Comparing habitat suitability and connectivity modeling methods for conserving pronghorn migrations

Area of Science:

• Ecology
• Conservation Biology
• Wildlife Management

Background:

• Terrestrial long-distance migrations are declining globally, with significant losses in North America.
• Habitat fragmentation poses a major threat to migratory species, necessitating research on effective corridor identification.
• Limited research exists comparing different modeling approaches for identifying migration corridors in fragmented landscapes.

Purpose of the Study:

• To compare the effectiveness of different habitat suitability and connectivity models in identifying pronghorn migration corridors.
• To evaluate the performance of maximum entropy (Maxent) and expert-based (Analytic Hierarchy Process) habitat suitability models.
• To assess the utility of least-cost modeling (LCM) and circuit theory for connectivity analysis.

Main Methods:

• Habitat suitability was modeled using Maxent and Analytic Hierarchy Process, incorporating variables like distance to wells, water, roads, NDVI, land cover, terrain shape, and fence presence.
• Connectivity was modeled using least-cost modeling and circuit theory, with habitat suitability outputs serving as cost surfaces.
• Pronghorn GPS movement data from spring and fall migrations were used to validate the identified corridors.

Main Results:

• The combination of Maxent habitat suitability models and LCM for connectivity analysis yielded the best results for identifying pronghorn migration corridors across both seasons.
• Maxent models outperformed expert-based models in predicting habitat suitability for pronghorn migrations.
• LCM corridors contained a higher number of pronghorn GPS locations compared to circuit theory corridors, indicating superior performance.

Conclusions:

• Maxent coupled with LCM provides a robust approach for identifying critical migration corridors in fragmented landscapes.
• Expert-based models offer a cost-effective alternative for corridor identification when species location data are scarce.
• A tiered approach to corridor conservation, considering varying widths and mitigation actions like fence removal, is recommended for effective conservation planning.