Relating the pathogen genetic variability with the virulence variability enables the host susceptibility and infectiousness to be defined with respect to the variants of the pathogen that can be observed (simultaneously or not) in a herd. Previous work carried out by partner 1 reported a large variability in shedding routes (milk, vaginal mucus, faeces) and patterns over time (intermittent, sporadic, persistent) (Guatteo et al., 2006; 2007). It also was shown, using an original modelling-based Bayesian approach, that some herds are characterised by a very low infection risk while others have a mild infection risk and a non-negligible intermittent shedding probability (Courcoul et al., 2010). Moreover, different clinical patterns at both the animal and herd levels were observed, possibly in relation to variability in the involved strains and their associated virulence. Therefore, we will investigate the contribution of strain variability to the diversity of the shedding patterns (in terms of intensity of shedding and shedding routes as described in T1b) observed for Cb infections in cattle herds. Animals located in herds involved in an intervention study currently under progress and designed to assess the effectiveness of control strategies will be selected on the basis of their observed clinical and infection patterns. The Cb isolates from these animals will be cultivated and typed and their virulence level assessed using both in vitro and in vivo (on mice) models. Associations between the reported Cb virulence levels and the shedding patterns observed in animals according to the identified strains then will be investigated.

Depending on the results, “strain-dependent” within-herd models will be implemented from an existing simulation model representing Cb spread within an infected dairy cattle herd (Courcoul et al., 2010). We will account for the observed related variability in shedding patterns. The developed models will differ with respect to the significance of the health states considered and values of transition parameters between these states.

If several strains are observed simultaneously in a herd, a methodological issue will emerge. The formulation of the mathematical model can be complicated by the increase in the number of variants co-circulating, especially if interactions occur between variants (e.g. cross-immunity). It then will be necessary to formalise, based on published generic formulations of such multi-strains systems (Andreasen et al., 1997), a within-herd model of Cb spread considering co-circulating variants.