Multiscale consensus habitat modeling for landscape level conservation prioritization

Area of Science:

• Ecology
• Conservation Biology
• Wildlife Management

Background:

• Wide-ranging carnivore populations face global threats from human-caused habitat fragmentation.
• Habitat quantification is a critical first step in conservation planning for fragmented populations.
• Species distribution models (SDMs) offer robust methods for habitat identification when appropriate scales and data are used.

Purpose of the Study:

• To identify critical black bear (Ursus americanus floridanus) habitat in Florida at both local and state scales.
• To assess potential future habitat fragmentation due to sea level rise and development.
• To inform conservation prioritization for this fragmented carnivore population.

Main Methods:

• Utilized presence-only location data and environmental variables to construct a consensus SDM.
• Employed Maxent and Mahalanobis distance algorithms for habitat modeling.
• Compared identified habitat areas with projections for sea level rise, human development, and existing protected areas.

Main Results:

• Local-scale models demonstrated superior performance compared to state-scale models.
• Identified 23,798 km² of habitat at the local scale and 45,703 km² at the state scale.
• Significant portions of identified habitat face threats: 10-14% from inundation, 7-16% from development, and 15-54% is currently unprotected.

Conclusions:

• Florida black bear habitat is susceptible to fragmentation, with local-scale models providing more accurate assessments.
• Sea level rise and development pose substantial risks to current and future black bear habitat.
• Urgent need for focused conservation strategies and improved connectivity between subpopulations to ensure long-term population stability.