Age-structured dynamic, stochastic and mechanistic simulation model of Salmonella Dublin infection within dairy herds

In the demand for a decision support tool to guide farmers wanting to control Salmonella Dublin (S. Dublin) in Danish dairy herds, we developed an age-structured stochastic, mechanistic and dynamic simulation model of S. Dublin in dairy herds, which incorporated six age groups (neonatal, preweaned calves, weaned calves, growing heifers, breeding heifers and cows) and five infection states (susceptible, acutely infected, carrier, super shedder and resistant). The model simulated population and infection dynamics over a period of 10 years in weekly time steps as: 1) population sizes of each of the six age-groups; 2) S. Dublin incidence and number of animals in each infection state; and 3) S. Dublin related morbidity and mortality in the acutely infected animals. The effects of introducing one infectious heifer on the risk of spread of S. Dublin within the herd and on the duration of infection were estimated through 1000 simulation iterations for 48 scenarios. The scenarios covered all combinations of three herd sizes (70, 200 and 400 cows), four hygiene levels indicating infectious contact parameters, and four herd susceptibility levels indicating different susceptibility parameters for the individual animals in each of the six age groups in the herd. The hygiene level was highly influential on the probability that the infection spread within the herd, duration of infection and epidemic size. The herd susceptibility level was also influential, but not likely to provide sufficient prevention and control of infection on its own. Herd size did not affect the probability of infection spread upon exposure, but the larger the herd the more important were management and housing practices that improve hygiene and reduce susceptibility to shorten durations of infection in the herd and to increase the probability of extinction. In general, disease and mortality patterns followed epidemic waves in the herds. However, an interesting pattern was seen for acute infections and abortions in adult cattle after the first 2 years of infection in herds with poor hygiene and high susceptibility. Repeated infections in young stock lead to a high proportion of resistant adult cattle, which lead to a dampening effect on acute infections in adults and associated abortions. Sensitivity analyses of 24 alternative scenarios showed that a super shedder state was not essential to mimic the infection dynamics and persistence patterns known from field studies, but a persistent carrier state was required in the model to mimic real life S. Dublin infections.