Graduation Year

2024

Document Type

Dissertation

Degree

Ph.D.

Degree Name

Doctor of Philosophy (Ph.D.)

Degree Granting Department

Marine Science

Major Professor

Cameron Ainsworth, Ph.D.

Committee Member

Steven Murawski, Ph.D.

Committee Member

Ernst Peebles, Ph.D.

Committee Member

Michael Schirripa, Ph.D.

Committee Member

Tracey Sutton, Ph.D.

Keywords

Atlantis, ecosystem modeling, trophic dynamics, vertical connectivity

Abstract

The focus of this dissertation was simulating trophic dynamics in the pelagic Gulf of Mexico (GOM) with an emphasis on ecosystem-based fishery management (EBFM) strategies for large pelagic species. Fisheries management in the United States largely utilizes single-species assessments, although managers are increasingly interested in EBFM approaches to better account for ecosystem effects on stocks such as predator/prey dynamics. Production in the surface layer of the pelagic zone is variable with environmental conditions and prey is often concentrated in patches by hydrodynamic features. Production is also transported vertically via diel vertical migration, during which deep-sea species provide potentially key forage resources for epipelagic predators. Locating prey in the pelagic zone can be a difficult task for visual predators, more so than in inshore habitats. Management of exploited pelagic predators could benefit from more investigation into how trophodynamics affect populations. The objectives of this work were to 1) explore spatial patterns and environmental drivers of pelagic feeding, 2) improve the estimation of pelagic diets in an ecosystem model of the GOM, and 3) evaluate ecosystem-based management strategies for a pelagic fish predator considering bottom-up trophic dynamics. I combined a hydrodynamically informed particle-tracking model with an Atlantis ecosystem model of the GOM to simulate feeding opportunities available to juvenile sea turtles depending on dispersal location. I found that food availability to juveniles was greatest offshore along the West Florida Shelf edge, and that hydrodynamic features including frontal zones facilitated retention in these areas. I then explored different statistical distributions to improve the representation of pelagic feeding dynamics in the GOM Atlantis model. To capture the mechanism of locating patchy prey in the pelagic environment, I applied zero inflated-type beta distributions to diet data that separately model the binomial process of encounter probability and the continuous beta distribution for diet composition. I found the zero-inflated beta model to be an improvement over a traditional beta distribution for fitting offshore predator diets. These improved diet estimates were incorporated into the GOM Atlantis model. Fitted diets showed that large offshore fishes heavily utilized deep-sea prey species, which could result in increased vulnerability to deep water disasters such as oil spills. I then used the GOM Atlantis model to conduct a management strategy evaluation that examined whether it was beneficial to consider the availability of mesopelagic fish prey in control rules for a large pelagic predator. Biomass and catch of the example predator, Thunnus albacares, were higher under ecosystem-based harvest control rules than under single-species and constant F management strategies. The benefit of EBFM increased as mesopelagic fish abundance decreased, suggesting that proactive EBFM for large pelagic fishes may be particularly strategic in the case of future deep-sea disturbances. This project demonstrates the utility of statistical and numerical modeling to simulate pelagic trophic dynamics and evaluate potential EBFM approaches.

Download

Share

COinS