Bayesian Analysis for Inference of an Emerging Epidemic: Citrus Canker in Urban Landscapes

Neri, Franco M.; Cook, Alex R.; Gibson, Gavin J.; Gottwald, Tim R.; Gilligan, Christopher A.

For other diseases, policy makers and disease control authorities may wish to wait to assess the likely severity of the infestation in order to consider the likely costs and benefits of control; delay may also be necessary to mobilise resources. Informed decision making invokes a series of questions about how to make inferences about the emerging epidemic: what type of epidemiological model can be used to characterise the epidemic and to predict future spread of disease? Where are the susceptible hosts and how are they distributed in the landscape? How is disease transmitted and what are the values of the epidemiological parameters for transmission and dispersal? How soon during the course of the epidemic can the parameters be reliably estimated? How should we take account of uncertainty? Here, we examine these questions using a combination of Bayesian statistical inference and a unique, spatially-and temporally-resolved data-set [28] for the invasion of a plant disease, Asiatic citrus canker, in Florida.