Critical parameters for modelling the spread of foot-and-mouth disease in wildlife.

A series of simulation experiments was conducted to determine how estimates of the latent and infectious periods, number of neighbours (contacts) and population size impact on the predicted magnitude and distribution of foot-and-mouth disease (FMD) outbreaks in white-tailed deer in southern Texas. Outbreaks were simulated using a previously developed and applied susceptible-latent-infected-recovered geographic automata model. There were substantial differences in the estimated predicted number of deer and locations infected, based on the model parameters used (3779-119 879 deer infected and 227-6526 locations affected). There were also substantial differences in the spatial risk of infection based on the model parameters used. The predicted spread of FMD was found to be most sensitive to the assumed latent period and the assumed number of contacts. How these parameters are estimated is likely to be critical in studies on the impact of FMD spread in situations in which wildlife reservoirs might potentially exist.