Estimation of foot-and-mouth disease windborne transmission risk from USA beef feedlots

Windborne spread of foot-and-mouth disease (FMD) requires specific epidemiological and meteorological conditions, thus modeling the risk of windborne spread involves integrating epidemiological and meteorological models. The objective of this study was to investigate the potential risk of windborne spread of FMD from an infected US feedlot using an integrated modeling approach, and to identify factors that determine this risk. To address this objective, we integrated a within-herd epidemiological model and an advanced atmospheric dispersion model, and calculated infection risk dependent on exposed herd size. A previously developed epidemiological model was used to simulate the spread of FMD through a typical U.S. feedlot, while the National Oceanic and Atmospheric Administration's (NOAA) HYSPLIT atmospheric dispersion model, which has been validated for FMD modeling, was used to model virus dispersion. Infection risk for exposed herds was calculated as a binomial probability accounting for dose and exposed herd size. We modeled risk of windborne spread from a typical 4000 head feedlot in the U.S. state of Iowa (IA), and a typical 48,000 head feedlot in the U.S. state of Kansas (KS) during winter and summer seasons. The risk of windborne spread of FMD varied based on weather/season conditions, estimated average viral shedding rate per head, size of infected herd, and size of exposed herd. In the baseline Kansas scenario (KS/10³/W), the median of the maximum daily risk of infecting a 1000-head exposed herd ranged from 58.16 % at 1 km to 0.78 % at 10 km (Table 4). In the baseline Iowa scenario (IA/10³/W), the median of the maximum daily risk of infecting a 1000-head exposed herd ranged from 21.78 % at 1 km to 0.05 % at 10 km (Table 4). The minimum control area recommended by the United States Department of Agriculture (USDA) in an FMD outbreak is 10 km from the infected premise. Our results indicate that significant risk of windborne spread may extend beyond 10 km in certain situations. This is particularly a concern in areas where there are large feedlots in relatively close proximity, such as in southwestern Kansas. Our model may be useful as a research tool in the absence of an outbreak and may help direct surveillance and response efforts in the event of an outbreak.