Movement-assisted localization from acoustic telemetry data.

Area of Science:

• Ecology
• Bioacoustics
• Wildlife tracking

Background:

• Acoustic telemetry is widely used for studying aquatic animals, generating large datasets on movement and distribution.
• Efficient analysis of acoustic data requires accurate localization of animal positions from receiver detections.
• Current localization methods often fail to utilize the full encounter history data available from receiver arrays.

Purpose of the Study:

• To develop a novel framework for acoustic telemetry localization by integrating explicit sub-models for movement, signal rate, and detection probability.
• To leverage spatial capture-recapture (SCR) methods to enhance localization accuracy and data utilization.
• To enable localization with fewer than three detections and predict positions even when individuals are not detected.

Main Methods:

• Analogous acoustic telemetry localization challenges to spatial capture-recapture (SCR) methods.
• Developed a movement-assisted localization framework integrating movement, signal rate, and detection probability sub-models.
• Applied generalizable Bayesian formulations of the SCR movement-assisted localization model.

Main Results:

• Movement-assisted localization models improved point-wise RMSE of localization estimates by over 50% compared to using only detection history.
• Achieved greater precision in estimated animal trajectories.
• Integrating a signal rate sub-model reduced biases in movement, signal rate, and detection parameter estimations.

Conclusions:

• Movement-assisted localization offers a flexible framework to maximize acoustic telemetry data utility.
• Conceptualizing localization within an SCR framework allows for broader applications and improved efficiency in studies of movement, resource selection, and space-use.
• Provides a unifying approach for modeling acoustic telemetry data.