Reconciling dynamic epidemiological models with long‐term outbreak data: The case of classical swine fever in Germany

Abstract Understanding the complex interplay between contact networks in social host species and individual movement decisions is essential for designing effective disease control strategies in wild animals. We use a spatially explicit eco‐epidemiological individual‐based model to investigate the effect of host movement decisions on disease spread and persistence and reconcile findings with the patterns of a long‐term outbreak dataset. Using alternative mechanistic host movement submodels, in which decisions where to move are affected by either landscape structure or density of conspecifics, we validate simulations of disease spread against the known long‐term patterns of spread of classical swine fever in wild boar in Northern Germany by applying the same sampling scheme as in the field. We compare simulated with observed data using three key metrics: age class distribution of infected hosts, speed of pathogen spread, and spatial distribution patterns of infected individuals. We found two main movement strategies matching the observed pathogen spread and spatial patterns: correlated, habitat‐driven movement and competition‐driven movement. Furthermore, the only movement strategy that was able to recreate the observed trend in the age class distribution of infected host individuals was the implicit movement, purely based on host density. Our results show the significant impact of habitat composition and host population density on disease outbreak dynamics.