MODELING PATHOGEN DYNAMICS FOR DISEASE CONTROL IN MARINE AQUACULTURE

Epidemics of pathogens continue to plague marine aquaculture throughout the world carrying much economic and environmental cost. Despite their significance little work has been done on modeling their epidemiology. Epidemic models hold the potential to better understand disease outbreaks, the extent of those outbreaks, and their control. Three important pathogens in marine aquaculture are White spot syndrome virus (WSSV) and Taura syndrome virus (TSV) of marine shrimp and Amyloodinium ocellatum, a dinoflagellate of finfish.  Models of these three pathogens exhibit a spectrum of variation and elaboration of epidemic models. WSSV and TSV are modeled by following the host population rather than the pathogen directly. A host population experiencing an outbreak of WSSV can be divided into Susceptible (S), Infected living (I), and infected Dead (D) hosts. In this simplest of models there is a straight line of transition of hosts from S to I to D. A more complex model is developed for TSV which has two additional host categories, Prepatent (E) and Chronically (C) infected hosts. The complication arises from TSV Infected hosts either dying from infection to become infected carcasses (D) or recovering to become Chronically infected (C). A. ocellatum is modeled by following the dynamics of the pathogen's life cycle stages directly. The stages are Trophont, Tomont, and Dinospore. The models' dynamics depend upon the value of parameters such as transmission from the various infected stages, pathogen induced mortality, recovery from acute infections, carcass decay, and carcass consumption control the dynamics of the models. The value of the parameters coupled with pathogen life history differences are responsible for the variation in potential spread of the various pathogens and their potential for control in aquaculture.

In addition to the development of the models, laboratory estimates have been made of the important parameters that control the dynamics of infections. By plugging these estimates into the models we can evaluate the potential of a particular pathogen to spread (basic reproduction number, R₀). Further we can investigate the relative importance of each coefficient to an epidemic through sensitivities and elasticities. From such analyses control methods can be focused on modifying parameters that will result in the greatest reduction in epidemic potential.