Transmission Dynamics of African Swine Fever Virus, South Korea, 2019

Figure 1. Spatial distribution of registered domestic pig farms in South Korea, indicating African swine fever–positive farms (IPs); ASFV-positive wild boars, confirmed during the study period (August 28–October 16, 2019); and pig farms visited by vehicles that had visited IPs >1 time during the study period. ASFV, African swine fever virus; IP, infected premises.

Figure 2. Timeline of reporting and culling of African swine fever virus IPs and control measures implemented during the African swine fever epidemic in South Korea, 2019. Reddish vertical shades represent movement restriction (standstill) imposed across the country (darker shades) or only in the affected municipalities (lighter shades). Numbers on the top represent the time when movement restriction was imposed. The colors of horizontal shades refer to IPs’ municipalities. IPs were numbered in the order of reporting dates. Over the course of the epidemic, six 48-hour standstill periods (bans on movements of livestock, persons, vehicles, and supplies to farms and slaughterhouses) were enforced across the country or in affected municipalities. IP, infected premises.

Figure 3. Number of farms registered in the government pig farm database that were emptied during the study period by culling or government purchase, South Korea, 2019. IP labels are shown on the day they were culled. Herds located within a 3-km radius around IPs and herds with an epidemiologic link to IPs (e.g., same ownership) were culled within 3 days after confirmation of African swine fever virus infection in IPs. As the epidemic developed, the government further depopulated remaining herds in affected municipalities as a preventive measure. *One farm stopped rearing pigs for reasons not associated with the epidemic. IP, infected premises.

Figure 4. Distribution of the number of African swine fever virus–infected premises connections through vehicle movements, South Korea, 2019. IP out-degree represents the number of other IPs to which an IP sent >1 vehicle; IP in-degree represents the number of other IPs from which an IP received >1 vehicle. With 1-day (A), 3-day (B), or 6-day (C) assumptions for the duration of vehicle infectiousness, only the movements made up to 20 days before an exit farm reported suspicion of ASFV infection and before an entry farm reported suspicion of infection were considered. In the boxplots, center horizontal lines represent medians, and box limits represent upper and lower quartiles. Upper and lower whiskers extend to the largest and smallest values within 1.5× interquartile ranges. The point represents an outlier. ASFV, African swine fever virus; IP, infected premises.

Figure 5. Spatial distribution of infected premises (IPs), non-IPs, African swine fever virus (ASFV)–positive wild boars, and potentially contaminating vehicle movements between IPs, South Korea, 2019. The duration of vehicle infectiousness was set to 3 days. Circles represent IPs; numbers represent the order of reporting dates. Pie charts show the estimated contribution of different transmission routes to the infection of each IP. Edge width is proportional to the number of potentially contaminating vehicle movements between IPs, weighted by the probability that an exit IP was infectious at the time of the vehicle departure. Edge arrows represent the direction of vehicle movements. Pig farm density is shown in reddish colors. Green squares represent the location of ASFV-positive wild boars; green-shaded ellipses represent spatial clusters.

Figure 6. Expected number of secondary farm cases of African swine fever (r) caused by 1 infected farm through the movements of vehicles, South Korea, 2019. r is computed as a function of the average daily number of vehicles visiting a farm (x-axis) and the average daily number of farms visited by a vehicle (y-axis). Different lines represent different thresholds for the proportion of iterations in which r was <1 (p = 1, 0.99, or 0.95). Vehicles were assumed to remain infectious for 3 days after leaving an infected farm. Appendix Figure 9 shows the results with different assumptions on the duration of vehicle infectiousness.