A computational framework for studying energetics and resource management in sea turtle migration and autonomous systems.

Sea turtles complete migrations across vast distances, covering entire ocean basins. To track these migrations, satellite tracking tags are attached to their shells. The impact of these tags must be considered to ensure that turtles' natural behavior is not artificially and adversely impacted through tag-related drag, and that the data collected by a small sample of sea turtles accurately represents the larger population. Additionally, it can be difficult to study animal energetics in the field over large migration distances. In this work, we modify a computational behavior model to study how satellite tracking tags affect turtle migration behavior. Our agent based model contains synthetic magnetic field environments that are used for navigation cues, an ocean current, resource distributions that represent locations of food, and an agent that attempts to migrate to several different goals. The agent loses energy as it progresses, and searches for the resource distributions to replenish itself. Our novel simulation framework demonstrates the relationship between an agent's available energy capacity, its energy consumption based on mechanical power expended, and its ability to navigate to all migratory goal points. This study can be utilized to (1) probe the impacts of an animal's energy capacity and foraging behavior on its resulting navigation and ecology, (2) guide future satellite tag designs, and (3) develop usage recommendations for a suitable tracking tag based on the type of experiment being conducted. Our model can be expanded beyond sea turtles to study other marine species (e.g., sharks, whales). Additionally, this model could be expanded to other domains within the marine environment. For example, it could be modified to examine design trade-offs in remotely operated vehicles (ROVs), which share many of the same operational constraints as sea turtles and other migratory species.