Geographical Information Systems: Landscape Patterns and Habitat Associations of Mammals
Type of Presentation
Poster Session
Location
University Library
Start Date
4-9-2026 2:00 PM
End Date
4-9-2026 3:15 PM
Description of Program
This presentation will utilize GIS to access various habitat types in the Upper Peninsula of Michigan. Most changes in habitat will be due to logging and forestry. Occurrences of different mammals will be analyzed with the corresponding habitat types to see if there are any patterns.
Abstract
Protected areas play a critical role in conserving biodiversity, supporting ecosystem services, and enhancing human wellbeing. Habitat loss and fragmentation, primarily driven by urbanization, industrialization, and logging, remain leading threats to global biodiversity, impacting mammal species. Fragmentation alters habitat structure, creating edge effects, shifting successional stages, and sometimes favoring disturbance-adapted or non-native species. Logging, a widespread anthropogenic disturbance, has both ecological costs and benefits. Logging roads and edge habitats influence mammal behavior and distribution, particularly among carnivores. In the Upper Peninsula of Michigan, logging is the primary driver of habitat loss. The Upper Peninsula is home to many large carnivores, including bears, canids, mustelids, and felids, which are especially vulnerable due to large home ranges and low population densities. This study employs camera trapping data to assess mammal occurrence across landscapes. Integrating GIS and ecological landscapes, this project evaluates spatial patterns and habitat characteristics. Specifically, I hypothesize that large mammal occurrence varies by habitat type and may shift spatiotemporally in response to logging activities. Five GIS data layers are used to test the hypothesis: landcover (LandSat), roads, human density (Tiger Files), wetlands, and digital elevation models (DEMs). Preliminary results suggest wolves and bears appear to be inversely correlated and roads seem to be a factor in occurrence of mammals. Mammals also appeared to be inversely correlated with clear cuts, as their presence was not detected in these areas.
Identify Grant
Funded in part by the NSF Optimization Computing Grant
Faculty / Staff Sponsor
Dr. John Yunger
Geographical Information Systems: Landscape Patterns and Habitat Associations of Mammals
University Library
Protected areas play a critical role in conserving biodiversity, supporting ecosystem services, and enhancing human wellbeing. Habitat loss and fragmentation, primarily driven by urbanization, industrialization, and logging, remain leading threats to global biodiversity, impacting mammal species. Fragmentation alters habitat structure, creating edge effects, shifting successional stages, and sometimes favoring disturbance-adapted or non-native species. Logging, a widespread anthropogenic disturbance, has both ecological costs and benefits. Logging roads and edge habitats influence mammal behavior and distribution, particularly among carnivores. In the Upper Peninsula of Michigan, logging is the primary driver of habitat loss. The Upper Peninsula is home to many large carnivores, including bears, canids, mustelids, and felids, which are especially vulnerable due to large home ranges and low population densities. This study employs camera trapping data to assess mammal occurrence across landscapes. Integrating GIS and ecological landscapes, this project evaluates spatial patterns and habitat characteristics. Specifically, I hypothesize that large mammal occurrence varies by habitat type and may shift spatiotemporally in response to logging activities. Five GIS data layers are used to test the hypothesis: landcover (LandSat), roads, human density (Tiger Files), wetlands, and digital elevation models (DEMs). Preliminary results suggest wolves and bears appear to be inversely correlated and roads seem to be a factor in occurrence of mammals. Mammals also appeared to be inversely correlated with clear cuts, as their presence was not detected in these areas.