Evaluating habitat suitability and connectivity for the endangered northernmost population of jaguars (Panthera onca)
2020 Spring. ; Includes bibliographical references. ; Understanding the distribution of wildlife habitat as well as landscape connectivity for threatened or endangered wildlife populations provides valuable information for conservation planning efforts. The northernmost population of jaguars (Panthera onca) is threatened with extinction, and has been extirpated from much of its historic range in the United States due to habitat loss/fragmentation, hunting, and poaching. Recent efforts by the United States government to expand U.S.-Mexico border infrastructure threaten to further fragment jaguar habitat in the borderlands. Models of habitat for jaguars in the United States have been developed, but they can be improved by using a finer analytical resolution, appropriately accommodating uncertainty by using a statistical framework, and using more appropriate data. Existing connectivity models have also been at coarse resolution, been broad in scope, and have not explicitly considered to effects of U.S.-Mexico border infrastructure. The Draft Recovery Plan for the jaguar was released by the United States Fish and Wildlife Service in 2017 and identified a need for additional research on jaguar habitat use and to identify key habitat and movement corridors in the U.S.-Mexico borderlands. To address these research needs, the goal of my thesis research was to provide updated and improved models of habitat and landscape connectivity for the northern population of jaguars in Sonora and Chihuahua, Mexico and Arizona and New Mexico, United States. For Chapter one, I developed a novel statistical model and applied it to predict and explain habitat selection by the northern population of jaguars. The study area for this chapter encompassed the Madrean Sky Islands (a complex of small mountain ranges in southeastern Arizona, southwestern New Mexico, and northwestern Mexico) and surrounding areas. Like many imperiled species and populations, data on jaguars are sparse, which limits our ability to gain insight into their ecology. To maximize inference, I developed a novel integrated Bayesian model that makes use of both presence-only and detection/non-detection data to model habitat selection. Results show that terrain ruggedness (+) and distance to riparian vegetation (-) are key predictors of habitat selection. There is a mean predicted 25,463 km2 of jaguar habitat in the study area. A mean predicted 40.6% of this habitat lies in the United states, suggesting that habitat in the U.S. could play an important role in the long term persistence of the northern jaguar population if jaguars are able to recolonize the region. In Chapter two, I turn focus to evaluating landscape connectivity, particularly in the context of U.S.-Mexico border infrastructure. I modeled landscape connectivity using Circuitscape, where the resistance surface was derived from the habitat suitability model from Chapter one in combination with human land use intensity and data on functional movement barriers (roads and border infrastructure). I evaluated the impacts of present-day border infrastructure as well as a likely future border scenario by running models for three different border infrastructure scenarios: 1) A scenario with no border infrastructure in place, 2) the present day border infrastructure scenario, and 3) a possible future scenario in which present day vehicle barriers are converted to pedestrian fencing. The resulting connectivity maps revealed that existing border infrastructure has far reaching consequences for habitat connectivity in the borderlands, and border wall expansion threatens to further isolate jaguar habitat in the United States from the breeding population in Mexico.
