UC Berkeley

Protecting ecosystems, corridors, and working landscapes is critical to addressing rapid global change. Given the pervasiveness of the human footprint, managers and policy-makers often seek a clearer picture of how development affects habitat conservation and species persistence. This dissertation used a combination of remote sensing, data synthesis, and empirical data collection to examine how protected areas have been impacted by development within and beyond their borders, as well as how terrestrial species have behaviorally responded to anthropogenic features in altered landscapes. This work measured ecosystem change spanning three scales, from global protected areas, to community-level ecological interactions, to individual-level change in animal behavior.

Habitat connectivity between protected areas and surrounding land influences protected area effectiveness. More isolated protected areas often exhibit reduced species movement, dispersal, and genetic diversity, resulting in wildlife decline. To quantify protected area isolation globally, in Chapter 2, I developed a habitat-edge detection approach to measure change between protected areas and surrounding land over time. I found a significant 20-year loss in habitat continuity along the borders of the world’s largest protected areas — the strongest evidence to date that protected areas have begun to resemble habitat islands in a human-dominated world. While habitat discontinuity may have negatively affected wild animals by disrupting patterns of movement or behavior, others may have used human-altered environments to gain resources or safety, with each type of response resulting in distinct outcomes for species interactions and wildlife persistence. In Chapter 3, I delved into this topic with a framework I used to examine the behavioral effects of human activity on 178 predator-prey pairs from published camera trap studies. I found four primary patterns of predator-prey response to people, highlighting the context-dependency of development on species interactions. While some wildlife species likely displayed a consistent behavioral response to disturbance, many species may have adjusted their behavior based on the level, frequency, or length of exposure to disturbance. To measure how behaviorally flexible species navigate landscapes with both development and natural features, in Chapter 4, I evaluated coyote (Canis latrans) habitat selection in a mixed-use agricultural landscape. I found that coyotes avoided development, but selected for roads and agricultural areas despite possible persecution and conflict. This study revealed a few of the risks, rewards, and tradeoffs that behaviorally plastic carnivores face in developed landscapes. To conclude, I summarize the conservation implications of this research, and call for incorporating the social, economic, and political drivers of land-use into wildlife ecology to promote effective and socially-conscious conservation into the future.