Degree

Doctor of Education (EdD)

Department

Educational Leadership

Document Type

Dissertation

Abstract

This dissertation investigates the temporal ecology of sperm whales (Physeter macrocephalus) in the northern Gulf of Mexico (Gulf of America), using multi-year passive acoustic recordings. Motivated by limited winter coverage in traditional visual surveys, this work examines how sperm whale presence and demographic composition inferred from their acoustic phonations vary across seasons and diel periods in a highly industrialized deep- water ecosystem. Two processed acoustic datasets are utilized: click-rate time series derived from raw acoustic data and size estimates, generated by the CABLE software package (Beslin, 2018), which infers whale body length from inter-pulse interval (IPI) measurements. First, several generalized linear models are compared to quantify seasonal and diel patterns in hourly click activity. Based on model selection using the Bayesian Information Criterion (BIC), zero-inflated negative binomial (ZINB) models provided the best fit for the click-rate data, capturing both overdispersion and the large number of zero-detection hours. This model structure offered the most parsimonious and biologically appropriate representation of sperm-whale acoustic activity. The estimated model parameters for different seasons are used to predict seasonal baseline activity which reveals strong winter and early-spring enhancements in baseline click rates and consistent daytime elevation in acoustic activity across all seasons. The second analytical component applied a multipath- based length-estimation method to classify detected individuals into broad demographic categories, Through CABLE (Cachalot Automatic Body Length Estimator). This analysis revealed consistent year-round presence of females and juveniles, with relatively stable seasonal patterns. In contrast, adult male detections were more variable and showed weaker seasonal structure overall, suggesting more wide-ranging movement or lower site fidelity. Both demographic groups exhibited a modest increase in predicted presence during the fall, indicating a shared seasonal enhancement in activity during that period. Across both datasets, long-term patterns align with prior Gulf studies linking cool-season activity to prey-field shifts, regional hydrography, and reduced low-frequency anthropogenic noise. By combining seasonal, diel, and demographic analyses, this dissertation contributes a new statistical modeling framework which links zero-inflated count models with autoregressive logistic presence models. This approach establishes a reproducible foundation for future PAM-based ecological and conservation research.

Date

13-1-2026

Committee Chair

Donna Wadsworth

DOI

https://proquest.com/docview/3292481434

First Committee Member

Amanda Mayeaux

Second Committee Member

Nathan Roberts

Download

Share

COinS