Files
Abstract
Outbreaks of infectious disease events in the marine environment often occur with little warning and can have severe consequences for their host populations. Infectious disease events in corals have increased over the last decade, coinciding with the continuing decline of the reef environment on a global scale. Understanding and predicting future disease events will be a critical step towards developing a response strategy. This dissertation provides a multi-scale approach towards understanding white pox disease, a disease that affects the import reef-building, yet critically endangered, coral, Acropora palmata. I first explore the diversity found within the surface mucus layer of A. palmata. I accomplish this through by measuring the alpha diversity from mucus samples collected from healthy, bleached, and diseased colonies. Species richness was greatest in samples from diseased corals. Seasonality was an important driver in distinguishing microbial communities. I also developed a statistical framework to identify factors influencing local disease transmission. Using this framework, I fit models to data collected during an outbreak of white pox disease and determined spatial diffusion provided the best fit. Using simulations, I then evaluated how censored surveillance data influenced model performance. Last, using biological and environmental data obtained over a 20-yr time period, I constructed a machine learning model that predicted disease occurrence in individual A. palamta colonies. This approach used a large set of environmental variables to predict disease presence or absence. Collectively, these results suggest that microbial communities are different between healthy and diseased, proximity to nearby infected is important for disease transmission, and disease event can be predicted by colony size, dissolved saturated oxygen, wind speed, and organic carbon.