Date of Award
Doctor of Philosophy (PhD)
Mark J. Butler IV
Frank P. Day
Agent-based models explicitly incorporate interactions and variation at the individual level, just as in the real world. As a result, agent-based models are realistic, intuitive, and generally less complex mathematically than their analytical counterparts. Their primary disadvantage is the large amount of detailed data required to construct and parameterize them. Although the use of agent-based simulation is increasing in ecology, they are highly specific, so are rarely used for development of theory. To demonstrate the flexibility and utility of this approach I developed a multi-species, agent-based, spatially-explicit model of the spiny lobster nursery of southern Florida that incorporates changing salinities, temperatures, and harmful algal blooms, and used it to evaluate the likely consequences of changes in hydrology due to the Comprehensive Everglades Restoration Program on the spiny lobster and associated hard-bottom communities. I found that water quality changes associated with the restoration effort will likely reduce lobster recruitment in Florida Bay 6–24% and will likely prevent the restoration of loggerhead sponges and vase sponges on which the lobster depends for shelter in Florida Bay. Then, I extended the model, incorporating disease dynamics appropriate to Panulirus argus Virus 1, including disease states, contact transmission, and two density-independent transmission mechanisms, to explore the effects of changes in host sociality on disease dynamics. I showed that when susceptible hosts avoid diseased conspecifics in a manner consistent with that demonstrated by P. argus in a system dominated by contact transmission, persistence of the disease requires an exogenous source. I also examined the effects of increased host aggregation on disease transmission. I found that in the absence of disease avoidance, outbreaks occurred rapidly, and by the end of the 10 year simulation, PaV1 was maintained continuously at unrealistically high levels. However, the disease avoidance behavior reduced simulated outbreak intensities and durations, and in the absence of other sources of PaV1, resulted in extinction of the disease within five years. Both the density independent infection of EBJs and the simulated arrival of infected cohorts of postlarvae maintained the disease in the larger population as a consequence of the long period between exposure and death.
Dolan, Thomas W..
"Spatially-Explicit Agent-Based Modeling of Ecosystem Change and Epizootiological Impacts on Caribbean Spiny Lobster, Panulirus argus"
(2011). Doctor of Philosophy (PhD), Dissertation, Biological Sciences, Old Dominion University, DOI: 10.25777/rx4s-mm17