Abstract
Movement is a central process in ecology. How organisms interact with critical resources and their environment and the way that these interactions shape movement patterns and spatial use is consequently of great interest to wildlife conservation planning and management. Agent-based modeling (ABM/ABMs) provides a flexible framework well-suited to incorporate fitness-seeking behaviors and decision-making and thus realistically simulate animal movement and spatial use. This dissertation aimed to develop, test, and validate an ABM for elephant movement and spatial use and to extend the ABM to explore how spatial variation in critical resources influence variation in herbivore movement, spatial use, and foraging efficiency. I first used GPS data from African savanna elephants (Loxodonta africana) in two southern African parks to identify seasonal and intraspecific differences in spatial use characteristics, including home range size and displacement distances. I then developed and validated a spatially-explicit ABM to simulate the movement and spatial use of elephants by linking the internal and external drivers of elephant movement through hierarchical behavior-based movement rules. To qualitatively explore model performance, I assessed whether the model was able to reproduce the twelve movement patterns present in the empirical data. For a quantitative comparison, I also used linear mixed models to identify differences in the movement characteristics between the simulated and empirical data. Finally, I used the validated ABM and artificial landscapes to explore how spatial differences in critical resources, including water and tree cover, translate into differences in spatial use and foraging efficiency of large, water-dependent herbivores. The results of my dissertation encourage the development, validation, and application of agent-based models in movement ecology, particularly for the exploration of ecological questions regarding animal movement.