UC Berkeley

Anthropogenic habitat fragmentation is a primary driver of species endangerment across the globe and has compounding effects on species diversity and ecosystem function. While efforts to enhance habitat connectivity are therefore essential to protecting biodiversity, a fundamental behavioral and ecological understanding of animal movement is first needed to successfully protect species movements. Understanding the role of behavior in determining animal movement patterns is essential to conservation planning, yet the extent to which an animal’s behavioral state (e.g. foraging, dispersing) influences its movements and resource selection has largely been ignored as part of conservation planning efforts. Further, because empirical studies of animal movement are most-often site- and species-specific, the processes underlying observed movement patterns are not well understood across taxa. This dissertation seeks to elucidate the processes that shape animal movements to advance the biological grounding of connectivity science and inform conservation efforts.

A systematic review of connectivity studies employing resource selection analysis examined how researchers have incorporated animal behavior into connectivity planning, and highlighted promising approaches for identifying wildlife corridors. The review revealed that most of the research done to date has superficially considered all animal location data as representative of resource selection patterns, despite recognition that an animal’s behavioral state can be an important component of space use. Those studies in the review that validated connectivity predictions with independent movement data indicate that general patterns of resource selection are not always a suitable proxy for movement preference during dispersal, and failure to recognize this distinction may have important consequences for species-specific efforts to preserve habitat connectivity.

Using high-resolution GPS and activity data from African wild dogs (Lycaon pictus), an Endangered species highly sensitive to habitat fragmentation, resource selection and responses to roads were measured across three behavioral states identified from activity data (hunting, resting, and traveling). The response of wild dogs to roads varied markedly with both the behavioral and landscape contexts in which roads were encountered, ranging between strong selection for and avoidance of roads depending on behavioral state. A comparison of these outputs to a full model that did not parse for behavior revealed that these patterns were not evident when all movement data were considered together in the full model. This study indicates that including behavioral information in resource selection models is critical to understanding wildlife responses to landscape features and suggests that successful application of resource selection analyses to conservation planning requires explicit examination of the behavioral contexts in which movement occurs.

The effects of behaviorally-mediated patterns of resource selection were then applied in a habitat connectivity modeling context. To illustrate the importance of behavioral information in connectivity assessments, behavior-specific predictions of connectivity were tested against long-distance dispersal movements of African wild dogs. Findings demonstrated that including only directed-movement behavior when measuring resource selection reveals far more accurate patterns of habitat connectivity than a model measuring resource selection independent of behavioral state. Results of this work suggest that connectivity studies that rely on resource selection analysis alone may be insufficient to target movement pathways and corridors for protection. This research highlights the value of incorporating animal behavior into connectivity planning.

To examine how basic movement processes scale up to produce emergent patterns for multiple species, movement data from over a dozen marine and terrestrial vertebrate species spanning three taxonomic classes, continents and orders of magnitude in body size were compared with computer-simulated idealized movement paths. This comparative approach revealed that similar movement patterns and properties recur in highly dissimilar ecological systems, and showed that a simple set of metrics can reliably classify broad-scale movement patterns such as migration, nomadism, and territoriality in disparate taxa. This classification system can be applied to inform predictions in multiple areas of ecological research, such as how an individual or species’ movement classification influences its response to climate change or its invasion potential in an exotic environment. In addition, this work provides researchers with a standardized set of movement metrics for expediently analyzing animal trajectories over time to detect any changes in movement pattern that may be indicative of environmental change.

Taken together, the body of work presented in this dissertation provides new approaches for researchers and practitioners to understand the ecology and conservation of animal movement, and in particular for measuring wildlife responses to widespread habitat alteration. Given limited conservation resources and rapidly changing environments, these contributions mark a key step in developing effective strategies to preserve critical wildlife movement processes.