Author ORCID Identifier
https://orcid.org/0009-0005-5048-6795
Defense Date
2026
Document Type
Dissertation
Degree Name
Doctor of Philosophy
Department
Systems Modeling and Analysis
First Advisor
Edward Boone
Second Advisor
Grace Chiu
Abstract
This dissertation develops stochastic approaches for reconstructing fish movement trajectories from satellite and acoustic telemetry data. A fundamental challenge in analyzing such data is the presence of gaps in tracking records, resulting from undetected locations and irregular, sparse observations. Addressing these gaps is critical for improving our understanding of fish movement behavior and for supporting effective marine conservation and management. The first part of the work focuses on satellite telemetry data. We propose an interpolation method to recover missing locations that occur when satellite signals are lost. Movement variability is then estimated using both maximum likelihood and Bayesian approaches, with performance assessed through simulation studies. The second part examines acoustic telemetry data where stationary receivers detect fish only within a limited range. This setting imposes spatial constraints, requiring that imputed trajectories neither cross land barriers nor enter the detection zones of unactivated receivers. To address this, we develop a spatiotemporal probabilistic model that incorporates these constraints, using a von Mises distribution to characterize directional movement, and includes a behavioral threshold to distinguish between directed movement and random exploration. Parameter estimation is performed within a full Bayesian framework using a hybrid Gibbs and Metropolis–Hastings MCMC algorithm, and the approach is validated through an oracle simulation study. Finally, we introduce a web-based application that translates the proposed Bayesian framework into an accessible tool for ecologists and fisheries managers, enabling trajectory reconstruction and interactive visualization without requiring specialized expertise in statistical modeling or Bayesian computation.
Rights
© The Author
Is Part Of
VCU University Archives
Is Part Of
VCU Theses and Dissertations
Date of Submission
5-4-2026
Included in
Applied Statistics Commons, Biometry Commons, Data Science Commons, Statistical Methodology Commons