The influence of farm connectedness on foot‐and‐mouth disease outbreaks in livestock

Abstract

We applied a previously published livestock foot‐and‐mouth disease (FMD) model to estimate host connectivity using a transmission kernel based on contact tracing and measured subsequent to an animal movement ban in the 2001 United Kingdom epidemic. Connectivity within county‐level farm landscapes were evaluated by considering the transmission kernel, host species composition, farm‐level susceptibility, farm‐level transmissibility, and distances between farms. The objectives were to (1) determine the impact of connectivity of the initially infected farm on the size of FMD outbreaks in four counties of differing connectivity; (2) compare FMD spread in counties that have different mean farm connectivity levels over all farms in the county; and (3) determine how connectivity relationships affect an example control measure. We used a spatially explicit stochastic model to simulate FMD outbreaks on livestock farm operations in four UK counties. Nine farms were selected for seeding infection, with each representing a different level of farm connectivity, measured by the sum of transmission rates between itself and all other farms. In each simulated outbreak, one farm was seeded with infection. In addition to the culling of infected premises within 24 h of being reported, as implemented in the FMD model, control of epidemic spread was implemented by culling farms within 1 and 3 km radii around infected premises. Increased levels of connectivity of the initially infected farm had a significant incremental effect on the epidemic length and the number of farms, cattle, and sheep lost. However, at higher levels of farm connectivity, these incremental effects were observed to plateau. Results showed significant variation in the level of overall farm connectivity between the counties, and counties with higher overall farm connectivity experienced increased simulated losses. Connectivity of the initially infected farm and mean connectivity among all farms in a county were strongly associated with effects of cull size, with disease control more effective at lower levels of farm connectivity. Host connectivity provides early information on the host‐pathogen landscape and could be used as an assessment tool for predicting epidemic risks, as well as enabling preemptive control strategies to limit the size of disease outbreaks.