Marine mammal monitoring has seen improvements in the last few decades with advances made to both the monitoring hardware and post-processing computation methods. The addition of tag-based hydrophones, Fastloc Global Positioning System (GPS) units, and an ever-increasing array of IMU sensors, coupled with the use of energetics proxies such as Overall Dynamic Body Acceleration (ODBA), has led to new insights into marine mammal swimming behavior that would not be possible using traditional secondary-observer methods. However, these advances have primarily been focused on and implemented in wild animal tracking, with less attention paid to the managed environment. This is a particularly important gap, as the cooperative nature of managed animals allows for research on swimming kinematics and energetics behavior with an intricacy that is difficult to achieve in the wild.

While proxy-based methods are useful for relative inter-or-intra-animal comparisons, they are not robust enough for absolute energetics estimates for the animals, which can limit our understanding of their metabolic patterns. Proxies such as ODBA are based on filtered on-animal IMU data, and measure the aggregate high-pass acceleration as an estimate for the magnitude of the animal’s activity level at a given point in time. Depending on its body structure and locomotive gait, tag placement on the animal and the specific filtering techniques used can significantly impact the results. Any relation made to energetics is then strictly a mapping:​ a relation that may apply well to an individual or group under specific experimental conditions, but is not generalizable. To address this gap, this dissertation presents new tag-based hardware and data processing methods for persistently estimating cetacean swimming kinematics and energetics, which are functional in both managed and wild settings.

Unfortunately, localization techniques for managed environments have not been thoroughly explored, so a new animal tracking method is required to spatially contextualize information on swimming behavior. State-of-the-art wild cetacean localization operates via sparse GPS updates upon animal surfacings, and can be paired with biologging-tag-based odometry for a continuous track. Such an approach is hindered by the structure and scale of managed environments:​ GPS suffers from increased error near and within buildings, and current odometry methods are insufficiently precise for habitat scales where locations of interest might be separated by meters, rather than kilometers (such as in the wild). There is then a need for a tracking method that uses an alternate source of absolute animal locations that can achieve the high precision necessary for meaningful results given the spatial scale. To this end, this dissertation presents a novel animal localization framework, based on tracking and sensor filtering techniques from the field of robotics that have been tailored for use in this setting.

To summarize, this dissertation targets two main gaps:​ 1) the lack of persistent, absolute estimates of animal swimming energetics and kinematics, and 2) the lack of a robust, precise localization method for managed cetaceans. To address these gaps, the hardware and animal tracking methods developed to enable the rest of the dissertation are first defined. Next, a physics-based approach to directly monitor cetacean swimming energetics is both presented and implemented to study animal propulsion patterns under varying effort conditions. Finally, a high-fidelity 3D monitoring framework is introduced for tracking institutionally-managed cetaceans, and is applied alongside the energetics estimation method to provide a first look at the potential of spatiallycontextualized animal monitoring.