Identifier

etd-01062017-142507

Degree

Master of Science (MS)

Department

Oceanography and Coastal Sciences

Document Type

Thesis

Abstract

Fish experiencing blooms of the cyanobacteria genera Microcystis and Anabaena acquire microcystin and saxitoxin through ingestion of contaminated food and absorption of dissolved toxin. Even low chronic doses induce sensory and motor impairment—the impact of which is unquantified in wild populations. Here, I introduce Lagrangian particle models for cyanobacteria and fish which test the hypotheses that impairment symptoms suppress movement and growth. This is implemented within the Finite-Volume Coastal Ocean Model (FVCOM). Cyanobacteria particles move vertically according to mixing and buoyancy (a function of cellular reservoirs). Fish navigate the horizontal domain, foraging in high growth areas, and fleeing when toxin increases. The framework is demonstrated here for the case of juvenile fish encountering Microcystis aeruginosa in an idealized Louisiana estuary. Self-shading reduces bloom growth, and causes algae to collect at the surface. Turbulent diffusivity is insufficient to break up this layer, so dissolved toxin becomes surface-intensified. Fish seek high growth areas in this environment, and dietary uptake increases. This triggers flight and swimming impairment. As cyanobacteria excrete microcystin, absorption forces fish to become intoxicated even in areas of lower toxic risk. Repeated flight means fish spend more time in suboptimal areas, with final growth reduced up to 6.6%. In vivo, this would be exacerbated by physiological stress and the metabolic cost of toxin removal. Collective movement (group diffusivity) is suppressed nearly 50% during wide-spread intoxication. Simulations show that within a certain parameter space, both movement and growth are suppressed relative to the control case as expected. However, additional experiments resulted in higher growth, indicating the methods are sensitive to model parameterization. Ultimately, these are sandbox cases, which will require carefully-designed lab and field experiments before predictive capability can be assumed.

Date

2016

Document Availability at the Time of Submission

Release the entire work immediately for access worldwide.

Committee Chair

Huang, Haosheng

Share

COinS