Modelling the spread of Bovine Viral Diarrhoea Virus (BVDV) in a managed metapopulation of cattle herds

In numerous epidemiological models developed within a metapopulation framework, it is assumed that a single infected individual introduced into a patch infects the whole patch and that the proportion of infected individuals into infected patches is consistent over time and among patches. If this approach is relevant for rapidly spreading pathogens, it is less appropriate for moderately spreading pathogens, like the Bovine Viral Diarrhoea Virus (BVDV), characterized by a variability in within-patch prevalence. Our objective is to study the respective influence of neighbouring relationships and animal movements on the spread of BVDV in a managed metapopulation of 100 cattle herds. Infection dynamics is represented by two coupled stochastic compartmental models in discrete-time: a within-herd and a between-herd models. Animal movements are mechanistically modelled. They largely influence the BVDV persistence, the prevalence in infected herds and the epidemic size. Neighbouring relationships only influence epidemic size. Whatever the neighbouring relationships, the infection does not persist in the metapopulation without animal movement between herds. The proposed model can be easily adapted for different herd contact structures.