Disease establishment within and among deer social groups: A neighborhood-based modeling approach

Disease transmission patterns can be greatly affected by social interactions, leading to uncertainty in the degree to which disease establishment and spread are linked to population density.  We developed a stochastic, neighborhood-based model of disease transmission within and among white-tailed deer (Odocoileus virginianus) social groups.  Relative contact rates within vs. between groups were based on data collected via global positioning system (GPS) collars from deer near Carbondale, Illinois, 2002-2006.  We also drew on published studies of how group size, home-range size, and home-range overlap vary with population density.  We then derived expressions for the basic reproductive number (R₀) of a disease, accounting for the finite numbers of group members and of overlapping social groups, assuming a random distribution of social group home ranges.  For directly (non-sexually) transmitted diseases we find that R₀, as traditionally defined, is weakly related to density due to the primacy of within-group transmission and relative stability of group size.  However, a group-based definition of R₀ (groups infected by the first infected group) showed a much stronger relationship with density.  Long infectious period relative to group fidelity complicates these results, depending on dispersal distance and whether emigrants join other groups or form their own through reproduction.  With indirect transmission, the breakdown of group-membership effects leads to consistent density dependence in R₀, although with a substantial time lag for highly persistent agents.  We evaluate the relative efficacy of control efforts to remove individuals or groups from the susceptible population, for directly and indirectly transmitted diseases.  Our findings highlight the interaction of transmission mode and social organization in understanding and managing disease in wild ungulate populations.