Advances in Statistical Methods for Fish Tracking: Complementary Insights from Acoustic Telemetry and Archival Biologging
Authors: Marine Gonse (Laboratoire L3i, La Rochelle Université, France)
Abstract
Over the past two decades, fish tracking technologies have transformed the study of aquatic ecology by enabling researchers to investigate movement dynamics, habitat use, behavioral interactions, and responses to anthropogenic pressures across spatial and temporal scales. Among the most widely used approaches, acoustic telemetry and archival biologging have emerged as complementary tools for monitoring fish movements in coastal and offshore environments.
This chapter provides an overview of recent advances in statistical methodologies developed to analyze these increasingly complex datasets and address major ecological and conservation questions. The chapter first introduces the principles underlying acoustic telemetry and archival tagging systems, including sensor technologies, data acquisition protocols, detection infrastructures, and the ecological scales at which these methods operate. Acoustic telemetry relies on networks of fixed or mobile receivers detecting uniquely coded transmitters, allowing reconstruction of fish movements and residency patterns. In contrast, archival tags continuously record environmental and behavioral variables such as depth, temperature, acceleration, or light levels, providing detailed information on fish physiology and movement in areas where acoustic infrastructures are unavailable. Building on these methodological foundations, the chapter reviews the main statistical frameworks currently used to process and interpret fish tracking data, including state-space models, hidden Markov models, occupancy and residency analyses, network-based approaches, and spatiotemporal habitat modelling. These methods are discussed in relation to key ecological applications such as identifying essential fish habitats, characterizing migration dynamics, and habitat connectivity. The integration of telemetry-derived indicators into marine spatial planning and ecosystem-based management is also examined. Recent studies increasingly use fish tracking data to evaluate the ecological effects of offshore wind farms, marine protected areas, habitat fragmentation, and commercial fisheries. Statistical approaches supporting conservation prioritization, impact assessment, and fisheries management are reviewed, with emphasis on the challenges posed by heterogeneous, incomplete, and multi-source observational data. Despite considerable progress, important methodological limitations remain. Current analytical frameworks often rely on strong parametric assumptions, simplified movement representations, or manually engineered features that may inadequately capture the complexity and multiscale nature of fish behavior. Furthermore, the growing volume and multimodality of telemetry data raise challenges related to scalability, uncertainty quantification, missing detections, and data fusion.
The chapter concludes with a discussion of emerging data-driven methodologies, including machine learning, deep learning, graph-based models, and self-supervised approaches for behavioral inference and movement prediction. These approaches offer promising opportunities to move beyond traditional hypothesis-driven analyses toward more adaptive and integrative frameworks for aquatic movement ecology.
Representative Publications