, they did not use corridors. The final predictive model of the probability of bear occurrence will assist managers by identifying areas where urban bears are most likely to live and areas that require direct management actions.

Share

COinS