Exposure assessments
The probability that domestic and wild pigs in South Korea are exposed to imported pig products (Pex ) was estimated as follows:
Pex = PW × PM × (SF ×IE + LF × LFw × Wa )
The pig products legally imported into South Korea may be consumed or discarded and become food waste at selling points and at all consumer and distribution levels. In South Korea, food waste at all levels including households, restaurants, meal places at schools and institutions and distribution stage are thrown separately and collected. The pay-as-you-throw food waste management system has been implemented in all sectors of the country since 2010 (Ministry of Environment, 2010). The proportion of food waste (PW ) vs supplied food was estimated using a normal distribution (Ministry of Environment, 2017a). The proportion of food (PW ) was then multiplied by the proportion of meat (PM ) in food waste, which was also parameterized using a normal distribution (Chang et al. , 2003).
Once discarded and collected, food waste has different destinations: feeding animals mainly for domestic pigs after heat treatment, production of biogas and fertilizer and disposal in landfills. This model assessed two potential pathways of exposure/contact of pig products with domestic pigs: 1) by swill feeding (SF ) and 2) wild boars having access to landfills (LF ). The swill feeding practice is applied for swine and poultry farms in South Korea according to the Waste Control Act (Act No. 15103, Enforcement on 29 May 2018) and the Standard and Criteria of Feed (Notification of the Ministry of Agriculture, Food and Rural Development (MAFRA), No. 2017-28, Enforcement on 1 Apr 2017). The proportion of food waste used for feed was estimated using a normal distribution based on the data of the Ministry of Environment (2015, 2016, 2017b). Most of the food waste recycled for feed is used mainly in pig farms after heat treatment at 80 °C for at least 30 min in authorized facilities. However, the process of heating does not ensure the inactivation of the ASFV, based on the conditions of 90 °C for at least 60 min and 100 °C (212 °F) for at least 30 min proposed by the OIE (2018a) and APHIS (2009), respectively. Thus, a certain degree of uncertainty about the inefficiency of the process (IE ) was included using a Pert distribution (0.05, 0.1, 0.2).
The other potential pathway of contact with wild boars accessing landfills was estimated with 3 parameters: (LF ) The probability that the food waste is disposed in landfills, (LFw ) The probability of wild boars being present near landfills, (Wa ) The probability of wild boars accessing and feeding in the landfills. Firstly, the proportion of meat in the food waste disposed in landfills (LF ) was obtained (Ministry of Environment, 2015; 2016; 2017b). The probability that wild boars are present around landfills (LFw ) was estimated using a Pert distribution (0.860, 0.919, 0.978) based on the 4th investigation on the Environment of South Korea (National Institute of Ecology, 2015) and lists of landfills and incineration facilities (Ministry of Environment, 2011). Wild boars were observed in a 10 km (minimum) and 20 km (maximum) radius from each landfill in ArcGIS for Desktop software (Esri Inc. Redlands, CA, USA) considering the natural home ranges of wild boar are 10-20 km. Data on the probability of wild boar having access and feeding in landfills is currently not available. Therefore, the probability that wild boars have access to landfills (Wa ) was assumed conservatively with Pert (0.05, 0.1, 0.2) as proposed by a previous study (Herrera-Ibatá et al. , 2017). The parameters and descriptions are shown in Table 3. Event scenarios for the two routes are detailed in Figure 1.